Octopus as predators of Haliotis laevigata on an

abalone sea ranch of south-western Australia

Submitted by

Claire Greenwell

This Thesis is presented for the degree of Bachelor of Science Honours

School of Veterinary and Life Sciences, Murdoch University, 2017

ii

Declaration

I declare that this Thesis is my own account of my research and contains as its main

content work, which has not previously been submitted for a degree at any tertiary

education institution.

Claire Nicole Greenwell

iii

Acknowledgements

Firstly I would like to thank Brad Adams, and the team at Ocean Grown Abalone.

Without your generous support this project would not have been possible. I consider

myself privileged to have had the opportunity to undertake this research in such a

unique ecosystem. To Mark Wall, thank you for all your efforts, from collecting

octopus samples, the explanation of processes, and taking us out diving. Your

enthusiasm, hard work, and positive attitude is inspiring. Thanks also to Steve Chase

and the dive team for your efforts and making us feel welcome on the ranch. I wish you

all the best for the future and look forward to watching the operation grow.

To my supervisors, Neil Loneragan, James Tweedley, and Ryan Admiraal thank you for

your friendship, support and encouragement. I have much admiration and respect for

you all. Ryan, thank you for your perseverance and help in tackling the data analysis.

Despite some pretty nightmarish coding, you have a way of making stats fun. I am

extremely grateful for your help. James, thank you for answering my countless

questions, and for the many hints and insights along the way. It has been great working

alongside you with your passion, enthusiasm and brilliant ideas. And to Neil, I would

like to express my deepest gratitude. Thank you for your kindness, for lending an ear,

and your words of wisdom. For keeping me on track, and expressing the importance of

work-life balance. You are an inspirational leader; the inclusiveness and encouragement

that you show towards your students and peers is extraordinary. I am fortunate to have

you as a mentor, and as a friend.

Thank you to Sarah Poulton, Jake Daviot, Jason Crisp, Theo Campbell, and Thea Linke

for making my days in the lab light hearted. To Brian Poh, thanks for your

encouragement, help with dissections, and handy hints along the way. Thank you to my

many helpers in the laboratory, specifically Brett Newmarch, Amber Bennett, Stefan

Brown, and Jenny Browne for your help in the processing of stable isotope samples.

Removing epiphytes is lacklustre work, but you each did it with a smile. Thanks to Sian

Glazier for guiding me through the isotope methods. Also to Mike van Keulen for your

iv

photography skills, and the use of your personal and microscope cameras, and also to

John Huisman for assisting in the identification of algal species. Special thanks goes to

the State Agriculture and Biotechnology Centre for granting access to their ball mill,

and to Chris Florides for providing training on operational procedures.

To all my family and friends, thank you for your love and support. To Wendy

Newmarch, thank you for taking the time to proof read my work. A special mention to

Alex Thornton – sharing the journey and bouncing ideas off each other has forged a

special friendship, and I thank you for that. I’d like to give a special thanks to my

partner Brett, not only for your efforts in the lab, but also for your tremendous support

throughout my Honours journey. It hasn’t been easy, but you made the difficult times

much lighter, so thank you.

Finally, I’d like to thank the Banksia Association for their generous Banksia Honours

Scholarship, and the School of Veterinary and Life Sciences for providing financial

support for the stable isotope analyses.

v

Abstract

Octopuses play key ecological roles within coastal marine environments around the

world. Their short-lived, rapid-growing lifecycle commands high feeding rates, which

has the potential to impose strong top-down controls on the benthic communities where

they are found. This Thesis investigated two major questions on the predatory role of

octopuses on an abalone sea ranch in south-western Australia: (i) does the distribution

of octopus on the sea ranch affect the mortality of Greenlip Abalone, Haliotis laevigata

(Chapter 2), and (ii) what is the diet and nutrient assimilation (δ13C and δ15N) of

Octopus cf. tetricus in this ecosystem (Chapter 3).

Data were collected by commercial divers over a 27-month period on octopus

abundance and the number of empty abalone shells at ~fortnightly to monthly intervals,

and the number of abalone surviving on artificial abalone habitats (“Abitats”) every six

months. These data were used to examine whether the number of empty shells, and the

counts of surviving abalone were related to the presence of octopus. Negative binomial

generalised linear models showed that the presence of octopus had a significant impact

on the number of empty abalone shells collected and estimated that the shell counts are

78% higher when O. cf. tetricus is present when adjusting for location (“Line”) and

Season. The Abitats provide an ideal habitat for octopuses, offering shelter and an

abundant food supply. The relationship between octopus and shell counts was also

influenced by the spatial position of the Abitats (Line) and Season, which may be linked

to environmental variability and the time at which abalone are seeded. The results from

time series (AR1) models for the relationship between octopus presence and abalone

survival were not significant, contradicting those from the negative binomial models.

This result is likely an artefact of the random sampling design for counting abalone to

estimate survival and uncertainty in the number of abalone seeded onto each Abitat. A

subsample of 110 shells collected by the divers was examined for evidence of octopus

predation. Twenty shells (18%) had a small, slightly ovoid hole with a bevelled edge,

consistent with the holes made by octopuses. These results confirm that octopuses are a

major source of mortality on the sea ranch, supporting the results of the negative

binomial models.

vi

Gastric tract and stable isotope analyses were undertaken to determine the

dietary composition of different sizes of O. cf. tetricus, and its significance as a predator

of H. laevigata on the abalone sea ranch. The crops and stomachs of 44 individuals

were examined to assess whether diet differed between these digestive organs, or with

increasing body size (< 300 g, 300 to 999 g, ≥ 1,000 g). A much higher proportion of

material could be identified in the crops (%Volume, %V = 96%) than in the stomachs

(26%). Molluscs contributed the largest %V in the crops (~31%) followed by

crustaceans (33%), with unidentified material only contributing ~4%. In contrast, 74%

of stomach contents were not identifiable, with molluscs comprising the majority of

identifiable material (~19%), followed by crustaceans (~6%). The dietary composition

of O. cf. tetricus did not change with increased body size (from 58.5 to 1596.1 g),

paralleling the findings of stable isotope analysis. However, only four octopus > 1,000 g

were available for analysis.

The nitrogen (δ15N) and carbon (δ13C) isotope signatures of octopus, fish,

abalone, and benthic primary producers were examined to understand the assimilation

of nitrogen and carbon by O. cf. tetricus, and the trophic position of this species within

the Flinders Bay sea ranch food web. Examination of δ15N values revealed that O. cf.

tetricus (8.08 ± 0.19) occupies a mid-trophic level, slightly below teleosts (10.10 ±

0.22) and loliginid squids (9.47 ± 0.27). The δ13C signature of O. cf. tetricus (-23.63 ±

0.42) was similar to three benthic teleosts, Coris auricularis and Opthalmolepis

lineolatus, and Upeneichthys vlamingii, highlighting the overlap in use of benthic food

resources by several consumers in this system. While gastric tract analyses suggest that

O. cf. tetricus is an important predator of H. laevigata, stable isotope analyses show a

less obvious trend. This is indicated by the enrichment of δ13C by ~4‰ between

abalone and octopus. The results from this Thesis provide information on the predatory

role of O. cf. tetricus, their significance as predators of H. laevigata, and their trophic

positioning within this coastal marine system.

vii

Table of Contents

Acknowledgements ......................................................................................................... iii Abstract ............................................................................................................................. v Chapter 1. General introduction ................................................................................... 1

1.1. Octopus cf. tetricus ................................................................................................ 1

1.2. Fisheries enhancement and application of artificial reefs ...................................... 3

1.3. Study aims .............................................................................................................. 5

Chapter 2. Distribution of octopus and mortality of Haliotis laevigata on an abalone sea ranch in south-western Australia .............................................................. 7 2.0. Abstract ...................................................................................................................... 7 2.1. Introduction ................................................................................................................ 8 2.2. Methods .................................................................................................................... 12

2.2.1. Study site ........................................................................................................... 12

2.2.2. Lease configuration ........................................................................................... 13

2.2.3. Culture and seeding of Abitats .......................................................................... 15

2.2.4. Monitoring abalone and predators .................................................................... 16

Predator surveys .......................................................................................................... 16

Abalone surveys .......................................................................................................... 18

2.2.5. Data exploration and statistical analyses ........................................................... 18

Relationship between octopus presence and empty abalone shells ............................ 19

Relationship between octopus presence and abalone survival .................................... 23

2.3. Results ...................................................................................................................... 24 2.3.1. Patterns of octopus distribution and empty abalone shells ............................... 24

2.3.2. Relationship between octopus presence and empty abalone shells .................. 29

2.3.3. Relationship between octopus presence and abalone survival .......................... 31

2.4. Discussion ................................................................................................................ 32 2.4.1. Patterns of octopus distribution and empty abalone shells ............................... 33

2.4.2. Relationship between octopus presence and empty abalone shells .................. 35

2.4.3. Relationship between octopus presence and abalone survival .......................... 37

viii

Chapter 3. The diet and nutrient assimilation (δ13C and δ15N) of Octopus cf. tetricus on an abalone sea ranch of south-western Australia .................................... 39 3.0. Abstract .................................................................................................................... 39 3.1. Introduction .............................................................................................................. 40 3.2. Materials and methods ............................................................................................. 44

3.2.1. Collection of Octopus cf. tetricus ..................................................................... 44

3.2.2. Quantification of Octopus cf. tetricus diet ........................................................ 44

3.2.3. Stable isotope preparation ................................................................................. 47

3.2.4. Analyses of dietary data .................................................................................... 49

Comparison of diet between gastric organs ................................................................ 49

Size related differences in diet .................................................................................... 51

Stable isotope composition of prey and primary producers ........................................ 52

3.3. Results ...................................................................................................................... 52 3.3.1. Comparison of diet between gastric organs ...................................................... 52

3.3.2. Size related differences in diet .......................................................................... 56

3.3.3. Stable isotope composition of prey and primary producers .............................. 58

3.4. Discussion ................................................................................................................ 61 3.4.1. Comparison of diet between gastric organs ...................................................... 62

3.4.2. Size related differences in diet .......................................................................... 63

3.4.3. Stable isotope composition of prey and primary producers .............................. 66

Chapter 4. General conclusions and recommendations ............................................ 70 4.1. Conclusions .............................................................................................................. 70 4.2. Future research and monitoring ............................................................................... 73

4.2.1. Monitoring abalone survival ............................................................................. 73

4.2.2. Dietary studies ................................................................................................... 73

4.2.3. Stable isotopes ................................................................................................... 74

4.3. Implications .............................................................................................................. 74 References ....................................................................................................................... 77

Appendix 1. Cephalopods and their role as predators and trophic intermediaries in the food webs of coastal marine systems ..................................................................... 88

1

Chapter 1. General introduction

This Thesis focuses on investigating the predation by octopus on Greenlip Abalone,

Haliotis laevigata, on a sea ranch in Flinders Bay, south-western Australia (Fig. 2.1). It

builds on an extensive literature review (Appendix 1) that evaluates the role of

cephalopods as predators in coastal systems, to examine octopus as predators in the

Flinders Bay system. This general introduction provides the overall framework for the

Thesis, leading into the research aims and key findings of the literature review.

Functioning as mesopredators in the global oceans, coleoid cephalopods

(i.e., cephalopods lacking an external shell) provide an important link between apex

predators and lower trophic levels, with the ability to influence the structure and

function of marine systems (Anderson 2000; Yick et al. 2012; Tanner et al. 2017).

Advanced behavioural, morphological, and physiological traits such as a short life-

history strategy, voracious feeding regimes, and a well-developed brain that facilitates

rapid learning, has enabled the coleoid cephalopods to thrive within the marine

environment (Hanlon and Messenger 1996; Boyle and Rodhouse 2005; Rodhouse et al.

2014). Their ability to modify their foraging behaviour in response to novel feeding

opportunities, the use of spatial memory, and active modification of habitats (by

octopuses) has enabled these molluscs to adapt readily to anthropogenic-driven changes

within coastal systems (Mather 1991; Rigby and Sakurai 2005; Finn et al. 2009). For

example, heavy fishing pressure of their teleost competitors, the utilisation of artificial

habitats, and ocean warming are having a positive influence on cephalopod abundance

(Polovina and Sakai 1989; André et al. 2010; Coll et al. 2013; Doubleday et al. 2016).

1.1. Octopus cf. tetricus

Octopus cf. tetricus is a relatively short-lived (≤ 18 months), medium sized (< 4 kg)

coastal octopus species endemic to south-western Australia (Amor et al. 2014; Caputi et

al. 2015; Hart et al. 2016). This species inhabits rocky reefs, seagrass beds and sandy

substrata at depths of up to 70 m, extending from Shark Bay in the north, to the South

Australian border (Fig. 1.1; Norman 2003; Amor et al. 2014; Hart et al. 2016). Until

2

recently it was believed that Octopus tetricus formed a single biological stock that

extended from the east coast to the west coast of Australia (Caputi et al. 2015).

However, phylogenetic analysis has since confirmed that the west coast species is

distinct from O. tetricus on the east coast and does not overlap in its distribution (Amor

et al. 2014; Caputi et al. 2015). Therefore, O. cf. tetricus on the west coast of Australia

is denoted with a cf. until the species is able to be reclassified (Caputi et al. 2015).

Figure 1.1. Known distribution of Octopus cf. tetricus on the west coast, and Octopus tetricus on the east coast of Australia. Adapted from Amor et al. (2014).

The lifecycle of O. cf. tetricus is characterised by a planktonic paralarval phase

(i.e., they are merobenthic), rapid juvenile growth, early maturity (~270 days for males

or 950 g, and ~380 days or 1800 g for females) and post-spawning mortality

(i.e., semelparity, Joll 1976; Leporati et al. 2015, see also Appendix 1, Fig. A1.1).

Females lay a single clutch of eggs containing ≤ 150,000 eggs and tend to them for ~30

days before the eggs hatch in large pulses over several days (Joll 1976; Leporati et al.

2015). Females are able to mate prior to reaching sexual maturity and store sperm for up

to 16 weeks, while males mate with multiple immature and mature females until they

reach senescence; increasing the probability of reproductive success (Joll 1976;

Leporati et al. 2015).

3

Octopus cf. tetricus is the most important commercially fished octopus species

in Western Australia (Hart et al. 2013b; Caputi et al. 2015). It is targeted in the

Developmental Octopus Fishery (Shark Bay to South Australian border), Cockburn

Sound (Line and Pot) Managed Fishery, and is landed as incidental by-catch in

significant quantities in the West Coast Rock Lobster Fishery (Shark Bay to Cape

Leeuwin; Hart et al. 2013b; Caputi et al. 2015).

1.2. Fisheries enhancement and application of artificial reefs

Aquaculture-based fisheries enhancement technologies are extensively utilised as

fisheries management tools around the world. Their primary goal is to improve harvests

and food security by rebuilding depleted stocks, increasing recruitment and

productivity, and supporting natural systems (Munro and Bell 1997; Bell et al. 2008;

Taylor et al. 2017). An estimated two-thirds of the world’s coastal fisheries are

exploited, overexploited, or in decline, and the global population continues to expand

(Botsford 1997; Leber et al. 2004; Worm et al. 2009; Ye et al. 2013) Aquaculture-based

enhancement, and the use of habitat enhancement structures that help to replenish

depleted stocks are critical to addressing productivity and easing the pressure on wild

fish populations (Leber et al. 2004; Taylor et al. 2017). Fisheries enhancement goes

beyond simply “good” harvest management, adopting practical solutions that help to

overcome ecological limitations in natural systems (Taylor et al. 2017). This includes

the use of artificial habitats to increase habitat availability for target species, and the

culture of hatchery-reared individuals to increase recruitment and improve stocking

densities (Munro and Bell 1997; Taylor et al. 2017; Broadley et al. 2017).

The release of cultured juveniles as a management intervention for improving

fisheries can be defined in three broad categories; stock-enhancement, restocking and

sea ranching, each with a distinct set of management objectives (Bell et al. 2008;

Loneragan et al. 2013; Taylor et al. 2017). Unlike restocking and stock enhancement,

animals released as part of sea ranching operations are not expected to contribute to

spawning biomass, with the simple aim of increasing yield by harvesting individuals at

4

a larger size through “put, grow, and take” operations (Bell et al. 2008; Loneragan et al.

2013). However, the effectiveness of fisheries enhancement is dependent on the ability

of an ecosystem to support additional cultured individuals (i.e., resource availability and

the carrying capacity of the population), which may be limiting, particularly in the

absence of suitable habitat (James et al. 2007; Loneragan et al. 2013). In such instances,

artificial habitats can play an important role in supporting increased productivity

(Bohnsack 1989; Briones-Fourzàn and Lozano-Álvarez 2001; Leitão 2013; Wu et al.

2016).

Artificial reefs are defined as natural or manufactured objects purposefully

submerged in aquatic systems with the aim of modifying biological, physical, or socio-

economic processes (Seaman 2000; Seaman 2008). They are common coastal features;

used globally to restore ecosystems and to increase the biomass of commercially and

recreationally important stocks (Leitão 2013; Cresson et al. 2014a; Mazzei and Biber

2015; Wu et al. 2016). With their often complex design, artificial reefs provide an

important substratum and enable the settlement of a diverse array of marine flora and

fauna which, in turn, provide an abundant food source for higher trophic level

consumers (Leitão 2013; Cresson et al. 2014b). Purposely designed structures used to

support species-specific culture have been effectively implemented in a number of sea

ranching operations (e.g., the sea cucumber Apostichopus japonicus and Haliotis spp.),

achieving the dual aim of fisheries enhancement and habitat improvement (James et al.

2007; Melville-Smith et al. 2013; Zhang et al. 2015; Xu et al. 2017).

Years of research and development have culminated in Australia’s first

commercial abalone (H. laevigata) sea ranch, operated by Ocean Grown Abalone

(OGA, www.oceangrownabalone.com.au). Established in the shallow coastal waters of

Flinders Bay on the south-west coast of Australia, the operation successfully combines

aquaculture-based enhancement with species-specific artificial reefs.

Despite significant advances in global fisheries enhancement and artificial reef

design technology, predation pressure remains a significant threat to the survival of

5

released juveniles (Støttrup et al. 2008; Hines et al. 2008; Bell et al. 2008; Poh et al.

2017). Octopus cf. tetricus has been identified as a potentially important predator of H.

laevigata (Hart et al. 2013a; Melville-Smith et al. 2013). Numerous incidences of

predation by octopus have been observed in stock enhancement trials and sea ranching

operations in south-western Australia (Hart et al. 2013a; Melville-Smith et al. 2013),

which could have significant negative impacts for the survival of released individuals.

1.3. Study aims

The primary objective of this study is to assess the predatory impact of O. cf. tetricus on

H. laevigata within the commercial sea ranch operated by OGA in Flinders Bay,

Western Australia. This has been achieved by investigating:

(i) the spatial and temporal pattern of distribution and abundance of O. cf. tetricus

and H. laevigata across the sea ranch (Chapter 2) to evaluate whether the

presence of octopus is correlated with abalone mortality;

(ii) the diet of O. cf. tetricus to understand how the dietary composition of this

predator changes according to the analysis of digestive organ (i.e., crops vs

stomachs), with increasing body size, and whether H. laevigata forms an

important component of the diet (Chapter 3); and

(iii) the assimilation of nutrients within the food web of Flinders Bay to determine

the source of carbon assimilated by O. cf. tetricus and its position in the food

web, using stable isotope analysis (Chapter 3).

The knowledge gained from these studies will provide a greater understanding of

the role of octopus in structuring coastal systems, particularly in Western Australia, and

their significance as predators of abalone in a rapidly growing sea ranching operation.

The understanding and directions for my Honours research were developed through the

extensive literature review (comprising 134 references) in Appendix 1, which

synthesised knowledge on the role of coleoid cephalopods as predators and trophic

intermediaries in the food webs of coastal marine systems. This review included a meta-

analysis of the diets of 24 coastal coleoid cephalopod species from the Loliginidae,

6

Octopodidae, Sepiidae, and Sepiolidae and highlighted distinct dietary patterns among

these families. These differences were related to variation in lifestyle between the

groups (i.e., benthic vs pelagic) and in general found a rapid shift in diet with ontogeny,

whereby a broader range of prey types, including mobile fish, and other cephalopods

were ingested by larger individuals (Appendix 1).

7

Chapter 2. Distribution of octopus and mortality of Haliotis laevigata on an abalone sea ranch in south-western Australia

2.0. Abstract

This study examined the distribution and abundance of octopus across an abalone,

Haliotis laevigata, sea ranching site situated in Flinders Bay, Western Australia to

understand how octopus may be impacting the production of abalone. Octopuses are

important predators among benthic communities, and the artificial habitats used for the

production of H. laevigata provide an ideal habitat for these predators, offering shelter

and an abundant food supply. The study used data collected during “predator” surveys

by commercial divers. A total of 64 predator surveys were conducted over a 27-month

period at ~fortnightly to monthly intervals, resulting in the removal of 654 octopuses

and collection of 17,666 empty abalone shells from 5,308 artificial abalone habitats

(“Abitats”). Octopuses were widely distributed across the sites, with some locations

(“Lines”) observing a greater number of octopuses, which may be associated with the

number of Abitats within a Line. The abundance of empty abalone shells was higher in

some areas, which may be explained by a combination of factors including

environmental variability and the spatial arrangement of Abitats. A subsample of 110

shells was examined for evidence of octopus predation; 18% of these had a small,

slightly ovoid hole with a bevelled edge, consistent with the boreholes made by octopus.

A negative binomial generalised linear model was used to assess the relationship

between the presence of octopus and empty abalone shell counts, and indicated that

octopus presence is associated with an estimated 78% increase in the number of empty

abalone shells per Abitat per day, adjusting for Line and Season. A time series model

was used to assess the relationship between the presence of octopus and abalone

survival (number of abalone on Abitats), but did not find a significant relationship.

These contradictory findings are quite possibly an artefact of the sampling method for

counting abalone, and uncertainty in the numbers of juvenile abalone seeded onto the

Abitats. They highlight the importance of undertaking repeated measures on the same

Abitats over time to better account for the variability that naturally exists between

Abitats to measure survival with greater precision.

8

2.1. Introduction

The global production of wild abalone, Haliotidae has been in decline since the 1960s,

stimulating significant investment in fisheries enhancement techniques (i.e., sea

ranching and restocking) in most abalone-producing countries of the world (e.g.,

Australia, Japan, Mexico, South Africa, and USA, Prince 2004). Increased demand for

seafood and anthropogenic stressors, particularly overfishing, climate change, and

habitat degradation have had significant negative ecological impacts on coastal marine

systems in recent times (Bell et al. 2008; Pearce et al. 2011; Taylor et al. 2017). Thus,

novel enhancement and farming techniques, including the use of artificial reefs that

provide suitable habitat to increase productivity, are fundamental to ensuring the

longevity of food security (Molony et al. 2003; Zhang et al. 2015; Wu et al. 2016;

Taylor et al. 2017). For example, artificial oyster-shell reefs have been successfully

implemented in China, enhancing production of the highly sought after sea cucumber

Apostichopus japonicus (Zhang et al. 2015); and advances in the collection, culture, and

management of the scallop Patinopecten yessoensis in Japan have seen the

reinvigoration of the collapsed fishery (Uki 2006).

However, enhancement efforts of coastal fisheries are not always successful

(e.g., Grimes 1998; Brown and Day 2002; Bell et al. 2005, 2008). While a series of

experimental trials have successfully demonstrated the proof of concept, efforts to

increase abalone production using artificial habitats have been limited (Schiel 1993;

James et al. 2007; Roberts et al. 2007; De Waal et al. 2013; Hart et al. 2013c). This is

due to ecological constraints including a lack of suitable habitat and strong density-

dependent processes in the larval phase (Shepherd et al. 2001; De Waal and Cook 2001;

Bell et al. 2008; Hart et al. 2013c). As Hart et al. (2013c) point out; enhancements are

unlikely to be successful when animals are released at sizes at which density-

dependence processes are influential. Thus a good grounding of species biology,

ecology, and the factors influencing the carrying capacity of a population (i.e.,

recruitment and mortality) are critical if fisheries enhancement is to be effective

(Shepherd et al. 2001; Hart et al. 2013c).

9

In addition to density-dependent effects, high mortality of transplanted or

cultured juveniles (i.e., seeds) presents a significant challenge and has the potential to

impede the economic viability of fisheries enhancement projects (McCormick et al.

1994; Shepherd et al. 2001; James et al. 2007; De Waal et al. 2013; Hart et al. 2013c).

Newly released seeds can have poor behavioural responses towards predators since

culture conditions lack the stimuli and environmental cues required for the development

of instinctive survival patterns that are essential for survival in the wild (Bell 1999;

James et al. 2007; Daniels and Watanabe 2010). Laboratory experiments on cultured

and wild Red Abalone, Haliotis rufescens, on artificial reefs show that wild individuals

moved swiftly into concealed positions and suffered significantly lower predation rates

by crabs, sea stars, and lobsters than cultured individuals, which initially tended to be

relatively inert in their response to predators (Schiel and Welden 1987). However, as

abalone acclimatised to the artificial reefs, their behavioural response to predators

improved, resulting in lower predation rates compared with abalone released soon after

being taken from the hatchery (Schiel and Welden 1987).

Crabs and fish, including wrasses, have been identified as predators of juvenile

abalone < 2 years old (Schiel and Welden 1987; Shepherd 1998; Shepherd and Clarkson

2001; James et al. 2007). In a long-term study of the Blue-throated Wrasse, Notolabrus

tetricus, at West Island in South Australia, abalone mortality was positively correlated

with wrasse density, with larger wrasses typically preying on larger abalone (Shepherd

and Turner 1985; Shepherd 1998; Shepherd and Clarkson 2001). Predation on abalone

may be amplified when juvenile densities are increased through the addition of cultured

juveniles to the environment due to opportunism by fishes. However, other studies by

Shepherd et al. (2001) and Hart et al. (2013a) did not identify an increase in predation

following releases of cultured abalone, suggesting that trends might be site specific.

In a sea ranching trial in Port Philip Bay, Victoria, the portunid Rough Rock

Crab, Nectocarcinus integrifrons, was observed foraging around “seeding modules”

(i.e., a series of large clean abalone shells strung together and used as a substratum for

10

seeds to attach) within 10 minutes of being placed onto the artificial reefs (James et al.

2007). Numerous crabs were also seen sheltering inside the seeding modules within a

few days (James et al. 2007) and thus there is a high potential for increased juvenile

abalone mortality, particularly within the first few hours of seeding (McCormick et al.

1994).

The addition of cultured seeds may also intensify predation by mobile predators

such as octopuses and rays, responding to increased prey abundance (Hinesa et al. 1997;

Shepherd 1998; Shepherd et al. 2001; Anderson 2001; Gende and Sigler 2006).

Octopuses are common predators in sub-tidal communities, targeting a range of prey

types including abalone, a preferred food item of octopus (Ambrose 1982; 1983). The

possession of a range of specialised morphological adaptations, i.e., prehensile arms

with strong suckers, a chitinous beak, tongue-like radula, and salivary papilla bearing

tiny teeth (see Appendix A1.3) make the octopuses formidable predators, and their rapid

growth rates stimulate voracious feeding regimes which have the potential to impose

strong top-down controls within these coastal environments (Rodhouse and Nigmatullin

1996; Boyle and Rodhouse 2005).

The stingrays Dasyatis brevicaudata and Myliobatis australis, and stingaree

Trygonoptera mucosa have also been identified as important predators of the Greenlip

Abalone, Haliotis laevigata, through visual observations of their behaviour and the

distinctive scrape markings on abalone shells caused by their blunt teeth (see James

2005). These predators are common in habitats containing high densities of abalone,

particularly in the period after seeding, which may be facilitated by high hunting

success, thus attracting more rays to the vicinity (Shepherd 1990; Shepherd et al. 2001;

James 2005).

The survival of cultured and transplanted juvenile abalone depends on individual

size and substrate type (Inoue 1976, cited in McCormick et al. 1994; James et al. 2007).

The relationship between survival and juvenile shell length is described by a logistic

curve with the survival of Giant Abalone, Haliotis gigantea, increasing from 10 to 70%,

11

as individuals grow from 10 to 30 mm in size (Inoue 1976, cited in McCormick et al.

1994; Hamasaki and Kitada 2008). Thus, the release of larger juveniles (> 30 mm shell

width [SW]) can increase survival rates (McCormick et al. 1994; Hamasaki and Kitada

2008). Similarly, the selection of structurally more complex natural habitats with

numerous rocky crevices or artificial reefs that incorporate crevices to protect juveniles

from predators has a positive influence on their survival (Shepherd and Turner 1985;

McCormick et al. 1994; Dixon et al. 2006; James et al. 2007).

Abalone sea ranching in Flinders Bay, Western Australia by Ocean Grown

Abalone (OGA) provides an example of world-class fisheries enhancement, combining

low environmental impact aquaculture-based enhancement of wild caught H. laevigata

and species-specific artificial reefs to increase abalone production. Established on a

commercial scale less than three years ago, the operation has been very successful,

producing ~40 t of abalone in 2017 with projections to increase production to 100 t by

2018 and to 200 t by 2020 (B. Adams, OGA, 2017 pers comm.). However, a sea

ranching trial that concluded in early January 2013 at the same location, found that

predation by O. cf. tetricus on abalone is a potential threat to the survival of seeded

abalone (Melville-Smith et al. 2013). Ocean Grown Abalone record the presence of

octopuses and the number of abalone shells on their whole lease in Flinders Bay at

intervals of two to four weeks, and count and measure a sub-sample of abalone across

the lease every six months. Electronic records on the counts of octopus, abalone shells,

and abalone survival are available from January 2015. These data provide information

for examining the relationship between the presence of octopus and abalone mortality.

In this study, data on the abundance and distribution of octopus (mainly O. cf.

tetricus), empty abalone shells, and counts of abalone on the Abitats were examined to

evaluate:

(i) the distribution and abundance of octopus across the three areas of Abitats in

the lease area for the sea ranching operation by OGA;

12

(ii) the distribution and abundance of empty abalone shells collected across the

three areas to test whether the number of empty shells is related to the

presence of octopus; and

(iii) whether the presence of octopus is related to higher abalone mortality, based

on abalone survival counts.

This commercial-scale abalone sea ranching operation provides a unique

opportunity to observe how octopus may be impacting the production of abalone in

south-western Australia.

2.2. Methods 2.2.1. Study site

Ocean Grown Abalone holds three 40 ha sites within their lease in Flinders Bay,

Augusta, on the south-west corner of Western Australia (~34°37 S, 115°19 E) where the

sea ranching of H. laevigata takes place (Fig. 2.1). The sites range from ~15 to 19 m in

depth, and the natural habitat consists of a sandy substratum interspersed with irregular

seagrass patches of mixed composition of Amphibolis antarctica, Amphibolis griffithii

and Halophila ovalis (Melville-Smith et al. 2013; pers. obs 2017.). Sites are referred to

as the Outer (most southern), Middle, and Inner Leases. The inner boundary of the

Outer site is located 3.69 km from the nearest land, while the Middle and Inner Lease

boundaries are 2.46 and 1.43 km from the nearest land (i.e., Cape Leeuwin). The Inner

Lease is situated ~0.20 km from the eastern boundary of the Middle Lease, and the

Outer Lease is situated ~1 km south-westerly of the Middle Lease (Fig. 2.1). The

southern part of Flinders Bay is characterised by fringing reefs colonised by

macroalgae, particularly Ecklonia spp., and natural reefs are situated to the east and

west of the lease sites (Melville-Smith et al. 2013).

13

Figure 2.1. Satellite image (Landsat 2016) showing the Lease (Inner, Middle and Outer) locations of the Greenlip Abalone (Haliotis laevigata) sea ranching sites in Flinders Bay, Western Australia. Inset denotes the location of Flinders Bay in Australia.

2.2.2. Lease configuration

A total of 5,308 Abitats were deployed on the seabed by divers to form a series of

artificial reef structures. Abitats were deployed in a hierarchy of scales: six to eight

Abitats make up a ‘Reef’, twelve to fourteen Reefs make up a ‘Bay’, and one to thirteen

Bays make up a ‘Line’ (Table 2.1). Nine Lines are situated on the Outer Lease, three on

the Middle Lease, and two on the Inner Lease (Fig. 2.2, Table 2.1). The Abitats were

designed to maximise the surface area and growth potential for H. laevigata, and each

has a surface area of 10 m2 (B. Adams, OGA, 2015 pers. comm.). The artificial reefs

run perpendicular to the prevailing south-westerly swells and are situated adjacent to

naturally occurring seagrass beds (B. Adams, OGA, 2015 pers. comm.). This design

provides a barrier to drifting algal wrack and thus maximises food availability for the

abalone (B. Adams, OGA, 2015 pers. comm.). Haliotis laevigata feed on drift algae,

predominantly epiphytic red algae attached to drifting seagrass and macroalgae

(Melville-Smith et al. 2013).

14

The Outer Lease was established over the period mid-2014 to early 2015,

followed by the Middle and Inner Leases in mid-2015 and late-2015 respectively. Once

the Abitats were positioned on the seafloor, they were left for a minimum conditioning

period of two weeks before any abalone was seeded.

Table 2.1. Number of Abitats, Reefs and Bays on each Line and Lease at the Greenlip Abalone, Haliotis laevigata, sea ranching site in Flinders Bay, Western Australia. Bold shows the total for each of the three Lease sites. Each reef has six to ten Abitats.

Lease and Line No. Bays No. Reefs No. Abitats

Outer (1) 33 429 2712 O1 11 135 810 O2 3 34 196 O3 2 26 103 O4 4 53 320 O5 3 48 259 O6 1 11 79 O7 5 70 380 O8 2 23 380 O9 2 29 185

Middle (2) 21 236 1580 M1 3 42 267 M2 5 35 290 M3 13 159 1023

Inner (3) 12 152 1016 I1 3 131 879 I2 10 21 137

Total 66 817 5308

15

Figure 2.2. Configuration of Lines within each of the three Leases at the Greenlip Abalone, Haliotis laevigata, sea ranching site in Flinders Bay, Western Australia. Note, Lines are not to scale.

2.2.3. Culture and seeding of Abitats

Wild, locally sourced broodstock are collected during October and November and

cultured in the commercial 888 Abalone production facility in Bremer Bay (B. Adams,

OGA, 2015 pers. comm.) according to standard H. laevigata aquaculture protocols

(Daume and Ryan 2004; Daume et al. 2004; Duame 2006). Over a period of ~18

months, hatchery-reared juveniles are grown to ~40 mm, the optimal size to be seeded

onto the Abitats (B. Adams, OGA, 2015 pers. comm.). At this size, H. laevigata readily

adapt to a natural macroalgal diet, are less cryptic than smaller abalone, and are better

equipped to withstand assaults from teleost and invertebrate predators, having

developed a muscular foot of sufficient size to clamp tightly to the surface of the Abitat

(B. Adams, OGA, 2015 pers. comm.). Prior to release, juvenile abalone are held in

quarantine for two weeks and samples are sent to the Fish Health Unit, Department of

16

Primary Industries and Regional Development to ensure they are free of pathogens and

diseases (B. Adams, OGA, 2015 pers. comm.).

The juvenile seed stock is transferred into custom-built release units, each

holding ~200 abalone, and then transported to Augusta in oxygenated tanks where they

are loaded onto commercial dive vessels (Fig. 2.3a, b). The release units are lowered

into the water column inside an aluminium cage (44 release units per cage) by a man-

operated loader crane, and divers individually position the release units onto the Abitats

(two units per Abitat), which are left to settle overnight (Fig 2.3c, d) (B. Adams, OGA,

2015 pers. comm.). The following day the lids of the release units are removed, and

over a period of three to four weeks the juvenile H. laevigata crawl out and colonise the

Abitats (B. Adams, OGA, 2015 pers. comm.). Approximately 50,000 H. laevigata are

seeded onto the Abitats over a period of four hours on any one stocking day.

The initial seeding of Abitats was carried out in the order that Lines were

deployed, with ~400 individuals being seeded onto each Abitat at the first deployment

(D1), followed by a reseeding of ~120 H. laevigata individuals onto each Abitat 12

months after the initial seeding (R1).

2.2.4. Monitoring abalone and predators Predator surveys

Haliotis laevigata are typically grown on the Abitats for a period of three years until

they reach approximately 130 to 140 mm shell width (~5 years old). During this time,

surface-supplied commercial divers inspect the Abitats on a weekly basis to observe the

general health of the abalone and to remove excessive growths of macroalgae, soft

corals and ascidians from the Abitats to increase the surface areas available to abalone.

Every 2 to 4 weeks, octopuses and other predators such as bailer shells (Melo spp.) are

removed from the area around the Abitats. Divers have constant voice communication

with the vessel and the vessel crew records their observations in real time directly onto

the OGA database.

17

(a)

(b)

(c)

(d)

Figure 2.3. Images of abalone seeding operations by Ocean Grown Abalone (OGA) including: (a) abalone seeding unit stocked with juvenile Haliotis laevigata, (b) oxygenated abalone transfer bins which hold the stocked abalone release units whilst being transferred from the production facility in Bremer Bay to Augusta, (c) abalone release units being lowered into the water column in an aluminium cage via crane, and (d) an abalone release unit being positioned onto the artificial abalone habitat (“Abitat”). Images courtesy of OGA.

Surveys of predators are conducted across the three leases, usually on a single

day. Due to time constraints, not all Bays or Lines can be surveyed and during periods

of harvest or reseeding predator surveys may not be conducted. Any octopuses sighted

around the Abitats are counted and removed with the number of octopuses recorded at

the Bay level (~33 to 148 Abitats, Table 2.1). Indices of food availability for

H. laevigata are recorded at the same time. The availability of algal wrack is

subjectively scored on a food index ranging from one to five, with five indicating a high

abundance of algal wrack. Any empty abalone shells observed around the Abitats are

also collected and recorded with the total number of shells recorded per Bay. The

18

frequency of surveys was lower in winter months and also lower in 2017 than in other

years because of logistic constraints (i.e., prevailing poor weather and sea conditions,

and other sea ranching duties). A total of 64 predator surveys were conducted between

20 January 2015 and 16 March 2017, during which time 654 octopuses were removed

and 17,666 abalone shells were collected.

Abalone surveys

Commercial divers monitor the growth and survival of H. laevigata twice annually at

intervals of approximately six months. The numbers of stocked abalone on five to ten

randomly selected Abitats in each Bay are counted. The stocked number of abalone is

then subtracted from the assumed number of abalone seeded onto the Abitats (400) to

establish abalone survival. Note that the same Abitats are not surveyed each time. The

average survival per Bay is then extrapolated according to the number of Bays on a Line

in order to determine an estimate of abalone survival per Line.

2.2.5. Data exploration and statistical analyses

Data exploration was carried out using the R statistical software package (R Core Team

2011), and the “ggplot2” (Wickham 2009) and “lubridate” (Garrett and Wickham 2011)

add-on packages to provide insights into potential trends and to identify the limitations

of the data, following the protocols described in Zuur et al. (2010). In order to elucidate

any trends in the distribution and abundance of octopus, the mean numbers of octopus

collected per survey per Line, and the mean numbers of octopus collected per survey

per Line, adjusted for the number of Abitats (number of octopus 1000-1 Abitats), were

calculated and plotted as bubble plots.

Empty abalone shells collected by divers during predator surveys between April

and July 2017 were measured to the nearest mm, grouped into 10 mm length classes and

plotted in a length frequency histogram. Descriptive statistics were then used to

determine the size distribution of these shells. Shells were inspected for any evidence of

predation, and any shells with conspicuous holes were examined in more detail. Shells

19

with holes were divided into four areas according to the position of the drill site

following Kojima (1992) and based on the underlying anatomy of abalone. The

frequency of holes in each area was counted. The four shell areas were:

• I: covers the digestive and respiratory organs, pericardial sac and heart;

• II: covers part of the mantle;

• III: covers the gonad, gut, and liver; and

• IV: covers the adductor muscle that attaches to the shell (Kojima 1992;

Morash and Alter 2016).

An exact multinomial test was employed to test whether the holes had a non-random

distribution over the four areas.

All inferential statistical analyses were performed using the R statistical

software package (R Core Team, 2017) and the “MASS” (Venables and Ripley 2002),

“pscl” (Zeileis et al. 2008), and “Xnomial” (Fox 2003) add-on packages.

Relationship between octopus presence and empty abalone shells

A series of count regression models were developed to test whether the presence of

octopus (i.e., “Octopus” > 0) results in a greater number of empty abalone shells

(“Shells”) being collected around the artificial reefs, compared with areas in which

octopus are absent (Table 2.2). Given that > 1 octopus accounted for only 8% of the

total number of observations, and very few counts of > 2 octopuses (2%) were found,

the presence of octopus (i.e., counts > 0) was selected as the independent variable.

Diagnostic plots were used to assess the suitability of linear regression models

for fitting the data. These showed major violations in the assumptions of linearity,

normality, and homoscedasticity. A Poisson log-linear generalised linear model was

also deemed inappropriate due to overdispersion, an artefact of the Poisson distribution

estimating one parameter for both the mean and the variance (Table 2.2). A negative

binomial generalised linear model was ultimately used, as it is appropriate for count

data and provides estimates of the mean for the parameter as well as a dispersion

20

parameter (Gardner et al. 1995, Table 2.2). A histogram of the frequency of empty shell

counts showed significant variability (Fig. 2.4). Shell counts ranged from 0 to over 100

with a modal count of 3-6 shells, a mean count of 9.9 and a median of 6 (i.e., the count

frequency distribution was heavily right skewed, Fig. 2.4), which is expected for count

data (see Gardner et al. 1995). However, the high proportion of zero counts is greater

than usual, with the proportion of zeros exceeding that predicted by standard models for

a set of observed characteristics (Baetschmann and Winkelmann 2013). Thus a zero-

inflated negative binomial generalised linear model was also considered (Table 2.2).

Table 2.2. Count models considered for assessing the relationship between the presence of octopus (“Octopus” > 0) and empty abalone shell (“Shells”) counts.

Model Summary Linear (Raw data, ln(x))

Untransformed and log transformed data did not satisfy the assumptions of linear regression.

Generalised linear model (Poisson) Data overdispersed, therefore model deemed inappropriate.

Generalised linear model (Negative binomial)

Most suitable model as indicated by Bayesian information criterion (BIC) and Akaike information criterion (AIC) estimates.

Generalised linear model (Zero-inflated negative binomial)

Less suitable than the negative binomial model, as indicated by BIC and AIC estimates.

Figure 2.4. Frequency distribution of empty abalone shells collected during predator surveys at the Greenlip Abalone, Haliotis laevigata sea ranching site in Flinders Bay, Western Australia between 20 January 2015 and 16 March 2017. The collection of > 150 shells were treated as outliers and removed for graphical and statistical purposes. Total shell counts = 1,663, 𝑥 = 9.93, number of surveys = 64.

21

To control for other variables that might be related to the number of empty

shells collected, Line and Season were included in the model. Likelihood ratio tests

were used for model comparison. A model-building approach was taken, adding one

covariate at each step to determine the most parsimonious model to explain the fit of the

data. The same approach was used to test for interaction effects, and if present, to

determine the extent of that interaction relative to the main effects. None of the

interaction effects were significant, thus a model consisting strictly of main effects was

selected. Line was considered an important predictor, to account for potential spatial

variability across the Leases, while Season was incorporated to account for natural

intra-annual variation.

Analyses were carried out both with and without potential outliers to assess their

impact. Five extreme observations were noted (i.e., where > 150 shells were collected

from a single Bay on any day) and considered as outliers. Models showed that the

inclusion of these potential outliers led to a stronger estimated relationship between the

presence of octopus and empty abalone shells, and therefore the more conservative

model excluding outliers is presented.

A Vuong test was used to assess which of the two models (i.e., a negative

binomial model or a zero-inflated negative binomial model) best fit the data (Table 2.2).

No significant difference was identified when comparing tests based on raw data.

However, BIC (i.e., Bayesian information criterion) and AIC (i.e., Akaike information

criterion) estimates indicated that the negative binomial model was significantly better

than the zero-inflated negative binomial model. For this reason, and to reduce model

complexity, a negative binomial model was employed.

The final negative binomial model, with a log link function, was used to model

the number of Shells collected as a function of the covariates [Equation (1)]. Fixed

covariates include Octopus > 0 (categorical with two levels), Line (categorical with 14

levels) and Season (categorical with four levels – Summer: December to February,

Autumn: March to May, Winter: June to August and Spring: September to November,

22

Table 2.3). The offset terms “Time Since Last Survey × Number of Abitats” were

included in the model to account for differences in exposure, as the number of shells

collected could be related to the time that has elapsed since the previous survey due to

the accumulation or loss of shells. The number of Abitats could be important since

octopus commonly display anti-social, intraspecific avoidance behaviour, with

competition for shelter sites resulting in incidents of conflict and cannibalism (Mather

and O’Dor 1991; Hanlon and Messenger 1996; Caputi et al. 2015). Thus, a greater

number of Abitats could potentially support a greater number of octopuses. As Time

Since Last Survey and the number of Abitats are both measured without error, they are

more appropriately included as offsets rather than as additional covariates.

To accurately interpret the regression coefficients and to account for differences

in exposure, as specified by the offset term, a back-transformation of logged counts was

carried out, and incident rate ratios calculated. Consequently, the incident rate

represents the number of empty abalone shells collected per Abitat per day (Table 2.3).

The full model can be written more explicitly as:

Yi ~ negative binomial (µi, k)

log(µi) = 𝛽0 + 𝛽1X1i + 𝛽2X2(2)i + 𝛽3X2(3)i + 𝛽4X2(4)i + 𝛽5X2(5)i +𝛽6X2(6)i +

𝛽 7X2(7)i + 𝛽 8X2(8)i + 𝛽 9X2(9)i + 𝛽 10X2(10)i + 𝛽 11X2(11)i + 𝛽 12X2(12)i +

𝛽13X2(13)i + 𝛽14X2(14)i + 𝛽15X3(2)i + 𝛽16X3(3)i + 𝛽17X3(4)i + log(Ti) + 𝜖i (1)

where µ denotes the mean response of the observation, k denotes the negative binomial

dispersion parameter, Yi denotes the number of Shells, X1 denotes Octopus > 0, X2

denotes the Line (2 = “O2”, 3 = “O3”, 4 = “O4”, 5 = “O5”, 6 = “O6”, 7 = “O7”, 8 =

“O8”, 9 = “O9”, 10 = “M1”, 11 = “M2”, 12 = “M3”, 13 = “I1”, 14 = “I2”), X3 denotes

the Season (2 = “Spring”, 3 = “Summer 2015”, 4 = “Winter”), log(Ti) denotes the offset

(i.e., the number of Abitats × Time Since Last Survey) and 𝜖i represents the error

associated with omitted or non-observable exogenous variables (Liu et al. 2005).

23

Results from the negative binomial model were then used to provide an estimate

of the total number of abalone shells associated with the presence of octopus for the 27-

month period in which 17,666 shells were collected. This estimate was then converted

to a percentage, representing the proportion of shells associated with octopus presence.

Relationship between octopus presence and abalone survival

Potential patterns in abalone survival counts over time were investigated using box plots

of the counts of abalone on Abitats. As survival counts were conducted over multiple

consecutive days (≤ 10), the survey dates for a given period were aggregated into four

periods: March and September of 2015, and the same months of 2016 (i.e., 4 survival

counts).

To assess the relationship between average abalone survival (“Survival Count”)

and the presence of octopus (“Octopus > 0”), an autoregressive (AR1) linear model

(i.e., a time series model where abalone survival counts are dependent on the preceding

survival count) was considered [Equation (2)]. The model included the average number

of abalone counted during survival surveys for abalone seeded at the first deployment

(D1), prior to harvesting for a given Bay. Average survival counts were chosen in

favour of absolute survival counts due to the variability in abalone survival counts

between Abitats.

A new variable “Time” was created, representing the number of days from the

first survival count. The first survival count for a given Bay was recorded as time point

0, and successive counts refer to the number of days since time 0. All octopuses

recorded within each Bay up to the first survival count (0 days), and then between

successive abalone counts, were aggregated to account for the total number of octopuses

present during a given time period. Survival counts from January 2017 were not

included in these analyses as, after this time, data were recorded at the Reef level (i.e.,

an aggregation of Abitats) instead of per Abitat.

24

Potential outliers (i.e., where abalone survival counts for a given Bay had

declined by a mean of > 150 abalone between successive surveys) were modeled

separately to assess the potential impact of those observations. Diagnostic plots used to

assess the suitability of simple linear models indicated major violations in the

assumptions of linearity, normality, and homoscedasticity. Data transformations did not

resolve these issues but the removal of extreme observations greatly improved the

model suitability. Thus the model presented excludes outliers. These extreme

observations are likely due to natural mortality of abalone in response to unfavorable

environmental conditions such as poor current flows and low food availability, as

opposed to octopus predation.

The full model can be written as:

Yi = 𝛽0 + 𝛽1X1i + 𝛽2X2i + 𝛽3X3i + 𝜖i (2)

where Yi denotes the number of Shells, X1 denotes Octopus > 0, X2 denotes the Previous

Average Shell Count, X3 indicates Time (number of days) since the first survey count,

and 𝜖i ~ N (0, 𝜎2) is the error term for some constant measure of variability 𝜎2.

2.3. Results

2.3.1. Patterns of octopus distribution and empty abalone shells

Overall, the total number of octopuses collected was greatest on the Outer Lease (422 =

65% of all octopus), which also had the greatest number of Abitats (2,712, Table 2.1),

followed by the Middle (154 octopus) and Inner Leases (78 octopus). Octopuses were

found across the Lines in each of the three Leases, with the greatest mean number of

octopuses sampled on Line O1 (0.47 octopus per survey), and lowest on O8 (0.13, Fig.

2.5). When adjusted to a standard number of Abitats per Line (i.e., octopus 1000-1

Abitats), the adjusted mean number of octopus was greatest at O6 (4.11), followed by I2

(3.04), and the lowest adjusted mean was recorded on O8 (0.33, Fig 2.6).

25

Figure 2.5. Bubble plot showing the mean number of octopus collected on each Line at the Greenlip Abalone, Haliotis laevigata, sea ranching site in Flinders Bay, Western Australia, during predator surveys conducted between 20 January 2015 and 16 March 2017. Note, Lease configurement not to scale.

Figure 2.6. Bubble plot showing the standardised mean number of octopus (i.e., the number of octopus per Abitat × 1000) collected on each Line at the Greenlip Abalone, Haliotis laevigata, sea ranching site in Flinders Bay, Western Australia during predator surveys conducted between 20 January 2015 and 16 March 2017. Total number of octopus collected = 654. Note, Lease configurement not to scale.

26

A total of 17,666 shells were collected between January 2015 and March 2016:

9,598 from the Outer Lease, 3,976 from the Middle and 4,092 from the Inner Lease.

These shells represent 0.3% of all cultured abalone seeded onto the Abitats. The average

number of shells collected per survey was highest on the Inner Lease (21.1), compared

with the Middle (6.3) and Outer (3.5) Leases. The average shell counts per Line were

highest on Lines I1 and I2, with 22.16 and 18.25 shells respectively (Fig. 2.7). When

adjusting for the number of Abitats per Line (shells 50-1 Abitats), shell counts were

greatest on Lines I2 and O6, with 6.66 and 3.66 shells respectively. The lowest adjusted

mean shell count occurred on Line O1 with 0.49 shells (Fig. 2.8).

A total of 110 empty H. laevigata shells were measured, and they ranged from

33 to 100 mm in shell length (SL) and had a mean SL of 55.9 mm (± 1SE = 1.2 mm)

(median SL = 54.0 mm). The modal length class was 50-59 mm, representing 35% of

all shells collected (Fig. 2.9a). Twenty shells (18%) had bevelled boreholes that were

characteristic of the holes made by octopuses (Fig. 2.10a, see Arnold and Arnold 1969;

Nixon and Elaine 1988; Kojima 1992). The boreholes were slightly ovoid at the top,

narrowing into a smaller aperture towards the inner surface of the shell (Fig. 2.10b).

Only one hole was found on each of the 20 shells with a borehole. Shells with

holes spanned the entire length range of all shells examined (33 to 100 mm), but their

mean SL of 62.0 ± 3.6 mm was slightly longer than for shells without a hole (54.6 ± 0.8

mm, N = 90). A total of 70% of shells with a drill hole were within the length range of

60 to 89 mm (Fig. 2.9b). The majority of the holes were concentrated in areas IV (40%)

and I (35%), with 8 and 7 holes respectively, and 10% and 15% in each of areas II and

III (Fig. 2.10b). An exact multinomial test showed that the frequency of holes did not

differ significantly among the four regions (P = 0.18).

27

Figure 2.7. Bubble plot showing the mean number of shells collected on each Line at the Greenlip Abalone, Haliotis laevigata, sea ranching site in Flinders Bay, Western Australia during predator surveys conducted between 20 January 2015 and 16 March 2017. Number of shells collected = 17,666, 64 surveys completed. Note, Lease configurement not to scale.

Figure 2.8. Bubble plot showing the standardised mean number of shells (i.e., number of shells per Abitat × 50) collected on each Line at the Greenlip Abalone, Haliotis laevigata, sea ranching site in Flinders Bay, Western Australia during predator surveys conducted between 20 January 2015 and 16 March 2017. Number of shells collected = 17,666, 64 surveys completed. Note, Lease configurement not to scale.

28

(a)

(b)

Figure 2.9. Shell length frequency distribution of (a) empty Haliotis laevigata shells collected from the Greenlip Abalone sea ranching site in Flinders Bay, Western Australia during predator surveys between 20 January 2015 and 16 March 2017, Total number of shells = 110, and (b) H. laevigata shells with boreholes, collected between April and July 2017, Number of shells with holes = 20.

(a)

(b)

Figure 2.10. Haliotis laevigata shells (a) with a bevelled borehole, characteristic of those made by octopus, and (b) location of the boreholes recorded on 20 shells. Shells collected from the Greenlip Abalone sea ranching site in Flinders Bay, Western Australia between April and July 2017. Note, no scale bar is shown on (b), and dots representing the location of boreholes are not in proportion to the size of the aperture. Shells with an borehole range from 33 to 100 mm in length.

29

2.3.2. Relationship between octopus presence and empty abalone shells

Negative binomial regression models suggest that the presence of octopus leads to a

78% (95% confidence limit [CL] = 49 - 126%;) increase in the number of empty shells

collected per Abitat per day relative to the absence of octopus, when adjusting for the

variables Line and Season (P < 0.001, Table 2.3). Line was also found to be an

important covariate associated with the number of empty shells collected. On the outer

Lease, the number of shells on Lines O7 and O2 were significantly higher (P < 0.001)

than the reference group (O1) with 113% (CL = 61 - 183%) and 87% (CL = 35 - 163%)

more shells per Abitat per day, respectively, adjusting for Octopus > 0 and Season

(Table 2.3). The number of shells collected on the Lines M1, M2, M3, O4, and O5 were

significantly lower than the reference group when adjusting for Octopus > 0 and Season

(P < 0.001, Table 2.3). Autumn had significantly higher shell counts compared with

Spring and Summer with 46% (CL = 16 – 84%) and 55% (CL = 25 – 92%) more shells

per Abitat per day respectively when adjusting for Octopus > 0 and Line (P < 0.001,

Table 2.3).

Based on a 78% increase in shell counts, the number of empty abalone shells

associated with the presence of octopus is 7,741 for the 27-month period in which shells

were collected. This represents 44% of all empty shells collected.

30

Table 2.3. Summary of the results for the negative binomial model to investigate variation in the number of abalone shells and presence of octopus. Number of Abalone Shells on Octopus > 0, Line, and Season, with the number of Abitats multiplied by the time since last survey as an offset in the model. Variable refers to terms included in the model; Estimate refers to the count for the coefficient; Incidence Rate refers to the exponentiated values of the regression coefficients; Std. Error = asymptotic standard errors of the regression coefficients; p-value = significance; Confidence Interval = 95% confidence interval of the exponentiated regression coefficients. Bold = P < 0.05; * indicates the reference level for each of the factors.

Confidence Interval

Variable Estimate Incidence Rate

Std. Error z-value P-value 2.5% 97.5%

(Intercept) -6.972 0.001 0.099 -70.736 < 0.001 0.001 0.001

Octopus > 0 0.576 1.778 0.086 6.687 < 0.001 1.493 2.126

Line (* O1)

O2 0.623 1.865 0.165 3.765 < 0.001 1.347 2.626

O3 -0.120 0.819 0.204 -0.979 0.328 0.550 1.257

O4 -1.658 0.190 0.306 -5.413 < 0.001 0.106 0.371

O5 -0.558 0.572 0.165 -3.385 < 0.001 0.412 0.808

O6 -0.377 0.686 0.264 -1.429 0.153 0.411 1.213

O7 0.754 2.127 0.134 5.635 < 0.001 1.611 2.832

08 -0.234 0.791 0.416 -0.564 0.573 0.366 2.053

O9 -0.459 0.632 0.248 -1.853 0.064 0.394 1.068

M1 -0.919 0.399 0.201 -4.556 < 0.001 0.269 0.610

M2 -2.096 0.123 0.353 -5.945 < 0.001 0.064 0.266

M3 -0.699 0.497 0.120 -5.828 < 0.001 0.388 0.638

I1 -0.046 0.954 0.151 -0.307 0.759 0.709 1.302

I2 0.266 1.305 0.274 0.974 0.329 0.782 2.346

Season (*Autumn)

2015 Spring -0.380 0.684 0.109 -3.494 < 0.001 0.543 0.861

2015 Summer -0.439 0.645 0.103 -4.263 < 0.001 0.522 0.796

2015 Winter 0.210 1.235 0.120 1.757 0.079 0.966 1.584

Null deviance: 2153.9 on 1596 degrees of freedom

Residual deviance: 1888.6 on 1579 degrees of freedom

Number of Fisher Scoring iterations: 1 Theta: 0.438

31

2.3.3. Relationship between octopus presence and abalone survival

The average H. laevigata survival counts varied greatly among Bays, particularly during

the March 2016 and September 2016 survey periods (Fig. 2.11). A number of outliers

were observed, where average counts were as low as 40 individuals (Fig. 2.11). Despite

significant overlap in the average counts between each of the periods, there is a clear

declining trend, with the mean survival count decreasing from 310 in March 2015 to

294 in September (5.2% decline), 261 in March 2016 and 236 in September (Fig. 2.11).

Assuming an average starting count of 400, this represents an average of 59% survival

over two years from September 2014 to September 2016. The greatest average decline

in numbers was recorded on the first count (23%) compared with subsequent 6 monthly

survivals of 88.8 to 94.8%. The median average survival count was slightly higher than

the mean in each period and was 318 abalone at 6 months (79.5% survival), 296 at 12

months, 273 at 18 months and 254 after two years (63.5% survival over 2 years,

Fig. 2.11).

Figure 2.11 Boxplots of six monthly average survival counts of Haliotis laevigata, surveyed by divers at the Greenlip Abalone sea ranching site in Flinders Bay, Western Australia. Survival counts consider abalone seeded at the first deployment (D1), prior to harvesting of a given Bay. N = 126 counts. Individual points represent outliers in the data (i.e., observations that fall 1.5 times below the lower interquartile range).

32

Autoregressive time series models (AR1) showed that the relationship between

the presence of octopus and abalone survival was not significant (P = 0.09, Table 2.4).

Successive abalone survival counts were significant in the model (P < 0.001), with

average survival increasing by 1.15 abalone compared with the previous average count,

when adjusting for Octopus > 0 and Time (Table 2.4).

Table 2.4. Summary of autoregressive AR(1) model output for the relationship between the Count of abalone shells remaining on Abitats and the presence of octopus. Average Count of abalone survival on Octopus > 0, Time, and Previous Average Count. Variable refers to terms included within the model; Estimate refers to the count for the coefficient; Std. Error refers to the standard errors asymptotic standard errors of the regression coefficients; p-value refers to the significance of the model parameter; Confidence Interval refers to the 95% confidence level of the regression coefficients. Bold face indicates a significance level of P < 0.05.

Confidence Interval

Variable Estimate Std. Error z-value P-value 2.50% 97.50%

(Intercept) -92.10 33.075 -2.785 0.007 -158.359 -25.845 Previous Average Count 1.145 0.10 11.552 < 0.001 0.946 1.344 Octopus > 0 23.194 13.352 1.737 0.088 -3.552 49.94 Time -0.049 0.058 -0.865 0.391 -0.165 0.066

Residual standard error: 32.11 on 56 degrees of freedom Adjusted R-squared: 0.751

2.4. Discussion

This study examined the relationship between octopus and abalone mortality to evaluate

the significance of octopus as predators of H. laevigata on a sea ranch of south-western

Australia. The data used for these analyses came from counts of octopus and empty

abalone shells (~2 to 4 weekly intervals) and counts of surviving abalone on Abitats,

recorded by commercial divers as part of the Ocean Grown Abalone monitoring

program for abalone and octopus. Direct evidence of predation was identified from the

occurrence of distinctive drill holes in a sub-sample of 110 empty H. laevigata shells of

H. laevigata, and the significance of predation was shown in negative binomial models.

33

2.4.1. Patterns of octopus distribution and empty abalone shells

Despite octopus abundance being highest on the Outer Lease, the number of surveys

and the number of Abitats was also higher in this area. The mean number of octopus

sampled was broadly similar among Lines. However, when adjusting for the number of

Abitats, there were notable differences identified among Lines. This may be a result of

observer bias, with fewer octopuses sighted on Lines comprising a greater number of

Abitats. For example, on the Inner Lease, the adjusted mean number of octopus was 6.6

times higher on Line I1 (137 Abitats) compared with I2 (879 Abitats). Lines O6 and O9

run perpendicular to one another, however octopus were 2.9 times higher on O6 (79

Abitats) compared with O9 (185 Abitats). In addition, the spatial layout of Abitats

within a Line may be a contributing factor. Octopuses have sophisticated sensory organs

used to detect predators, and a complex camouflage system allowing them to remain

inconspicuous (see Appendix 1, A1.4.1, A1.4.2). Therefore, they have the potential to

be overlooked by divers.

Environmental variability and the spatial layout of the Abitats likely contribute

to lower shell counts on some Lines. Typically, larger numbers of empty shells were

found on Lines comprising fewer Abitats. For example, on the Outer Lease shell counts

were greatest on the Line O6, which comprises the fewest number of Abitats (79), and

lowest on Line O1, which comprises the largest number of Abitats (810). On the Inner

Lease, shell counts were 5.3 times higher on Line I2 (137 Abitats) than on Line I1 (879

Abitats). Where divers are required to cover a larger distance, less time can be spent at

any one Bay, which may result in lower shell recovery. In areas of high current flows,

there may be a greater propensity for shells to be buried or dispersed away from the

Abitats, which could be amplified during storm events leading to lower shell counts

(Hines and Pearse 1982). The dispersal of shells, even just a few metres away from the

Abitats may reduce the number of shells that are collected, as divers are likely to be

focused on the area directly surrounding the Abitats.

Among the sample of empty shells examined, 18% had drill holes which were

characteristic of the holes made by octopus (Pilson and Taylor 1961; Kojima 1992). The

34

holes had a distinctive bevel around the aperture and were slightly ovoid; features that

readily distinguish octopus predation from other sources (Pilson and Taylor 1961;

Kojima 1992). Drilling by octopuses has been documented for several other coastal

species (e.g., Octopus bimaculatis, Pilson and Taylor 1981; Ambrose 1983; Tegner and

Butler 1985; and Octopus vulgaris, Ambrose 1983; Kojima 1992; Smith 2003),

however this appears to be the first account for Octopus cf. tetricus, the most commonly

encountered octopus species in Flinders Bay. The percentage of shells with boreholes

contrasts with the findings of Ambrose (1983), Kojima (1992), and Smith (2003), who

identified that < 10% (3.4 to 7.7%) of Haliotis sp. shells had been drilled by octopus.

These findings demonstrate that O. cf. tetricus will resort to drilling a tightly attached

abalone when they are unable to simply pull abalone off of the substratum (McQuaid

1994; Smith 2003). The presence of drill holes in the shells of H. laevigata suggest that

O. cf. tetricus produce cephalotoxins that are capable of paralysing their prey when

injected (Ghiretti 1959; Nixon 1984).

The presence of drill holes suggests that at least one-fifth of all abalone

mortality on the sea ranch is attributable to octopus predation. This percentage likely

represents the minimum mortality of H. laevigata by octopus but confirms that

octopuses are a major source of mortality. Octopus may be able to pull small abalone

directly off the reef without the need to bore into the shell, or alternatively, they may

target abalone in a “feeding” position. When H. laevigata feed, they elevate their shell,

the forepart of the foot, and extend their tentacles to catch drifting algae (Shepherd

1973). Whilst in this position, H. laevigata may be more vulnerable to predation, as

their typical stronghold on the reef substratum is reduced.

Predation by octopus is likely to be related to the size of the octopus, with larger

individuals targeting larger abalone. Size-related octopus predation has been

documented in South Africa, where O. vulgaris predates on the mussel Perna perna and

larger octopuses targeted larger prey, gaining a higher energy return per attack than on

smaller prey (McQuaid 1994). While the distribution of borehole locations was not

35

significantly different from a random distribution (which may be related to small

sample sizes), holes in the shells of H. laevigata were concentrated over the adductor

attachment, the heart and the pericardial sac (Kojima 1992; Morash and Alter 2016).

The injection of cephalotoxin into these regions is perhaps the most efficient location to

target, as it is close to the muscle responsible for holding the abalone on the substratum.

Octopus have excellent cognitive abilities, with the ability to learn from past

experiences (Anderson et al. 2008b). Therefore the areal distribution of boreholes is

likely to be the result of a learnt behaviour, with octopus targeting the areas which

succumb to cephalotoxin most rapidly (Wodinsky 1969; Anderson et al. 2008b).

2.4.2. Relationship between octopus presence and abalone shells

The presence of octopus was associated with a greater number of empty abalone shells

(adjusted for the number of Abitats and time since the previous survey) than when

octopuses were absent. The artificial abalone habitats used for the production of

H. laevigata in Flinders Bay likely provide octopus with two important resources: the

abalone on the Abitats provide a concentrated potential food source, and the Abitats

provide shelter for octopus. The rapid growth rates of octopus command voracious

feeding regimes, and thus their predatory impact on benthic communities can be

significant (Boyle and Rodhouse 2005; Eddy et al. 2017). While octopuses adapt

rapidly to new environments (Rigby and Sakurai 2005) they are vulnerable to predation

from teleosts, particularly at smaller sizes (Aronson 1991; Mather and O’Dor 1991).

Thus, the selection of ‘homes’ is driven by a need to avoid predators in areas that

provide adequate resources (i.e., food, or sites for the brooding of eggs, Aronson 1991;

Mather 1991; Mather 1994; Anderson 1997). Octopuses are commonly observed among

the Abitats, sheltering inside established dens surrounded by empty abalone shells, and

feeding on ranched H. laevigata (M. Wall, OGA, 2017 pers. comm.).

Despite the association between the presence of octopus and the number of

empty shells collected, the recovery of shells in other studies is discernibly low (i.e., 3

to 10% of all seeded abalone (McCormick et al. 1994; McShane and Naylor 1997;

36

Dixon et al. 2006; Goodsell et al. 2006). Empty shells have the potential to be buried,

swept away by currents, or crushed by crabs so that shells cannot be distinguished from

the substratum (Hines and Pearse 1982; Schiel and Welden 1987; McCormick et al.

1994; Catchpole et al. 2001; Dixon et al. 2006; James et al. 2007). Laboratory studies

demonstrate that the loss of shells due to crabs and lobsters can be as high as 71%

(Schiel and Welden 1987). Consequently, shell counts are likely to provide unreliable

estimates of mortality, and the impact of octopuses, based on empty abalone shells may

be underestimated (Tegner and Butler 1985; McCormick et al. 1994; Dixon et al. 2006).

However, a dedicated study investigating shell recovery rates would be required to

understand whether shell counts provide a reliable index of total abalone mortality

within the Flinders Bay system. While estimates are only likely to provide the minimum

mortality of H. laevigata, the collection of shells can be useful in elucidating the cause

of death (Schiel and Welden 1987; McCormick et al. 1994; James 2005; Dixon et al.

2006).

The results from the negative binomial models estimated that 44% of all empty

abalone shells observed were associated with the presence of octopus. This estimate is

26% higher than the proportion of shells found with octopus drill holes, calculated from

a sub-sample of 110 empty abalone shells. The higher proportion of shells associated

with octopus presence estimated by the model suggests that octopus can remove

abalone from the Abitats by direct force. Based on shell recovery rates of 3% and 10%

(McCormick et al. 1994; McShane and Naylor 1997; Dixon et al. 2006; Goodsell et al.

2006), the number of empty abalone shell associated with the presence of octopus could

range from 77,410 to 258,033 shells for the 27-month period over which shells were

collected

The high spatial variability in shell counts observed among Lines (compared

with the reference level) may be associated with natural environmental variability

across the Lease or associated with the number of Abitats, as previously discussed. The

significantly lower number of shells collected during spring and summer compared with

37

autumn may be related to environmental variability or to the numbers of abalone seeded

within a particular season. For example, roughly ten times fewer H. laevigata were

seeded during summer compared with autumn, and, consequently, predation rates and

the number of empty shells collected are also likely to be lower. Cultured juveniles are

highly vulnerable to predation immediately after seeding, due to poor behavioural

responses towards predators (Schiel and Welden 1987; McCormick et al. 1994).

Cultured abalone are typically sluggish in their response to predators and take longer to

firmly attach to the reef substratum compared with wild individuals (Schiel and Welden

1987). This behaviour is likely a reflection of their juvenile history, typified by an

absence of predators (Schiel and Welden 1987). The acclimatisation to the wild can take

several days, and consequently, newly released individuals are likely to be naïve to

predation (Schiel and Welden 1987). Additionally, abalone experience high levels of

stress during the transplantation process, which stimulates mucus secretions that are

highly attractive to predators (Tegner and Butler 1989). Consequently, predator

abundance and abalone mortality may increase immediately following seeding

(Shepherd 1990; McCormick et al. 1994; James et al. 2007). Furthermore, elevated

stress levels following transplantation may reduce fitness, leaving weakened abalone

susceptible to scavengers (Tegner and Butler 1989). Evidence for elevated mortality in

the post-release period was identified from abalone survival counts, whereby survival

was lowest at the first six monthly survival count compared with subsequent counts.

2.4.3. Relationship between octopus presence and abalone survival

The relationship between the presence of octopus and abalone survival counts was not

significant, contrasting the findings found for the presence of octopus and empty

abalone shell counts. In addition, the relationship between the presence of octopus and

abalone survival was contrary to what would be expected, given that octopuses are

voracious predators and appear to be preying upon abalone. These findings are likely to

be an artefact of random sampling design (where random Abitats are sampled at each

time point but without a unique identifier to indicate whether the same Abitat has been

resampled at multiple time points), and significant variation in the number of abalone

38

seeded on individual Abitats. When abalone are seeded, two seeding modules are placed

onto each Abitat, with each module estimated to contain ~200 individuals. However, the

actual number of abalone has the potential to vary significantly from this estimate, since

abalone are not individually counted during the packing process. Rather, the packing of

seeding modules is based on weight. Due to natural variation in growth, the size and

weight of abalone differ considerably among individuals, and thus the total number of

abalone per module is likely to vary greatly, and as a consequence, the “number” seeded

per Abitat (“400”) is an approximation only. In addition, survival counts are based on

randomly selected Abitats within a Bay, and the Abitats counted on subsequent surveys

may differ from those counted previously. Thus, changes in abalone survival based on

this counting methodology, cannot be estimated with precision over time.

Quantification of the number of abalone seeded on each Abitat and repeated counts on

the same Abitats over time, would allow changes in survival to be better understood.

Given the significant variability between abalone survival counts, and the

counterintuitive relationship between the presence of octopus and abalone survival,

estimates from the model cannot be considered reliable.

39

Chapter 3. The diet and nutrient assimilation (δ13C and δ15N) of Octopus cf. tetricus on an abalone sea ranch in south-western Australia

3.0. Abstract

This study used gastric tract (crop and stomach) and stable isotope analyses to evaluate

the dietary composition of the commercially important Octopus cf. tetricus, and its

significance as a predator of Greenlip Abalone, Haliotis laevigata, on artificial abalone

habitats (“Abitats”) in Western Australia. The contents of the crop and stomach from 44

octopuses, ranging from 58.5 to 1596.1 g (mantle length = 52 to 152 mm) were

compared with understand the dietary composition that could be determined from

examining each organ. A total of 13 taxa were identified in the crop contents and 10 in

the stomach. Within the stomach contents, 74% of the total percentage volume (%V) of

material was unidentifiable, with molluscs comprising the majority of identified items

(total %V = 19%: ~6% octopus, 2% abalone and 2% bivalves). In contrast, < 4% of

material in the crop contents was unidentifiable; providing a much higher taxonomic

resolution of dietary items than stomachs. Among the crop contents, molluscs

contributed the largest %V (~31% abalone, ~13% non-cephalopod molluscs, ~6%

octopus and ~4% bivalves), followed by crustaceans (33%). Dietary composition based

on crops did not differ significantly among three size classes of octopus (< 300 g, 300-

999 g, ≥ 1,000 g), paralleling the findings of stable isotope analysis. Examination of the

nitrogen stable isotope (δ15N) of octopus, fish, abalone and macrophytes revealed that

O. cf. tetricus occupies a mid trophic level position, slightly below teleosts and loliginid

squids. The carbon stable isotope (δ13C) signature of O. cf. tetricus was similar to the

teleosts Coris auricularis, Opthalmolepis lineolatus and Upeneichthys vlamingii, which

may be explained by an overlap in the use of food resources among these benthic

consumers. Gastric tract analyses suggest that O. cf. tetricus is an important predator of

H. laevigata. This trend was less obvious from stable isotope analysis, due to the

relative enrichment of δ13C in O. cf. tetricus compared with H. laevigata. These

findings may be explained by the contribution of other prey items to the diet of octopus,

such as crustaceans and other cephalopods (as identified by gastric tract analysis).

40

3.1. Introduction

Octopuses play important ecological roles within coastal marine environments around

the world, both as predators and prey of a diverse range of species (for detail, see

Appendix 1). Due to rapid growth rates which stimulate voracious feeding regimes, and

the ability to adapt rapidly to their surroundings, octopuses can impose significant

predatory impacts; controlling the trophic structure and functioning of the communities

in which they live (Nixon 1987; Rodhouse and Nigmatullin 1996; Boyle and Rodhouse

2005; André et al. 2010).

Octopus cf. tetricus is a coastal merobenthic species (i.e., they possess a

planktonic larval phase), endemic to the waters of south-western Australia (Caputi et al.

2015; Hart et al. 2016). The limited dietary information available for this species

indicates that O. cf. tetricus predates on Western Rock Lobster, Panulirus cygnus,

caught in lobster traps/pots (Hart et al. 2016), and its diet is likely to comprise

predominantly crustaceans and molluscs, like the diets of other coastal octopuses (Joll

1977; Mangold 1983; Villanueva 1993; Smith 2003; Villegas et al. 2014).

The dietary composition of octopuses is influenced by ontogeny, and as a

consequence, size (Ambrose 1984; Ambrose 1986; Villanueva 1993; Smith 2003; Boyle

and Rodhouse 2005; Regueira et al. 2017). As octopuses increase in body size, the

breadth of their diet typically increases; reflecting an increase in arm length relative to

mantle size, and growth of the brachial crown that allows them to capture larger, more

mobile benthic organisms (Villanueva et al. 1995; Rodhouse and Nigmatullin 1996;

Smith 2003; Boyle and Rodhouse 2005). For example, a study of the Musky Octopus

Eledone moschata in the Adriatic Sea (Croatia), found that teleosts and cephalopods

became increasingly important in the diet as body size increased, while the relative

contribution of crustaceans declined (Šifner and Vrgoč 2009).

Foraging behaviour and habitat type are also important factors influencing the

dietary composition of octopuses (Ambrose 1984; Hewitt et al. 2008). Past foraging

success can drive the modification of prey selection (Ambrose 1984; Mather and O’Dor

41

1991; Anderson 1997; Smith 2003; Gasalla et al. 2010), while the strong affinity of

some prey taxa towards a particular set of environmental and biological characteristics

affects the functional composition of a community, and therefore prey availability

(Ambrose 1984; Hewitt et al. 2008).

Following prey capture, the flesh of crustaceans and molluscs is delicately

removed from the exoskeleton or shell with the use of the beak, suckers and radula

(Boyle and Rodhouse 2005). Larger prey are more easily gleaned than small prey

(Boucher-Rodoni and Mangold 2009). A cocktail of digestive compounds are released

from the posterior salivary gland, which help to break down the prey prior to ingestion

(Nixon 1984; Grisley and Boyle 1987; Grisley and Boyle 1990). The partly digested

prey enters the oesophagus, which feeds into the crop diverticulum that briefly stores

the food (Boyle and Rodhouse 2005; Boucher-Rodoni and Mangold 2009). Digestive

enzymes supplied by the digestive gland further break down the prey, which then passes

to the stomach in instalments (Boucher-Rodoni and Mangold 2009). Rapid digestion

results in little, often unrecognisable food in the stomach, with material beyond the

stomach being visually unidentifiable (Grisley and Boyle 1990; Boyle and Rodhouse

2005; Regueira et al. 2017). When only a small amount of prey is consumed (< 1% of

the total octopus weight), prey can be digested with the first release of enzymes

(Boucher-Rodoni and Mangold 2009).

Dietary studies of octopuses are commonly undertaken by examining prey items

found in the gastric tract (i.e., the crop and stomach) and analysing middens (i.e.,

counting the prey remains discarded around an octopus’ den, Nixon 1987; Villanueva

1993; Mather and Nixon 1995; Smith 2003). Gastric tracts typically contain a broader

range of prey taxa and a higher occurrence of soft bodied organisms than those found in

middens (Boyle and Rodhouse 2005; Jackson et al. 2007; Villegas et al. 2014). This is

despite the fact that prey remains are highly fragmented due to their mastication by the

octopus’ beak, the selective rejection of hard parts during consumption, and rapid

digestion. Conversely, hard-shelled prey (i.e., gastropods and bivalves), are typically

42

over-represented in midden counts as water currents and benthic scavengers

(e.g., brachyurans) rapidly remove the remains of lightweight prey such as crustaceans,

polychaetes and teleosts (Ambrose and Nelson 1983; Smith 2003).

In addition to these analyses of gastric tract and midden contents, biochemical

techniques such as stable isotope analysis have been used increasingly in cephalopod

research of their trophic ecology. These chemical tracers facilitate an understanding of

the long-term assimilation of nutrients into tissues; helping disentangle trophic

relationships and overcoming some of the inherent bias of gastric tract analyses,

including the ‘last meal’ bias (Jackson et al. 2007; Layman et al. 2012). Since stable

isotopes summarise long-term dietary intake and do not identify individual items in the

diet, their use in dietary studies is most valuable when combined with gastric tract

analysis. This allows the breadth of diet to be evaluated whilst enabling the longer term

assimilation of carbon and nitrogen to be quantified (Jackson et al. 2007; Layman et al.

2012; Phillips et al. 2014).

Limited information is known about the trophic position of octopuses in south-

western Australia (Caputi et al. 2015), and the trophic functioning of artificial reefs is a

relatively new area of research (Cresson et al. 2014a). Recent research suggests that

artificial reefs can enhance primary productivity and, as a consequence, support

secondary production and increase yields in small-scale fisheries (Cresson et al. 2014a;

Wu et al. 2016). Given the increasing importance of cephalopods in the marine

environment (Doubleday et al. 2016) and their innate ability to adapt to new

surroundings (Mather 1994; Rigby and Sakurai 2005; Finn et al. 2009), improving our

understanding of the trophic relationships of octopus within the food web is

fundamental to assessing their ecological role in marine communities (Livingston 2003;

Navarro et al. 2013).

Artificial reefs are typically deployed to enhance the production of small-scale

coastal fisheries worldwide (Cresson et al. 2014a; Taylor et al. 2017). Throughout Asia,

this is combined with aquaculture-based enhancement (i.e., the release of cultured

43

individuals in the marine environment), to enhance the yields of commercially valuable

stocks such as the sea cucumber Apostichopus japonicus, Haliotis sp., and bivalve

shellfish including the Manila Clam, Venerupis philippinarum (Goodsell et al. 2006;

James et al. 2007; Becker et al. 2008; Melville-Smith et al. 2013; Zhang et al. 2015;

Taylor et al. 2017). In recent times, species-specific artificial reefs have been used to

support the sea ranching of cultured Haliotis laevigata in south-western Australia (see

Chapter 2, Figure 2.2). However, high stocking densities may increase the abundance of

opportunistic predators such as fishes, crabs, and octopuses responding to an increase in

prey abundance, which could have negative impacts for the survival of released H.

laevigata (McCormick et al. 1994; Shepherd and Clarkson 2001; James et al. 2007;

Melville-Smith et al. 2013; Hart et al. 2013a). In a three-year abalone sea ranching trial

in Flinders Bay, O. cf. tetricus was identified as a potentially important predator of H.

laevigata, with numerous incidences of predation observed during the study period

(Melville-Smith et al. 2013).

The aims of this study were to examine the contents of the gastric tract, and the

assimilation of carbon and nitrogen by O. cf. tetricus collected from artificial reefs in

Flinders Bay, Western Australia (used to grow cultured H. laevigata, see Chapter 2.2) to

determine:

(i) whether the dietary composition of O. cf. tetricus differs between the crops

and stomachs and, if so, which organ provides the best dietary information;

(ii) whether the dietary composition of O. cf. tetricus changes with increasing

body size;

(iii) whether O. cf. tetricus consumes H. laevigata, and if so, what proportion of

the diet it represents; and

(iv) the assimilation of nutrients (C and N) by O. cf. tetricus and H. laevigata

within the marine food web of Flinders Bay to understand the long-term

dietary trends of O. cf. tetricus and where this species fits within the food

web, using stable isotope analysis.

44

3.2. Materials and methods

3.2.1. Collection of Octopus cf. tetricus

Individuals of O. cf. tetricus were caught by commercial abalone divers from the Ocean

Grown Abalone (OGA) lease sites, situated in Flinders Bay, Western Australia (see

Chapter 2.2). Samples were collected by hand during daylight hours between March and

August 2017 at depths of ~20 m. Samples were frozen within 7 h after capture, i.e., as

soon as possible with the constraints of commercial operations on the abalone lease (see

Chapter 2.2).

3.2.2. Quantification of Octopus cf. tetricus diet

After defrosting, the total weight (g), total length (mm), and mantle length (mm) of each

O. cf. tetricus was measured and its gastric tract removed. The crop was separated from

the rest of the tract and its fullness was estimated visually on a scale of 1 to 10, with 10

being the equivalent of 100% full. Any material found in the oesophagus, before it

entered the crop was grouped with the crop contents. All contents from the stomach

were removed and examined separately. Contents that could not be identified from the

crop and stomach were rinsed through a 500 𝜇m sieve to remove silt and excess liquid

that would hinder the identification of the remaining contents (e.g., Smith 2003;

Regueira et al. 2017). The material was then sorted under a dissecting microscope and

identified to the lowest taxonomic level possible, based mainly on the examination of

hard parts, namely the radula and odontophore of abalone; and crustacean exoskeletons,

legs and chelipeds. Cephalopod flesh was recognised by its distinctive texture, the

characteristic chromatophores on the skin’s surface and the presence of black ink in the

gastric tract, while bivalves could be recognised by their characteristic feeding

structures (i.e., the ctenidium, Fig. 3.1). Where no identifiable features were present to

discriminate between abalone and bivalves, the material was assigned as ‘non-

cephalopod mollusc’.

The volumetric contributions of each prey item found within the crop and the

stomach of O. cf. tetricus was estimated by eye and expressed as percentage by volume

45

(%V). Prey in an advanced state of digestion, or so highly masticated that it could not be

identified was labelled ‘unidentified’ (Table 3.2).

Octopus were assigned to three weight classes; i.e., < 300 g, 300-999 g, and ≥

1000 g, in an attempt to measure size related changes in dietary composition (following

Smith 2003). The size class categories were selected to evaluate the diets of juvenile,

sub-adult and mature or large individuals. Since males and females mature at different

weights, ~950 and ~1,800 g respectively (Leporati et al. 2015), a lower limit of 1,000 g

was set as a baseline to account for differences between sexes.

46

(a) Odontophore of Haliotis laevigata

(b) Radula of Haliotis laevigata

(c) Crop contents

(d) Stomach contents

(e) Octopus tissues

(f) Rinsed octopus tissue

Figure 3.1. Dietary items found in the crops and stomachs of Octopus cf. tetricus: (a) odontophore of Haliotis laevigata; (b) radula of H. laevigata; (c) contents of crop including brachyurans and H. laevigata, as evidenced by chelipeds and odontophore; (d) stomach contents showing the advanced state of digestion compared with the crop; (e) cephalopod tissue found in the crop before rinsing; and (f) cephalopod tissue found in the crop after rinsing. Numbers in images denote the octopus sample number, and the contents (c) and (d) are from the same animal.

47

3.2.3. Stable isotope preparation

A range of consumer and primary producer species that were visually abundant around

the artificial reefs, were collected by hand or hand spear by SCUBA divers at depths of

~19 m during May 2017 and frozen (Table 3.1).

Each fish was thawed, identified to species and the total length (mm) measured.

A portion of the dorsal muscle tissue was removed for stable isotope analysis and the

skin removed using a scalpel, dissection scissors, and forceps. The total weight and

shell length of each H. laevigata was measured, and a section of the muscular foot was

removed (following Guest et al. 2008). A segment of the mantle tissue of O. cf. tetricus

was extracted, and the thin outer layer of tissue containing the chromatophore organs

was removed.

Algae and seagrass were identified to at least family level, rinsed with deionised

water to remove salts, and then epiphytes were removed using a razor blade to minimise

the influence of epiphytes on the isotopic ratios. Tissue was selected haphazardly from

the lower, midline and tip of each algal frond (following Guest et al. 2008).

Each consumer tissue sample was placed in a pre-cleaned open glass vial, while

algae samples were dried in clean aluminum foil trays and oven dried at 60 °C for up to

72 h. Dried samples were then broken down manually into small pieces using a mortar

and pestle and placed into 2.0 mL Eppendorf tubes. Standard measures of 0.5 mL for

consumer tissue and 1.0 mL for primary producer tissue were placed into each tube with

a 5 mm ball bearing. A Qiagen TissueLyser II (ball mill), operated at 27 shakes per

second, was used to grind each of the tissue samples into fine powders.

Samples were sent to the West Australian Biogeochemistry Centre at the

University of Western Australia and analyzed on a Delta V Plus continuous flow

isotope-ratio mass spectrometer. The isotope ratios of nitrogen 15N/14N and carbon 13C/12C were calculated and presented as a delta (δ) value, representing the relative per

mil (‰) difference between the sample and the recognized international standard (Pee

Dee belemnite carbonate and atmospheric nitrogen for δ13C and δ15N, respectively).

48

Table 3.1. The number of primary producer and consumer samples collected and analysed for stable isotopes of δ13C and δ15N from the Greenlip Abalone (Haliotis laevigata) sea ranching site in Flinders Bay, Western Australia in May 2017. The total numbers of samples of major groupings are shown in parentheses. * refers to location (Line) from which abalone were sampled (see Chapter 2.2.2).

Taxa Number of samples

Primary producers (18) Seagrass (3)

Amphibolis antarctica 3 Macroalgae (15)

Rhodophyta (9) Graciliara sp. 3 Plocamium mertensii 3 Spyridia sp. 3

Phaeophyceae (6) Sargassum sp. 3 Sargassum sp. 3

Consumers (61) Abalone (24) Haliotis laevigata 1 Year O1* 4 Haliotis laevigata 1 Year O5* 4 Haliotis laevigata 2 Years O1* 4 Haliotis laevigata 2 Years O5* 4 Haliotis laevigata 3 Years O1* 4 Haliotis laevigata (Seeds 0+) 4

Cephalopods (14) Loliginidae sp. 2

Octopus (12) O. cf. tetricus (< 300 g) 4 O. cf. tetricus (300 – 999 g) 4 O. cf. tetricus (≥ 1,000 g) 4

Teleosts (20) Coris auricularis 4 Chrysophrys auratus 4 Neatypus obliquus 4 Pseudocaranx georgianus 2 Opthalmolepis lineolatus 4 Upeneichthys vlamingii 2

Tunicata sp. (3)

49

3.2.4. Analyses of dietary data

Multivariate analyses were carried out using the PRIMER v7 multivariate statistical

package (Clarke et al. 2014a) and the Permutational Multivariate Analysis of Variance

(PERMANOVA)+ add-on module (Anderson et al. 2008a) to evaluate how the dietary

composition of O. cf. tetricus varied among body size classes (i.e., weight class) and

between gastric organs (i.e., crop vs. stomach). The null hypothesis of no significant

difference in dietary composition between a priori groups was rejected when a

significance level (P) was ≤ 0.05.

Comparison of diet between gastric organs

In order to elucidate any broad trends in dietary composition between gastric organs

(accounting for potential changes with increasing body size), the mean volumetric

percentage contributions of each of the 13 dietary categories found in the crops and

stomachs of O. cf. tetricus in each weight class (Table 3.2) were calculated and plotted

as stacked bar graphs.

Preliminary analyses were employed to determine whether the dietary

composition of O. cf. tetricus differed according to the Gastric Organ (2 levels; Crop

and Stomachs), taking into account any potential confounding effect of Weight Class (3

levels; < 300, 300-999 and > 1,000 g). This analysis included unidentified material,

unlike traditional multivariate analysis of diet, where unidentified material is eliminated

as a dietary category (e.g., Platell and Potter 2001; Platell et al. 2010; French et al.

2012; Coulson et al. 2015). The rationale for this was to place focus on what was

unknown about the diet, and to make comparisons between the gastric organs.

As the gastric tracts of individual octopus may contain only a small number of

dietary categories, two octopuses collected at the same time from the same location may

differ markedly in their dietary composition. This can mask subtle but ‘real’ trends in

diet. To reduce this potential effect, samples for each weight class for each dietary

organ were randomly sorted into groups of between three and five, depending on the

50

total number of octopus in the samples (Platell and Potter 2001; Lek et al. 2011;

Coulson et al. 2015).

The dietary categories ‘Octopus cf. tetricus’ and ‘unidentified cephalopods’

were grouped for this and subsequent multivariate analyses, as the latter category

comprised only a very small proportion of the diet. Algae and shell grit/sediment were

removed from the analysis, since the ingestion of these items is likely to be an

unintentional consequence of prey capture (Boyle and Rodhouse 2005). The data for the

remaining dietary categories were then standardised (i.e., re-converted to percentage

contribution). The percentage volumetric contribution of the different dietary categories

in each group of replicates were averaged and square-root transformed to down-weight

abundant items and remove any tendency for the main dietary constituents to be

excessively dominant (see Lek et al. 2011). These averaged and transformed data were

then used to construct a Bray–Curtis resemblance matrix, which was in turn, subjected

to PERMANOVA (Anderson et al. 2008a). A two-way crossed PERMANOVA was

used to elucidate whether there was a significant interaction between the Gastric Organ

and Weight Class, and if so, to determine the extent of that interaction relative to the

main effects.

Two-way Analyses of Similarities (ANOSIM; Clarke and Green 1988) were

used to test which factors, if any, were significant. In this and subsequent ANOSIM

tests, both R-statistics and their associated significance level (P) are cited. The

magnitude of the R-statistic ranges from ~0 to 1, with 0 indicating that the average

similarity among and within groups does not differ, and 1 indicating that all samples

within each group are more similar to each other than to any of the samples from other

groups (Clarke et al. 2014a). A non-metric Multidimensional Scaling (nMDS)

ordination plot was constructed from the same Bray-Curtis resemblance matrix used

above, to provide a visual representation of the similarity among groups of samples.

A shade plot (i.e., a visual representation of a frequency matrix, Clarke et al.

2014b), was constructed to elucidate the dietary categories responsible for any

51

differences in the dietary composition of O. cf. tetricus based on samples of crops and

stomachs. On this plot, white space for a dietary category demonstrates that the category

was never recorded, while the depth of shading from grey to black is linearly

proportional to the square-root transformed percentage volumetric contribution of that

category. To improve clarity, dietary categories that contributed < 10% of the overall

volume of any group of samples (i.e., barnacles, and isopods) were excluded from the

shade plot. The dietary categories (y-axis) are ordered to optimise the seriation statistic

ρ, by non-parametrically correlating their resemblances to the distance structure of a

linear sequence (Clarke et al. 2014a). The order of the samples (x-axis) was seriated

independently of the dietary categories, but employing a constrained seriation by

Gastric Organ. Thus, crop samples were kept together and separate from those of

stomachs and vice versa, but samples within each gastric organ were ordered to

maximise ρ (Clarke et al. 2014a; Coulson et al. 2015).

Size related differences in diet

Having determined which gastric organ of O. cf. tetricus provided the best

representation of dietary composition, the data from that organ was used to determine

the presence and strength of any weight related changes in dietary composition

(i.e., crops and/or stomachs). For these analyses the %V contribution of the unidentified

material was removed, the data were then re-standardised and samples were pooled into

groups of three to four samples from the same weight class, where possible, and

averaged.

These pre-treated data were then square root transformed and used to create a

Bray–Curtis resemblance matrix, which was subjected to one-way ANOSIM to

determine whether dietary composition differed among weight classes. This same

matrix was used to construct an nMDS ordination plot, which provided a visual

representation of the similarity of samples. A shade plot was used to determine the basis

for any differences in diet among weight classes and the trends in dietary categories

responsible for those differences.

52

Stable isotope composition of prey and primary producers

Mean values (± 1 SE) of δ15N and δ13C were calculated for each taxonomic group of

consumers and primary producers collected. Prior to undertaking analyses, Shapiro-

Wilk tests showed that the stable isotope data were not normally distributed. Therefore,

a non-parametric Kruskal-Wallis test was used to determine whether the δ13C and δ15N

signatures differed significantly among O. cf. tetricus and other consumers, and

between primary producers. When significant differences were detected, a post-hoc

Dunn’s multiple comparisons test was performed to determine which a priori groups

differed from each other. The Dunn’s test is appropriate for groups with unequal

numbers of observations (Zar 2010). These inferential statistical analyses were

performed using the R statistical software package (R Core Team 2017) and the “FSA”

(Ogle 2017) add-on package.

To calculate the number of trophic levels present within the system, the mean

δ15N value for all primary producers was calculated and assigned the trophic level of

one (following Linke 2011; Como et al. 2017). The mean δ15N value for abalone was

calculated and assigned the trophic level of two, and higher sequential trophic levels

were estimated by adding the δ15N fractionation between trophic levels, assumed to be

3.4‰ (Vander Zanden and Rasmussen 2001).

3.3. Results 3.3.1. Comparison of diet between gastric organs

Of the 44 O. cf. tetricus sampled, 34 contained food in either the crop and/or the

stomach. Prey items found in the gastric tract were allocated to one of 13 dietary

categories (Table 3.2). All 13 dietary categories were found in the crops, compared with

10 in the stomachs. Of the stomach contents, 74% of all material was unidentifiable,

compared with only 4% in the crops (Table 3.2, Fig. 3.2). For crops alone, molluscs

made the largest contribution by volume percentage (%V), comprising 54% of the diet

of O. cf. tetricus (Table 3.2, Fig. 3.2). One third (~31%) of the molluscs was attributed

to abalone, ~13% to non-cephalopod molluscs, ~6% to octopus and ~4% bivalves

53

(Table 3.2, Fig. 3.2). Crustaceans (~33%) were the second largest dietary category by

%V, of which ~8% were brachyurans (comprising at least three species; Table 3.2). For

stomachs alone, molluscs comprised the majority (73%) of the identifiable contents

(19% of the total stomach contents), of which ~6% was octopus, 2% abalone and 2%

bivalves (Table 3.2). A total of ~6% of the diet in the stomach comprised crustaceans

(Table 3.2). Barnacles, brachyurans and isopods were absent from the stomach contents

(Table 3.2, Fig. 3.2).

Table 3.2. Mean volumetric contribution (%V) of the major dietary components found in the crops and stomachs of Octopus cf. tetricus, collected from the Greenlip Abalone, Haliotis laevigata, sea ranching site in Flinders Bay, Western Australia between March and July 2017. * = dietary items excluded from the multivariate analysis of crop and stomach dietary composition.

Dietary Category Crops Stomachs

Molluscs (total) 53.90 19.16 Abalone 31.10 2.21 Bivalve 3.59 2.21 Non-cephalopod mollusc 13.24 8.31 Cephalopod 5.97 6.44 Octopus cf. tetricus 5.68 5.85 Unidentified cephalopod 0.29 0.59

Crustaceans (total) 33.49 5.96 Amphipod 0.03 0.29 Barnacle 0.03 0.00 Brachyuran 7.97 0.00 Isopod 0.29 0.00 Unidentified crustacean 25.16 5.66

*Algae 3.15 0.06 *Shell fragment/Sediment 5.43 0.56 *Unidentified Material 4.04 74.26 Samples examined (N) 44 44 Samples with prey (N, %) 34, 77.3% 34, 77.3%

Two-way PERMANOVA detected a significant difference in the dietary

composition of O. cf. tetricus between Gastric Organ (crops vs stomach), but not among

Weight Classes, and the Gastric Organ × Weight Class interaction was not significant

(Table 3.3). A two-way ANOSIM demonstrated that the contents of the crops and

54

stomachs differed significantly and that the magnitude of those differences was

moderate (R = 0.53, P < 0.001). This is clearly evident on the nMDS ordination plot,

with the points representing the crops (to the left of the nMDS) typically very well

separated from those of the stomachs (to the right of Fig. 3.3).

Figure 3.2. Stacked bar graphs showing the mean percentage volumetric contributions of the various dietary categories to sequential weight classes (g) in the crops and stomachs of Octopus cf. tetricus collected from the Greenlip Abalone, Haliotis laevigata, sea ranching site in Flinders Bay, Western Australia between March and July 2017. Number of crops and stomachs examined are shown above the stacked bar graphs for each weight class. N = 34

Table 3.3. Degrees of freedom (df), mean squares (MS), pseudo-F ratios, and significance levels (P) for PERMANOVA test, employing a Bray-Curtis similarity matrix derived from the mean percentage volumetric contributions of the eight main dietary categories to the crops and stomachs of Octopus cf. tetricus collected from the Greenlip Abalone, Haliotis laevigata, sea ranching site in Flinders Bay, Western Australia between March and July 2017. Bold face denotes significant difference.

Factor df MS %MS Pseudo-F P Gastric Organ 1 21560 83.33 12.78 0.001 Weight Class 2 2404.10 9.29 1.98 0.089 Gastric Organ × Weight Class 2 695.55 2.69 0.57 0.732 Residual 18 1212.70 4.69

0%

20%

40%

60%

80%

100%

Volumetric

Con

tribution

WeightClass(g)

CropsStomachs

Unidentifiedmaterial

Shellfragment/sediment

Algae

Un-identifiedcrustacean

Isopod

Brachyuran

Barnacle

Amphipod

Cephalopod

Noncephalopodmollusc

Bivalve

Abalone

1515414164

55

Figure 3.3. nMDS ordination plot, constructed from a Bray–Curtis similarity matrix of the volumetric contributions of the eight main dietary categories from the � Crops and � Stomachs of Octopus cf. tetricus, collected from the Greenlip Abalone, Haliotis laevigata, sea ranching site in Flinders Bay, Western Australia between March and July 2017.

The shade plot indicates that the differences between crops and stomachs were

due to the large volumetric contributions of unidentified material in the stomachs, while

abalone and brachyurans had a greater contribution in the crops than stomachs (Fig. 3.2

and 3.4). The contribution of crustaceans was also higher in the crops than stomachs

(Fig. 3.4). Crustaceans occurred in all but one group of crops, while abalone were

present in all but two of the 12 groups. Unidentified material was present in every group

of stomachs, with three samples containing only unidentified material (Fig. 3.4).

56

Figure 3.4. Shade plot showing the mean square-root transformed percentage volumetric contributions of eight main dietary categories found in the crops and stomachs of Octopus cf. tetricus collected from the Greenlip Abalone, Haliotis laevigata, sea ranching site in Flinders Bay, Western Australia between March and July 2017. White represents an absence of prey, and grey to black represents the increasing contribution of a dietary category to the total diet.

This analysis demonstrates that the examination of different gastric organs can

lead to dramatically different results in the dietary compositions of O. cf. tetricus. Size

related changes in the diet of O. cf. tetricus was therefore undertaken using only the

data obtained from crops, as this was considered to provide a more-reliable indicator of

dietary composition.

3.3.2. Size related differences in diet

One-way ANOSIM demonstrated that the dietary composition of O. cf. tetricus did not

differ between weight classes (R = 0.05, P = 0.31), which is evident from the nMDS

plot, with the interspersion of points representing each of the three weight classes

(Fig. 3.5). Points from the smallest size class are towards the centre-right of the nMDS

plot, with those for the largest octopus above and below them.

57

Figure 3.5. nMDS ordination plot, constructed from a Bray–Curtis similarity matrix that employed the volumetric contributions of the nine main dietary categories to the crops of sequential weight classes (p < 300, p 300–999, r ≥ 1,000 g) of Octopus cf. tetricus collected from the Greenlip Abalone, Haliotis laevigata, sea ranching site in Flinders Bay, Western Australia between March and July 2017.

The shade plot demonstrates that abalone and crustaceans, including

brachyurans are abundant and frequently consumed components of the diet of O. cf.

tetricus regardless of body size (Fig. 3.6, see also Fig. 3.2). Other molluscs such as

cephalopods and bivalves, and small crustaceans including amphipods and isopods,

vary in their relative contributions between weight classes, but the presence of these

dietary categories is less consistent among groups of samples and they contribute less to

the overall diet (Fig. 3.6).

58

Figure 3.6. Shade plot showing the mean square-root transformed percentage volumetric contributions of each of the nine main dietary categories found in the crops and stomachs of Octopus cf. tetricus in sequential weight classes (p < 300, p 300–999, r ≥ 1,000 g) collected from the Greenlip Abalone, Haliotis laevigata, sea ranching site in Flinders Bay, Western Australia between March and July 2017. White represents an absence of prey, and grey to black represents the increasing contribution of a dietary category to the total diet.

3.3.3. Stable isotope composition of prey and primary producers

The mean δ15N values for O. cf. tetricus ranged from 7.92 ± 0.30 (≥ 1,000 g) to 8.27 ±

0.36‰ (300-999 g, Table 3.4, Fig. 3.6). The mean δ15N values for abalone (3.77 ± 0.09

to 3.99 ± 0.01‰) sampled from different locations and of different ages were from 3.93

to 4.93‰ lower than those for octopus, with the lowest δ15N recorded for the newly

seeded H. laevigata (3.34 ± 0.06‰, Table 3.4, Fig. 3.7). The δ15N values for the teleosts

ranged from 9.02 ± 0.14 (Upeneichthys vlamingii) to 11.54 ± 0.54‰ (Chrysophrys

auratus), while loliginid squids were 9.49 ± 0.27‰ (Table 3.4, Fig. 3.7). Considerable

differences were also identified between primary producers, with the mean δ15N values

ranging from 1.80 ± 0.18 (the seagrass Amphibolis antarctica) to 3.76 ± 0.05‰ (the red

alga Plocamium mertensii, Table 3.4, Fig 4.7).

59

The δ15N values did not differ significantly among size classes of O. cf. tetricus

(Kruskal-Wallis, X22 = 0.35, P = 0.84, Table 3.4). However, significant differences were

found between O. cf. tetricus and other consumer groups (Kruskal-Wallis, X32 = 48.93,

P < 0.001), including teleosts (10.10 ± 0.22‰, Dunn’s Test, Z11, 19 = -2.72, P = 0.03)

and abalone (3.77 ± 0.05‰, Z11, 23 = -3.08, P < 0.01, Table 3.4).

Table 3.4. Stable nitrogen (δ15N) and carbon (δ13C) values and for primary producers and consumers collected from the Greenlip Abalone (Haliotis laevigata) sea ranching site in Flinders Bay, Western Australia in May 2017, and prepared for stable isotope analysis. N = sample size. The total numbers of samples of major groupings are shown in parentheses. * refers to location (Line) from which abalone were sampled (see Chapter 2.2.2)

δ15N ‰ δ13C ‰ Taxon N Mean SE Mean SE Primary producers (85) 2.90 0.17 -28.08 1.49

Seagrass (3) 1.80 0.18 -17.85 0.14 Amphibolis antarctica 3 1.80 0.18 -17.85 0.14

Macroalgae (15) 3.12 0.15 -30.13 1.20 Rhodophyta (9) 3.09 0.23 -32.71 1.46

Graciliara sp. 3 2.25 0.19 -26.92 0.44 Plocamium mertensii 3 3.76 0.05 -36.02 0.24 Spyridia sp. 3 3.28 0.05 -35.19 0.20

Phaeophyceae (6) 3.15 0.15 -26.26 0.19 Sargassum sp. 3 3.44 0.02 -26.50 0.20 Sargassum sp. 3 2.87 0.19 -26.01 0.28

Consumers Abalone (24) 3.77 0.05 -27.41 0.26

Haliotis laevigata Y1 O1* 4 3.77 0.09 -27.55 0.15 Haliotis laevigata Y1 O5* 4 3.99 0.01 -28.27 0.15 Haliotis laevigata Y2 O1* 4 3.90 0.08 -27.69 0.45 Haliotis laevigata Y2 O5* 4 3.80 0.03 -28.43 0.23 Haliotis laevigata Y3 O1* 4 3.83 0.09 -27.55 0.43 Haliotis laevigata (0+ Seeds) 4 3.34 0.06 -25.00 0.08

Cephalopods (14) 8.28 0.22 -23.37 0.34 Loliginidae sp. 2 9.49 0.27 -21.82 0.20

Octopus (12) 8.08 0.19 -23.63 0.42 O. cf. tetricus (< 300 g) 4 8.05 0.43 -23.86 0.76 O. cf. tetricus (300 – 999 g) 4 8.27 0.36 -24.03 0.49 O. cf. tetricus (≥ 1,000 g) 4 7.92 0.30 -23.00 0.51

Teleosts (14) 10.10 0.22 -23.23 0.47 Coris auricularis 4 9.66 0.02 -23.00 0.10 Chrysophrys auratus 4 11.54 0.54 -21.53 0.44 Neatypus obliquus 4 10.20 0.08 -27.05 0.14 Pseudocaranx georgianus 2 10.41 0.02 -21.34 0.26 Opthalmolepis lineolatus 4 9.37 0.20 -22.56 0.06 Upeneichthys vlamingii 2 9.02 0.14 -22.71 0.30

Tunicata sp. (3) 4.48 0.42 -22.41 0.28

60

The δ15N values differed significantly between the teleosts (X52 = 17.47, P <

0.05): C. auratus and Opthalmolepis lineolatus (Z3,3 = -3.17, P = 0.02), and C. auratus

and Pseudocaranx georgianus (Z3,3 = 0.64, P = 0.02, Table 3.4). Significant differences

in δ15N were also identified among the primary producers (X22 = 6.49, P = 0.04):

between Phaeophyceae (3.15 ± 0.15‰) and A. antarctica (Z2,2 = 2.25, P = 0.05), and

Rhodophyta (3.09 ± 0.23‰) and A. antarctica; (Z2,2 = 2.44, P = 0.04, Table 3.4).

Considering the baseline values for δ15N and a fractionation of 3.4‰ for trophic

levels above herbivores (i.e., abalone, Vander Zanden and Rasmussen 2001), the food

web spanned ~4 trophic levels between primary producers and teleosts (Fig. 3.7).

Octopus cf. tetricus occupied a mid trophic level of <3.5, while teleosts and loliginid

squid occupied a trophic level of ≥ 3.5. H. laevigata occupied the trophic level of two,

just above the primary producers (Fig. 3.7).

Figure 3.7. Mean (± 1 SE) carbon and nitrogen isotope values of abalone Haliotis laevigata; primary producers Amphibolis antarctica, Gracilliara sp., Sargassum sp.; cephalopods Loliginidae sp., and Octopus cf. tetricus; and teleosts Coris auricularis, Neatypus obliquus, Opthalmolepis lineolatus, Chrysophrys auratus, Pseudocaranx georgianus, and Upeneichthys vlamingii. Samples collected from the Greenlip Abalone, H. laevigata, sea ranching site in Flinders Bay, Western Australia during May 2017. TL denotes trophic level, with the trophic baseline (TL 1) derived from the mean δ15N primary producer value.

H.laevigataH.laevigata(seeds)

A.antartica

N.obliquus

U.vlamingii

Graciliarasp.

C.auricularisO.lineolatus

O.cf.tetricus

C.auratus

Sargassumsp.

Tunicatasp.

P.georgianusLoliginidaesp.

Spyridiasp.P.mertensii

0

2

4

6

8

10

12

14

-40 -35 -30 -25 -20 -15

δ15N

(‰)

δ13C(‰)

Legend

�Seagrass�Rhodophyta�Phaeophyceae�Abalone�Tunicata�Cephalopods�Teleosts

TL1TL2

TL4

TL3

61

The mean δ13C values of O. cf. tetricus size classes varied by only ~1‰ ranging

from -23.00 ± 0.51 (≥ 1,000 g) to -24.03 ± 0.49‰ (300-999 g, Table 3.4). The δ13C

value of H. laevigata seeds (-25.00 ± 0.08‰) was slightly more enriched than older age

classes, which ranged from -27.55 ± 0.15‰ to -28.43 ± 0.23‰ (Table 3.4, Fig. 3.7).

Among the teleosts, Neatypus obliquus had the lowest δ13C signature of -27.05 ±

0.14‰, while C. auratus had the highest at -21.53 ± 0.44‰ (Table 3.4, Fig. 3.7).

Among the primary producers, the red alga P. mertensii was the most depleted in δ 13C

(-36.02 ± 0.24‰), and the seagrass A. antarctica was most enriched (-17.85 ± 0.14‰,

Table 3.4, Fig. 3.7).

The δ13C values of O. cf. tetricus did not differ significantly among size classes

(X22 =1.5, P = 0.47). However, the δ13C signature of O. cf. tetricus differed significantly

from potential other consumer groups (X32 = 36.7, P < 0.001). A subsequent multiple

comparisons test revealed that the only significant difference was between O. cf.

tetricus (-23.63 ± 0.42‰) and abalone (-27.41 ± 0.26‰, Z11,23 = -3.73; P < 0.001).

Significant differences in δ13C were also identified among the teleosts (X52 = 17.5, P <

0.01): between N. obliquus and C. auratus (Z3,3 = -3.56, P < 0.01), and N. obliquus and

P. georgianus (Z3,3 = -3.03, P = 0.35, Table 3.4). The δ13C values differed significantly

among the primary producers (X22 = 12.26, P < 0.01): between the Rhodophyta (-32.71

± 1.46‰) and the seagrass A. antarctica (Z2,2 = -3.25, P < 0.01), and Rhodophyta and

Phaeophyceae (-26.26 ± 0.19‰, Z2,2 = 2.27, P = 0.05, Table 3.4).

3.4. Discussion

This study used gastric organ content (crops and stomachs) and stable isotope analyses

to evaluate the feeding of O. cf. tetricus, and its significance as a predator of H.

laevigata on artificial abalone habitats in a sea ranch of south-western Australia. It also

provides the first comparative assessment of using different dietary organs for

evaluating the diet of octopuses, identifying significant differences in the relative

contributions of taxon between crops and stomachs. Multivariate statistical analyses

were employed to understand how the dietary composition of O. cf. tetricus changes

62

with increasing body size. However, no difference in diet was identified. The isotopes

δ15N and δ13C were examined to understand the long-term dietary trends of O. cf.

tetricus and the trophic level occupied by this species in the food web.

3.4.1. Comparison of diet between gastric organs

The dietary composition of O. cf. tetricus differed greatly between the crop and

stomach. Much greater taxonomic resolution was found in the crop contents than the

stomachs, which in terms of volume were dominated by highly digested, unidentifiable

material. Abalone made the greatest %V contribution to the crops, but was negligible in

the stomachs. Consequently, the contribution of abalone to the diet of O. cf. tetricus

would be greatly underestimated if combining dietary information on crops and

stomachs, or using stomachs alone. This highlights the difficulties of gastric tract

analysis in assessing dietary composition among octopuses.

As part of the ingestion process of octopuses, prey is masticated by the beak,

and hard parts may be selectively excluded (Boyle and Rodhouse 2005; Boucher-

Rodoni and Mangold 2009). Food moves through the oesophagus and is briefly stored

in the highly distendible crop, before passing to the stomach where the bulk of digestion

takes place (Boucher-Rodoni and Mangold 2009, Appendix 1). The rapid digestion

process means that prey identification from the stomach contents can be difficult, with

the potential for soft-bodied organisms to be under-represented due to their shorter gut

retention time than robust hard parts, such as the exoskeletons of crustaceans (Breiby

and Jobling 1985; Pierce et al. 1994; Boyle and Rodhouse 2005; Regueira et al. 2017).

Consequently, crop contents provide a better representation of the diet as prey, while in

a masticated form, is typically less digested compared with stomach contents, and the

bias towards hard-shelled prey is less. In a study of the Flying Squid Todarodes

sagittatus (which lack a crop) in the Norwegian Sea, > 50% of individuals contained

prey in the stomach that was in an advanced state of digestion (i.e., of a pulp-like

consistency, Breiby and Jobling 1985). This precluded the complete identification of

63

prey, particularly where hard structures were absent, and consequently soft bodied prey

were likely to be underestimated (Breiby and Jobling 1985).

3.4.2. Size related differences in diet

No significant ontogenetic changes were found in the dietary composition of O. cf.

tetricus. This is atypical for octopus as crustaceans are frequently consumed by

juveniles, but as they increase in body size, teleosts and other cephalopods become

more important (Nixon 1987; Villanueva 1993; Smith 2003; Villegas et al. 2014). In

addition to abalone, in this study O. cf. tetricus consumed mainly crustaceans including

brachyurans, and bivalves. The non-significant difference between size classes in the

present study was also reflected in the stable isotope analysis. This may be an artefact of

few large individuals (>1,000 g) being available for study and no octopus ≥ 2,000 g

being sampled. The lack of ontogenetic differences may also be due to a low diversity

of prey in the sea ranching lease of Flinders Bay, or the high abundance of H. laevigata

available to O. cf. tetricus.

However, the absence of these larger individuals could be related to the regular

removal of octopuses from the ranch, preventing growth to the upper size limits for this

species (Leporati et al. 2015). Alternatively, the absence of very large individuals may

be associated with an offshore migration of mature individuals (Leporati et al. 2015).

On the west coast of Australia, it is believed that mature O. cf. tetricus females migrate

offshore to rocky-reefs, which provide a suitable substratum to lay and brood their eggs,

and are followed by males in search of mates (Anderson 1997; Leporati et al. 2015).

The diversity of prey found in the crops (13 prey categories) was lower than that

of other gastric tract analysis studies (Cortez et al. 1995; Grubert et al. 1999; Šifner and

Vrgoč 2009), however sample sizes were also much smaller in the current study,

particularly for octopus ≥ 1,000 g (total N = 44, Nlarge = 4). For example, in a study of

the Maori Octopus, Octopus maorum (N = 137), on the eastern shelf of Australia, 14

taxon were identified, of which 10 were identified to species level (Grubert et al. 1999),

and in a study of the Gould Octopus, Octopus mimus (N = 947), in the Pacific Ocean

64

(Chile), 25 different prey items from five taxonomic groups were identified (Cortez et

al. 1995) over 7 and 12 months respectively.

While the breadth of diet may increase with a larger number of samples, the low

diversity in the diet of O. cf. tetricus may be due to three other main factors. Firstly, the

ability to effectively discriminate between unique molluscan and crustacean species due

to mastication and the gleaning of prey items is likely to be a contributing factor (Boyle

and Rodhouse 2005). In the study of O. mimus, mollusc shells were collected from

around octopus middens during sampling, which aided in the identification of prey

when occasional shell fragments were located in the gastric tract (Cortez et al. 1995). A

similar approach could be implemented and extended to include a survey of benthic

invertebrates to assist in future identification of potential prey items. While at least three

species of brachyurans were present in the diet of O. cf. tetricus, prey fragments were

either too small or too fragmented to accurately identify species. In addition, non-

cephalopod molluscan tissue was commonly encountered, but could not be identified

due to the absence of distinguishing morphological features, such as an odontophore,

radula or ctenidia. Complementary DNA-based approaches could be applied to aid in

the identification of digested prey items, with the potential to provide a more

comprehensive understanding of diet (Braley et al. 2010; Pompanon et al. 2012).

Midden counts could also be used to complement gastric tract analysis, helping

to reveal additional prey sources and a more complete picture of the diet of O. cf.

tetricus (see Smith 2003). While midden counts are biased towards hard-shelled prey

such as bivalves and gastropods (Ambrose 1984; Smith 2003), such analyses would

assist in determining dietary breadth. For example, a four-year study of prey remains in

the middens of the California Two Spot Octopus, Octopus bimaculatus, in the North

Pacific Ocean (California) resulted in 55 different species being identified (Ambrose

1984). While the vast majority of prey identified were gastropods and bivalves (~78%),

crustaceans, teleosts, and other cephalopods were also found (Ambrose 1984).

65

The second reason for the low diversity of the crop contents of O. cf. tetricus

may be a function of the readily available food sources associated with the sea ranching

Abitats in Flinders Bay. Haliotis laevigata are likely to attract other animals, which are

also potential prey of octopus, including crabs (Carter et al. 1985; James et al. 2007); a

preferred prey of octopuses (Ambrose 1984; Ambrose 1986; García García and

Valverde 2006). Consequently, O. cf. tetricus feeding on abalone and crabs around the

Abitats expend relatively little energy in obtaining their nutritional needs, as they do not

need to move far from their dens to find suitable prey. Octopuses have been identified

as time saving foragers, spending much of their time within their dens (Mather and

O’Dor 1991; Scheel et al. 2016). Thus, the occupation of these artificial habitats

appears to be well suited to the time minimising foraging strategy of O. cf. tetricus.

Thirdly, the low diversity in the diet of O. cf. tetricus may be an artefact of the

husbandry practices carried out as part of the sea ranching operations. Abitats are

routinely scraped clean to remove the growth of algae and sessile invertebrates to

increase the surface area available to abalone (see Chapter 2, 2.2.4). Due to the

considerable distance of the Abitats from the source of colonists (i.e., natural reefs in

Flinders Bay), and the relatively short timeframe between Abitat cleaning, the diversity

of sessile invertebrates growing on the Abitats is likely to be comparatively lower than

what may be expected if successional processes were left to equilibrate (Svane and

Petersen 2001).

Shade plots highlight the importance of molluscs and crustaceans to the diet of

O. cf. tetricus, concurrent with the findings for other coastal octopuses (Ambrose and

Nelson 1983; Ambrose 1984; Ambrose 1986; Nixon 1987; Smith 2003; Scheel et al.

2016). However, the relative contribution of various prey items differs considerably

between species (see Appendix 1, A1.7.1), which likely represents a compromise

between prey availability and prey preferences (Ambrose 1984). For example, while

crustaceans are a preferred prey source of O. bimaculatus in California, their abundance

66

is relatively low in the local environment, and thus gastropods are consumed most

frequently (Ambrose 1984).

The contribution of abalone to the diet of O. cf. tetricus is greater than that

observed in the gastric tract of other species (Grubert et al. 1999; Smith 2003; Villegas

et al. 2014), which is likely to be the result of O. cf. tetricus exploiting the readily

available abalone on the Abitats in Flinders Bay. Octopuses adapt readily to novel

feeding opportunities (Rigby and Sakurai 2005), and since H. laevigata are very

abundant on the Abitats, it is no surprise that the majority of the diet is attributable to

abalone. Ambrose (1984) noted a similar finding for O. bimaculatus situated within

close proximity to mussel beds, with these individuals feeding almost exclusively on

mussels. Although, O. cf. tetricus is likely to consume crustaceans whenever the

opportunity arises due to their high energetic returns and low handling time (from

capture to ingestion, Ambrose 1984, McQuaid 1994).

Crustaceans are an important prey for cephalopods due to their low lipid content

(1.08 g per 100 g; García García and Valverde 2006; United States Department of

Agriculture 2016a). Unlike other active marine predators, which use lipids for their

long-term energy stores, cephalopods utilise amino acids as their principal energy

source, which supports rapid growth rates (Lee 1995; Boyle and Rodhouse 2005). As a

result, lipid digestibility is low in cephalopods (O’Dor et al. 1984; Lee 1995) and high

lipid diets reduce their growth rates (García García and Aguado Giménez 2002).

Abalone is lower in lipids than crustaceans (0.76 g per 100 g) and has a high caloric

content (United States Department of Agriculture 2016b). Thus, the combination of

crustaceans and abalone, and their associated macronutrient composition may be

sufficient to meet the nutritional needs of O. cf. tetricus. Based on crop content analysis,

O. cf. tetricus appears to be an important predator of H. laevigata.

3.4.3. Stable isotope composition of prey and primary producers

The δ15N signatures of O. cf. tetricus were similar between size classes, reflecting the

similar dietary composition between the groups (Table 3.4). The δ15N signature of O. cf.

67

tetricus is lower than that of O. maorum (12.6‰) on the south-eastern Australian

continental shelf, but similar to the Southern Keeled Octopus, Octopus Berrima, at the

same location (9.0‰, Davenport and Bax 2002). Dietary studies have shown that O.

maorum consumes large volumes of cephalopods and fishes (Grubert et al. 1999),

which explains the higher δ15N signature in this species. However, no dietary

information is available for O. berrima. The lower δ15N value of O. cf. tetricus is likely

explained by the high consumption of H. laevigata and potentially crustaceans, but also

the low volumes of teleosts consumed.

The δ15N values of O. cf. tetricus were enriched by more than one trophic level

compared with H. laevigata, but depleted compared with the teleosts and loliginid

squids, highlighting distinct differences in diet among the groups (Cherel et al. 2011).

The low δ15N signature of H. laevigata highlights its role as a primary consumer, and its

exclusive macroalgal diet (Guest et al. 2008). Generally, the δ15N signature increased

with fish size, with the highest values for the largest individuals caught (C. auratus),

which may be explained by the ability of larger fish to capture larger prey from a wider

range of sources (Davenport and Bax 2002). The δ15N value of Pink Snapper, C.

auratus, was consistent with the values attained for this species on the southeast

Australian continental shelf (11.3‰; Davenport and Bax 2002). The diet of C. auratus

in south-western Australia comprises echinoderms, teleosts, crustaceans and molluscs,

including cephalopods (French et al. 2012). The trophic shift for vertebrate consumers

is higher than herbivores or those that feed on invertebrates alone (McCutchan et al.

2003). Thus, the consumption of teleosts by C. auratus likely explains the enrichment

of δ15N compared with O. cf. tetricus and other teleosts. The δ15N value for Maori

Wrasse, O. lineolata, in the current study (9.4‰) was lower than that identified by

Davenport and Bax (13.1‰, 2002). Such differences may be indicative of local

variation in dietary composition, driven by intraspecific competition and resource

availability (Anseeuw et al. 2010; Lek et al. 2011).

68

The δ15N values indicate that the food web of Flinders Bay spanned ~4 trophic

levels, with O. cf. tetricus occupying a mid-trophic position, slightly below the teleosts

and loliginid squids. The δ15N signatures of O. cf. tetricus and loliginid squid parallel

Lozano-Montes et al. (2011) who found that the trophic level for cephalopods (octopus

and squid) on the west coast of Australia range from 3.1 to 3.6. The enrichment of δ15N

by more than 3.4‰ between H. laevigata and O. cf. tetricus indicates that octopuses are

assimilating δ15N from other dietary sources in addition to abalone, and they are feeding

on consumers at trophic levels 2 and 3. This enrichment likely reflects the consumption

of other prey taxon such as cephalopods and crustaceans, as indicated by complimentary

crop content analyses. It is likely that crustaceans and echinoderms are at the

intermediate trophic level, although echinoderms are present in low numbers around the

Abitats (M. Wall, Ocean Grown Abalone, 2017 pers. comm.). In a study by Lepoint et

al. (2000), these taxon occupied a mid trophic position between fishes and non-

cephalopod molluscs, with a similar position expected for the Flinders Bay sea ranch

ecosystem. The relatively low enrichment of δ15N between abalone and primary

producers is consistent with Guest et al. (2008), with herbivores exhibiting significantly

lower fractionation compared with high trophic level carnivores (Vander Zanden and

Rasmussen 2001; McCutchan et al. 2003).

Evidence from δ13C isotopes of consumers and the crop contents of O. cf.

tetricus suggests that carbon in the Flinders Bay system is derived largely from benthic

sources. Octopus cf. tetricus had similar δ13C values among size classes reflecting the

similarity of diet, as confirmed by gut content analysis. Based on a δ13C fractionation of

~1‰ between trophic levels (Layman et al. 2012; Cresson et al. 2014b), the enrichment

of ~4‰ between H. laevigata and O. cf. tetricus suggests that an intermediary food web

constituent, such as crustaceans, is likely to be an important component of the diet.

Bayesian mixing models such as Stable Isotope Analysis in R could be utilised to

identify the relative contributions of the various nutrients assimilated by O. cf. tetricus

(Parnell et al. 2010, see Chapter 4.2).

69

Haliotis laevigata seeds were slightly more enriched than abalone seeded > 1

year earlier, which may be explained by differences in food availability. Under culture

conditions, seeds are reared on diatom plates for ~9 months and are then weaned onto

commercially formulated feeds (Melville-Smith et al. 2013). After being seeded onto

the Abitats, abalone switch to a natural macroalgal diet. In a feeding study of Disk

Abalone, Haliotis discus discus, off northern Japan a decrease in δ13C was associated

with a shift in diet from diatoms to juvenile Phaeophyceae and/or small Rhodophyta,

while an increase in δ13C represented a shift in diet to adult Phaeophyceae (Won et al.

2010).

The δ13C values of H. laevigata were intermediate between the Rhodophyta and

Phaeophyceae values, suggesting a mixed diet. While the δ13C signatures of Sargassum

and Gracilliara are similar to that of H. laevigata, the slight depletion of δ13C in

H. laevigata suggests that other macroalgal species are also important components of

the diet. At least 57 species of macroalgae were identified on the Abitats over a six-

month period from July 2015 to February 2016 (Melville-Smith 2016). Diversity was

highly seasonal, with species richness and abundance notably higher in the winter

months (Melville-Smith 2016). Therefore, the diet of H. laevigata is likely to change

seasonally, which may have important implications for the interpretation of δ13C

sources if the longer-term assimilation of nutrients is not considered (see Chapter 4.2).

The δ13C values of O. cf. tetricus were similar to the teleosts King Wrasse, C.

auricularis, O. lineolatus, and Bluespotted Goatfish, U. vlamingii, indicating an overlap

in the primary carbon sources utilised among these species (Davenport and Bax 2002).

These teleost species are benthic feeders and utilise similar food resources to O. cf.

tetricus, explaining the similar δ13C signatures among consumers (Currie and Sorokin

2010; Lek et al. 2011). For example, a study from south-western Australia found that O.

lineolatus fed mainly on large benthic crustaceans and gastropods, while C. auricularis

fed mainly on small crustaceans and gastropods (Lek et al. 2011).

70

Chapter 4. General conclusions and recommendations 4.1. Conclusions

The study on the distribution and abundance of octopus, and the mortality of Greenlip

Abalone, Haliotis laevigata, based on data collected by commercial divers (Chapter 2),

demonstrated that octopuses are broadly distributed across the Ocean Grown Abalone

(OGA) sea ranch in Flinders Bay, Western Australia. The presence of octopus had a

significant impact on the survival of abalone, with an estimated 78% more empty shells

recorded when octopus are present than when absent, adjusting for location (“Line”)

and Season. The abundance of octopus was typically greater on Lines that comprised

fewer abalone habitats (“Abitats”), which may be related to detection bias and the time

divers spend within a Bay searching for octopus and counting shells. Bays with greater

numbers of Abitats provide greater areas of shelter for octopus, potentially making their

detection more difficult, particularly because of their cryptic adaptations. While the

number of empty shells was higher in some locations than others, this may be related to

factors other than octopus predation alone. Environmental influences, such as storm

events and currents likely play an important role in the dispersal of shells, and the

number of Abitats within a Line may result in partially buried or dispersed shells being

overlooked due to the time constraints imposed on divers. The abundance of other

abalone predators, such as crayfish and crabs, will also contribute to the number of

empty shells in an area.

The boreholes found in the sample of H. laevigata shells collected by OGA

divers were consistent with those made by other octopus species (e.g., Octopus

bimaculatis, Pilson and Taylor 1981; Ambrose 1983; and Octopus vulgaris, Ambrose

1983; Kojima 1992), providing evidence of drilling by Octopus cf. tetricus and the

possession of cephalotoxins in this species. Boreholes were found in shells across the

entire length range (33 to 110 mm), which indicates that octopus prey on all sizes of

abalone. An exact multinomial test showed that the frequency of boreholes did not

differ significantly among the four regions, which may be related to small sample sizes

(total of 20 shells with holes). However, 75% of holes were concentrated over the

71

adductor muscle, pericardial sac and the heart, organs that are most sensitive to the

injection of toxin. The areal distribution of boreholes may be related to a learnt

behaviour, whereby octopuses learn the most efficient location to drill over time, with

the adductor muscle potentially succumbing to the cephalotoxins of octopus more

quickly than other locations (Kojima 1992).

Negative binomial generalised linear models demonstrated that the presence of

octopus results in a 78% increase in the number of empty abalone shells collected per

Abitat per day after adjusting for Line and Season. There was also a significant

association between the number of empty shells and Line, as well as between the

number of shells and Season. In contrast, results from time series models used to assess

the relationship between the presence of octopus and abalone survival (counts every six

months) were not significant, which is counterintuitive to what was expected. This

finding may be an artefact of the random sampling design for estimating abalone

survival (i.e., counting abalone on random Abitats within a Bay at each time point,

rather than taking abalone counts from the same Abitats each time), and uncertainty in

the numbers of live abalone seeded onto each Abitat. This highlights the importance of

quantifying the number of juvenile abalone being seeded onto unique Abitats, and

undertaking repeated measures on the same Abitats over time in order to reduce

variability among survival counts and to allow abalone survival to be measured with

greater precision.

The study on the diet and nutrient assimilation (δ13C and δ15N) of O. cf. tetricus

(Chapter 3) demonstrates that O. cf. tetricus, like other coastal octopuses, feeds

predominantly on molluscs and crustaceans (Nixon 1987; Grubert et al. 1999; Smith

2003). Pronounced differences in diet were found between the contents of the crops and

stomachs, with the crops containing a much high proportion of identifiable items than

the stomachs, which contained large proportions of highly digested, unidentifiable

material. No differences in dietary composition were identified between the weight

classes of octopus, paralleling the results of the stable isotope analyses. However, few

72

octopus > 1,000 g were examined and no octopus > 1,600 g. As O. cf. tetricus grows to

a maximum size of ~4 kg (Hart et al. 2016), ontogenetic differences may be more

pronounced if larger individuals (i.e., > 2,000 g) are sampled and included in dietary

analyses.

The relatively low diversity of prey items in the crops and stomachs of O. cf.

tetricus is likely due to a combination of factors. Firstly, prey items are highly

masticated as part of the ingestion process making identification to species level

extremely difficult. A reference collection of benthic epifauna, particularly molluscs

and crustaceans, collected from the local environment, may enhance species

identification in the future. Secondly, abalone is available in excess and concentrated on

the Abitats within the sea ranching lease. The predation on abalone and organisms

attracted to the structures, such as brachyurans, provides an opportunity for in-situ

foraging by octopus and suits a time minimising foraging strategy. Finally, the routine

cleaning of Abitats to remove algae and sessile invertebrates growing on their surface is

likely to reduce species diversity, and ultimately, the range of prey available to O. cf.

tetricus.

The most important dietary component in the crops of octopus was abalone and

to a lesser extent crustaceans. These results clearly show that O. cf. tetricus is exploiting

the high abundance of abalone, but also feeds on crustaceans whenever the opportunity

arises, likely due to their low handling time from capture to ingestion. In addition, the

high amino acid, low lipid macronutrient combination of abalone and crustaceans is

likely to adequately satisfy the dietary needs of octopus, whilst supporting rapid growth.

The examination of δ15N and δ13C stable isotopes highlights the importance of

incorporating information obtained from gastric contents for explaining trends among

consumers and assessing trophic interactions. Observations from gastric tract analyses

suggest that O. cf. tetricus is an important predator of H. laevigata. However, this trend

was less obvious from stable isotope analyses due to the relative enrichment of δ13C

(~4‰) between the two species. These findings are likely due to the contribution of

73

other prey items to the diet of octopus, such as crustaceans and other cephalopods (as

identified by gastric tract analysis). The δ15N values of O. cf. tetricus show that the

species occupies a mid-trophic level, slightly below the teleosts and loliginid squid, and

is more than one trophic level above H. laevigata.

4.2. Future research and monitoring 4.2.1. Monitoring abalone survival

Monitoring abalone after seeding would be valuable to the sea ranching operation.

Remote underwater video systems could be utilised to understand the behavioural

responses of both abalone seeds and predators, including octopus, immediately after the

deployment of the abalone seeds in their release units (Fig. 2.3a). A subsequent census

of survival by divers shortly after seeding (within two weeks) could allow initial

predation rates to be quantified.

The information on abalone survival would be greatly enhanced by quantifying

the number of abalone seeded onto the Abitats for future deployments and counting

abalone on the same uniquely identified Abitats within each Bay to estimate survival on

each Abitat. This will allow patterns of variability in abalone survival to be identified

across the ranch and survival estimates to be measured with greater precision.

4.2.2. Dietary studies

Future dietary studies of O. cf. tetricus in this area would benefit from collecting larger

numbers of octopus, particularly those > 1,000 g and placing them on ice immediately

following their capture to slow their digestive processes (Pierce et al. 1994; Boyle and

Rodhouse 2005; Boucher-Rodoni and Mangold 2009). This would potentially reduce

the proportion of “unidentified” food items in the stomachs.

The collection of additional octopus over a longer period and larger weight

range may identify subtle differences in diet between size classes, particularly since the

diversity of diets among octopuses appears to be dependent on the timeframe over

which information is collected (Nixon 1987; Mather and O’Dor 1991). Furthermore, the

74

collection of samples throughout the year may enable seasonal variation in the diet to be

identified. Comparison with diets of octopus from areas outside the lease would allow

the relative contribution of abalone to the diet to be evaluated under natural conditions

to help evaluate their selection of abalone on the sea ranching release.

4.2.3. Stable isotopes

Further sampling of source and consumer pools at temporal scales that reflect the

assimilation and turnover rates of nutrients should be carried out to allow seasonal

trends in nitrogen and carbon assimilation to be investigated (Post 2002; Layman et al.

2012). The collection of a broader range of prey types for stable isotope analysis,

including crustaceans, would also be important to provide a complete depiction of the

Flinders Bay food web, and the nutrient assimilation by O. cf. tetricus. Bayesian mixing

models such as Stable Isotope Analysis in R, could be incorporated to identify the

relative contributions of the various nutrients assimilated by octopus (Parnell et al.

2010). These mixing models provide powerful insights into the relative contribution of

sources to consumers and are most effective when combined with gastric tract analyses,

allowing the proportional contributions estimated from isotopes to be interpreted

(Layman et al. 2012; Phillips et al. 2014). In addition to analyzing other consumers,

assessing the δ15N and δ13C isotopes of detritus, particulate organic matter (POM) and

epiphyte signatures may provide further insight into nutrient sources utilised by abalone

(see e.g., Guest et al. 2008).

4.3. Implications

The results from this Thesis highlight the importance of O. cf. tetricus as a predator of

H. laevigata on the sea ranching lease in Flinders Bay, south-western Australia. Firstly,

the identification of distinctive boreholes in the shells of abalone provides an insight

into the feeding behaviour of octopuses. If octopuses are not able to remove abalone

from the Abitats by direct force, they resort to drilling, which accounted for 18% of the

sub-sample of shells investigated. This percentage likely represents the minimum

predatory impact on H. laevigata, but confirms that octopuses are a major source of

75

mortality. It is possible that smaller abalone, or those in a “feeding” position (Shepherd

1973), can be pulled directly from the Abitats. When H. laevigata feed, they elevate the

forepart of their foot and their shell to catch drifting algae (Shepherd 1973). Whilst in

this position their stronghold on the reef substratum is reduced, rendering them

vulnerable to predation.

Secondly, negative binomial models provide further evidence of the deleterious

impacts of octopus to the survival of cultured abalone. The proportion of empty abalone

shells associated with the presence of octopus was 44% of all empty shells collected.

This figure may be underestimated, as it is likely that shell recovery is lower than the

total mortality of abalone. Finally, crop contents suggest that H. laevigata forms a major

component of the diet of O. cf. tetricus, highlighting the ability of these opportunistic

and cunning predators to adapt to novel feeding opportunities.

No larger octopus (> 1,600 g) were collected during the four months in which

octopus were collected by commercial divers. However, the absence of these larger

individuals could be related to the removal of octopuses from the ranch, limiting the

numbers of larger individuals. Alternatively, the absence of larger individuals may be

associated with an offshore migration of mature individuals in search of suitable reefs

for the brooding of eggs (Leporati et al. 2015). Consequently, the size classes examined

in this study may be representative of the local population of O. cf. tetricus in this

ecosystem. Further sampling of octopus within the lease and at locations away from the

sea ranch is needed to test this hypothesis.

While this study provides an assessment of the dietary composition of O. cf.

tetricus in the Flinders Bay sea ranch ecosystem and the importance of H. laevigata to

the diet in this system, these findings cannot be extended to populations inhabiting

natural environments. The high stocking densities of abalone on the Abitats is likely to

drive the foraging behaviour of O. cf. tetricus and subsequently the composition of its

diet. The Abitats also provide potential shelter for octopus, and the combination of

abalone abundance and shelter may concentrate octopus within the sea ranching lease.

76

Sampling of O. cf. tetricus from locations beyond the sea ranch would provide

information on what constitutes “natural” densities and diet, and to what extent the diet

and nutrient assimilation of O. cf. tetricus on the sea ranch differ from other “natural”

octopus populations.

The knowledge gained from these studies provides a greater understanding of

the significance of O. cf. tetricus as predators of abalone, and the trophic positioning of

this important mesopredator within the food web of a rapidly expanding sea ranching

operation. This information can be extended to future sites, including the new lease

being established in Esperance to improve operational procedures. It also highlights the

value of collecting information outside the sea ranch to understand the population size

and structure of O. cf. tetricus, their densities, and diets under natural conditions. This

will enable a greater understanding of how octopuses have responded to the high

densities of abalone on the Abitats and the provision of additional shelter within the

OGA lease.

77

References

Ambrose, R. F. (1986). Effects of octopus predation on motile invertebrates in a rocky subtidal community. Marine Ecology. Progress Series 30, 261–273.

Ambrose, R. F. (1984). Food preferences, prey availability, and the diet of Octopus bimaculatus Verrill. Journal of Experimental Marine Biology and Ecology 77, 29–44. doi:http://dx.doi.org/10.1016/0022-0981(84)90049-2

Ambrose, R. F., and Nelson, B. V (1983). Predation by Octopus vulgaris in the Mediterranean. Marine Ecology 4, 251–261. doi:10.1111/j.1439-0485.1983.tb00299.x

Amor, M. D., Norman, M. D., Cameron, H. E., and Strugnell, J. M. (2014). Allopatric speciation within a cryptic species complex of Australasian octopuses. PLoS ONE 9. doi:10.1371/journal.pone.0098982

Anderson, F. E. (2000). Phylogenetic relationships among loliginid squids (Cephalopoda: Myopsida) based on analyses of multiple data sets. Zoological Journal of the Linnean Society 130, 603–633. doi:10.1111/j.1096-3642.2000.tb02203.x

Anderson, M. J., Gorley, R. N., and Clarke, K. R. (2008a). ‘PERMANOVA+ for PRIMER: Guide to software and statistical methods’. (PRIMER-E Ltd: Plymouth UK.)

Anderson, R. C., Wood, J. B., and Mather, J. A. (2008b). Octopus vulgaris in the Caribbean is a specializing generalist. Marine Ecology Progress Series 371, 199–202. doi:10.3354/meps07649

Anderson, T. J. (1997). Habitat selection and shelter use by Octopus tetricus. Marine Ecology Progress Series 150, 137–148. doi:10.3354/meps150137

Anderson, T. W. (2001). Predator responses, prey refuges, and density-dependent mortality of a marine fish. Ecology 82, 245–257. doi:10.1890/0012-9658(2001)082[0245:PRPRAD]2.0.CO;2

André, J., Haddon, M., and Pecl, G. T. (2010). Modelling climate-change-induced nonlinear thresholds in cephalopod population dynamics. Global Change Biology 16, 2866–2875. doi:10.1111/j.1365-2486.2010.02223.x

Anseeuw, D., Guelinckx, J., and Snoeks, J. (2010). Differences in stable isotope composition within and among zooplanktivorous Utaka cichlid populations from Lake Malawi. African Journal of Ecology 48, 378–385. doi:10.1111/j.1365-2028.2009.01124.x

Arnold, J. M., and Arnold, K. O. (1969). Some aspects of hole-boring predation by Octopus vulgaris. Integrative and Comparative Biology 9, 991–996. doi:10.1093/icb/9.3.991

Aronson, R. B. (1991). Ecology, paleobiology and evolutionary constraint in the octopus. Bulletin of Marine Science 49, 245–255.

Baetschmann, G., and Winkelmann, R. (2013). Modeling zero-inflated count data when exposure varies: With an application to tumor counts. Biometrical Journal 55, 679–686. doi:10.1002/bimj.201200021

Becker, P., Barringer, C., and Marelli, D. C. (2008). Thirty years of sea ranching Manila Clams (Venerupis philippinarum): Successful techniques and lessons learned. Reviews in Fisheries Science 16, 44–50. doi:10.1080/10641260701790259

Bell, J. D. (1999). Transfer of technology on marine ranching to small island states. In ‘Marine ranching: global perspectives with emphasis on the Japanese experience. Status and prospects on the development and improvement of coastal fishing ground.’ pp. 53–65. (FAO Inland Water Resources and Aquaculture Service, Fishery Resources Division.: Rome.)

Bell, J. D., Leber, K. M., Blankenship, H. L., Loneragan, N. R., and Masuda, R. (2008).

78

A new era for restocking, stock enhancement and sea ranching of coastal fisheries resources. Reviews in Fisheries Science 16, 1–9. doi:10.1080/10641260701776951

Bohnsack, J. A. (1989). Are high densities of fishes at artificial reefs the result of habitat limitation or behavioral preference? Bulletin of Marine Science 44, 631–645.

Botsford, L. W. (1997). The management of fisheries and marine ecosystems. Science 277, 509–515. doi:10.1126/science.277.5325.509

Boucher-Rodoni, R., and Mangold, K. (2009). Experimental study of digestion in Octopus vulgaris (Cephalopoda: Octopoda). Journal of Zoology 183, 505–515. doi:10.1111/j.1469-7998.1977.tb04202.x

Boyle, P. R., and Rodhouse, P. G. K. (2005). ‘Cephalopods: Ecology and fisheries’. (Blackwell Science Ltd: Ames, Iowa.) doi:10.1002/9780470995310

Braley, M., Goldsworthy, S. D., Page, B., Steer, M., and Austin, J. J. (2010). Assessing morphological and DNA-based diet analysis techniques in a generalist predator, the arrow squid Nototodarus gouldi. Molecular Ecology Resources 10, 466–474. doi:10.1111/j.1755-0998.2009.02767.x

Breiby, A., and Jobling, M. (1985). Predatory role of the Flying Squid (Todarodes sagittatus) in North Norwegian Waters. Northwest Alantic Fisheries Organisation. Science 9, 125–132.

Briones-Fourzàn, P., and Lozano-Álvarez, E. (2001). Effects of artificial shelters (Casitas) on the abundance and biomass of juvenile spiny lobsters Panulirus argus in a habitat-limited tropical reef lagoon. Marine Ecology Progress Series 221, 221–232. doi:10.3354/meps221221

Broadley, A. D., Tweedley, J. R., and Loneragan, N. R. (2017). Estimating biological parameters for penaeid restocking in an Australian temperate estuary. Fisheries Research 186, 488–501. doi:10.1016/j.fishres.2016.09.007

Brown, C., and Day, R. L. (2002). The future of stock enhancements: Lessons for hatchery practice from conservation biology. Fish and Fisheries 3, 79–94. doi:10.1046/j.1467-2979.2002.00077.x

Caputi, N., Feng, M., Pearce, A., Benthuysen, J., Denham, A., Hetzel, Y., Matear, R., Jackson, G. D., Molony, B., Joll, L. M., and A, C. (2015). Management implications of climate change effect on fisheries in Western Australia, Part 2: Case studies. FRDC Project No. 2010/535. Fisheries Research Report No. 261. Department of Fisheries, Western Australia. Perth.

Carter, J. W., Jessee, W. N., Foster, M. S., and Carpenter, A. L. (1985). Management of artificial reefs designed to support natural communities. Bulletin of Marine Science 37, 114–128.

Catchpole, E. a, Freeman, S. N., Morgan, B. J., and Nash, W. J. (2001). Abalone I: analyzing mark-recapture-recovery data incorporating growth and delayed recovery. Biometrics 57, 469–477. doi:10.1111/j.0006-341X.2001.00469.x

Cherel, Y., Gasco, N., and Duhamel, G. (2011). Top predators and stable isotopes document the cephalopod fauna and its trophic relationships in Kerguelen waters. The Kerguelen Plateau: marine ecosystem and fisheries, 99–108. Available at: http://www.cebc.cnrs.fr/publipdf/2011/CSFI_2011.pdf

Clarke, K. R., Gorley, R. N., Somerfield, P. J., and Warwick, R. M. (2014a). ‘Change in marine communities: an approach to statistical analysis and interpretation’ 3rd ed. (PRIMER-E Ltd: Plymouth UK.)

Clarke, K. R., and Green, R. H. (1988). Statistical design and analysis for a ‘biological effects’ study. Marine Ecology 46, 213–226. doi:10.3354/meps046213

Clarke, K. R., Tweedley, J. R., and Valesini, F. J. (2014b). Simple shade plots aid better long-term choices of data pre-treatment in multivariate assemblage studies.

79

Journal of the Marine Biological Association of the United Kingdom 94, 1–16. doi:10.1017/S0025315413001227

Coll, M., Navarro, J., Olson, R. J., and Christensen, V. (2013). Assessing the trophic position and ecological role of squids in marine ecosystems by means of food-web models. Deep-Sea Research Part II: Topical Studies in Oceanography 95, 21–36. doi:10.1016/j.dsr2.2012.08.020

Como, S., van der Velde, G., and Magni, P. (2017). Temporal variation in the trophic levels of secondary consumers in a Mediterranean coastal lagoon (Cabras Lagoon, Italy). Estuaries and Coasts, 1–15. doi:10.1007/s12237-017-0265-7

Cortez, T., Castro, B. G., and Guerra, Á. (1995). Feeding dynamics of Octopus mimus (Mollusca: Cephalopoda) in northern Chile waters. Marine Biology 123, 497–503. doi:10.1007/BF00349228

Coulson, P. G., Platell, M. E., Clarke, K. R., and Potter, I. C. (2015). Dietary variations within a family of ambush predators (Platycephalidae) occupying different habitats and environments in the same geographical region. Journal of Fish Biology 86, 1046–1077. doi:10.1111/jfb.12612

Cresson, P., Ruitton, S., and Harmelin-Vivien, M. (2014a). Artificial reefs do increase secondary biomass production: Mechanisms evidenced by stable isotopes. Marine Ecology Progress Series 509, 15–26. doi:10.3354/meps10866

Cresson, P., Ruitton, S., Ourgaud, M., and Harmelin-Vivien, M. (2014b). Contrasting perception of fish trophic level from stomach content and stable isotope analyses: A Mediterranean artificial reef experience. Journal of Experimental Marine Biology and Ecology 452, 54–62. doi:10.1016/j.jembe.2013.11.014

Currie, D. R., and Sorokin, S. J. (2010). A preliminary evaluation of the distribution and trophodynamics of demersal fish from Spencer Gulf. doi:10.1080/3721426.2010.10887143

Daniels, H. V, and Watanabe, W. O. (2010). ‘Practical flatfish culture and stock enhancement’. (John Wiley & Sons Ltd: Iowa, U.S.A.)

Daume, S., Huchette, S., Ryan, S., and Day, R. W. (2004). Nursery culture of Haliotis rubra: The effect of cultured algae and larval density on settlement and juvenile production. Aquaculture 236, 221–239. doi:10.1016/j.aquaculture.2003.09.035

Daume, S., and Ryan, S. (2004). Nursery culture of the abalone Haliotis laevigata : larval settlement and juvenile production using cultured algae or formulated feed. Journal of Shellfish Research 23, 1019–1026.

Davenport, S. R., and Bax, N. J. (2002). A trophic study of a marine ecosystem off southeastern Australia using stable isotopes of carbon and nitrogen. Canadian Journal of Fisheries and Aquatic Sciences 59, 514–530. doi:10.1139/f02-031

Dixon, C. D., Day, R. W., Huchette, S. M. H., and Shepherd, S. A. (2006). Successful seeding of hatchery-produced juvenile greenlip abalone to restore wild stocks. Fisheries Research 78, 179–185. doi:10.1016/j.fishres.2005.11.023

Doubleday, Z. A., Prowse, T. A. A., Arkhipkin, A., Pierce, G. J., Semmens, J., Steer, M., Leporati, S. C., Lourenço, S., Quetglas, A., Sauer, W., and Gillanders, B. M. (2016). Global proliferation of cephalopods. Current Biology 26, R406–R407. doi:10.1016/j.cub.2016.04.002

Duame, S. (2006). Growth and survival of juvenile greenlip abalone (Haliotis laevigata) feeding on germlings of the macroalgae ulva sp . Journal of Shellfish Research, 1–18. doi:10.2983/0730-8000(2006)25[239:gasojg]2.0.co;2

Eddy, T. D., Lotze, H. K., Fulton, E. A., Coll, M., Ainsworth, C. H., de Araújo, J. N., Bulman, C. M., Bundy, A., Christensen, V., Field, J. C., Gribble, N. A., Hasan, M., Mackinson, S., and Townsend, H. (2017). Ecosystem effects of invertebrate fisheries. Fish and Fisheries 18, 40–53. doi:10.1111/faf.12165

80

Finn, J. K., Tregenza, T., and Norman, M. D. (2009). Defensive tool use in a coconut-carrying octopus. Current Biology 19. doi:10.1016/j.cub.2009.10.052

Fox, J. (2003). Effect displays in R for generalised linear models. Journal of Statistical Software 8, 1–27.

French, B., Platell, M. E., Clarke, K. R., and Potter, I. C. (2012). Ranking of length-class, seasonal and regional effects on dietary compositions of the co-occurring Pagrus auratus (Sparidae) and Pseudocaranx georgianus (Carangidae). Estuarine, Coastal and Shelf Science 115, 309–325. doi:10.1016/j.ecss.2012.09.004

García García, B., and Aguado Giménez, F. (2002). Influence of diet on ongrowing and nutrient utilization in the common octopus (Octopus vulgaris). Aquaculture 211, 171–182. doi:10.1016/S0044-8486(01)00788-8

García García, B. G., and Valverde, C. (2006). Optimal proportions of crabs and fish in diet for common octopus (Octopus vulgaris) ongrowing. Aquaculture 253, 502–511. doi:10.1016/j.aquaculture.2005.04.055

Gardner, W., Mulvey, E. P., and Shaw, E. C. (1995). Regression analyses of counts and rates: Poisson, overdispersed poisson, and negative binomial models. Psychological Bulletin 118, 392–404. doi:10.1037/0033-2909.118.3.392

Garrett, G., and Wickham, H. (2011). Dates and times made easy with lubridate. Journal of Statistical Software 40, 1–25. Available at: http://www.jstatsoft.org/v40/i03/.

Gasalla, M. A., Rodrigues, A. R., and Postuma, F. A. (2010). The trophic role of the squid Loligo plei as a keystone species in the South Brazil Bight ecosystem. ICES Journal of Marine Science 67, 1413–1424. doi:10.1093/icesjms/fsq106

Gende, S. M., and Sigler, M. F. (2006). Persistence of forage fish ‘hot spots’ and its association with foraging Steller sea lions (Eumetopias jubatus) in southeast Alaska. Deep-Sea Research Part II: Topical Studies in Oceanography 53, 432–441. doi:10.1016/j.dsr2.2006.01.005

Ghiretti, F. (1959). Cephalotoxin: the crab-paralysing agent of the posterior salivary glands of cephalopods. Nature 183, 1192–1193.

Goodsell, P. J., Underwood, A. J., Chapman, M. G., and Heasman, M. P. (2006). Seeding small numbers of cultured black-lip abalone (Haliotis rubra Leach) to match natural densities of wild populations. Marine and Freshwater Research 57, 747–756. doi:10.1071/MF06039

Grimes, C. B. (1998). Marine stock enhancement: sound management or techno-arrogance? Fisheries 23, 18–23. doi:10.1577/1548-8446(1998)023<0018

Grisley, M. S., and Boyle, P. R. (1987). Bioassay and proteolytic activity of digestive enzymes from octopus saliva. Comparative Biochemistry and Physiology -- Part B: Biochemistry and 88, 1117–1123. doi:10.1016/0305-0491(87)90014-9

Grisley, M. S., and Boyle, P. R. (1990). Chitinase, a new enzyme in octopus saliva. Comparative Biochemistry and Physiology -- Part B: Biochemistry and 95, 311–316. doi:10.1016/0305-0491(90)90081-4

Grubert, M. A., Wadley, V. A., and White, R. W. G. (1999). Diet and feeding strategy of Octopus maorum in southeast Tasmania. Bulletin of Marine Science 65, 441–451.

Guest, M. A., Nichols, P. D., Frusher, S. D., and Hirst, A. J. (2008). Evidence of abalone (Haliotis rubra) diet from combined fatty acid and stable isotope analyses. Marine Biology 153, 579–588. doi:10.1007/s00227-007-0831-9

Hamasaki, K., and Kitada, S. (2008). The enhancement of abalone stocks: lessons from Japanese case studies. Fish and Fisheries 9, 243–260. doi:10.1111/j.1467-2979.2008.00280.x

Hanlon, R. T., and Messenger, J. B. (1996). ‘Cephalopod behaviour’. (Cambridge

81

University Press: Cambridge.) Hart, A. M., Fabris, F., Strain, L. W. S., Davidson, M., and Brown, J. (2013a). Stock

enhancement in Greenlip Abalone Part II: Population and ecological effects. Reviews in Fisheries Science 21, 310–320. doi:10.1080/10641262.2013.812505

Hart, A. M., Leporati, S. C., Marriott, R. J., and Murphy, D. (2016). Innovative development of the Octopus cf. tetricus fishery in Western Australia. FRDC Project No 2010/200. Fisheries Research Report No. 270. Perth.

Hart, A. M., Murphy, D., Leporati, S. C., and Joll, L. M. (2013b). Octopus fishery status report. In ‘Status Reports of the Fisheries and Aquatic Resources of Western Australia 2011/12: The State of the Fisheries’. (Eds W. J. Fletcher and K. Santoro.) pp. 111–115. (Department of Fisheries: Perth.)

Hart, A. M., Strain, L. W. S., Fabris, F., Brown, J., and Davidson, M. (2013c). Stock enhancement in Greenlip Abalone Part I: Long-term growth and survival. Reviews in Fisheries Science 21, 299–309. doi:10.1080/10641262.2013.812503

Hewitt, J. E., Thrush, S. F., and Dayton, P. D. (2008). Habitat variation, species diversity and ecological functioning in a marine system. Journal of Experimental Marine Biology and Ecology 366, 116–122. doi:10.1016/j.jembe.2008.07.016

Hines, A. H., Johnson, E. G., Young, A. C., Aguilar, R., Kramer, M. A., Goodison, M., Zmora, O., and Zohar, Y. (2008). Release strategies for estuarine species with complex migratory life cycles: Stock enhancement of Chesapeake blue crabs (Callinectes sapidus). Reviews in Fisheries Science 16, 175–185. doi:10.1080/10641260701678090

Hines, A. H., and Pearse, J. S. (1982). Abalones, shells, and sea otters: dynamics of prey populations in central California. Ecology 63, 1547–1560. doi:10.2307/1938879

Hinesa, A. H., Whitlatch, R. B., Thrush, S. F., Hewitt, J. E., Cummings, V. J., Dayton, P. K., and Legendre, P. (1997). Nonlinear foraging response of a large marine predator to benthic prey: eagle ray pits and bivalves in a New Zealand sandflat. Journal of Experimental Marine Biology and Ecology 216, 191–210. doi:10.1016/S0022-0981(97)00096-8

Jackson, G. D., Bustamante, P., Cherel, Y., Fulton, E. A., Grist, E. P. M., Jackson, C. H., Nichols, P. D., Pethybridge, H., Phillips, K., Ward, R. D., and Xavier, J. C. (2007). Applying new tools to cephalopod trophic dynamics and ecology: Perspectives from the Southern Ocean Cephalopod Workshop, February 2-3, 2006. Reviews in Fish Biology and Fisheries 17, 79–99. doi:10.1007/s11160-007-9055-9

James, D. S. (2005). Environmental site assessment for abalone ranching on artificial reef. PhD Thesis. Available at: http://dro.deakin.edu.au/view/DU:30023267 [accessed 30 July 2017]

James, D. S., Day, R. W., and Shepherd, S. A. (2007). Experimental abalone ranching on artificial reef in Port Phillip Bay, Victoria. Journal of Shellfish Research 26, 687–695. doi:10.2983/0730-8000(2007)26[687:EAROAR]2.0.CO;2

Joll, L. M. (1977). Growth and Food Intake of Octopus tetricus (Mollusca: Cephalopoda) in Aquaria. Marine and Freshwater Research 28, 45–56. doi:10.1071/MF9770045

Joll, L. M. (1976). Mating, egg-laying and hatching of Octopus tetricus (Mollusca: Cephalopoda) in the laboratory. Marine Biology 36, 327–333. doi:10.1007/BF00389194

Kojima, H. (1992). Octopus predation on the abalone Haliotis discus discus. In ‘First International Symposium on Abalone Biology, Fisheries, and Culture. La Paz, Mexico 21-25 November 1989’. (Eds S. A. Shepherd, M. J. Tegner, and S. A. Guzman del Proo.) pp. 21–29. (Blackwell Publishing Ltd: Oxford.)

Layman, C. A., Araujo, M. S., Boucek, R., Hammerschlag-Peyer, C. M., Harrison, E.,

82

Jud, Z. R., Matich, P., Rosenblatt, A. E., Vaudo, J. J., Yeager, L. A., Post, D. M., and Bearhop, S. (2012). Applying stable isotopes to examine food-web structure: An overview of analytical tools. Biological Reviews 87, 545–562. doi:10.1111/j.1469-185X.2011.00208.x

Leber, K. M., Kitada, S., Blankenship, H. L., and Svasand, T. (2004). ‘Stock enhancement and sea ranching: developments, pitfalls and opportunities’ 2nd ed. (Blackwell Publishing Ltd: Oxford, UK.)

Lee, P. G. (1995). Nutrition of cephalopods: Fueling the system. Marine and Freshwater Behaviour and Physiology 25, 35–51. doi:10.1080/10236249409378906

Leitão, F. (2013). Artificial reefs: from ecological processes to fishing enhancement tools. Brazilian Journal of Oceanography 61, 77–81. doi:10.1590/S1679-87592013000100009

Lek, E., Fairclough, D. V., Platell, M. E., Clarke, K. R., Tweedley, J. R., and Potter, I. C. (2011). To what extent are the dietary compositions of three abundant, co-occurring labrid species different and related to latitude, habitat, body size and season? Journal of Fish Biology 78, 1913–1943. doi:10.1111/j.1095-8649.2011.02961.x

Lepoint, G., Nyssen, F., Gobert, S., Dauby, P., and Bouquegneau, J.-M. (2000). Relative impact of a seagrass bed and its adjacent epilithic algal community in consumer diets. Marine Biology 136, 513–518. doi:10.1007/s002270050711

Leporati, S. C., Hart, A. M., Larsen, R., Franken, L. E., and De Graaf, M. (2015). Octopus life history relative to age, in a multi-geared developmental fishery. Fisheries Research 165, 28–41. doi:10.1016/j.fishres.2014.12.017

Linke, T. (2011). Trophic interactions among abundant members of the fish fauna in a permanently-open and a seasonally-open estuary in South-western Australia. PhD Thesis. Murdoch University. Available at: http://researchrepository.murdoch.edu.au/id/eprint/12662/2/02Whole.pdf

Liu, H., Davidson, R. A., Rosowsky, D. V., and Stedinger, J. R. (2005). Negative binomial regression of electric power outages in hurricanes. Journal of Infrastructure Systems 11, 258–267. doi:10.1061/(ASCE)1076-0342(2005)11:4(258)

Livingston, R. J. (2003). ‘Trophic organisation in coastal systems’. (CRC Press: Boca Raton.)

Loneragan, N. R., Jenkins, G. I., and Taylor, M. D. (2013). Marine stock enhancement, restocking, and sea ranching in Australia: Future directions and a synthesis of two decades of research and development. Reviews in Fisheries Science 21, 222–236. doi:10.1080/10641262.2013.796810

Lozano-Montes, H. M., Loneragan, N. R., Babcock, R. C., and Jackson, K. (2011). Using trophic flows and ecosystem structure to model the effects of fishing in the Jurien Bay Marine Park, temperate Western Australia. Marine and Freshwater Research 62, 421–431. doi:10.1071/MF09154

Mangold, K. (1983). Eledone moschata. In ‘Cephalopod life cycles. Species Accounts’. (Ed P. R. Boyle.) pp. 387–400. (Academic Press: London, Orlando.)

Mather, J. A. (1991). Foraging, feeding and prey remains in middens of juvenile Octopus vulgaris (Mollusca: Cephalopoda). Journal of Zoology 224, 27–39. doi:10.1111/j.1469-7998.1991.tb04786.x

Mather, J. A. (1994). ‘Home’ choice and modification by juvenile Octopus vulgaris (Mollusca: Cephalopoda): Specialized intelligence and tool use? Journal of Zoology 233, 359–368. doi:10.1111/j.1469-7998.1994.tb05270.x

Mather, J. A., and Nixon, M. (1995). Octopus vulgaris (Cephalopoda) drills the chelae

83

of crabs in Bermuda. Journal of Molluscan Studies 61, 405–406. doi:10.1093/mollus/61.3.405

Mather, J. A., and O’Dor, R. K. (1991). Foraging strategies and predation risk shape the natural history of juvenile Octopus vulgaris. Bulletin of Marine Science 49, 256–269.

Mazzei, V., and Biber, P. (2015). Autotrophic net productivity patterns at four artificial reef sites in the Mississippi Sound. Hydrobiologia 749, 135–154. doi:10.1007/s10750-014-2160-6

McCormick, T. B., Herbinson, K., Mill, T. S., and Altick, J. (1994). A review of abalone seeding, possible significance and a new seeding device. Bulletin of Marine Science 55, 680–693.

McCutchan, J. H., Lewis, W. M., Kendall, C., and McGrath, C. C. (2003). Variation in trophic shift for stable isotope ratios of carbon, nitrogen, and sulfur. Oikos 102, 378–390. doi:10.1034/j.1600-0706.2003.12098.x

McQuaid, C. D. (1994). Feeding behaviour and selection of bivalve prey by Octopus vulgaris Cuvier. Journal of Experimental Marine Biology and Ecology 177, 187–202. doi:10.1016/0022-0981(94)90236-4

McShane, P. E., and Naylor, J. R. (1997). Direct estimation of natural mortality of the New Zealand abalone, Haliotis iris. New Zealand Journal of Marine and Freshwater Research 31, 135–137. doi:10.1080/00288330.1997.9516750

Melville-Smith, R. (2016). Investigating critical biological issues for commercial greenlip abalone sea ranching in Flinders Bay, Western Australia. FRDC Project Number: 2014/214. Milestone report 4. Australian Capital Territory.

Melville-Smith, R., Adams, B., Wilson, N. J., and Caccetta, L. (2013). Sea ranching trials for commercial production of greenlip (Haliotis laevigata) abalone in Western Australia : an outline of results from trials conducted by Ocean Grown Abalone Pty Ltd. North Fremantle. Available at: http://www.oceangrown.com.au/OGA FRDC research-report.pdf

Molony, B. W., Lenanton, R., Jackson, G., and Norriss, J. (2003). Stock enhancement as a fisheries management tool. Reviews in Fish Biology and Fisheries 13, 409–432. doi:10.1007/s11160-005-1886-7

Morash, A. J., and Alter, K. (2016). Effects of environmental and farm stress on abalone physiology: perspectives for abalone aquaculture in the face of global climate change. Reviews in Aquaculture 8, 342–368. doi:10.1111/raq.12097

Munro, J. L., and Bell, J. D. (1997). Enhancement of marine fisheries resources. Reviews in Fisheries Science 5, 185–222. doi:10.1080/10641269709388597

Navarro, J., Coll, M., Somes, C. J., and Olson, R. J. (2013). Trophic niche of squids: Insights from isotopic data in marine systems worldwide. Deep-Sea Research Part II: Topical Studies in Oceanography 95, 93–102. doi:10.1016/j.dsr2.2013.01.031

Nixon, M. (1987). Cephalopod diets. In ‘Cephalopod life cycles. Comparative Reviews’. (Ed P. R. Boyle.) pp. 201–219. (Academic Press: London, Orlando.)

Nixon, M. (1984). Is there external digestion by octopus? Journal of Zoology 202, 441–447. doi:10.1111/j.1469-7998.1984.tb05094.x

Nixon, M., and Elaine, M. (1988). Drilling by Octopus vulgaris (Mollusca Cephalopoda) in the Mediterranean. J. Zool., Lond. 216, 687–716. doi:10.1111/j.1469-7998.1988.tb02466.x

Norman, M. D. (2003). ‘Cephalopods: a world guide’ 2nd ed. (Conch Books: Hackenheim, Germany.)

O’Dor, R. K., Mangold, K., Boucher-Rodoni, R., Wells, M. J., and Wells, J. (1984). Nutrient bsorption, storage and remobilization in octopus. Marine & Freshwater Behaviour & Physiology 3, 239–258. doi:10.1080/10236248409387049

84

Ogle, D. H. (2017). FSA: Fisheries stock analysis. Available at: https://github.com/droglenc/FSA

Parnell, A. C., Inger, R., Bearhop, S., and Jackson, A. L. (2010). Source partitioning using stable isotopes: Coping with too much variation. PLoS ONE 5. doi:10.1371/journal.pone.0009672

Pearce, A., Lenanton, R., Jackson, G., Moore, J., Feng, M., and Gaughan, D. (2011). The ‘marine heat wave’ off Western Australia during the summer of 2010/11. Fisheries Research Report No. 222. Perth, Western Australia.

Phillips, D. L., Inger, R., Bearhop, S., Jackson, A. L., Moore, J. W., Parnell, A. C., Semmens, B. X., and Ward, E. J. (2014). Best practices for use of stable isotope mixing models in food-web studies. Canadian Journal of Zoology 92, 823–835. doi:10.1139/cjz-2014-0127

Pierce, G. J., Boyle, P. R., Hastie, L. C., and Santos, M. (1994). Diets of squid Loligo forbesi and Loligo vulgaris in the northeast Atlantic. Fisheries Research 21, 149–163. doi:10.1016/0165-7836(94)90101-5

Pilson, M. E., and Taylor, P. B. (1961). Hole drilling by octopus. Science 134, 1366–8. doi:10.1126/science.134.3487.1366-a

Platell, M. E., Hesp, S. A., Cossington, S. M., Lek, E., Moore, S. E., and Potter, I. C. (2010). Influence of selected factors on the dietary compositions of three targeted and co-occurring temperate species of reef fishes: Implications for food partitioning. Journal of Fish Biology 76, 1255–1276. doi:10.1111/j.1095-8649.2010.02537.x

Platell, M. E., and Potter, I. C. (2001). Partitioning of food resources amongst 18 abundant benthic carnivorous fish species in marine waters on the lower west coast of Australia. Journal of Experimental Marine Biology and Ecology 261, 31–54. doi:10.1016/S0022-0981(01)00257-X

Poh, B., Tweedley, J. R., Chaplin, J. A., Trayler, K. M., and Loneragan, N. R. (2017). Estimating predation rates of restocked individuals: The influence of timing-of-release on metapenaeid survival. Fisheries Research. doi:10.1016/j.fishres.2017.09.019

Polovina, J. J., and Sakai, I. (1989). Impacts of artificial reefs on fishery production in Shimamaki, Japan. Bulletin of Marine Science 44, 997–1003.

Pompanon, F., Deagle, B. E., Symondson, W. O. C., Brown, D. S., Jarman, S. N., and Taberlet, P. (2012). Who is eating what: Diet assessment using next generation sequencing. Molecular Ecology 21, 1931–1950. doi:10.1111/j.1365-294X.2011.05403.x

Post, D. M. (2002). Using stable isotopes to estimate trophic position: Models, methods, and assumptions. Ecology 83, 703–718. doi:10.2307/3071875

Prince, J. D. (2004). The Decline of Global Abalone (Genus Haliotis) Production in the Late Twentieth Century: Is there a Future? In ‘Stock Enhancement and Sea Ranching’. (Eds K. M. Leber, S. Kitada, H. L. Blankenship, and T. Svasand.) pp. 427–443. (Blackwell Publishing Ltd: Oxford, UK.) doi:10.1002/9780470751329.ch31

R Core Development Team (2011). ‘R: A Language and environment for statistical computing’. (The R Foundation for Statistical Computing.: Vienna, Austria.) Available at: http://www.r-project.org/

Regueira, M., Guerra, Á., Fernández-Jardón, C. M., and González, A. F. (2017). Diet of the horned octopus Eledone cirrhosa in Atlantic Iberian waters: ontogenetic and environmental factors affecting prey ingestion. Hydrobiologia 785, 159–171. doi:10.1007/s10750-016-2916-2

Rigby, P. R., and Sakurai, Y. (2005). Multidimensiorial tracking of giant Pacific

85

octopuses in Northern Japan reveals unexpected foraging behaviour. Marine Technology Society Journal 39, 64–67.

Roberts, R. D., Keys, E. F., Prendeville, G., and Pilditch, C. A. (2007). Viability of abalone (Haliotis iris) stock enhancement by release of hatchery-reared seed in Marlborough, New Zealand. Journal of Shellfish Research 26, 697–703. doi:10.2983/0730-8000(2007)26[697:VOAHIS]2.0.CO;2

Rodhouse, P. G. K., and Nigmatullin, C. M. (1996). Role as Consumers. Philosophical Transactions of the Royal Society of London B: Biological Sciences 351, 1003–1022. doi:10.1098/rstb.1996.0090

Rodhouse, P. G. K., Pierce, G. J., Nichols, O. C., Sauer, W., Arkhipkin, A. I., Laptikhovsky, V. V, Lipinski, M. R., Ramos, J. E., Gras, M., Kidokoro, H., Sadayasu, K., Pereira, J., Lefkaditou, E., Pita, C., Gasalla, M. A., Haimovici, M., Sakai, M., and Downey, N. (2014). Chapter Two - Environmental effects on cephalopod population dynamics: Implications for management of fisheries. In ‘Advances in Marine Biology’. (Ed E. A. G. Vidal.) pp. 99–233. (Academic Press: Oxford: United Kingdom.) doi:http://dx.doi.org/10.1016/B978-0-12-800287-2.00002-0

Scheel, D., Leite, T. S., Mather, J. A., and Langford, K. (2016). Diversity in the diet of the predator Octopus cyanea in the coral reef system of Moorea, French Polynesia. Journal of Natural History, 1–19. doi:10.1080/00222933.2016.1244298

Schiel, D. R. (1993). Experimental evaluation of commercial-scale enhancement of abalone Haliotis iris populations in New Zealand. Marine Ecology Progress Series 97, 167–181. doi:10.3354/meps097167

Schiel, D. R., and Welden, B. C. (1987). Responses to predators of cultured and wild red abalone, Haliotis rufescens, in laboratory experiments. Aquaculture 60, 173–188. doi:10.1016/0044-8486(87)90286-9

Seaman, W. (2000). ‘Artificial reef evaluation with application to natural marine habitats’. (CRC Press: Boca Raton, U.S.A.) doi:1016/S0165-7836(03)00126-7

Seaman, W. (2008). Coastal artificial habitats for fishery and environmental management and scientific advancement. In ‘Fisheries for Global Welfare and Environment, 5th World Fisheries Congress 2008’. pp. 335–349 doi:335-349

Shepherd, S. A. (1973). Studies on southern Australian abalone (Genus Haliotis). i. Ecology of five sympatric species. Marine and Freshwater Research 24, 217–258. doi:10.1071/MF9730217

Shepherd, S. A. (1990). Studies on southern australian abalone (Genus Haliotis). XII. long-term recruitment and mortality dynamics of an unfished population. Marine and Freshwater Research 41, 475–492. doi:10.1071/MF9900475

Shepherd, S. A. (1998). Studies on southern Australian abalone (genus Haliotis) - XIX: Long-term juvenile mortality dynamics. Journal of Shellfish Research 17, 813–825. doi:none 4/2009

Shepherd, S. A., and Clarkson, P. S. (2001). Diet, feeding behaviour, activity and predation of the temperate Blue-throated Wrasse, Notolabrus tetricus. Marine and Freshwater Research 52, 311–322. doi:10.1071/MF99040

Shepherd, S. A., Preece, P. A., and White, R. G. W. (2001). Tired nature’s sweet restorer? Ecology of abalone (Haliotis spp.) stock enhancement in Australia. In ‘Workshop on Rebuilding Abalone Stocks in British Columbia Canadian Special Publication of Fisheries and Aquataquatic Sciences No. 130’. (Ed A. Campbell.) pp. 84–98. (NRC Research Press.) doi:doi:10.1139/9780660182926

Shepherd, S. A., and Turner, J. A. (1985). Studies on southern Australian abalone (genus Haliotis). VI. Habitat preference, abundance and predators of juveniles. Journal of Experimental Marine Biology and Ecology 93, 285–298.

86

doi:10.1016/0022-0981(85)90245-X Šifner, K. S., and Vrgoč, N. (2009). Diet and feeding of the musky octopus, Eledone

moschata, in the northern Adriatic Sea. Journal of the Marine Biological Association of the United Kingdom 89, 413. doi:10.1017/S0025315408002488

Smith, C. D. (2003). Diet of Octopus vulgaris in False Bay, South Africa. Marine Biology 143, 1127–1133. doi:10.1007/s00227-003-1144-2

Støttrup, J. G., Overton, J. L., Paulsen, H., Möllmann, C., Tomkiewicz, J., Pedersen, P. B., and Lauesen, P. (2008). Rationale for restocking the Eastern Baltic Cod stock. Reviews in Fisheries Science 16, 58–64. doi:10.1080/10641260701678231

Svane, I., and Petersen, J. K. (2001). On the problems of epibioses, fouling and artificial reefs, a review. Marine Ecology 22, 169–188. doi:10.1046/j.1439-0485.2001.01729.x

Tanner, A. R., Fuchs, D., Winkelmann, I. E., Gilbert, M. T. P., Pankey, M. S., Ribeiro, Â. M., Kocot, K. M., Halanych, K. M., Oakley, T. H., da Fonseca, R. R., Pisani, D., and Vinther, J. (2017). Molecular clocks indicate turnover and diversification of modern coleoid cephalopods during the Mesozoic Marine Revolution. Proceedings of the Royal Society of London B: Biological Sciences 284. Available at: http://rspb.royalsocietypublishing.org/content/284/1850/20162818#ref-9 [accessed 29 July 2017]

Taylor, M. D., Chick, R. C., Lorenzen, K., Agnalt, A. L., Leber, K. M., Blankenship, H. L., Haegen, G. Vander, and Loneragan, N. R. (2017). Fisheries enhancement and restoration in a changing world. Fisheries Research 186, 407–412. doi:10.1016/j.fishres.2016.10.004

Tegner, M. J., and Butler, R. A. (1989). Introduction: The rationale for abalone seeding. In ‘Handbook of culture of abalone and other marine gastropods’. (Ed K. O. Hahn.) pp. 157–182. (CRC Press: Boca Raton, U.S.A.)

Tegner, M. J., and Butler, R. A. (1985). The survival and mortality of seeded and native Red Abalones Haliotis-Rufescens on the Palos-Verdes Peninsula California USA. California Fish Game 71(3): 15, 150–163.

Uki, N. (2006). Stock enhancement of the Japanese scallop Patinopecten yessoensis in Hokkaido. Fisheries Research 80, 62–66. doi:10.1016/j.fishres.2006.03.013

United States Department of Agriculture (2016a). National nutrient database: Crustaceans, crab, raw. Available at: https://ndb.nal.usda.gov/ndb/foods/show/4618?fgcd=&manu=&lfacet=&format=Abridged&count=&max=50&offset=&sort=default&order=asc&qlookup=crab+raw&ds=&qt=&qp=&qa=&qn=&q=&ing= [accessed 13 September 2017]

United States Department of Agriculture (2016b). National nutrient database: Mollusks, abalone, raw.

Venables, W. N., and Ripley, B. D. (2002). ‘MASS: Modern applied statistics with S’. (Springer: New York.) Available at: http://www.stats.ox.ac.uk/pub/MASS4

Villanueva, R. (1993). Diet and mandibular growth of Octopus magnificus (Cephalopoda). South African Journal of Marine Science 13, 121–126. doi:10.2989/025776193784287239

Villanueva, R., Nozais, C., and Boletzky, S. (1995). The planktonic life of octopuses. Nature 377, 107. doi:10.1038/377107a0

Villegas, E. J. A., Ceballos-Vázquez, B. P., Markaida, U., Abitia-Cárdenas, A., Medina-López, M. A., and Arellano-Martínez, M. (2014). Diet of Octopus bimaculatus Verril, 1883 (Cephalopoda: Octopodidae) in Bahía De Los Ángeles, Gulf of California. Journal of Shellfish Research 33, 305–314. doi:10.2983/035.033.0129

De Waal, S., Balkhair, M., Al-Mashikhi, A., and Khoom, S. (2013). Investigating the translocation and seeding of wild Haliotis mariae Wood, 1828, in the Sultanate of

87

Oman. Journal of Shellfish Research 32, 315–323. doi:10.2983/035.032.0210 De Waal, S., and Cook, P. (2001). Quantifying the physical and biological attributes of

successful ocean seeding sites for farm-reared juvenile abalone (Haliotis midae). Journal of Shellfish Research 20, 857–861.

Wickham, H. (2009). ‘ggplot2: Elegant graphics for data analysis’. (Springer-Verlag: New York.) Available at: http://ggplot2.org

Wodinsky, J. (1969). Penetration of the shell and feeding on gastropods by octopus. Integrative and Comparative Biology 9, 997–1010. doi:10.1093/icb/9.3.997

Won, N.-I., Kawamura, T., Takami, H., Noro, T., Musashi, T., and Watanabe, Y. (2010). Ontogenetic changes in the feeding habits of the abalone Haliotis discus hannai: field verification by stable isotope analyses. Canadian Journal of Fisheries and Aquatic Sciences 67, 347–356. doi:http://dx.doi.org/10.1139/F09-187

Worm, B., Hilborn, R., Baum, J. K., Branch, T. A., Collie, J. S., Costello, C., Fogarty, M. J., Fulton, E. A., Hutchings, J. A., Jennings, S., Jensen, O. P., Lotze, H. K., Mace, P. M., McClanahan, T. R., Minto, C., Palumbi, S. R., Parma, A. M., Ricard, D., Rosenberg, A. A., Watson, R., and Zeller, D. (2009). Rebuilding global fisheries. Science 325, 578–585. doi:10.1126/science.1173146

Wu, Z., Zhang, X., Lozano-Montes, H. M., and Loneragan, N. R. (2016). Trophic flows, kelp culture and fisheries in the marine ecosystem of an artificial reef zone in the Yellow Sea. Estuarine, Coastal and Shelf Science 182, 86–97. doi:10.1016/j.ecss.2016.08.021

Xu, Q., Zhang, L., Zhang, T., Zhang, X., and Yang, H. (2017). Functional groupings and food web of an artificial reef used for sea cucumber aquaculture in northern China. Journal of Sea Research 119, 1–7. doi:10.1016/j.seares.2016.10.005

Ye, Y., Cochrane, K., Bianchi, G., Willmann, R., Majkowski, J., Tandstad, M., and Carocci, F. (2013). Rebuilding global fisheries: The World Summit oal, costs and benefits. Fish and Fisheries 14, 174–185. doi:10.1111/j.1467-2979.2012.00460.x

Yick, J. L., Barnett, A., and Tracey, S. R. (2012). The trophic ecology of two abundant mesopredators in south-east coastal waters of Tasmania, Australia. Marine Biology 159, 1183–1196. doi:10.1007/s00227-012-1899-4

Vander Zanden, M. J., and Rasmussen, J. B. (2001). Variation in δ15N and δ13C trophic fractionation: Implications for aquatic food web studies. Limnology and Oceanography 46, 2061–2066. doi:10.4319/lo.2001.46.8.2061

Zar, J. H. (2010). ‘Biostatistical analysis’ 5th ed. (Pearson Prentice Hall: Upper Saddle River, NJ.)

Zeileis, A., Kleiber, C., Jackman, S., and URL, J. of S. S. 27(8). (2008). pscl: Regression models for count data in R. Journal of Statistical Software 27, 1–25. Available at: http://www.jstatsoft.org/v27/i08/

Zhang, L., Zhang, T., Xu, Q., Qiu, T., Yang, H., and Liu, S. (2015). An artificial oyster-shell reef for the culture and stock enhancement of sea cucumber, Apostichopus japonicus, in shallow seawater. Aquaculture Research 46, 2260–2269. doi:10.1111/are.12383

Zuur, A. F., Ieno, E. N., and Elphick, C. S. (2010). Zuur et al (2010) A protocol for data exploration to avoid common statistical problems.pdf. Methods in Ecology and Evolution 1, 3–14. doi:10.1111/j.2041-210X.2009.00001.x

Appendix 1. Cephalopods and their role as predators and trophic intermediaries in the food webs of coastal marine systems

A1.0. Abstract

Cephalopods are formidable predators that display complex, flexible behaviours that

have allowed them to thrive in the world’s oceans. In recent decades their biomass has

increased significantly, which is likely to have considerable ecological implications,

particularly in dynamic, shallow water coastal environments. In contrast to the landings

from most global finfish fisheries, which are at their limits, those for cephalopods have

increased markedly since the 1990s. Their short life history strategy characterised by

rapid growth stimulates voracious feeding regimes, and their large global biomass has

the potential to impose significant top-down controls on a broad range of commercially

and ecologically important species. This review summarises the role of coleoid

cephalopods (i.e., the Loliginidae, Octopodidae, Sepiidae, and Sepiolidae), as predators

in coastal marine food webs. The morphological and physiological traits and feeding

behaviours that contribute to their success as predators are summarised from the

existing literature. Equipped with a range of unique morphological features and highly

developed sensory organs, cephalopods are well adapted for a predatory lifestyle.

Information on the diet composition of coastal coleoid cephalopods was synthesised

from published literature. Distinct dietary patterns were found, with crustaceans,

teleosts, and molluscs common in the gastric tracts of all members. However, the

relative contributions of the various prey taxa varied considerably among individual

families, reflecting the position they occupy within the water column. Loliginidae had

the most distinct diet, with teleosts, and cephalopods making a greater contribution to

their diets than those of the other families. A pronounced shift in diet with ontogeny is a

ubiquitous feature across the four families, with increased body size facilitating

cannibalism and intraspecific predation. Finally, their role as predators and trophic

intermediaries in coastal marine ecosystems is described. Further studies that utilise

biochemical tracers are needed to evaluate trophic relationships, long-term dietary

trends, and their response to environmental variability. This information will help

inform reliable assessments of cephalopods and other fished species, and ensure that

cephalopod fisheries are developed in an ecologically sustainable way.

89

Table of Contents

Appendix 1. Cephalopods and their role as predators and trophic intermediaries in the food webs of coastal marine systems ..................................................................... 88 A1.0. Abstract ................................................................................................................. 88 A1.1. Introduction ........................................................................................................... 90

A1.1.1. Focus of the review ........................................................................................ 95

A1.2. Biodiversity and zoogeography ............................................................................. 95 A1.3. Feeding morphology and physiology .................................................................... 96

A1.3.1. Sensory organs ............................................................................................... 97

A1.3.2. Chromatophore organs ................................................................................... 98

A1.3.3. Feeding apparatus ........................................................................................... 99 A1.4. Feeding behaviour ............................................................................................... 100

A1.4.1. Foraging behaviour ...................................................................................... 100

A1.4.2. Diel feeding activity ..................................................................................... 101

A1.4.3. Shoaling ........................................................................................................ 102

A1.4.4. Prey selection ............................................................................................... 103

A1.5. Growth rates and food intake .............................................................................. 106 A1.6. Diets .................................................................................................................... 107

A1.6.1. Broad dietary trends ..................................................................................... 107

A1.6.2. Ontogenetic dietary shifts ............................................................................. 110

A1.6.3. Cannibalism .................................................................................................. 110

A1.6.4. Influence of sampling method on interpretation of diets ............................. 111

A1.6.5. Dietary differences among families ............................................................. 114

A1.7. Predation in marine systems ................................................................................ 117 A1.7.1. Predation impacts ......................................................................................... 117

A1.7.2. Trophic relationships .................................................................................... 119

A1.8. Conclusions and directions for future research ................................................... 120 A1.9. References ........................................................................................................... 123

90

A1.1. Introduction

Cephalopods are exclusively marine molluscs distributed widely throughout the world’s

oceans (Boyle & Rodhouse, 2007). They are found in practically all habitats from the

polar seas to the tropics; in neritic and open oceanic waters from the littoral zone

through to abysmal benthic environments up to 5,000 m deep (Amaratunga & Caddy,

1983; Boyle & Rodhouse, 2007). The class Cephalopoda comprises approximately 800

species from two subclasses: the Coleoidea, represented by the Octopoda, Sepiida,

Sepiolida, Teuthida, and Vampyromorphida; and the Nautiloidea, represented by the

Nautilida; a primitive group distinguished by their distinctive external shell (Boyle and

Rodhouse 2005; Rodhouse et al. 2014).

Cephalopods share a number of basic features in their body organisation such as

the presence of a radula (i.e., a file-like feeding organ), an internal or external shell, and

a central nervous system that is arranged around the oesophagus (Boyle & Rodhouse,

2007). However, the two subclasses have a number of distinct biological characteristics,

the most obvious being the complete loss of an external shell among the coleoids (Boyle

and Rodhouse 2005). The coleoids are monocyclic, single-season breeders characterised

by high growth rates, early maturity, and post-spawning mortality (i.e., they are

semelparous, Boyle & Boletzky, 1996; Boyle & Rodhouse, 2005). They have a short

life-history strategy characterised by rapid growth, with most medium size coastal

species senescencing after one to two years of life (Fig. A1.1). Due to their rapid growth

rates, the demands on protein synthesis are high (Lee 1995). Unlike other active marine

predators that utilise lipids for their long-term energy stores, cephalopods synthesize

amino acids very efficiently, which has facilitated the evolution of rapid growth rates

throughout the lifecycle (Lee 1995; Boyle and Rodhouse 2005).

91

Figure A1.1. Generalised stages in the lifecycle of a benthic octopus, adapted from the descriptions found in Boyle and Rodhouse (2005); Promboon et al. (2011) and Caputi et al. (2015). Images sourced from Carefoot (2017) and Integration and Application Network (2017).

Cephalopods are considered to be the most evolved of all invertebrates,

possessing a range of highly-developed chemical, tactile, and visual senses, which

facilitate learning and aid in the capture and selection of prey (Lee 1995; Darmaillacq et

al. 2004). They are very versatile predators, with the ability to change their prey when

their preferred prey is unavailable. Their diet also varies with ontogeny and these two

characteristics of cephalopod feeding lead to their occupation of several trophic levels

and ecological niches during their lifecycle (Rodhouse et al. 2014).

Predation is a central feature of community organisation, playing a key role in

the structure and function of marine systems. Predators have the ability to control prey

populations directly by reducing abundance, but also through indirect processes which

disrupt competition and succession (Underwood and Fairweather 1989). Seminal works

examining the interactions among rocky intertidal invertebrates (Paine 1966; 1974), and

sea otters in North Pacific kelp-forest ecosystems (Estes and Duggins 1995) revealed

the significance of predation (i.e., “top-down processes”), and introduced the term

“keystone species” in recognition of the impact of those species on community structure

92

and function, far beyond their relative biomass in the community. The impacts of

predation within marine food webs can vary considerably between communities

(Heithaus et al. 2008), dependent on a range of factors including predation intensity, the

metabolic needs of the predator, dietary preferences, and prevailing environmental

conditions (Ambrose 1984; Williams et al. 2004). For short-lived, opportunistic, high

turn-over species such as cephalopods, their predatory impact on prey populations

varies greatly; both spatially and over time as their populations fluctuate in biomass

(Rodhouse and Nigmatullin 1996; Heithaus et al. 2008).

An example of the potential scale of predation impacts by cephalopods is

provided by the Humboldt Squid, Dosidicus gigas, during their migration to the Gulf of

California (Ehrhardt 1991; Rodhouse and Nigmatullin 1996). During their migration,

their diet shifts from one dominated by myctophids (25%) and crabs (20%) early in the

migration in winter (January), to one comprising predominantly sardines (80%) and

mackerel (18%) towards the end of the migration in summer (July, Ehrhardt 1991). In

1980 and 1981, D. gigas consumed an estimated 60,000 metric tonnes of Californian

Sardines, Sardinops sagax caerule (Ehrhardt 1991), and the increased mortality through

heavy predation pressure is thought to have been a factor contributing to greatly reduced

commercial sardine landings during this period (Ehrhardt 1991).

Cephalopods are important species within marine food webs, both as predators

of invertebrates and fish, and as prey for higher trophic level consumers such as marine

mammals, predatory fish and seabirds (Boyle and Rodhouse 2005). Rich in protein,

coastal cephalopods have comparable mean calorific values (i.e., 22.7 ± 1.5 kJ/g dry wt)

to teleosts (24.7 ± 0.9 kJ/g for demersal fishes and 26.6 ± 0.9 kJ/g for pelagic fishes;

Blaber and Bulman 1987), and their soft, unarmored and highly digestible body renders

them vulnerable to predation (Boyle and Rodhouse 2005). Recent evidence suggests

that their biomass and landings from commercial fisheries are increasing as a result of

changing ecological conditions, such as ocean warming, and heavy fishing pressure on

predators of cephalopods (Fig. A1.3, André et al. 2010; Coll et al. 2013; Doubleday et

93

al. 2016). Although, overexploitation and unfavourable environmental conditions such

as a severe El Niño event in 2015 have resulted in temporal declines of squids in recent

times, particularly in the waters around Chile, Peru, Argentina, and U.S.A. (FAO

2016a). Improving our understanding of cephalopod predator-prey interactions is

fundamental to assessing their ecological role within the marine environment, and

evaluating how ecosystems may respond to environmental perturbations (Livingston

2003; Navarro et al. 2013).

Cephalopods, particularly the squids, are also commercially important,

supporting various industrial, local and artisanal fisheries across the globe, with major

fisheries occurring around Argentina, China, India, Japan, Peru, New Zealand, and

Russia (Fig 2.2; Rodhouse et al. 2014). In recent decades, the economic value of

cephalopods has grown significantly. Estimated global landings have increased from 1

million tonnes in the mid 1960s to over 3 million tonnes since 1995, and 4 million

tonnes in some years from 2005 to 2014 (Fig. A1.3, FAO 2016). Landings peaked in

2014 at 4.75 million tonnes (Fig. A1.3, FAO 2016). Overall, squids contribute to

approximately three-quarters of the total catch, with the remainder of landings

comprising cuttlefish and octopus (Fig 2.3). This trend is consistent with traditional

teleost fisheries, whose tonnage has also quadrupled since the 1950s (Worm and Branch

2012). Unlike finfish, which are at their global limits of exploitation (Worm and Branch

2012; Bell et al. 2016), cephalopods represent one of the few marine resources with

potential for increased sustainable harvest at local levels (Clarke 1996b; Caddy and

Rodhouse 1998; Hunsicker et al. 2010).

94

Figure A1.2. Global cephalopod landings in 2014. White indicates no catch, and blue (1.5e-21-4.1e-7 t/km2) to red represents an increase in cephalopod landings (1.2e-2–6.2e+1 t/km2; Pauly et al. 2015).

Figure A1.3. Trends in global landings of cephalopod species groups from 1950 to 2014, nei = not enough information (taken from FAO 2016).

95

A1.1.1. Focus of the review

This review focuses on synthesising the current knowledge of coastal cephalopod diets,

and the role of cephalopods as predators and trophic intermediaries in coastal marine

food webs, as their biology and ecology has been reviewed extensively (Boyle 1990;

Rocha et al. 2001; Boyle and Rodhouse 2005). It also focuses exclusively on coastal

coleoid species (i.e., the Loliginidae, Octopodidae, Sepiidae, and Sepiolidae) because of

their relatively similar biology, and the intensity of anthropogenic impacts, such as

fishing, nutrient enrichment, and habitat modification on their populations in coastal

systems. As nautiloids are very distinct from coleoids in their biology, life history

strategy, and behaviour, they ‘must be considered separately’ (Hanlon and Messenger

1996), and are not examined in this review. Thus, in this review, the term cephalopod

refers to the coastal coleoid species only.

The review of coastal cephalopods and their place in food webs considers the

following major topics: (i) biodiversity and zoogeography; (ii) feeding morphology and

physiology; (iii) feeding behaviour; (iv) growth rates and food intake; (v) diets; and (vi)

predation in marine systems, and concludes with a summary of the major findings and

directions for future research. The section on cephalopod feeding includes a meta-

analysis of the diets of 24 species from the families Loliginidae, Octopodidae, Sepiidae,

and, Sepiolidae. This meta-analysis, which is based on 28 studies provides an

understanding of how dietary composition differs among families and two main dietary

methodologies: the examination of gastric tracts and octopus middens.

A1.2. Biodiversity and zoogeography

Among the ~ 800 species of cephalopod, diversity is greatest in shallow coastal waters

and in the deep oceans beyond the mesopelagic zone (Boyle & Rodhouse 2005; Coelho,

1985). In shallow coastal zones, benthic octopuses (Octopodidae) dominate the rocky

habitats of temperate and tropical waters, and the Sepiidae (cuttlefish) and Sepiolidae

(bobtail squid) are abundant in habitats with soft bottoms where they can burrow or

remain inconspicuous near the substratum (Boyle & Rodhouse 2005). The true squids

96

(i.e., the Loliginidae) are mostly pelagic and are commonly found in coastal waters and

on continental shelves, but the distribution and depth at which they occur varies widely

among its members (Coelho 1985; Clarke 1996b; Boyle and Rodhouse 2005). The

Ommastrephidae squids are generally found in the epipelagic and mesopelagic zones of

shelf waters (Coelho 1985; Clarke 1996b; Boyle and Rodhouse 2005).

The number of cephalopod species varies greatly among genera and families,

with over half of the described species being attributed to monotypic genera, and the

majority of these occur within a diverse range of families of mainly deep-sea origin

including the highly unusual Vampyroteuthidae, Opisthoteuthidae, and Cirroteuthidae

which occupy the mesopelagic, bathypelagic, and abyssal waters of the world’s oceans

(Boyle and Rodhouse 2005). In contrast, the genera Octopus and Sepia each have over

100 species, while other coastal forms from the Loliginidae and Sepiolidae each

contribute a considerable number of species to the class (Voss 1977; Boyle and

Rodhouse 2005). The faster, more active coastal cephalopods are thought to have

evolved to compete successfully with teleost fishes through the loss of the external shell

(Vermeij 1994; Boyle and Rodhouse 2005). This allowed the exploitation of a high-

energy lifestyle and invasion of productive coastal regions from the relatively

unfavourable deep sea, where the majority of original cephalopod fauna were restricted

(Vermeij 1994; Boyle and Rodhouse 2005).

A1.3. Feeding morphology and physiology

Cephalopods possess a range of advanced behavioural, morphological, and

physiological traits that contribute to their success as predators within the marine

environment (Rocha et al. 2001; Boyle and Rodhouse 2005). They have the largest,

most complex nervous system of all invertebrates (Young 1971) including a well-

developed brain that features a chemo-tactile memory system. This facilitates rapid

learning and cognition, and the development of complex plastic behaviours (Hanlon and

Messenger 1996; Rodhouse et al. 2014). Extensive research on octopuses highlights

their highly adaptable and intelligent nature. This includes: the active selection of home

97

sites and subsequent modification of habitat; the use of defensive tools such as shells

and coconuts for protection against predators; the use of spatial memory and visual

landmarks in navigation; and the modification of foraging behaviour when presented

with novel feeding opportunities (Mather 1991b; Mather 1994; Rigby and Sakurai 2005;

Finn et al. 2009). For example, data from acoustically tagged Giant Pacific Octopus,

Octopus doeflini, off the coast of Japan revealed that this normally benthic species

undertakes distinct, large vertical movements within the water column (Rigby and

Sakurai 2005). Observations by SCUBA divers confirmed that this vertical migration

was in response to the presence of a pelagic gillnet, and octopuses were scaling the net

and feeding opportunistically on the ensnared fish (Rigby and Sakurai 2005); a highly

adaptive behaviour demonstrating clear intelligence.

A1.3.1. Sensory organs

Cephalopod sensory organs are very sophisticated, with the ability to detect prey

through sight, scent, sound, and distant touch (Budelmann 1996). Despite possessing

only a single colour photoreceptor in their retina and therefore being colour blind

cephalopods have acute vision (Marshall and Messenger 1996; Hanlon and Messenger

1996). A strong rectilinear arrangement of photoreceptor cells in the retina allows

polarised light to be discriminated and subsequently, camouflaged prey to be detected

(Shashar et al. 2000; Boyle and Rodhouse 2005). Polarisation sensitivity is utilised

during foraging as a means to detect counter shaded camouflage by light-reflecting fish

and prey that utilise transparency as a predator avoidance mechanism (Shashar et al.

2000).

Coleoids possess an epidermal line system, analogous to the lateral line in fish

that is sensitive to particle motion, giving the sense of distant touch and aiding in the

detection of prey (Hanlon and Messenger 1996; Boyle and Rodhouse 2005). The

epidermal line is sensitive to water movements as low as 0.06 𝜇m, which is equivalent

to a cuttlefish buried in the sand being able to detect a fish of 1 m in body length from

30 m away (Budelmann 1996). In addition, specialised receptor cells on the surface of

98

the suckers enhance chemotactile sensitivity, with up to 10,000 chemoreceptor cells

present in any one sucker of an octopus (up to 16 million per individual; Budelmann

1996). Complex balance and orientation organs known as statocysts, which parallel the

inner ear of the vertebrates, are situated in the cartilage on either side of the brain

(Williamson 1995; Boyle and Rodhouse 2005). These highly advanced organs provide

the animal with critical sensory information, which is incorporated into a range of

behaviours including locomotion, eye movement control, and body-colour patterning

(Williamson 1995) as well as providing the information required for stabilization and

the regulation of activities within a three dimensional space (Budelmann 1996).

A1.3.2. Chromatophore organs

Cephalopods “have evolved the most sophisticated system of colour control and pattern

formation known” in the animal kingdom (Boyle and Rodhouse 2005). The colouring

and patterning of cephalopods result from an arrangement of three layers of pigmented

cells called chromatophore organs, an important feature of cephalopod physiology

(Hanlon and Messenger 1996; Boyle and Rodhouse 2005; Osorio 2014). The

chromatophore organs are used frequently to provide camouflage to avoid predation and

ambush unsuspecting prey, and to facilitate communication between conspecifics

(Boyle and Rodhouse 2005; Hanlon 2007). Located under the skin at a density of 50

chromatophores per mm2, these organs form part of the neuro-musculature system and

consist of intracellular, pigment containing, cytoelastic sacs which are arranged radially

within the muscle fibres (Boyle and Rodhouse 2005; Hanlon 2007; Osorio 2014). The

pigment sacs become increasingly visible as the chromatophore muscles contract,

imparting colour pigment against a patch of the skin’s surface approximately 300 𝜇m

across (Boyle and Rodhouse 2005; Osorio 2014). When muscles relax, the pigment sacs

retract and the colouration against the skin is reduced, exposing the light coloured tissue

below (Boyle and Rodhouse 2005; Osorio 2014). However, the appearance of the skin’s

surface is further complicated by differing pigment hues, static reflecting mirror-like

iridophores, and matte white leucophores situated beneath the chromatophores that

99

scatter light outwards (Boyle and Rodhouse 2005; Osorio 2014). These features

combine to produce the complex patterning, texture, and colouration observed within

the group (Boyle and Rodhouse 2005; Osorio 2014). The chromatophore organs are

controlled directly by the brain and each muscle is innervated by motor-neurons that

convey signals, without synapse, that allow for rapid (0.2 to 2.02 seconds) and complex

changes to take place in response to the surrounding environmental conditions (Boyle

and Rodhouse 2005; Hanlon 2007). Body patterning is highly complex, and often the

intricacy of patterning is associated with habitat complexity, reproductive strategies,

predation, and socialisation (Hanlon and Messenger 1996; Boyle and Rodhouse 2005).

A1.3.3. Feeding apparatus

Cephalopods are well adapted for a predatory lifestyle, possessing a range of unique

apparatus to capture and handle a broad selection of prey (Rodhouse and Nigmatullin

1996). Most notable are the prehensile arms which bear numerous strong suckers that

are used for pulling and overcoming large prey, and the extensible tentacles of the

Decapodiformes (i.e., cuttlefish and squid), used for seizing prey (Rodhouse and

Nigmatullin 1996; Hanlon and Messenger 1996). The arm appendages form the

versatile brachial crown, which serves as the functional mouth and aids in the capture,

manipulation, and immobilisation of large prey (Boletzky 1983; Rodhouse and

Nigmatullin 1996). The suckers, equipped with chemo-sensory receptors, are

particularly important among the octopuses who use them for the ‘blind’ detection of

prey whilst foraging in crevices and under ledges where potential prey remain unseen

(Hanlon and Messenger 1996).

The buccal mass is a complex structure comprising a chitinous beak and

associated musculature used for tearing food into small pieces; a tongue-like radula

equipped with rows of small chitinous teeth used for drilling and rasping; and salivary

papilla bearing tiny teeth that can be protruded from the mouth (Kojima 1992; Mather

and Nixon 1995; Boyle and Rodhouse 2005). Some cephalopods, particularly the

octopuses, also possess posterior salivary glands that produce cephalotoxin; a

100

glycoprotein, capable of paralysing prey when injected (Ghiretti 1959; Nixon 1984).

The cephalotoxin can also stimulate external digestion in octopuses, aiding in the

breakdown of membranes and musculo-skeletal attachments of prey (Nixon 1984). The

chemical composition of cephalotoxins varies among species, and the details of their

properties and primary structures are not yet fully understood (Ueda et al. 2008).

A1.4. Feeding behaviour A1.4.1. Foraging behaviour

Cephalopods utilise a range of hunting strategies whilst pursuing prey, including active

searching, ambush, luring, pursuit, stalking, speculative pouncing, and even hunting in

disguise whereby individuals mimic seaweed or herbivorous fish (Hanlon and

Messenger 1996). Hunting strategies vary considerably among species, although

flexibility in the use of hunting forms is a common feature (Boyle & Rodhouse, 2007).

The method adopted is dependent on the surrounding environment, physical structure of

the substratum, and an individual’s positioning within the water column (Forsythe and

Hanlon 1997). For example, benthic coastal octopuses such as the Common Octopus,

Octopus vulgaris, and the Brazil Reef Octopus, Octopus insularis, commonly use

ambush predation, remaining inconspicuous on the substratum, searching for prey using

their acute eyesight, then attack potential prey as they move within striking distance

(Leite et al. 2009). However, they also are active searchers commonly adopting tactile

foraging behaviours such as (i) “crawl”, where they use their strong suckers to move

along the substratum searching for prey using their tactile and chemical senses; (ii)

“poke”, where they use their arms and chemo-tactile senses to explore crevices likely to

contain prey; and (iii) “web-over” where an octopus extends its arms and web over a

small part of the substratum, and then uses its arm tips to explore the enclosed area for

prey (Mather 1991a; Leite et al. 2009).

The diversity of feeding strategies that may be used by a cephalapod is

demonstrated by the Carribbean Reef Squid, Sepioteuthis sepioidea, which adopts four

flexible modes of hunting (Moynihan and Rodaniche 1982; Hanlon and Messenger

101

1996). During the day, small schools of squid typically wait for small fishes to come

towards them, and then ambush them one by one with their extendable tentacles

(Moynihan and Rodaniche 1982; Hanlon and Messenger 1996). When the squids are

dispersed at night, they actively search and stalk their prey, nearing ever closer before

striking (Moynihan and Rodaniche 1982; Hanlon and Messenger 1996). They also hunt

in disguise at night, using floating seagrasses and seaweed to remain camouflaged, often

with a head-down posture and then ambush unsuspecting prey that approach too closely

(Moynihan and Rodaniche 1982; Hanlon and Messenger 1996). Sepioteuthis sepioidea

also engage in speculative hunting, using their arms to disturb the sediments to flush out

small invertebrates and fishes (Moynihan and Rodaniche 1982).

Cephalopods hunt and locate their prey visually, and octopuses are able to target

sedentary or slow moving organisms such as mussels and abalone with precision.

Sedentary prey may be attacked and captured by pulling or drilling (Mather 1991a). For

example, O. vulgaris will commonly attempt to force a mussel open in the first instance,

with each pull usually lasting for just a few seconds (McQuaid 1994). However, if

pulling is not successful within five or ten minutes, the octopus resorts to drilling a hole

in the mollusc’s shell using its salivary papilla (McQuaid 1994; Mather and Nixon

1995). When feeding on large mussels, octopus may give a single cursory pull and

immediately resort to drilling, adopting a strategy that minimises the time required to

successfully overcome the prey item (McQuaid 1994).

A1.4.2. Diel feeding activity

As opportunists, cephalopods show a degree of flexibility in the times they are most

active in response to their surrounding environment (Meisel et al. 2003). While some

species are characterised as typically either diurnal, nocturnal, or crepuscular feeders,

individuals may undertake foraging excursions outside these periods (Mather 1991a;

Forsythe and Hanlon 1997). In some cases they are active at different times in different

locations. For example, O. vulgaris has a distinct nocturnal pattern of foraging activity

in the Mediterranean Sea and in the laboratory (Altman 1967; Kayes 1974; Brown et al.

102

2006). However, O. vulgaris off the coast of Bermuda in the Atlantic Ocean is very

active during the day (Mather 1988). This contrast may be due to prey availability,

minimising the risks of predation, or a strategy to reduce the potential for interspecific

competition. Thus, individual populations of the same species may utilise different

ecological cues to maximise their foraging opportunities when prey is most abundant

(Mather 1988; Meisel et al. 2003). Resource partitioning has been demonstrated in three

co-occurring Hawaiian octopuses, Octopus cyanea (diurnal), Octopus hawaiians

(crepuscular), and Octopus ornatus (nocturnal, Houck 1982). Although these species are

found within close proximity (~ 30 m) of each other on shallow reef flats, the temporal

partitioning of their foraging patterns appears to be an effective mechanism for reducing

interspecific interactions (Houck 1982; Meisel et al. 2003). Distinct differences in diel

activity, as well as habitat preferences may also result in dietary variation between

species (i.e., resource partitioning), since different prey are likely to be active at

different times within a 24-hour period (Houck 1982).

A1.4.3. Shoaling

The behavioural adaptation of shoaling is commonly observed among the squids and is

likely to have evolved as a mechanism for reducing predation whilst increasing an

individual’s feeding success (Neill and Cullen 1974; Krause 1994; Boyle and Rodhouse

2005). Although cooperative hunting has not been documented among the cephalopods

(Hanlon and Messenger 1996). Shoaling varies greatly over temporal and spatial scales,

but can have significant impacts on local prey densities (Boyle and Rodhouse 2005).

Sepioteuthis sepioidea is a particularly interesting example of a shoaling species.

They are highly gregarious and form large, closely packed groups that comprise

juveniles, sub-adults, and mature adults of both sexes, which congregate near the

substrate; a situation that is unique among the cephalopods (Moynihan and Rodaniche

1982; Hanlon and Messenger 1996). When exposed to a potential threat, all members of

the group turn to face it, “close ranks and assume stereotyped line and crescent

formations” (Moynihan and Rodaniche 1982). With the largest individuals positioned in

103

the wings or at the front of a progressively diminishing diagonal line, they advance

toward the threat (Moynihan and Rodaniche 1982). It has been suggested that shoaling

could increase observational learning within a population, fostering group traditions,

which may be beneficial to short-lived species (Moynihan and Rodaniche 1982; Hanlon

and Messenger 1996).

A1.4.4. Prey selection

Cephalopods are voracious active predators that hunt exclusively live prey (Boyle &

Rodhouse, 2007). They are flexible carnivores, able to exploit a wide range of prey

types (Boyle & Rodhouse, 2007; Nixon, 1987). Whilst foraging, an active predator can

opt to consume a potential prey item as it is encountered, or disregard it in favour of

alternative prey (McQuaid 1994). The response of a predator to a potential prey item

depends on past experience with that prey type, the characteristics of the prey (i.e., prey

defence mechanisms), the calorific value of the prey, and its own prey preferences

(Shettleworth et al. 1993; Darmaillacq et al. 2004).

Optimal foraging theory suggests that a predator should select prey that

maximises fitness, with behaviours that optimise net energy intake conferring a

selective advantage (Hughes 1980; McQuaid 1994; Krebs 2008). Prey selection should,

therefore, be based on a balance between net energy gain and handling time (Hughes

1980; Anderson et al. 2008b). However, due to competition, the risk of predation, or the

requirement for specific resources, foraging behaviours may deviate from those

considered to be optimal (Hughes 1980). In theory, a predator should only pursue a

potential prey if it is unlikely that the predator would be able to secure a prey of higher

calorific value in the same time that it takes to capture and ingest the lower value prey

item (Hughes 1980). This predicts that high-value prey will be consumed whenever the

opportunity arises, with diet composition expanding to include other less valuable prey

when high-value prey is scarce, or when specific nutrients are sought (Hughes 1980).

Thus, the diet of these actively searching mobile predators likely represents a

104

compromise between prey availability and prey preferences, with the dietary

composition for any particular species likely to vary in space and time (Ambrose 1984).

Cephalopods are frequently described as opportunistic or generalist predators

due to the broad range of prey items found in their diet, and the differences in the diet of

conspecifics between geographical regions (Rodhouse and Nigmatullin 1996; Boyle and

Rodhouse 2005). However, some species appear to be much more selective in their

feeding than others, providing evidence that cephalopods may actively seek specific

prey types and focus on specific prey characteristics (e.g., items of a particular size,

Ambrose 1984; Ambrose 1986; Kojima 1992; McQuaid 1994; Anderson et al. 2008).

This contrasts with the view of Rodhouse and Nigmatullin (1996), who suggest that

few, if any, are selective in their choice of prey.

Binary-choice laboratory experiments reveal that the California Two Spot

Octopus, Octopus bimaculatus, has a strong preference for crabs, with individuals

ignoring all other prey types until crabs have been depleted (Ambrose 1984). However,

in the subtidal habitats surrounding Santa Catalina Island in California, U.S.A.,

O. bimaculatus feeds mainly on gastropods (> 70%), with crustaceans forming just 5%

of the diet since crustaceans are present at very low densities (Ambrose 1984). These

findings demonstrate that while octopuses have strong prey preferences, their diet can

be limited by prey availability, but they will consume their preferred prey types

whenever they are encountered (Ambrose 1984).

Octopus vulgaris in the southern Caribbean exhibit marked prey selectivity, with

strong differences detected between individuals (Anderson et al. 2008b). A total of

seventy-five prey taxa were identified from 38 octopus dens, however, the remains

surrounding individual dens varied significantly, even when dens were separated by a

short < 50 m distance (Anderson et al. 2008). For example, the middens adjacent to the

den of one individual consisted exclusively of a single species of bivalve, Pinna carnea,

while the midden at a den located only 50 m away contained the remains of six taxa:

three crustacean, two bivalve, and one gastropod species (Anderson et al. 2008b). These

105

small-scale spatial differences may be the result of learnt behaviours, with octopus

targeting the species that succumb to their foraging technique most easily (Anderson et

al. 2008b). The fact that octopuses can be ‘specialising generalists’ may enable

individuals to capitalise on local prey abundance and maximise energy intake, with

foraging efficiency and learning leading to increased specialisation (Anderson et al.

2008b).

Selective predation of abalone and bivalves by octopus based on size provides

evidence for the selection of prey that maximises energy intake. Kojima (1992)

examined the predation of octopus, mainly O. vulgaris, on the abalone Haliotis discus

discus by examining shells for boreholes. Octopus predation increased with shell size

from 2.5% of small shells (30-39 mm shell width [SW]) to 44% of large shells (100-109

mm SW), which is likely driven by prey profitability (Kojima 1992). While the direct

removal of H. discus discus was unable to be quantified, it is possible that smaller

individuals could simply be pulled from the reef by octopus, without the need to bore

into the shell. Shells less than 30 mm in length comprised only 16% of the total number

of shells. However, since the direct removal of abalone from the substratum is likely to

be less energetically costly than drilling, octopus may predate on juvenile abalone when

the opportunity arises. Juvenile abalone are cryptic and reside in interstitial reef cavities

(Shepherd and Turner 1985; Prince et al. 1988, 2008), yet foraging for prey in crevices

using chemotactile exploration (Mather 1991a), which may allow octopuses to detect

and prey on smaller abalone.

Experiments on captive held O. vulgaris in South Africa, which feed mainly on

the mussel Perna perna, demonstrate a non-random selection of mussels when offered

an excess of prey (McQuaid 1994). Large octopus selected larger mussels with higher

energetic returns, calculated by dividing the energetic value of the prey, (i.e., mussel

flesh dry mass), by the total handling time (i.e., the amount of time required to consume

the prey item, McQuaid 1994). Larger mussels have a higher energetic value, however,

shell thickness is greater and the diameter of the adductor mussels holding the shell

106

closed is stronger, making larger mussels more energetically costly for smaller octopus

to overcome (McQuaid 1994). Therefore, it is not surprising that prey size selection

varies among octopuses based on body size, with larger octopuses targeting larger

mussels, while smaller individuals prefer smaller mussels (McQuaid 1994).

A1.5. Growth rates and food intake

Cephalopod growth has been modelled as a two-step process, characterised by an initial

exponential growth phase whereby the relative instantaneous growth rates remain

constant, followed by a logarithmic phase in which growth slows gradually and

concludes with sexual maturation, spawning, and death (Forsythe and Van Heukelem

1987; Semmens et al. 2004; Boyle and Rodhouse 2005). Although this two-step growth

phase is less evident in the data collected from field studies, probably due to a lack of

data on the difficult to sample small juveniles and paralarvae (Forsythe and Van

Heukelem 1987; Semmens et al. 2004; Boyle and Rodhouse 2005).

Laboratory studies demonstrate the rapid non-asymptotic rates at which

cephalopods grow (i.e., 3-10% of body weight d-1), and the approximately linear

relationship that exists between daily food intake and growth when food is provided in

excess (Joll 1977; Lee 1995; Semmens et al. 2004). In a laboratory study of the

relationship between food intake and growth in Octopus cf. tetricus, one individual

gained ~1,350 g in a period of 69 days (19.6 g d-1), increasing in size from ~750 to

2,100 g, while another gained ~800 g in 45 days (17.8 g d-1), increasing from ~500 to

1,300 g (Joll 1977).

Temperature and feeding rates are key factors influencing growth, particularly in

the early growth phase, where elevated temperatures promote higher growth rates (Joll

1977; Forsythe 2004). The Forsythe Hypothesis postulates that monthly cohorts that

experience gradually warming temperatures grow more rapidly than cohorts that hatch

in cooler conditions (Forsythe 2004). This hypothesis, which has received strong

endorsement (Jackson et al. 1997; Villanueva 2000; Hatfield et al. 2001; Forsythe et al.

107

2001; Vidal et al. 2002), predicts that elevated temperatures will accelerate cephalopod

lifecycles. Therefore anthropogenic climate change is likely to have a positive influence

on cephalopod biomass within the global oceans (Forsythe 2004). At least for as long as

the optimal thermal range of a species is not exceeded (Doubleday et al. 2008).

A1.6. Diets A1.6.1. Broad dietary trends

In order to determine the dietary composition of coastal cephalopods, a meta-analysis of

24 species across 28 studies was conducted encompassing species within the

Loliginidae, Octopodidae, Sepiidae and Sepiloidae. To ensure the consistency of

taxonomic resolution between families, each prey item was allocated to one of seven

main dietary groups, namely bivalves, cephalopods, crustaceans, gastropods,

polychaetes, teleosts, and ‘other’ prey, which includes taxonomic categories outside

these groups or prey that could not be identified. Dietary composition was expressed as

percent number (%N, i.e., the frequency of an individual prey item, expressed as a

percentage of all prey items). Note that in a few cases, prey items were grouped into one

of the seven main dietary groups when higher taxonomic resolution was presented.

While dietary information is limited to a few common species, such as Loligo

forbesi, Loligo vulgaris, Octopus vulgaris, Sepia elegans, and Sepia officinalis,

conspicuous trends were identified among families (Table A1.1). Crustaceans, teleosts,

and molluscs including other cephalopods typify the diet of these coastal families.

However, the relative contributions of the various prey taxa varied considerably among

the individual families. For example, crustaceans, teleosts, and shelled molluscs, such as

bivalves and gastropods, dominate the diets of benthic octopuses (Table A1.1).

Typically, the demersal sepiolids and sepiids consume large amounts of crustaceans and

teleosts, while the semi-pelagic loliginid squids tend to consume large proportions of

fishes, cephalopods, and crustaceans (Table A1.1).

Table A1.1. Percentage contribution (% number of total prey items) of the seven main prey groups in the gastric tracts of species within the Octopodidae, Sepiidae, Sepiloidae, and Loliginidae, and midden counts within the Octopodidae. B = bivalves, CE = cephalopods, crustaceans (CR), gastropods (G), polychaetes (P), teleosts (T), and (O) other prey. N (sample size) refers to the number of gastric tracts examined that contained food, or the number of middens counted. Location = broad geographic region of the study. Lifestyle = position in the water column adults of the species occupy. Family Species B CE CR G P T O N Location Lifestyle

Mid

dens

Oct

opod

idae

Enteroctopus dofleini1 19 77 1.22 2.78 193 Gulf of Alaska Benthic Octopus bimaculatus2 5.83 6.7 83.5 3.97 North Pacific Benthic Octopus bimaculatus2 6.7 5.03 73.18 15.09 North Pacific Benthic Octopus bimaculatus3 62.5 1.3 32.5 0.6 3.1 261 Gulf of California Benthic Octopus cyanea4 47 32 17 4 56 South Pacific Benthic Octopus insularis5 17.53 70 12.47 117 South East Atlantic Benthic Octopus insularis6 15.8 68.4 13.2 2.6 12 South East Atlantic Benthic Octopus insularis6 83.4 0.2 6.6 9.4 0.4 72 South East Atlantic Benthic Octopus insularis6 2.4 97.6 16 South East Atlantic Benthic Octopus vulgaris7 72 19 9 7 Balearic Sea Benthic Octopus vulgaris8 51 30 19 38 Caribbean Sea Benthic Octopus vulgaris9 30.63 21.74 47.59 0.04 41 North West Atlantic Benthic Octopus vulgaris10 12.8 12.1 71.5 3.6 245 South Atlantic Benthic

Gas

tric

trac

ts

Oct

opod

idae

Eledone cirrhosa11 5.98 70.64 0.13 2.09 17.73 3.43 498 North West Atlantic Benthic Eledone moschata12 0.6 10.8 57.1 2.1 1.3 28 0.1 799 North Adriatic Sea Benthic Octopus bimaculatus3 5.3 3.6 51.5 19.2 1.5 4.8 14.1 261 Gulf of California Benthic Octopus hubbsorum13 1 3 88 4 2 2 226 North East Pacific Benthic Octopus magnificus14 2.41 0.81 62.77 2.41 6.6 23.83 1.17 102 South West Atlantic Benthic Octopus maorum15 2.68 78.19 18.12 1.01 72 Tasman Sea Benthic Octopus mimus16 9.09 67.29 12.62 11 485 South Pacific Benthic Octopus vulgaris10 51.6 28.6 8.2 8.5 3.1 250 South Atlantic Benthic

109

1Vincent et al. (1998); 2Ambrose (1984); 3Villegas et al. (2014); 4Scheel et al. (2016); 5Leite et al. (2009); 6Leite et al. (2016); 7Ambrose and Nelson (1983); 8Anderson et al. (2008); 9Kuhlmann and McCabe (2014); 10Smith 2003; 11Regueira et al. (2017); 12Šifner and Vrgoč (2009); 13Lopez-Uriarte and Rios-Jara (2009); 14Villanueva (1993); 15Grubert et al. (1999); 16Cortez et al. (1995); 17Hunsicker and Essington (2006); 18Dos Santos and Haimovici (2002); 19Collins (1994); 20Pierce et al. (1994); 21Wangvoralak et al. (2011); 22Gasalla et al. (2010); 23Coelho et al. (1997); 24Sauer and Lipiński (1991); 25Guerra (1985); 26Castro and Guerra (1990); 27Vafidis et al. (2009); 28Alves et al. (2006).

Table A1.1 continued Family Species B CE CR G P T O n Location Lifestyle

Gas

tric

trac

ts

Lol

igin

idae

Doryteuthis pealeii17 49.45 27.73 22.82 1383 North West Atlantic Pelagic Doryteuthis sanpaulensis18 6.4 23.3 36.4 33.9 313 South West Atlantic Pelagic Loligo forbesi19 7.49 26.4 73.7 1715 Irish, Celtic Seas Pelagic Loligo forbesi20 6.9 17.35 0.13 75.62 1322 South East Atlantic Pelagic Loligo forbesi21 0.43 7.54 20.78 3.2 68.04 0.01 360 Northern North Sea Pelagic Loligo plei22 7.5 36.9 7.5 48.1 264 South West Atlantic Pelagic Loligo vulgaris23 19 7.2 72.7 1.1 280 North East Atlantic Semi-pelagic Loligo vulgaris23 25.6 19.8 51 3.6 346 North East Atlantic Semi-pelagic Loligo vulgaris20 18.51 15.22 6.87 59.4 1837 South East Atlantic Semi-pelagic Loligo vulgaris24 41.91 22.11 7.26 22.77 5.95 284 South East Atlantic Semi-pelagic

Sepi

idae

Sepia elegans25 80 20 72 North West Atlantic Demersal Sepia elegans26 83 0.2 15.1 1.7 366 North West Atlantic Demersal Sepia elegans27 1.64 80.99 3.31 12.4 1.66 41 Aegean Sea Demersal Sepia officinalis25 1.8 67.5 30.7 150 North West Atlantic Demersal Sepia officinalis26 0.9 57.1 0.3 39.7 2 623 North West Atlantic Demersal Sepia officinalis28 11.89 12.52 46.43 2.06 1.59 25.52 339 North West Atlantic Demersal Sepia officinalis27 62.26 5.66 32.08 24 Aegean Sea Demersal Sepia orbignyana27 6.35 77.78 1.59 14.28 27 Aegean Sea Demersal

Sepi

olid

ae Rondeletiola minor27 55 15 25 5 14 Aegean Sea Benthic

Rossia macrosoma27 1.92 80.77 1.92 1.92 13.46 0.1 18 Aegean Sea Benthic Sepietta intermedia27 5 75 15 5 15 Aegean Sea Benthic Sepietta neglecta27 62.5 18.75 18.75 14 Aegean Sea Benthic Sepietta oweniana27 59.46 6.76 16.22 17.56 25 Aegean Sea Benthic Sepietta robusta27 7.41 88.89 3.7 14 Aegean Sea Benthic

110

A1.6.2. Ontogenetic dietary shifts

Large changes in diet with ontogeny have been well documented among the

cephalopods; a reflection of the changes that occur in the brachial crown with growth

(Boyle and Rodhouse 2005). While the minimum prey length typically does not change,

often the maximum prey size and diversity of prey increases with increasing body size

(Smith 2003; Boyle and Rodhouse 2005). As is the case with many marine predators,

including their teleost competitors, larger individuals are better equipped to consume a

wider range of prey types (Gosline 1996; Consoli et al. 2010).

The potential scale of ontogenetic changes in diet are shown by O. vulgaris in

False Bay South Africa, with small individuals (0-300 g) frequently consuming large

numbers of small crustaceans (e.g., amphipods and isopods), and moderate numbers of

winkles and large crustaceans (e.g., crabs and lobsters; Smith 2003). As individuals

increase in size (301-1,000 g), the number of small crustaceans and winkles consumed

decreases and the number of larger prey items, such as large crustaceans, abalone, and

teleosts increases (Smith 2003). The diet of the largest individuals (>1,000 g) is most

diverse, with high numbers of abalone, teleosts, small and large crustaceans, and other

octopuses (Smith 2003).

Among the squids, piscivory increases markedly with growth, while the

importance of crustaceans and cephalopods appears to be variable, likely due to

temporal and spatial differences in prey availability (Table A1.1; Pierce et al. 1994;

Coelho et al. 1997; Gasalla et al. 2010).

A1.6.3. Cannibalism

Cannibalism and interspecific predation are a ubiquitous feature among cephalopods

(Table A1.1; Ibáñez and Keyl 2010). Cannibalism is most frequent during periods of

low food abundance and under high population densities, such as large squid

aggregations and when the territories of benthic octopuses are reduced, leading to

increased encounters (Aronson 1986; Boyle and Rodhouse 2005; Ibáñez and Keyl

111

2010). The rare phenomena of sexual cannibalism has also been noted for Octopus

cyanea (Hanlon and Forsythe 2008). In the majority of interactions, however,

differences in intra-cohort body size appears to be an important feature, with incidents

of cannibalism typically increasing in larger individuals.

A1.6.4. Influence of sampling method on interpretation of diets

Until recently, the primary method used to understand the trophic role and dietary

composition of cephalopods was gastric tract analysis and midden counts, for benthic

octopuses (e.g., Nixon 1987; Mather 1991a; Smith 2003; Jackson et al. 2007) (Table

A1.1). Gastric tract analysis is based mainly on the examination of hard structures, such

as bones, otoliths, gill lamellae, polychaete setae, cephalopods lenses and beaks, and

crustacean exoskeleton to identify prey to at least order for invertebrates, but often

family or species for vertebrates (Nixon 1987; Boyle and Rodhouse 2005; Jackson et al.

2007). Soft tissue components can be under-represented in studies based on gastric

tracts due to the metabolism of cephalopods being rapid (Pierce et al. 1994; Smith 2003;

Boucher-Rodoni and Mangold 2009). Midden counts are based on the collection of

shells and the remains of any prey discarded by octopus around their dens.

In order to determine whether dietary composition varies between sampling

methodologies, data was collected from studies utilising gastric tracts and midden

counts (Table A1.1). The Octopodidae are the only coleoid family that discards their

prey in middens close to their dens, therefore only species belonging to this family were

included in this analysis. The selected dietary data from Table A1.1 (i.e., percentage

contribution by numbers) were square root transformed and used to construct a Bray-

Curtis resemblance matrix. This matrix was, in turn, subjected to one-way Analysis of

Similarities (ANOSIM; Clarke and Green, 1998) test to determine whether dietary

composition differed significantly between methods. It was also used to produce a non-

metric multidimensional scaling plot (nMDS; Clarke, 1993) to provide a visual

representation of the similarity among samples, and the dietary items responsible for

any differences among a priori groups using shade plots (Clarke et al 2014).

112

One-way ANOSIM demonstrated that the diet composition of the Octopodidae

differed significantly between the two methodologies (R = 0.69, P = 0.001). This is

shown on the associated nMDS plot, with the points representing diet derived from

gastric tracts forming a discrete group on the right hand side of the plot, clearly

separated from the midden counts on the left hand side (Fig. A1.4). The points from the

midden counts were also more widely dispersed, indicating greater variation in the diet

than those from gastric tracts (Fig. A1.4). Hard parts in middens can accumulate over

time which may account for the greater variation in items found in midden contents.

Gastric tract analysis shows that crustaceans and teleosts are most important to the diets

of octopuses, but for midden counts hard-shelled prey such as bivalves and gastropods

are most important (Fig. A1.5).

This analysis demonstrates that sampling method can lead to dramatically

different results, even within a species, highlighting the limitations of each technique.

For example, the gastric tracts of O. vulgaris contained a high percentage by number

(%N) of crustaceans (51.6%), teleosts (8.5%) and polychaetes (8.2%) in their gastric

tracts, but a lower percentage of shelled molluscs (28.6%, Table A1.1; Smith 2003). In

contrast, middens contained a high %N of shelled molluscs (84.3%), and a low %N of

crustaceans (12.1%), teleosts, and polychaetes (< 3.6% total, Table A1.1; Smith 2003).

Hard parts of prey such as exoskeletons, bones, and setae have a higher gut

retention time than soft tissues, and thus crustaceans, teleosts, and polychaetes are likely

to be over-represented in gastric tract analysis (Pierce et al. 1994; Smith 2003).

Conversely, midden counts are heavily biased towards shelled molluscs, and against

soft-bodied organisms. In dynamic shallow-water environments, lightweight prey

discards, such as fish skeletons are readily lost from midden piles, and small, soft-

bodied organisms that are consumed whole leave behind no trace, resulting in an over-

estimation of the importance of shelled molluscs when the remains from middens are

counted (Ambrose and Nelson 1983; Mather 1991a; Smith 2003).

113

Figure A1.4. Non-metric multi-dimensional scaling (nMDS) ordination plot, derived from a Bray–Curtis similarity matrix of the square-root transformed percentage contribution (% number of total prey items) of the seven main prey groups in the gastric tracts (p) and midden piles (q) of octopus species.

Figure A1.5. Shade plot showing the mean square-root transformed percentage contribution (% number of total prey items) in the gastric tracts (p) and midden piles (q) of octopus species. White represents an absence of prey, and grey to black represents the increasing contribution of a dietary category to the total diet. Prey groups (x-axis) are bivalves (B), cephalopods (CE), crustaceans (CR), gastropods (G), polychaetes (P), teleosts (T), and (O) other prey.

While gastric tract analysis typically uncovers a greater range of prey taxon, this

approach can result in biases in diet composition, particularly when fragments of large

prey are discarded and small prey are consumed whole (Table A1.1; Nixon 1987; Boyle

114

and Rodhouse 2005). Furthermore, prey types with few hard parts such as flatworms, or

molluscs where the hard shell has been discarded, may go undetected altogether (Nixon

1987).

Biochemical measures of diet, such as stable isotope analysis and fatty acid

trophic markers provide complementary techniques to track the origins and pathways of

organic molecules to understand the trophic role of cephalopods within the food web

(Graeve et al. 2002; Jackson et al. 2007). When food is digested by a consumer,

nutrients and distinct fatty acid ratios are assimilated into the animal’s tissues producing

unique isotopic and fatty acid signatures (Petersen and Fry 1987; Graeve et al. 2002;

Jackson et al. 2007; Navarro et al. 2013). These signatures provide precise information

with respect to trophic positioning (i.e., an organisms position within a food web), and

the trophic pathways within a marine ecosystem (Navarro et al. 2013). They are also

useful in that they offer spatial and temporal insights into trophic relationships, as well

as providing information on long-term dietary habits of cephalopods, helping alleviate

the bias associated with rapid digestion and the selective removal of hard parts (Layman

et al. 2012; Navarro et al. 2013).

A1.6.5. Dietary differences among families

The information from the 24 species representing the four families detailed in Table

A1.1 (i.e., members of the Loliginidae, Octopodidae, Sepiidae, and Sepiloidae) was

subjected to the same suite of multivariate techniques described above (i.e., nMDS,

ANOSIM, and shade plots).

One-way ANOSIM demonstrated that the dietary composition of species in each

of the four families differed significantly (R = 0.49, P = 0.001). Pairwise comparisons

between families demonstrated that the dietary compositions of the loliginid squids

differed significantly from each of the other families (P = 0.001). This is shown clearly

on the nMDS plot where the points for the loliginid squid form a discrete group on the

right-hand side of the plot, distinct from the other families on the left-hand side (Fig.

115

A1.6). While less marked with some overlap between the groups, significant (P ≤ 0.04)

differences were detected between each of the other families except between the

Octopodidae and the Sepiloidae (Fig. A1.6, Table A1.2).

Figure A1.6. nMDS ordination plot derived from a Bray–Curtis similarity matrix of the square-root transformed percentage contribution (% number of total prey items) of the seven main prey groups in the gastric tracts of species within the Octopodidae (p), Sepiidae (q), Sepiloidae (¢), and Loliginidae (¿).

Non-metric MDSStandardise Samples by TotalTransform: Square rootResemblance: S17 Bray-Curtis similarity

FamilyOctopodidaeSepiidaeSepiolidaeLoliginidae

2D Stress: 0.18

116

Table A1.2. Pairwise R statistics and significance level (P) values derived from a one-way ANOSIM test on the square-root transformed percentage contribution (% number of total prey items) of the seven main prey groups in the gastric tract of species within the Octopodidae, Sepiidae, Sepiloidae, and Loliginidae. Grey shading indicates no significant difference between families (P = > 0.05).

Groups R statistic P value Octopodidae vs Sepiidae 0.17 0.040 Octopodidae vs Sepiolidae 0.06 0.213 Octopodidae vs Loliginidae 0.83 0.001 Sepiidae vs Sepiolidae 0.23 0.021 Sepiidae vs Loliginidae 0.82 0.001 Sepiolidae vs Loliginidae 0.93 0.001

The shade plot shows that the loliginid squids consume larger proportions of

teleosts and cephalopods (mainly other squids), with few bivalves or polychaetes

compared with the other families (Fig. A1.7). Crustaceans are the most important

dietary component in the octopodids, sepiids and sepiloids, followed by teleosts, while

the relative importance of each of the remaining prey categories varies among families.

These differences in diet composition among families are likely to be a

reflection of lifestyle (i.e., the position they occupy within the water column), and

resource availability. Not surprisingly, benthic octopuses consume large numbers of

crustaceans (e.g., brachyurans), teleosts, gastropods, and bivalves as these prey items

are typically abundant in benthic environments (Fig. A1.7). Similarly, the low

contributions of benthic prey, such as shelled molluscs and polychaetes, and high %N of

teleosts, crustaceans (e.g., prawns), and other cephalopods (e.g., other squids) to the diet

of the Loliginidae is representative of their semi-pelagic lifestyle (Fig. A1.7). The

presence of both demersal and pelagic fishes in the diets of the loliginids indicates,

however, that they feed throughout the water column (Pierce et al. 1994; Dos Santos

and Haimovici 2002; Gasalla et al. 2010).

117

Figure A1.7. Shade plot showing the mean square-root transformed percentage contribution (% number of total prey items) of the seven main prey groups, bivalves (B), cephalopods (CE), crustaceans (CR), gastropods (G), polychaetes (P), teleosts (T), and (O) ‘other’ prey, in the gastric tract of members of the Octopodidae (p), Sepiidae (q), Sepiloidae (¢), and Loliginidae (¿). White represents an absence of prey, and grey to black represents the increasing contribution of a dietary category to the total diet.

A1.7. Predation in marine systems A1.7.1. Predation impacts

Predation is a fundamental ecological process controlling the structure and functioning

of all ecosystems on earth. It has a negative impact on prey survival and abundance, and

is distinguished from other forms of foraging behaviour by the disfigurement or

complete elimination of another animal (Curio 1976). Some predators (i.e., keystone

species), play key roles within an ecosystem and are essential for maintaining species

richness and community organisation (Underwood and Fairweather 1989; Gasalla et al.

2010). The ability of cephalopods to adapt to their environment and exert a strong

influence on marine communities has embedded their description as a keystone species

within the literature (Boyle and Rodhouse 2005; Gasalla et al. 2010; Rosa et al. 2012;

Pierce and Portela 2014). Rapid growth coupled with high metabolic rates has resulted

in flexible yet voracious feeding regimes, and their opportunistic, plastic life-history

strategy enables them to respond to environmental fluctuations rapidly, which can have

significant impacts on both the higher and lower trophic levels in the food web

G

B

O

CR

P

T

CE

0123456789 Family

OctopodidaeSepiidaeSepiolidaeLoliginidae

118

(Rodhouse and Nigmatullin 1996; Boyle and Rodhouse 2005; André et al. 2010). As a

consequence, their predatory impact on prey populations can vary both spatially and

temporally (Rodhouse and Nigmatullin 1996), and their large biomass within the global

oceans is likely to have a significant impact on the teleost, crustacean, and molluscan

prey that form the great majority of their diet (Nixon 1987; Table A1.1).

Despite being referred to as keystone species, few studies have quantified

experimentally the scale of cephalopod impacts on the ecological function of coastal

systems. Recent ecological modelling has utilised a keystoneness index (sensu Libralato

et al. 2006) and this highlights the importance of squid trophic interactions in the

Southern Brazil Bight ecosystem (Gasalla et al. 2010). The predatory impact of

D. gigas on the Californian sardine S. sagax caerulea, as previously mentioned,

provides an illustration of the potential scale a cohort can exert on a prey population;

and although a mesopelagic species, it highlights the potential top-down controls that

high turnover coastal species pose in these dynamic shallow water environments.

A study of O. vulgaris, commonly found in the intertidal and sub-tidal

communities off southern California, showed that when present in high densities, it

dramatically reduces the density and diversity of motile invertebrate prey (Ambrose

1986). However an 80% reduction in O. vulgaris density over a five year period

resulted in a five-fold increase in prey abundance (Ambrose 1986). Unlike other

intertidal communities, where species richness has been positively correlated with

predation pressure (e.g., Paine 1966; Peterson et al. 1979), high octopus density appears

to have a negative impact on species richness (Ambrose 1986). In this system, the

absence of a competitive dominant may explain why predation did not result in

increased species richness (Ambrose 1986). Additionally, being a motile invertebrate

community may reduce the effectiveness of predation, since organisms are able to

actively seek refuge and move to more favourable locations (Ambrose 1986). Several

species of crabs, shrimps, and bivalves remained at relatively low densities throughout

the duration of the study, which is thought to have been driven by selective predation,

119

leading to a reduction in the abundance of preferred species despite octopus occurring in

low numbers (Ambrose 1986).

A1.7.2. Trophic relationships

Cephalopods are considered important intermediaries within the marine food web,

facilitating a transfer of energy between first and second order consumers, and higher

trophic level predators at a range of stages and sizes throughout their lifecycle (Dos

Santos and Haimovici 2002; Boyle and Rodhouse 2005). Highly muscular and offering

a rich source of protein (> 90% of the live weight), little of the cephalopod body is

resistant to digestion, providing a high intrinsic food value for consumers (Semmens et

al. 2004; Boyle and Rodhouse 2005). However, the indigestible hard parts, particularly

the chitinous beak, provide an important tool for disentangling trophic relationships and

understanding their importance within the marine food web (Boyle and Rodhouse

2005). Analysis of the stomach contents of cephalopod predators highlights their global

importance within the diets of many vertebrate species (e.g., Clarke 1996a; Cherel and

Hobson 2005; Cherel et al. 2009; Staudinger et al. 2013). For example, global seabird,

pinniped, and cetacean populations are estimated to consume between 12.5 and 24

million tonnes annually (Staudinger et al. 2013), which is at least twice the biomass of

global cephalopod landings (Fig. A1.3). The high volume of paralarvae and juvenile

cephalopods (> 55%) in the diet of commercially valuable finfishes such as the

Yellowfin Tuna, Thunnus albacares, and Blackfin Tuna, Thunnus atlanticus, provides

evidence of the importance of cephalopods as trophic intermediaries, even from a small

size (Staudinger et al. 2013). As juveniles, cephalopods form an important food

component for many predatory fishes, however, as these molluscs increase in size their

relationships with other taxa shift, and the hunters become the hunted.

Cephalopods occupy a range of trophic levels, which differ considerably

between species, life-stage, location, and ecosystem type (Livingston 2003; Navarro et

al. 2013; Rodhouse et al. 2014). Analysis of trophic relationships highlights the diverse

nature of cephalopod feeding regimes and their ability to feed at multiple levels, as

120

evidenced by distinct isotopic nitrogen (𝛿15 N) signatures. It is these distinct 𝛿15 N

signatures that allow for the relative trophic position of a consumer to be characterised

(Vander Zanden and Rasmussen 1999; Layman et al. 2012). The mean (± 1 s.d.)

trophic level of squids determined from synthesized stable isotope data (and adjusted

for isotopic variability), in temperate coastal and shelf areas has been estimated at 3.61

± 0.76, slightly higher than those in tropical areas (3.23 ± 0.59; Navarro et al. 2013).

These results are broadly similar to that of Lozano-Montes et al. (2011) who found that

the mean trophic level of squid and octopus in temperate waters on the west coast of

Western Australia ranged from 3.1 to 3.6 (± 0.50). Navarro et al. (2013) found that the

mean trophic level of squid varied considerably among oceans, being greatest in

Antarctica (4.11 ± 0.99), followed by the Arctic (4.01 ± 0.49), Indian Ocean (3.29 ±

0.35), Pacific Ocean (3.46 ± 0.91), and least in the Mediterranean (2.65 ± 0.24).

Body size is an important feature of trophic positioning, with increased size

resulting in a progressive increase in 𝛿15N; denoting a shift in diet to higher trophic

levels with growth (Cherel et al. 2009; Ruiz-Cooley et al. 2010). Similarly, isotopic

carbon (𝛿15 C) ratios vary among primary producers, offering insights into the original

sources of dietary carbon and enabling variation in habitat use to be identified (Cherel et

al. 2009; Ruiz-Cooley et al. 2010). Although this method is commonly used to

determine trophic level, the analysis of dietary composition of coastal cephalopods has

not been determined using this approach.

A1.8. Conclusions and directions for future research

This review summarises the current knowledge of coastal cephalopod diets and the

characteristics of their biology that contribute to their success as predators within global

marine ecosystems. Their live fast, die young life history strategy commands voracious

feeding routines, which is supported by their ability to adapt to new and novel foraging

opportunities. With a range of highly sophisticated sensory and chromatophore organs,

coupled with a range of unique morphological accessories, such as prehensile arms with

strong suckers, a chitinous beak for tearing food, a tongue-like radula, salivary papilla

121

bearing tiny teeth, and some with cephalotoxins they are well-adapted predators.

Flexibility in foraging behaviours and the times an individual is most active enables the

cephalopods to strike a balance between energy intake and energy expenditure,

consistent with the principles of optimal foraging theory (e.g., Hughes 1980). They

show intelligence; a feature that is unique among the invertebrates, with the ability to

learn from past experiences; enhancing foraging efficiency and the opportunity for

specialisation on high energy content prey.

The methodology adopted in dietary studies was shown to significantly

influence the results of these studies. Multivariate analyses of coastal octopuses

demonstrated distinct differences between the dietary compositions determined from

gastric tract analysis and midden counts, with large volumes of prey taxa with hard parts

such as crustaceans and teleosts identified within gastric tracts, while middens are

heavily biased towards hard-shelled prey, such as bivalves and gastropods. The meta-

analysis of 24 species, representing four families of coastal cephalopods (Octopodidae,

Loliginidae, Sepiidae, and Sepiloidae) demonstrated the importance of crustaceans and

teleosts to the diet of all families, but also highlighted distinct differences between

families with respect to relative prey contributions, particularly for the semi-pelagic

Loliginidae. Such differences are likely a consequence of the variation in lifestyle that

occurs between the groups. A rapid shift in diet with ontogeny is a ubiquitous feature

among the cephalopods. An increase in body size facilitates the capture and ingestion of

a wider range of prey types, and the incorporation of larger, more mobile species such

as teleosts and other cephalopods, including their conspecifics, i.e., cannibalism.

Cephalopods play key trophic roles within marine systems, supporting the

transfer of energy between lower order consumers and higher trophic level predators.

They can exert considerable predatory impacts within a community, which is enhanced

by their ability to respond rapidly to favourable environmental perturbations such as

increased water temperatures that promote growth. They occupy a range of trophic level

122

positions, however, this varies considerably during the course of an individuals life

history and the ecosystem type in which they inhabit.

The trophodynamics of cephalopod communities are yet to be fully explored,

with most research limited to a few commercially valuable species, particularly pelagic

squids, which dominate the cephalopod landings from global fisheries (Rodhouse et al.

2014). Additionally, most of the work on cephalopod diets and trophic analysis is

restricted to a few commonly occurring and easily accessible species, with little

information known about the long-term dietary trends and nutrient assimilation in

coastal systems (Rodhouse et al. 2014). In view of the current unprecedented level of

exploitation of cephalopods, further studies that utilise biochemical tracers such as

stable isotopes and fatty acids, are needed to evaluate trophic interactions, diet

composition, cephalopod lifecycles and their response to environmental variability. This

information will help inform reliable assessments of cephalopods and other fished

species, and ensure that cephalopod fisheries are developed in an ecologically

sustainable way (Boyle and Rodhouse 2005; Jackson et al. 2007; Hunsicker et al. 2010;

Rodhouse et al. 2014). It will also provide the basis for developing more realistic

marine ecosystem models, and assessing the ecological role of cephalopods in the

marine environment.

123

A1.9. References

Altman, J. (1967). The behaviour of Octopus vulgaris Lam. in its natural habitat: a pilot study. Underwater Association Report 1966–1967, 77–83.

Alves, D. M., Cristo, M., Sendão, J., and Borges, T. C. (2006). Diet of the cuttlefish Sepia officinalis (Cephalopoda: Sepiidae) off the south coast of Portugal (eastern Algarve). Journal of the Marine Biological Association of the UK 86, 429. doi:10.1017/S0025315406013312

Amaratunga, T., and Caddy, J. F. (1983). The role of cephalopods in the marine ecosystem. Advances In Assessment Of World Cephalopod Resources 231, 379–415.

Ambrose, R. F. (1986). Effects of octopus predation on motile invertebrates in a rocky subtital community. Marine Ecology. Progress Series 30, 261–273.

Ambrose, R. F. (1984). Food preferences, prey availability, and the diet of Octopus bimaculatus Verrill. Journal of Experimental Marine Biology and Ecology 77, 29–44. doi:http://dx.doi.org/10.1016/0022-0981(84)90049-2

Ambrose, R. F., and Nelson, B. V (1983). Predation by Octopus vulgaris in the Mediterranean. Marine Ecology 4, 251–261. doi:10.1111/j.1439-0485.1983.tb00299.x

Anderson, R. C., Wood, J. B., and Mather, J. A. (2008). Octopus vulgaris in the Caribbean is a specializing generalist. Marine Ecology Progress Series 371, 199–202. doi:10.3354/meps07649

André, J., Haddon, M., and Pecl, G. T. (2010). Modelling climate-change-induced nonlinear thresholds in cephalopod population dynamics. Global Change Biology 16, 2866–2875. doi:10.1111/j.1365-2486.2010.02223.x

Aronson, R. B. (1986). Life history and den ecology of Octopus briareus Robson in a marine lake. Journal of Experimental Marine Biology and Ecology 95, 37–56. doi:10.1016/0022-0981(86)90086-9

Bell, J. D., Watson, R. A., and Ye, Y. (2016). Global fishing capacity and fishing effort from 1950 to 2012. Fish and Fisheries. doi:10.1111/faf.12187

Blaber, S. J. M., and Bulman, C. M. (1987). Diets of fishes of the upper continental slope of eastern Tasmania: content, calorific values, dietary overlap and trophic relationships. Marine Biology 95, 345–356. doi:10.1007/BF00409564

Boletzky, S. (1983). Sepia officinalis. In ‘Cephalopod life cycles. Species Accounts’. (Ed P. R. Boyle.) pp. 31–52. (Academic Press: London.)

Boucher-Rodoni, R., and Mangold, K. (2009). Experimental study of digestion in Octopus vulgaris(Cephalopoda: Octopoda). Journal of Zoology 183, 505–515. doi:10.1111/j.1469-7998.1977.tb04202.x

Boyle, P. R. (1990). Cephalopod biology in the fisheries context. Fisheries Research 8, 303–321. doi:10.1016/0165-7836(90)90001-C

Boyle, P. R., and Boletzky, S. (1996). Cephalopod Populations: Definition and Dynamics. Philosophical Transactions of the Royal Society B: Biological Sciences 351, 985–1002. doi:10.1098/rstb.1996.0089

Boyle, P. R., and Rodhouse, P. G. K. (2005). ‘Cephalopods: ecology and fisheries’. (Blackwell Science Ltd: Ames, Iowa.) doi:10.1002/9780470995310

Brown, E. R., Piscopo, S., De Stefano, R., and Giuditta, A. (2006). Brain and behavioural evidence for rest-activity cycles in Octopus vulgaris. Behavioural Brain Research 172, 355–359. doi:10.1016/j.bbr.2006.05.009

Budelmann, B. U. (1996). Active marine predators: The sensory world of cephalopods. Marine and Freshwater Behaviour and Physiology 27, 59–75. doi:10.1080/10236249609378955

Caddy, J. F., and Rodhouse, P. G. K. (1998). Cephalopod and groundfish landings:

124

Evidence for ecological change in global fisheries? Reviews in Fish Biology and Fisheries 8, 431–444. doi:10.1023/A:1008807129366

Carefoot, T. (2017). Octopuses and their relatives: Reproduction. Available at: http://www.asnailsodyssey.com/LEARNABOUT/OCTOPUS/octoEgg.php [accessed 14 April 2017]

Castro, B. G., and Guerra, Á. (1990). The diet of Sepia officinalis (Linnaeus, 1758) and Sepia elegans (D’Orbigny, 1835) (Cephalopoda, Sepioidea) from the Ria de Vigo (NW Spain). Scientia Marina 54, 375–388.

Cherel, Y., and Hobson, K. A. (2005). Stable isotopes, beaks and predators: a new tool to study the trophic ecology of cephalopods, including giant and colossal squids. Proceedings of The Royal Society B 272, 1601–7. doi:10.1098/rspb.2005.3115

Cherel, Y., Ridoux, V., Spitz, J., and Richard, P. (2009). Stable isotopes document the trophic structure of a deep-sea cephalopod assemblage including giant octopod and giant squid. Biology letters 5, 364–367. doi:10.1098/rsbl.2009.0024

Clarke, M. R. (1996a). Cephalopods as Prey. III. Cetaceans. Philosophical Transactions of the Royal Society B: Biological Sciences 351, 1053–1065. doi:10.1098/rstb.1996.0093

Clarke, M. R. (1996b). The Role of Cephalopods in the World’s Oceans: General Conclusion and the Future. Philosophical Transactions of the Royal Society B: Biological Sciences 351, 1105–1112. doi:10.1098/rstb.1996.0096

Coelho, M. L. (1985). Review of the influence of oceanographic factors on cephalopod distribution and life cycles. NAFO Scientific Council Studies 9, 47–57.

Coelho, M. L., Domingues, P., Balguerias, E., Fernandez, M., and Andrade, J. P. (1997). A comparative-study of the diet of Loligo vulgaris (Lamarck, 1799) (Mollusca, Cephalopoda) from the South Coast of Portugal and the Saharan Bank (Central-East Atlantic). Fisheries Research 29, 245–255. doi:10.1016/S0165-7836(96)00540-1

Coll, M., Navarro, J., Olson, R. J., and Christensen, V. (2013). Assessing the trophic position and ecological role of squids in marine ecosystems by means of food-web models. Deep-Sea Research Part II: Topical Studies in Oceanography 95, 21–36. doi:10.1016/j.dsr2.2012.08.020

Collins, M. (1994). Diet of the squid Loligo forbesi Steenstrup (Cephalopoda: Loliginidae) in Irish waters. ICES Journal of Marine Science 51, 337–344. doi:10.1006/jmsc.1994.1034

Consoli, P., Battaglia, P., Castriota, L., Esposito, V., Romeo, T., and Andaloro, F. (2010). Age, growth and feeding habits of the bluemouth rockfish, Helicolenus dactylopterus dactylopterus (Delaroche 1809) in the central Mediterranean (southern Tyrrhenian Sea). Journal of Applied Ichthyology 26, 583–591. doi:10.1111/j.1439-0426.2010.01467.x

Curio, E. (1976). ‘The ethology of predation’. (Springer-Verlag: Berlin; New York.) Darmaillacq, A.-S., Dickel, L., Chichery, M.-P., Agin, V., and Chichery, R. (2004).

Rapid taste aversion learning in adult cuttlefish, Sepia officinalis. Animal Behaviour 68, 1291–1298. doi:10.1016/j.anbehav.2004.01.015

Doubleday, Z. A., Pecl, G. T., Semmens, J. M., and Danyushevsky, L. (2008). Using stylet elemental signatures to determine the population structure of Octopus maorum. Marine Ecology Progress Series 360, 125–133. doi:10.3354/meps07389

Doubleday, Z. A., Prowse, T. A. A., Arkhipkin, A., Pierce, G. J., Semmens, J., Steer, M., Leporati, S. C., Lourenço, S., Quetglas, A., Sauer, W., and Gillanders, B. M. (2016). Global proliferation of cephalopods. Current Biology 26, R406–R407. doi:10.1016/j.cub.2016.04.002

Ehrhardt, N. M. (1991). Potential impact of a seasonal migratory jumbo squid

125

(Dosidicus gigas) stock on a Gulf of California sardine (Sardinops sagax caerulea) population. Bulletin of Marine Science 49, 325–332. doi:10.1017/CBO9781107415324.004

Estes, J. A., and Duggins, D. O. (1995). Sea otters and kelp forests in Alaska: Generality and variation in a community ecological paradigm. Ecological Monographs 65, 75–100. doi:10.2307/2937159

FAO (2016a). Globefish highlights. Available at: http://www.fao.org/3/a-i6592e.pdf FAO (2016b). The state of world fisheries and aquaculture. Available at:

http://www.fao.org/3/a-i5798e.pdf Finn, J. K., Tregenza, T., and Norman, M. D. (2009). Defensive tool use in a coconut-

carrying octopus. Current Biology 19. doi:10.1016/j.cub.2009.10.052 Forsythe, J. W. (2004). Accounting for the effect of temperature on squid growth in

nature: from hypothesis to practice. Marine and Freshwater Research 55, 331. doi:10.1071/MF03146

Forsythe, J. W., and Hanlon, R. T. (1997). Foraging and associated behavior by Octopus cyanea Gray, 1849 on a coral atoll, French Polynesia. Journal of Experimental Marine Biology and Ecology 209, 15–31. doi:10.1016/S0022-0981(96)00057-3

Forsythe, J. W., and Van Heukelem, W. F. (1987). Growth. In ‘Cephalopod life cycles. Comparative Reviews’. (Ed P. R. Boyle.) pp. 135–156. (Academic Press: London.)

Forsythe, J. W., Walsh, L. S., Turk, P. E., and Lee, P. G. (2001). Impact of temperature on juvenile growth and age at first egg-laying of the Pacific reef squid Sepioteuthis lessoniana reared in captivity. Marine Biology 138, 103–112. doi:10.1007/s002270000450

Gasalla, M. A., Rodrigues, A. R., and Postuma, F. A. (2010). The trophic role of the squid Loligo plei as a keystone species in the South Brazil Bight ecosystem. ICES Journal of Marine Science 67, 1413–1424. doi:10.1093/icesjms/fsq106

Ghiretti, F. (1959). Cephalotoxin: the crab-paralysing agent of the posterior salivary glands of cephalopods. Nature 183, 1192–1193.

Gosline, W. A. (1996). Structures associated with feeding in three broad-mouthed, benthic fish groups. Environmental Biology of Fishes 47, 399–405. doi:10.1007/BF00005053

Graeve, M., Kattner, G., Wiencke, C., and Karsten, U. (2002). Fatty acid composition of Arctic and Antarctic macroalgae: Indicator of phylogenetic and trophic relationships. Marine Ecology Progress Series 231, 67–74. doi:10.3354/meps231067

Grubert, M. A., Wadley, V. A., and White, R. W. G. (1999). Diet and feeding strategy of Octopus maorum in southeast Tasmania. Bulletin of Marine Science 65, 441–451.

Guerra, Á. (1985). Food of the cuttlefish Sepia officinalis and S. elegans in the Ria de Vigo (NW Spain) (Mollusca: Cephalopoda). Journal of Zoology 207, 511–519. doi:10.1111/j.1469-7998.1985.tb04947.x

Hanlon, R. (2007). Cephalopod dynamic camouflage. Current Biology 17, R400–R404. doi:10.1016/j.cub.2007.03.034

Hanlon, R. T., and Forsythe, J. W. (2008). Sexual cannibalism by Octopus cyanea on a Pacific coral reef. Marine and Freshwater Behaviour and Physiology 41, 19–28. doi:10.1080/10236240701661123

Hanlon, R. T., and Messenger, J. B. (1996). ‘Cephalopod behaviour’. (Cambridge University Press: Cambridge.)

Hatfield, E. M., Hanlon, R. T., Forsythe, J. W., and Grist, E. P. (2001). Laboratory testing of a growth hypothesis for juvenile squid Loligo pealeii (Cephalopoda: Loliginidae). Canadian Journal of Fisheries and Aquatic Sciences 58, 845–857.

126

doi:10.1139/f01-030 Heithaus, M. R., Frid, A., Wirsing, A. J., and Worm, B. (2008). Predicting ecological

consequences of marine top predator declines. Trends in Ecology and Evolution 23, 202–210. doi:10.1016/j.tree.2008.01.003

Houck, B. A. (1982). Temporal spacing in the activity patterms of three Hawaiian schallow-water octopods. The Nautilus 96, 152–156.

Hughes, R. N. (1980). Optimal foraging theory in the marine context. Annual Review of Marine Biology and Oceanography 18, 423–481.

Hunsicker, M. E., and Essington, T. E. (2006). Size-structured patterns of piscivory of the longfin inshore squid (Loligo pealeii) in the mid-Atlantic continental shelf ecosystem. Canadian Journal of Fisheries and Aquatic Sciences 63, 754–765. doi:10.1139/f05-258

Hunsicker, M. E., Essington, T. E., Watson, R. A., and Sumaila, U. R. (2010). The contribution of cephalopods to global marine fisheries: can we have our squid and eat them too? Fish and Fisheries 11, 421–438. doi:10.1111/j.1467-2979.2010.00369.x

Ibáñez, C. M., and Keyl, F. (2010). Cannibalism in cephalopods. Reviews in Fish Biology and Fisheries 20, 123–136. doi:10.1007/s11160-009-9129-y

Integration and Application Network (2017). Octopus. Available at: http://ian.umces.edu/imagelibrary/ [accessed 14 April 2017]

Jackson, G. D., Bustamante, P., Cherel, Y., Fulton, E. A., Grist, E. P. M., Jackson, C. H., Nichols, P. D., Pethybridge, H., Phillips, K., Ward, R. D., and Xavier, J. C. (2007). Applying new tools to cephalopod trophic dynamics and ecology: Perspectives from the Southern Ocean Cephalopod Workshop, February 2-3, 2006. In ‘Reviews in Fish Biology and Fisheries’. pp. 79–99 doi:10.1007/s11160-007-9055-9

Jackson, G. D., Forsythe, J. W., Hixon, R. F., and Hanlon, R. T. (1997). Age, growth, and maturation of Lolliguncula brevis (Cephalopoda: Loliginidae) in the northwestern Gulf of Mexico with a comparison of length-frequency versus statolith age analysis. Canadian Journal of Fisheries and Aquatic Sciences 54, 2907–2919. doi:10.1139/f97-192

Joll, L. M. (1977). Growth and Food Intake of Octopus Tetricus (Mollusca: Cephalopoda) in Aquaria. Marine and Freshwater Research 28, 45–56. doi:10.1071/MF9770045

Kayes, R. J. (1974). The daily activity pattern of Octopus vulgaris in a natural habitat. Marine Behaviour and Physiology 2, 337–343. doi:10.1080/10236247309386935

Kojima, H. (1992). Octopus predation on the abalone Haliotis discus discus. In ‘First International Symposium on Abalone Biology, Fisheries, and Culture. La Paz, Mexico 21-25 November 1989’. (Eds S. A. Shepherd, M. J. Tegner, and S. A. Guzman del Proo.) pp. 21–29. (Blackwell Publishing Ltd: Oxford.)

Krause, J. (1994). Differential fitness returns in relation to spatial position in groups. Biological Reviews 69, 187–206. doi:10.1111/j.1469-185X.1994.tb01505.x

Krebs, C. J. (2008). ‘The ecological world view’. (CSIRO Publishing: Collingwood, Vic.)

Kuhlmann, M. L., and McCabe, B. M. (2014). Diet specialization in Octopus vulgaris at San Salvador, Bahamas. Marine Ecology Progress Series 516, 229–237. doi:10.3354/meps11070

Layman, C. A., Araujo, M. S., Boucek, R., Hammerschlag-Peyer, C. M., Harrison, E., Jud, Z. R., Matich, P., Rosenblatt, A. E., Vaudo, J. J., Yeager, L. A., Post, D. M., and Bearhop, S. (2012). Applying stable isotopes to examine food-web structure: An overview of analytical tools. Biological Reviews 87, 545–562.

127

doi:10.1111/j.1469-185X.2011.00208.x Lee, P. G. (1995). Nutrition of cephalopods: Fueling the system. Marine and

Freshwater Behaviour and Physiology 25, 35–51. doi:10.1080/10236249409378906

Leite, T. S., Batista, A., Lima, F., Barbosa, J., and Mather, J. A. (2016). Geographic variability of Octopus insularis diet: from oceanic island to continental populations. Aquatic Biology 25, 17–27. doi:10.3354/ab00655

Leite, T. S., Haimovici, M., and Mather, J. A. (2009). Octopus insularis (Octopodidae), evidences of a specialized predator and a time-minimizing hunter. Marine Biology 156, 2355–2367. doi:10.1007/s00227-009-1264-4

Libralato, S., Christensen, V., and Pauly, D. (2006). A method for identifying keystone species in food web models. Ecological Modelling 195, 153–171. doi:10.1016/j.ecolmodel.2005.11.029

Livingston, R. J. (2003). ‘Trophic organisation in coastal systems’. (CRC Press: Boca Raton.)

Lopez-Uriarte, E., and Rios-Jara, E. (2009). Reproductive biology of Octopus hubbsorum (Mollusca: Cephalopoda) along the Central Mexican Pacific Coast. Bulletin of Marine Science 84, 109–121.

Lozano-Montes, H. M., Loneragan, N. R., Babcock, R. C., and Jackson, K. (2011). Using trophic flows and ecosystem structure to model the effects of fishing in the Jurien Bay Marine Park, temperate Western Australia. Marine and Freshwater Research 62, 421–431. doi:10.1071/MF09154

Marshall, N. J., and Messenger, J. B. (1996). Colour-blind camouflage. Nature 382, 408–409.

Mather, J. A. (1988). Daytime activity of juvenile Octopus vulgaris in Bermuda. Malacologia 29, 69–76.

Mather, J. A. (1991a). Foraging, feeding and prey remains in middens of juvenile Octopus vulgaris (Mollusca: Cephalopoda). Journal of Zoology 224, 27–39. doi:10.1111/j.1469-7998.1991.tb04786.x

Mather, J. A. (1994). ‘Home’ Choice and Modification by Juvenile Octopus vulgaris (Mollusca: Cephalopoda): Specialized Intelligence and Tool Use? Journal of Zoology 233, 359–368. doi:10.1111/j.1469-7998.1994.tb05270.x

Mather, J. A. (1991b). Navigation by spatial memory and use of visual landmarks in octopuses. Journal of Comparative Physiology A 168, 491–497. doi:10.1007/BF00199609

Mather, J. A., and Nixon, M. (1995). Octopus vulgaris (Cephalopoda) drills the chelae of crabs in Bermuda. Journal of Molluscan Studies 61, 405–406. doi:10.1093/mollus/61.3.405

McQuaid, C. D. (1994). Feeding behaviour and selection of bivalve prey by Octopus vulgaris Cuvier. Journal of Experimental Marine Biology and Ecology 177, 187–202. doi:10.1016/0022-0981(94)90236-4

Meisel, D. V, Byrne, R. A., Kuba, M., Griebel, U., and Mather, J. A. (2003). Circadian rhythms in Octopus vulgaris. Berliner Paläobiol. Abh 3, 171–177.

Moynihan, M., and Rodaniche, A. F. (1982). ‘The behaviour and natural history of the Caribbean Reef Squid Sepioteuthis sepioidea’. (Verlag Paul Parey: Berlin.)

Navarro, J., Coll, M., Somes, C. J., and Olson, R. J. (2013). Trophic niche of squids: Insights from isotopic data in marine systems worldwide. Deep-Sea Research Part II: Topical Studies in Oceanography 95, 93–102. doi:10.1016/j.dsr2.2013.01.031

Neill, S. R. J., and Cullen, J. M. (1974). Experiments on whether schooling by their prey affects the hunting behaviour of cephalopods and fish predators. Journal of Zoology 172, 549–569. doi:10.1111/j.1469-7998.1974.tb04385.x

128

Nixon, M. (1987). Cephalopod diets. In ‘Cephalopod life cycles. Comparative Reviews’. (Ed P. R. Boyle.) pp. 201–219. (Academic Press: London, Orlando.)

Nixon, M. (1984). Is there external digestion by Octopus? Journal of Zoology 202, 441–447. doi:10.1111/j.1469-7998.1984.tb05094.x

Osorio, D. (2014). Cephalopod behaviour: Skin flicks. Current Biology 24. doi:10.1016/j.cub.2014.06.066

Paine, R. T. (1966). Food Web Complexity and Species Diversity. The American Naturalist 100, 65–75. doi:10.1086/282400

Paine, R. T. (1974). Intertidal community structure experimental studies on the relationship between a dominant competitor and its principal predator. Oecologia (Beri.) 15, 93–120.

Pauly, D., Zeller, D., and Sea Around Us Project (2015). Global cephalopod landings. Available at: http://www.seaaroundus.org/data/#/spatial-catch?funcgroups=25 [accessed 5 June 2017]

Peterson, C. (1979). The Importance of Predation and Competition in Organizing the Intertidal Epifaunal Communities of Barnegat Inlet, New Jersey. Oecologia 39, 1–24.

Pierce, G. J., Boyle, P. R., Hastie, L. C., and Santos, M. (1994). Diets of squid Loligo forbesi and Loligo vulgaris in the northeast Atlantic. Fisheries Research 21, 149–163. doi:10.1016/0165-7836(94)90101-5

Pierce, G. J., and Portela, J. (2014). Fisheries production and market demand. In ‘Cephalopod culture’. (Eds J. Iglesias, L. Fuentes, and R. Villanueva.) pp. 41–58. (Springer: New York.)

Prince, J. D., Peeters, H., Gorfine, H., and Day, R. W. (2008). The novel use of harvest policies and rapid visual assessment to manage spatially complex abalone resources (Genus Haliotis). Fisheries Research 94, 330–338. doi:10.1016/j.fishres.2008.07.016

Prince, J. D., Sellers, T. L., Ford, W. B., and Talbot, S. R. (1988). Recruitment, growth, mortality and population structure in a southern Australian population of Haliotis rubra (Mollusca: Gastropoda). Marine Biology 100, 75–82. doi:10.1007/BF00392957

Promboon, P., Nabhitabhata, J., and Duengdee, T. (2011). Life cycle of the marbled octopus, Amphioctopus aegina (Gray) (Cephalopoda: Octopodidae) reared in the laboratory. Scientia Marina 75, 811–821. doi:10.3989/scimar.2011.75n4811

Regueira, M., Guerra, Á., Fernández-Jardón, C. M., and González, A. F. (2017). Diet of the horned octopus Eledone cirrhosa in Atlantic Iberian waters: ontogenetic and environmental factors affecting prey ingestion. Hydrobiologia 785, 159–171. doi:10.1007/s10750-016-2916-2

Rigby, P. R., and Sakurai, Y. (2005). Multidimensiorial tracking of giant Pacific octopuses in Northern Japan reveals unexpected foraging behaviour. Marine Technology Society Journal 39, 64–67.

Rocha, F., Guerra, Á., and González, A. F. (2001). A review of reproductive strategies in cephalopods. Biol. Rev 76, 291–304. doi:10.1017/S1464793101005681

Rodhouse, P. G. K., and Nigmatullin, C. M. (1996). Role as Consumers. Philosophical Transactions of the Royal Society of London B: Biological Sciences 351, 1003–1022. doi:10.1098/rstb.1996.0090

Rodhouse, P. G. K., Pierce, G. J., Nichols, O. C., Sauer, W., Arkhipkin, A. I., Laptikhovsky, V. V, Lipinski, M. R., Ramos, J. E., Gras, M., Kidokoro, H., Sadayasu, K., Pereira, J., Lefkaditou, E., Pita, C., Gasalla, M. A., Haimovici, M., Sakai, M., and Downey, N. (2014). Chapter Two - Environmental effects on cephalopod population dynamics: Implications for management of fisheries. In

129

‘Advances in marine biology’. (Ed E. A. G. Vidal.) pp. 99–233. (Academic Press: Oxford: United Kingdom.) doi:http://dx.doi.org/10.1016/B978-0-12-800287-2.00002-0

Rosa, R., Pimentel, M. S., Boavida-Portugal, J., Teixeira, T., Trübenbach, K., and Diniz, M. (2012). Ocean Warming Enhances Malformations, Premature Hatching, Metabolic Suppression and Oxidative Stress in the Early Life Stages of a Keystone Squid. PLoS ONE 7, e38282. doi:10.1371/journal.pone.0038282

Ruiz-Cooley, R. I., Villa, E. C., and Gould, W. R. (2010). Ontogenetic variation of δ13C and δ15N recorded in the gladius of the jumbo squid Dosidicus gigas: Geographic differences. Marine Ecology Progress Series 399, 187–198. doi:10.3354/meps08383

Dos Santos, R. A., and Haimovici, M. (2002). Cephalopods in the trophic relations off southern Brazil. Bulletin of Marine Science 71, 753–770.

Sauer, W. H. H., and Lipiński, M. R. (1991). Food of squid Loligo vulgaris Reynaudii (Cephalopoda: Loliginidae) on their spawning grounds off the Eastern Cape, South Africa. South African Journal Of Marine Science 10, 193–201. doi:10.2989/02577619109504631

Scheel, D., Leite, T. S., Mather, J. A., and Langford, K. (2016). Diversity in the diet of the predator Octopus cyanea in the coral reef system of Moorea, French Polynesia. Journal of Natural History, 1–19. doi:10.1080/00222933.2016.1244298

Semmens, J. M., Pecl, G. T., Villanueva, R., Jouffre, D., Sobrino, I., Wood, J. B., and Rigby, P. R. (2004). Understanding octopus growth: Patterns, variability and physiology. In ‘Marine and Freshwater Research’. pp. 367–377 doi:10.1071/MF03155

Shashar, N., Hagan, R., Boal, J. G., and Hanlon, R. T. (2000). Cuttlefish use polarization sensitivity in predation on silvery fish. Vision Research 40, 71–75. doi:10.1016/S0042-6989(99)00158-3

Shepherd, S. A., and Turner, J. A. (1985). Studies on southern Australian abalone (genus Haliotis). VI. Habitat preference, abundance and predators of juveniles. Journal of Experimental Marine Biology and Ecology 93, 285–298. doi:10.1016/0022-0981(85)90245-X

Shettleworth, S. J., Reid, P. J., and Plowright, C. M. S. (1993). The psychology of diet selection. In ‘Diet selection. An interdisciplinary approach to foraging behaviour’. (Ed R. N. Hughes.) pp. 56–77. (Blackwell Scientific Oublications: Oxford.)

Šifner, K. S., and Vrgoč, N. (2009). Diet and feeding of the musky octopus, Eledone moschata, in the northern Adriatic Sea. Journal of the Marine Biological Association of the United Kingdom 89, 413. doi:10.1017/S0025315408002488

Smith, C. D. (2003). Diet of Octopus vulgaris in False Bay, South Africa. Marine Biology 143, 1127–1133. doi:10.1007/s00227-003-1144-2

Staudinger, M. D., Juanes, F., Salmon, B., and Teffer, A. K. (2013). The distribution, diversity, and importance of cephalopods in top predator diets from offshore habitats of the Northwest Atlantic Ocean. Deep-Sea Research Part II: Topical Studies in Oceanography 95, 182–192. doi:10.1016/j.dsr2.2012.06.004

Ueda, A., Nagai, H., Ishida, M., Nagashima, Y., and Shiomi, K. (2008). Purification and molecular cloning of SE-cephalotoxin, a novel proteinaceous toxin from the posterior salivary gland of cuttlefish Sepia esculenta. Toxicon 52, 574–581. doi:10.1016/j.toxicon.2008.07.007

Underwood, A. J., and Fairweather, P. G. (1989). Supply-side ecology and benthic marine assemblages. Trends in Ecology and Evolution 4, 16–20. doi:10.1016/0169-5347(89)90008-6

Vafidis, D., Kallianiotis, A., Chartosia, N., and Koukouras, A. (2009). The Sepioidea

130

(Cephalopoda, Mollusca) fauna of the Aegean Sea: Comparison with the neighbouring seas and notes on their diet composition. Journal of Biological Research 11, 57–71.

Vermeij, G. J. (1994). The Evolutionary Interaction Among Species: Selection, Escalation, and Coevolution. Annual Review of Ecology and Systematics 25, 219–236. doi:10.1146/annurev.es.25.110194.001251

Vidal, E. A. G., DiMarco, F. P., Wormuth, J. H., and Lee, P. G. (2002). Influence of temperature and food availability on survival, growth and yolk utilization in hatchling squid. In ‘Bulletin of Marine Science’. pp. 915–931

Villanueva, R. (1993). Diet and mandibular growth of Octopus magnificus (Cephalopoda). South African Journal of Marine Science 13, 121–126. doi:10.2989/025776193784287239

Villanueva, R. (2000). Effect of temperature on statolith growth of the European squid Loligo vulgaris during early life. Marine Biology 136, 449–460. doi:10.1007/s002270050704

Vincent, T. L. S., Scheel, D., and Hough, K. R. (1998). Some Aspects of Diet and Foraging Behavior of Octopus dofleini Wülker, 1910 in its Northernmost Range. Marine Ecology 19, 13–29. doi:10.1111/j.1439-0485.1998.tb00450.x

Voss, G. (1977). Present status and new trends in cephalopod systematics. In ‘Symposia of the Zoological Society of London’. (Eds M. Nixon and J. B. Messenger.) pp. 49–60. (Academic Press: London.)

Wells, M. J., O’Dor, R. K., Mangold, K., and Wells, J. (1983). Diurnal changes in activity and metabolic rate in octopus vulgaris. Marine Behaviour and Physiology 9, 275–287. doi:10.1080/10236248309378598

Williams, T. M., Estes, J. A., Doak, D. F., and Springer, A. M. (2004). Killer appetites: Assessing the role of predators in ecological communities. Ecology 85, 3373–3384. doi:10.1890/03-0696

Williamson, R. (1995). The statocysts of cephalopods. In ‘Cephalopod Neurobiology: Neuroscience Studies in Squid, Octopus and Cuttlefish’. (Eds J. N. Abbott, R. Williamson, and L. Maddock.) pp. 503–520. (Oxford University Press: Oxford.)

Worm, B., and Branch, T. A. (2012). The future of fish. Trends in Ecology and Evolution 27, 594–599. doi:10.1016/j.tree.2012.07.005

Young, J. Z. (1971). ‘The anatomy of the nervous system of Octopus vulgaris’. (Oxford University Press: Oxford: United Kingdom.)

Vander Zanden, J. M., and Rasmussen, J. B. (1999). Primary consumer δ13C and δ15N and the trophic position of aquatic consumers. Ecology 80, 1395–1404. doi:10.2307/177083