cbd-ts-43-en

7/31/2019 cbd-ts-43-en

1/68

43D T S N.S f

D

Forest Resilience,Biodiversity,and Climate Change

A Synthesis of the Biodiversity/Resilience/

Stability Relationship in Forest Ecosystems

7/31/2019 cbd-ts-43-en

2/68

Forest resilience, Biodiversity, and climatechange

A Synthesis o the Biodiversity/Resilience/Stability Relationshipin Forest Ecosystems

Ian Tompson, Canadian Forest Service, 1219 Queen St. East, Sault Ste. Marie, Ontario, Canada P6A 2E5Brendan Mackey, Te Australian National University, Te Fenner School o Environment and Society, College oMedicine, Biology and Environment, Canberra AC, 200 Australia

Steven McNulty, USDA Forest Service, Southern, Global Change Program, 920 Main Campus Dr., Suite 300,Raleigh, NC 27606 USAAlex Mosseler, Canadian Forest Service, 1350 Regent Street South, Fredericton, New Brunswick, Canada E3B 5P7

7/31/2019 cbd-ts-43-en

3/68

2

Forest Resilience, Biodiversity, and Climate Change

Acknowledgements

Tis document was prepared by Ian Tompson (Canadian Forest Service), Brendan Mackey (Te AustralianNational University and Te Fenner School o Environment and Society), Steven McNulty (USDA ForestService), and Alex Mosseler (Canadian Forest Service).

Te authors would like to thank the ollowing persons and organizations or providing comments on anearlier dra o this paper: Z. Bennadji (FAO), A. Campbell (UNEP-WCMC), . Christophersen (CBD), P.Dewees (World Bank), S. Ferrier (CSIRO), V. Kapos (UNEP-WCMC), P. Mayer (IUFRO), R. Nasi (CIFOR),J. Parrotta (U.S. Forest Service), O.L. Phillips (University o Leeds), J.A. Prado (FAO), C. Saint-Laurent(IUCN), B. Schamp (Algoma University), the International ropical imber Organization (IO), and theUnited Nations Framework Convention on Climate Change (UNFCCC).

Tis document has been produced with the nancial assistance o the Government o Norway. Te viewsexpressed herein can in no way be taken to reect the ocial opinion o the Government o Norway.

Published by the Secretariat o the Convention on Biological Diversity. ISBN 92-9225-137-6

Copyright 2009, Secretariat o the Convention on Biological Diversity

Cover photo credits (rom top to bottom): A. Couillaud, A. Zemdega, P. Skubisz, . Gage

Te designations employed and the presentation o material in this publication do not imply the expressiono any opinion whatsoever on the part o the copyright holders concerning the legal status o any country,territory, city or area or o its authorities, or concerning the delimitation o its rontiers or boundaries. Tispublication may be reproduced or educational or non-prot purposes without special permission rom thecopyright holders, provided acknowledgement o the source is made.

c:Tompson, I., Mackey, B., McNulty, S., Mosseler, A. (2009). Forest Resilience, Biodiversity, and ClimateChange. A synthesis o the biodiversity/resilience/stability relationship in orest ecosystems. Secretariat othe Convention on Biological Diversity, Montreal. echnical Series no. 43, 67 pages.

F u p :Secretariat o the Convention on Biological DiversityWorld rade Centre413 St. Jacques, Suite 800Montreal, Quebec, Canada H2Y 1N9Phone: 1 (514) 288 2220

Fax: 1 (514) 288 6588E-mail: [email protected]: http://cbd.int

7/31/2019 cbd-ts-43-en

4/68

3


Table of Contents

Glossary .......................................................................................................................................................................4Foreword ......................................................................................................................................................................6

Summary or Policy-makers .....................................................................................................................................7

1. Introduction ............................................................................................................................................................9

1.1 Forests, climate, and climate change ......................................................................................................9

1.2 Denitions o and related to resilience ................................................................................................10

1.3 Components o biodiversity and denitions ......................................................................................12

1.4 Issues o scale and resilience .................................................................................................................13

2. Genetic diversity and resilience to change ........................................................................................................13

3. Te relationships among biodiversity, productivity and unction, and resilience and stability .......... ......17

3.1 Teoretical background .........................................................................................................................17

3.2 Evidence o a diversity productivity relationship in orests ................... ...................... ..................... 18

3.2.1 Diversity-productivity relationships and orest resilience ..................... ...................... .............. 21

3.3 Diversity and stability ............................................................................................................................22

3.3.1 Diversity and invasion o ecosystems ...........................................................................................23

3.3.2 Diversity and insect pests ...............................................................................................................24

3.3.3 Diversity and stability o processes in orests ..............................................................................25

3.4 Summary o diversity-resilience processes .........................................................................................25

4. Resilience, biodiversity, and orest carbon dynamics ......................................................................................25

4.1 Forests and the global carbon cycle .....................................................................................................25

4.2 Biodiversity and resilience o orest-carbon dynamics......................................................................27

5. Case studies o orest resilience and comparisons under climate change by orest biome ...................... ...305.1 Boreal orest biome ................................................................................................................................30

5.1.1 Climate change and boreal orest resilience.................................................................................31

5.1.2 Case-study: western North American lodgepole pine ...............................................................31

5.1.3 Case-study: North American boreal mixedwoods .....................................................................31

5.2 emperate orest biome .........................................................................................................................34

5.2.1 emperate Forests and Environmental Stressors ........................................................................34

5.2.2 Case-study: Moist evergreen temperate orests ...........................................................................34

5.2.3 Case-study: southern Europe ........................................................................................................37

5.2.4 Case-study: eastern North American deciduous orests ............................................................37

5.3 ropical orests ........................................................................................................................................39

5.3.1 Climate change and tropical orest resilience ..............................................................................415.3.2 Case study: Amazon rain orest .....................................................................................................41

5.4 Summary among orest biomes ............................................................................................................43

6. Conclusions and ecological principles ..............................................................................................................43

6.1 Ecological principles to oster orest ecosystem resilience and stability under climate change ...44

7. Literature cited ......................................................................................................................................................46

7/31/2019 cbd-ts-43-en

5/68

4


Glossary

t d suAdaptation Adjustment in natural or human systems in response to actual or ex-

pected climatic stimuli or their eects, which moderates harm or ex-ploits benecial opportunities.

UNFCCC

Adaptivecapacity

Te ability o a species to become adapted (i.e., to be able to live andreproduce) to a certain range o environmental conditions as a result ogenetic and phenotypic responses.

Dobzhansky1968

Biodiversityor biologicaldiversity

Te variability among living organisms rom all sources including ter-restrial, marine, and other aquatic ecosystems and the ecological com-plexes o which they are part; this includes diversity within species,among species, and o ecosystems.

CBD

Biomass Organic material both above ground and below ground, and both liv-ing and dead, e.g., trees, crops, grasses, tree litter, roots, etc. FAO 2006

Biome A regional ecosystem with a distinct assemblage o vegetation, animals,microbes, and physical environment oen reecting a certain climateand soil

Helms 1998

Carbonsequestration

Te process o removing carbon rom the atmosphere and depositingit in a reservoir.

UNFCCC

Deorestation Te direct human-induced conversion o orested land to non-orestedland.

UNFCCC -MarrakechAccords

Ecologicalresilience

Te ability o a system to absorb impacts beore a threshold is reachedwhere the system changes into a dierent state.

Gunderson 2000

Ecosystem A community o all plants and animals and their physical environment,unctioning together as an interdependent unit.

Helms 1998

Ecosystem A dynamic complex o plant, animal and micro-organism communitiesand their non-living environment interacting as a unctional unit.

CBD

EcosystemServices (alsoecosystemgoods andservices)

Te benets people obtain rom ecosystems. Tese include provision-ing services such as ood, water, timber, and bre; regulation servicessuch as the regulation o climate, oods, disease, wastes, and waterquality; cultural services such as recreation, aesthetic enjoyment, andspiritual ulllment; and supporting services such as soil ormation,photosynthesis, and nutrient cycling.

MillenniumEcosystemAssessment

Engineeringresilience

Te capacity o a system to return to its pre-disturbance state Gunderson 2000

ForestDegradation

Changes within the orest which negatively aect the structure or unc-tion o the stand or site, and thereby lower the capacity to supply prod-ucts and/or services

FAO 2001

ForestDegradation

A degraded orest is a secondary orest that has lost, through humanactivities, the structure, unction, species composition or productivitynormally associated with a natural orest type expected on that site.Hence, a degraded orest delivers a reduced supply o goods and ser-vices rom the given site and maintains only limited biological diversity.Biological diversity o degraded orests includes many non-tree compo-nents, which may dominate in the under-canopy vegetation.

UNEP/CBD

7/31/2019 cbd-ts-43-en

6/68

5


Forest state Most commonly considered in terms o the dominant assemblage otree species orming an ecosystem at a location, the unctional roles

those species play, and the characteristic vegetation structures (height,layers, stems density, etc.) at maturity.

Tis document

Functionalgroups

Assemblages o species perorming similar unctional roles within anecosystem, such as pollination, production, or decomposition (i.e., tro-phic groups), hence providing some redundancy.

Hooper andVitousek 1997

Fundamentalniche

A geographic area with the appropriate set o abiotic actors in whicha species could occur.

Hutchinson1957

GeneticDiversity

Any variation in the nucleotides, genes, chromosomes, or whole ge-nomes o organisms.

Tis document

Mitigation In the context o climate change, a human intervention to reduce thesources or enhance the sinks o greenhouse gases.

UNFCCC

Modiednatural orest

Forest/other wooded land o naturally regenerated native species wherethere are clearly visible indications o human activities. Includes, butis not limited to, selectively logged-over areas, naturally regeneratingareas ollowing agricultural land use, areas recovering rom human-induced res, areas where it is not possible to distinguish whether theregeneration has been natural or assisted.

FAO 2006

Monotypicstand

A orest stand containing one tree species. Tis document

Plantation Forest/other wooded land o introduced species and in some cases na-tive species, established through planting or seeding, mainly or pro-duction o wood or non-wood goods

FAO 2006

Primary

orest

Forest/other wooded land o native species, where there are no clearly

visible indications o human activities and the ecological processes arenot signicantly disturbed.

FAO 2006

Productivityorproduction

Te rate at which biomass is produced per unit area by any class oorganisms.

Helms 1998

Resilience Te capacity o an ecosystem to return to the pre-condition state ol-lowing a perturbation, including maintaining its essential characteris-tics taxonomic composition, structures, ecosystem unctions, and pro-cess rates.

Holling 1973

Resistance Te capacity o the ecosystem to absorb disturbances and remain large-ly unchanged.

Holling 1973

Silviculture Te art o producing and tending a orest by manipulating its estab-lishment, composition and growth to best ulll the objectives o theowner. Tis may, or may not, include timber production.

Helms 1998

Succession Progressive changes in species composition and orest communitystructure caused by natural processes (nonhuman) over time.

Helms 1998

Stability Te capacity o an ecosystem to remain more or less in the same statewithin bounds, that is, the capacity to maintain a dynamic equilibriumin time while resisting change.

Holling 1973

7/31/2019 cbd-ts-43-en

7/68

6


Foreword

Te worlds orest ecosystems provide environmental services that benet, directlyor indirectly, all human communities, including watershed protection, regionalclimatic regulation, bre, ood, drinking water, air purication, carbon storage,recreation, and pharmaceuticals.

Forests harbour an estimated two thirds o all terrestrial species, and a ascinat-ing array o ecological processes. Te ecological stability, resistance, resilience, andadaptive capacities o orests depend strongly on their biodiversity. Te diversity ogenes, species, and ecosystems coners on orests the ability to withstand externalpressures, and the capacity to bounce back to their pre-disturbance state or adapt

to changing conditions. Tis review explores these relationships based on published scientic literature.

Tis publication is a direct response to a request by the ninth meeting o the Conerence o the Parties to theCBD to explore the links between biodiversity, orest ecosystem resilience, and climate change. Its ndingsare relevant or the uture implementation o the CBD, but also the United Nations Framework Conventionon Climate Change (UNFCCC), the Forest Instrument o the United Nations Forum on Forests (UNFF),and other international and regional orest-related agreements. It provides a compelling rationale or theconservation and sustainable use o biodiversity in any orest-based climate change mitigation and adapta-tion eorts.

In the present debate on climate change, the carbon storage capacity o orests and their role in mitigationis receiving increasing attention. While the international climate change negotiations have now recognizedthe value o ecosystem-based adaptation, in reality ecosystem-based mitigation and adaptation are two sideso the same coin. Protecting primary orests and restoring managed or degraded orest ecosystems make avital contribution to both reducing anthropogenic emissions and aiding societal adaptation to unavoidable

climate change. It is the resilience inherent to intact orest ecosystems - ully unctional units o plants, ani-mals, micro-organisms, and ungi that provides the best insurance against climate change and prospectsor ensuring orests meet the needs o present and uture generations.

P

hoto:CDB

Ahmed DjoghlaExecutive SecretarySecretariat o the Conventionon Biological Diversity

7/31/2019 cbd-ts-43-en

8/68

7


Summary for Policy-makers

Resilience is the capacity of a forest to withstand (absorb) external pressures and return, over time, to itspre-disturbance state. When viewed over an appropriate time span, a resilient orest ecosystem is able tomaintain its identity in terms o taxonomic composition, structure, ecological unctions, and process rates.

Te available scientic evidence strongly supports the conclusion that the capacity of forests to resistchange, or recover ollowing disturbance, is dependent on biodiversity at multiple scales.

Maintaining and restoring biodiversity in forests promotes their resilience to human-induced pressuresand is thereore an essential insurance policy and saeguard against expected climate change impacts. Bio-diversity should be considered at all scales (stand, landscape, ecosystem, bioregional) and in terms o allelements (genes, species, communities). Increasing the biodiversity in planted and semi-natural orests willhave a positive eect on their resilience capacity and oen on their productivity (including carbon storage).

Te resilience of a forest ecosystem to changing environmental conditions is determined by its biologicaland ecological resources, in particular (i) the diversity o species, including micro-organisms, (ii) the geneticvariability within species (i.e., the diversity o genetic traits within populations o species), and (iii) the re-gional pool o species and ecosystems. Resilience is also inuenced by the size o orest ecosystems (generally,the larger and less ragmented, the better), and by the condition and character o the surrounding landscape.

Primary forests are generally more resilient (and stable, resistant, and adaptive) than modied naturalorests or plantations. Tereore, policies and measures that promote their protection yield both biodiversityconservation and climate change mitigation benets, in addition to a ull array o ecosystem services. Never-theless, it must be recognized that certain degraded orests, especially those with invasive alien species, maybe stable and resilient, and these orests can become serious management challenges i attempts are made tore-establish the natural ecosystem to recover original goods and services.

Some forest ecosystems with naturally low species diversity nevertheless have a high degree of resilience,such as boreal pine orests. Tese orests, however, are highly adapted to severe disturbances, and their domi-nant tree species have a broad genetic variability that allows tolerance to a wide range o environmentalconditions.

Te carbon pool is largest in old primary forests, especially in the wet tropics, which are stable forest sys-tems with high resilience.

Te permanence of eorts under UNFCCC negotiations, such as reducing emissions from deforestationand orest degradation (REDD), and o other orest-based climate change mitigation and adaptation policiesand measures, is linked to the resilience o orests, and thus to orest biodiversity. REDD activities thereore

should take biodiversity conservation into consideration, as this will help maintain orest ecosystem resil-ience and the long-term stability o the carbon pool.

Te regional impacts of climate change, especially interacting with other land use pressures, might be suf-cient to overcome the resilience o even some large areas o primary orests, pushing them into a perma-nently changed state. I orest ecosystems are pushed past an ecological tipping point, they could be trans-ormed into a dierent orest type, and, in extreme cases, a new non-orest ecosystem state (e.g. rom orestto savannah). In most cases, the new ecosystem state would be poorer in terms o both biological diversityand delivering ecosystem goods and services.

Plantations and modied natural forests will face greater disturbances and risks for large-scale losses dueto climate change than primary orests, because o their generally reduced biodiversity. Te risks can partlybe mitigated by adhering to a number o orest management recommendations:

o Maintain genetic diversity in orests by avoiding practices that select only certain trees or harvestingbased on site, growth rate, or orm.

7/31/2019 cbd-ts-43-en

9/68

8


o Maintain stand and landscape structural complexity, using natural orests and processes as models.

o Maintain connectivity across orest landscapes by reducing ragmentation, recovering lost habitats(orest types), expanding protected area networks, and establishing ecological corridors.

o Maintain unctional diversity and eliminate the conversion o diverse natural orests to monotypic orreduced-species plantations.

o Reduce non-natural competition by controlling invasive species and reduce reliance on non-native treecrop species or plantation, aorestation, or reorestation projects.

o Manage plantation and semi-natural orests in an ecologically sustainable way that recognizes andplans or predicted uture climates. For example, reduce the odds o long-term ailure by apportioningsome areas o assisted regeneration or trees rom regional provenances and rom climates that approxi-

mate uture climate conditions, based on climate modelling.

o Maintain biodiversity at all scales (stand, landscape, bioregional) and o all elements (genes, species,communities) by, or example, protecting tree populations which are isolated, disjunct, or at margins otheir distributions, source habitats, and reuge networks. Tese populations are most likely to representpre-adapted gene pools or responding to climate change and could orm core populations as conditionschange.

o Ensure that there are national and regional networks o scientically designed, comprehensive, ad-equate, and representative protected areas. Build these networks into national and regional planning orlarge-scale landscape connectivity.

7/31/2019 cbd-ts-43-en

10/68

9


1. Introduction

Tis paper reviews the concepts o ecosystem resil-ience, resistance, and stability in orests and their re-lationship to biodiversity, with particular reerenceto climate change.

Te report is a direct response to a request by theninth meeting o the Conerence o the Parties to theCBD, in decision IX/51, to explore the links betweenbiodiversity, orest ecosystem resilience, and climatechange. Forests are emphasized because they aremajor reservoirs o terrestrial biodiversity and con-tain about 50% o the global terrestrial biomass car-bon stocks (IPCC 2007, FAO 2000). Emissions romdeorestation and degradation remain a signicant(ca. 18-20%) source o annual greenhouse gas emis-sions into the atmosphere (IPCC 2007), and there-ore the conservation, appropriate management andrestoration o orests will make a signicant contri-bution to climate change mitigation. Further, orestshave a certain natural capacity to adapt to climatechange because o their biodiversity. Some animalshave important roles in ecosystem processes and or-ganization, such as pollination, seed dispersal, andherbivory, and the loss o these species has clearnegative consequences or ecosystem resilience (e.g.,

Elmqvist et al. 2003). Here, however, we limit ourdiscussion to botanical aspects o orests, with theexception o some discussion o insect pests and dis-eases as these inuence orest resilience and stability.

Forests have many unique properties, related totheir high rates o primary productivity and biodi-versity, which distinguish them ecologically romother ecosystems. Such properties include biologi-cal structures that develop in vertical and horizontallayers o live and dead plants, complex processes atmultiple vertical levels rom within soil layers up to

the canopy, the capacity or sel-renewal in the aceo constant small and large disturbances, co-evolvedplant-animal and plant-plant interactions, and theinuence orest landscapes can have on micro- andregional climates, especially in closed-canopy tropi-cal orests. Forests are comprised o multiple ecosys-tems that are associated with variable edaphic andmicroclimate conditions across broad landscapes.

In the annex to decision II/9, the Conerence othe Parties to the Convention on Biological Diver-

1. Decision IX/5 requests the Executive Secretary to: Collect,

compile and disseminate information on the relation between

forest ecosystem resistance and resilience, forest biodiversity,

and climate change, through the clearing-house mechanism and

other relevant means.

sity recognized that Forest biological diversity resultsrom evolutionary processes over thousands and evenmillions o years which, in themselves, are driven byecological orces such as climate, fre, competitionand disturbance. Furthermore, the diversity o orestecosystems (in both physical and biological eatures)results in high levels o adaptation, a eature o or-est ecosystems which is an integral component o theirbiological diversity. Within specifc orest ecosystems,the maintenance o ecological processes is dependentupon the maintenance o their biological diversity.

Humans are having long-term cumulative impactson Earths ecosystems through a range o consump-tive, exploitive, and indirect mechanisms, even to

the extent o inuencing the global climate (IPCC2007). Te major impacts o humans on orest eco-systems include loss o orest area, habitat ragmen-tation, soil degradation, depletion o biomass and as-sociated carbon stocks, transormation o stand ageand species composition, species loss, species intro-ductions, and the ensuing cascading eects, such asincreasing risk o re (Uhl and Kauman 1999, Ger-wing 2002). As a result, there has long been globalconcern about the long-term capacity o orests tomaintain their biodiversity and associated rates osupply o goods and services (including carbon stor-

age, ood, clean water, and recreation). Tis concernhas been amplied ollowing observed impacts oc-curring to global orests as a result o climate change(e.g., Phillips 1997, Kellomaki et al. 2008, Phillips etal. 2009, Malhi et al. 2009).

1.1 Forests, climate, and climate change

Superimposed on the many other anthropogenicimpacts on orest ecosystems noted above is human-orced global climate change. Climate has a majorinuence on rates o photosynthesis and respira-tion (Woodward et al. 1995, Kueppers et al. 2004,Law et al. 2007), and on other orest processes, act-ing through temperature, radiation, and moistureregimes over medium and long time periods. Cli-mate and weather conditions also directly inuenceshorter-term processes in orests, such as requencyo storms and wildres, herbivory, and species mi-gration (Gundersen and Holling 2002). As the glob-al climate changes, orest ecosystems will changebecause species physiological tolerances may be ex-ceeded and the rates o biophysical orest processeswill be altered (Olesen et al. 2007, Kellomaki et al.2008, Malhi et al. 2008).

Forests can be useully conceived as complex, sel-organizing systems with multiple natural processes

7/31/2019 cbd-ts-43-en

11/68

10


that respond autonomously to internal and externaldrivers. For example, as available water becomeslimiting, the height and density o the tree cano-pies is reduced because o basic ecophysiologicalrelationships governing environmental controls onplant growth (Berry and Roderick 2002). I climatechange results in a signicant reduction in wateravailability, then the orest system will naturallychange species composition (or state see deni-tion below). For example, the vegetation will reacha threshold beyond which the vegetation structure isnot suciently tall and dense to comprise a orest,along with the concomitant changes in the dominanttaxonomic composition o the plant community(Stephenson 1990). Under severe drying conditions,

orests may be replaced by savannahs or grasslands(or even desert), while under increased temperature,open taiga can be replaced by closed boreal orests(assuming that there is sucient moisture to sup-port plant growth during the newly extended grow-ing season) (e.g., Price and Scott 2006, Kellomaki etal. 2008).

Forests can also inuence regional climates, depend-ing on their extent and this is particularly true o theAmazon orest (Betts et al. 2008, Phillips et al. 2009).Hence, numerous eedbacks exist between climate

and orests as the climate changes (Bonan et al. 2003,Callaghan et al. 2004, Euskirchen et al. 2009). Teseeedbacks are mediated through changes to albedo(Euskirchen et al. 2009), altered carbon cycle dynam-ics (Heath et al. 2005, Phillips et al. 2009), energyuxes and moisture exchange (Wildson and Agnew1992, Bonan et al. 2003), and herbivory, resulting inincreased res (Ayres and Lomardero 2000). Hence,maintaining orest resilience can be an importantmechanism to mitigate and adapt to climate change.

1.2 Defnitions o and related to resilience

We discuss several closely related terms throughoutthis paper and dene them here, including resilience,resistance, state, and stability. We dene as the capacity o an ecosystem (i.e., orest type, inthis paper) to return to the original state ollowing aperturbation, maintaining its essential characteristictaxonomic composition, structures, ecosystem unc-tions, and process rates (Holling 1973). Similarly,Walker and Salt (2006) dened resilience as the ca-pacity o a system to absorb disturbance and still re-tain its basic unction and structure, and thereore itsidentity (i.e., recognizable as the same by humans).

A orest ecosystem can respond in dierent ways todisturbances and perturbations. Depending on the

capacity o orests to cope with the degree o change,the characteristic taxonomic composition, veg-etation structure, and rates o ecosystem processesmay or may not be altered; that is, the resilience othe orest ecosystem may or may not be overcome.Forest characteristics can be used individually orin combination to dene a orest ecosystem .Most commonly, a orest state is considered in termso the dominant assemblage o tree species ormingan ecosystem at a location, along with the unctionalroles those species play, and the characteristic veg-etation structures (height, layers, stems density, etc.)at maturity. So, a given mature orest type has a par-ticular suite o characteristics that identiy its state.(Note that we use the terms system and ecosystem

synonymously throughout.)

A dierence has been made in the scientic litera-ture between engineering resilience and ecologi-cal resilience (Holling 1973, Peterson et al. 1998,Gunderson 2000, Walker et al. 2004). Engineeringresilience is related to the capacity o a system to re-turn to its more-or-less exact pre-disturbance state,and the assumption is that there is only one steadystate. Te latter concept has also been more recentlyreerred to as equilibrium dynamics. Ecological re-silience is dened as the ability o a system to ab-

sorb impacts beore a threshold is reached where thesystem changes into a dierent state altogether. Forexample, in the case o increasing climatic drought,a resilient orest ecosystem according to the engi-neering denition is one that would recover romdrought stress, with little or no change in speciescomposition. I the ecological denition is used, thenit is acknowledged that more than one stable systemstate is possible, with resilience being the measureo a orest ecosystems capacity to withstand a pro-longed drought beore being converted into a dier-ent vegetation ecosystem (e.g., non-orest); though itmight go through several other dierent but stable

orest states with new species compositions, beorethe conversion to grassland. Many o those succes-sive orest states might be able to provide most orall o the goods and services provided by the initialstate, and all would be recognizable as a orest type.Tis is also reerred to as non-equilibrium dynamics.

Forests are engineering resilient in the sense thatthey may recover, aer a period o time, rom a cata-strophic disturbance to their original, pre-distur-bance state maintaining, more-or-less, the originalspecies composition. Te main ecosystem states o

interest are dened by the dominant oristic (tree)composition and stand structure. However, it isalso useul to consider the question o ecological

7/31/2019 cbd-ts-43-en

12/68

11


Drever et al. 2006). Maintaining or restoring orestresilience is oen cited as a necessary societal ad-aptation to climate change (e.g., Millar et al. 2007,Chapin et al. 2007). Drever et al. (2006) noted theimportance o clariying the questions: resilience owhat and resilience to what? Here, the o what areparticular characteristics o orest ecosystems (e.g.,carbon sequestration, water use/yield), and the towhat are environmental and human-caused dis-turbances, especially climate change. For example,an individual species physiological tolerances maybe exceeded by natural environmental change orhuman-caused events. Consequently, the speciescomposition o a orest may change while other eco-system characteristics persist.

Forests are generally to change, that is, theychange little within bounds as a result o non-cata-strophic disturbances, such as chronic endemic in-sect herbivory or minor blowdown and canopy gapscreated by the death o individual or small groupso trees. Forests may also be resistant to certain en-vironmental changes, such as weather patterns overtime, owing to redundancy at various levels amongunctional species (as discussed urther below, re-dundancy reers to the overlap and duplication inecological unctions perormed by the diversity o

genomes and species in an ecosystem). Ecosystemsmay be highly resilient but have low resistance to agiven perturbation. For example, grasslands are notresistant, but are highly resilient, to re. However,most well-developed orests, especially primary oldorests, are both resilient and resistant to changes(e.g., Holling 1973, Drever et al. 2006).

Resistance is related to the concept o b inthe sense that, in response to minor perturbations, aorest ecosystem remains within a range o variationaround a specied ecosystem state. Stability reectsthe capacity o an ecosystem to maintain a dynamic

equilibrium over time while resisting change to adierent state. A stable ecosystem p when ithas the capacity to absorb disturbances and remainlargely unchanged over long periods o time.

Species stability reers to consistent species composi-tion over time. Drever et al. (2006) suggested thatorest types that naturally progress through succes-sional compositional changes are not necessarilychanging state. On the other hand, a orest that wasonce dominated by a certain suite o species and thathas changed as a result o new environmental condi-

tions or human intererence has changed ecosystemstates. For example, i a harvested boreal spruce-pine-dominated orest regenerates to a mixedwood,

F u b xw c u

p p , w u b uu p p

bu 50-80 .

resilience with respect to the capacity o a orest tocontinue to provide certain (most or all) ecosystemgoods and services, even i the orest compositionand structure are permanently altered by distur-bances.

Resilience is an emergent property o ecosystemsthat is conerred at multiple scales by genes, species,unctional groups o species (see denition below),and processes within the system (Gunderson 2000,

Credit:A.Mosseler

Credit:A.Mosseler

Credit:A.Mosseler

7/31/2019 cbd-ts-43-en

13/68

12


or i selective logging or disease eliminates speciesrom a orest system, we would suggest that the sys-tem has indeed changed states. Tat is, even thoughit is still a orest, the ecosystem state, as dened bythe dominant taxonomic composition o the canopytrees, has changed, along with various processes suchas rates o growth and types o pollination. Further-more, in this new state, some or many o the goodsand services will also have changed and there maybe eects on other elements o biodiversity resultingrom changes in the provision o habitats and there-ore the persistence o dependent animal species.Ecosystems may change states in response to distur-bances, and the new state may or may not supply thesame goods and services as the original state. Fur-

ther, i species diversity is positively related to stabili-ty and resilience o orest systems, then species losseswill likely have consequences or the long-term pro-duction o goods and services. Consequently, thereis considerable interest in developing the capacityto understand and predict the mechanisms associ-ated with resilience as it relates to the ways in whichorests ecosystems respond to degradation, loss ospecies, and climate change (e.g., Kinzig et al. 2001,Scherer-Lorenzen et al. 2005).

1.3 Components o biodiversity and defnitions

Biodiversity is oen considered, especially withinthe orest management community, as simply a listo species present at a location. Te term can alsobe used in the context o providing habitats or spe-cies o some particular value o interest to people,and in this sense biodiversity is a good producedby the ecosystem. While bencompassesboth these latter meanings, it is actually a broaderterm intended to encompass various measures o theull richness o lie on Earth. As dened by the Con-vention on Biological Diversity, biological diver-sity means the variability among living organisms

rom all sources including terrestrial, marine, andother aquatic ecosystems and the ecological com-plexes o which they are part; this includes diversitywithin species, among species, and o ecosystems.Allen and Hoekstra (1992) dened biodiversityeven more broadly to include the variety o lie atmultiple scales o ecological organization, includinggenes, species, ecosystems, landscapes, and biomes.

Here we consider biodiversity in terms o speciccomponents that are particularly relevant to or-est ecosystems and equate them with the scale

at which they are classied and mapped by hu-mans. In so doing, we reer to standard metricsincluding genetic diversity and species richness

that relate to the dominant plant and animal spe-cies that characterize a given orest ecosystem.We also reer to terms that describe the vegetationstructure (height, density, complexity) and age.

We make reerence to unctional redundancy,unctional types or species, and unctional groups.Several studies have established that resilience inecosystems is related to the biological diversity inthe system and the capacity that it coners to main-tain ecosystem processes (Walker 1995, Petersonet al. 1998, Loreau et al. 2001, Hooper et al. 2005,Drever et al. 2006, Bodin and Wimen 2007). Mostecosystem processes are controlled by, or are theresult o, biodiversity. However, not all species are

necessarily equally important in maintaining theseprocesses (Walker 1992, 1995, Diaz et al. 2003) andthere is some redundancy at multiple levels withinmost ecosystems (Hooper et al. 2005). Fuup are assemblages o species perorming simi-lar unctional roles within an ecosystem, such aspollination, production, or decomposition, henceproviding the ecosystem with a level o redundancy(e.g., see Hooper et al. 2002). As discussed urtherbelow, unctional diversity is not necessarily corre-lated with species richness (Diaz and Cabido 2001,Hooper et al. 2005). Functional species that domi-

nate ecosystem processes are not inevitably the mostnumerous species in the system (e.g., Hooper andVitousek 1997, Diaz et al. 2003), and it is importantto understand which species are contributing mostto maintaining the ows o goods and services imanagement or protection is an objective. We areespecially interested in unctional diversity (withinunctional groups) in ecosystems because evidencehas accumulated, especially in grassland systems,which implicates a relationship between unctionaldiversity and ecosystem properties, including resil-ience and the related system attributes o stabilityand resistance (Diaz and Cabido 2001, Hooper et

al. 2005). Under changed conditions, however, spe-cies that had a limited or no unctional role (pas-senger species) may become unctionally dominant(driver species), hence buering the ecosystemagainst large changes and conerring resilience; thatis, passengers can become the drivers (Walker 1995).Tis variable response has also been termed unc-tional response diversity and is critical to ecosystemresilience (Chapin et al. 1997, Elmqvist et al. 2003).Loss o unctional species in the absence o redun-dancy has negative consequences or the ecosystemto the point o ecosystem collapse (Chapin et al.

1997). Hooper et al. (2005) noted that there is a clearneed or continued research into the relationshipbetween species richness and ecosystem stability.

7/31/2019 cbd-ts-43-en

14/68

13


1.4 Issues o scale and resilience

Proper scaling is essential in the application o atheoretical ramework. Most ecosystems are subjectto disturbance regimes that occur across a range otemporal and spatial scales. Single communities inorests may occur across several tens to hundredso hectares, while orests may be considered acrosshundreds to thousands o square kilometres. Forest

stands may change continuously as a result o small-scale chronic disturbances that do little to aect thesystem, or they may change considerably at largescales owing to severe disturbances. Scaling is an im-portant actor in dening ecosystem resilience, butscale and resilience are oen investigated or dierentpurposes. Resilience studies generally ocus on howand why individual ecosystems maintain or changestates, while scaling studies oen examine ecologicalphenomena assuming steady-state ecosystems (Hol-ling 1973). However, resilience is a scale-dependentphenomenon. Ecosystems are both temporally andspatially resilient when ecological interactions rein-orce each other to reduce the impact o disturbancesover time. Tis condition can be achieved througha range o mechanisms including species unctionalredundancy, or osetting dierences among species.

At larger scales in orests, there is also a level o po-tential role or species-level beta diversity (i.e. spa-tial turnover in species composition o communi-ties) in enhancing ecosystem resilience in the aceo large-scaled environmental change. Regionalspecies pools provide a level o redundancy at largescales that may coner resilience i the capacity to

migrate across the landscape persists. Tis con-cept has not been well-examined in the literature.

Dening resilience requires a temporal componentthat is related to disturbance requency and recov-ery o the ecosystem. For most orests, we tend toconsider resilience over many decades to centuries.While some existing terrestrial ecosystems seemto have persisted largely unchanged or thousandso years (Hopper and Gioia 2004), environmen-tal change and disturbance o sucient magnitudeeventually alter all ecosystems. Resilient orest eco-systems, in response to a disturbance, ollow a suc-cessional pathway that returns the ecosystem toits pre-disturbance state, at least structurally andunctionally. Tis is particularly the case or orestsdominated by small-scaled disturbances. A distur-bance may be suciently severe to reorganize an

ecosystem into a state, which in the short term (i.e.,decades), may have a dierent resistance, but in thelong term (i.e., centuries) may be equally as resil-ient as the original state. Furthermore, in the verylong-term, the altered state o the ecosystem maysimply be part o a long-term dynamical process.

O course, ecosystems and orests are comprised oassemblages o individual species. Across regions,individual species ranges reect their physiologicaland ecological niches, with the latter reecting theconditions where they have, among other things, a

competitive advantage (Hutchinson 1958). Specieswith broad physiological niche requirements maybe highly resilient to even signicant global climatechange. Likewise, species with a narrow ecologi-cal niche may be more resilient than they appear, ichanged conditions provide them with an advantageat the expense o competitors. In either case, thisonly applies to species which have large enough genepools and the ability to migrate. Where populationsizes and genetic diversity have been reduced, and/orthe mobility o species is restricted through habitatragmentation or by natural lack o species mobility,the likelihood o successul adaptation to environ-

mental change, such as climate change, is diminished.

2. Genetic diversity and resilience tochange

While resilience can be attributed to many levels oorganization o biodiversity, the genetic compositiono species is the most undamental. Molecular genet-ic diversity within a species, species diversity withina orested community, and community or ecosystemdiversity across a landscape and bioregion representexpressions o biological diversity at dierent scales.Te basis o all expressions o biological diversity isthe genotypic variation ound in populations. Teindividuals that comprise populations at each level

m b p ,w p w-p -ub p -

. h, k bu .

Credit:B.J.S

tocks,CanadianForestService

7/31/2019 cbd-ts-43-en

15/68

14


persal mechanisms, constitute the undamental de-terminants o potential species responses to change(Pease et al. 1989, Halpin 1997). In the past, plantshave responded to dramatic changes in climate boththrough adaptation and migration (Davis and Shaw2001).

Te capacity or long-distance migration o plants byseed dispersal is particularly important in the evento rapid environmental change. Most, and probablyall, species are capable o long-distance seed disper-sal, despite morphological dispersal syndromes thatwould indicate morphological adaptations primarilyor short-distance dispersal (Cwyner and MacDon-ald 1986, Higgins et al. 2003). Assessments o mean

migration rates ound no signicant dierences be-tween wind and animal dispersed plants (Wilkinson1997, Higgins et al. 2003). Long-distance migrationcan also be strongly inuenced by habitat suitabil-ity (Higgins and Richardson 1999) suggesting thatrapid migration may become more requent and vis-ible with rapid changes in habitat suitability underscenarios o rapid climate change. Te discrepancybetween estimated and observed migration ratesduring re-colonization o northern temperate orestsollowing the retreat o glaciers can be accountedor by the underestimation o long-distance disper-

sal rates and events (Brunet and von Oheimb 1998,Clark 1998, Cain et al. 1998, 2000). Nevertheless,concerns persist that potential migration and ad-aptation rates o many tree species may not be ableto keep pace with projected global warming (Davis1989, Huntley 1991, Dyer 1995, Collingham et al.1996, Malcolm et al. 2002). However, these modelsreer to undamental niches and generally ignore theecological interactions that also govern species dis-tributions.

o ecological organization are subject to natural se-lection and contribute to the adaptive capacity or re-silience o tree species and orest ecosystems (Mull-er-Starck et al. 2005). Diversity at each o these levelshas ostered natural (and articial) regeneration oorest ecosystems and acilitated their adaptation todramatic climate changes that occurred during thequaternary period (review by: DeHayes et al. 2000);this diversity must be maintained in the ace o antici-pated changes rom anthropogenic climate warming.

Genetic diversity (e.g., additive genetic variance)within a species is important because it is the basisor the natural selection o genotypes within popu-lations and species as they respond or adapt to en-

vironmental changes (Fisher 1930, Pitelka 1988,Pease et al. 1989, Burger and Lynch 1995, Burdonand Trall, 2001, Etterson 2004, Reusch et al. 2005,Schaberg et al. 2008). Te potential or evolutionarychange has been demonstrated in numerous long-term programmes based on articial selection (Fal-coner 1989), and genetic strategies or reorestationin the presence o rapid climate change must ocuson maintaining species diversity and genetic diversi-ty within species (Ledig and Kitzmiller 1992). In theace o rapid environmental change, it is importantto understand that the genetic diversity and adap-

tive capacity o orested ecosystems depends largelyon in situ genetic variation within each populationo a species (Bradshaw 1991). Populations exposedto a rate o environmental change exceeding the rateat which populations can adapt, or disperse, maybe doomed to extinction (Lynch and Lande 1993,Burger and Lynch 1995). Genetic diversity deter-mines the range o undamental eco-physiologicaltolerances o a species. It governs inter-speciccompetitive interactions, which, together with dis-

i p u cbb, p p j ub

b u. T p b p uu u bu. T p e yuqu n F, Pu r, w k w

hu hu 1989.

Credit:J.Parrotta

7/31/2019 cbd-ts-43-en

16/68

15


One o the best approaches, when dealing with anuncertain uture, is diversication because no singleapproach will t all situations, and this applies alsoto the development o orest management strategies(Ledig and Kitzmiller 1992, Millar et al. 2007). Inthe biological realm, maintaining species and ge-netic diversity addresses the need to be prepared orwhatever environmental changes might happen, andthis is undamental to the concept o resilience. Spe-cies have two main means by which they adapt tochange: they can either disperse by seed or vegeta-tive propagules in the direction o a more avourableenvironment, or they can change their gene requen-cies to avour genotypes (genetic constitutions) thatare better adapted to the changed environment (Bur-

don and Trall 2001, Reusch et al. 2005). Speciesmay also adapt through phenotypic plasticity, i theirgenotype entails a range o permissible responses(with respect to the species morphological, physi-ological, behavioural or lie history strategies andtraits) that are suited to the new conditions (Nusseyet al. 2005).

Seed and pollen dispersal, and gene requencychanges can occur simultaneously and interact inthe process o adaptation. For instance, dispersaloen promotes gene ow among highly ragmented

tree populations; thereby maintaining within-popu-lation levels o genetic diversity and preventing thegenetic dri and loss o genetic diversity that canoccur through inbreeding within small, isolated orragmented tree populations (Hall et al. 1996, Younget al. 1996, Nason and Hamrick 1997, Cascante etal. 2002, Rajora et al. 2002, Fuchs et al. 2003, Mos-seler et al. 2004, Degen et al. 2006, Clouthier et al.2007, OConnell et al. 2007, Farwig et al. 2008). Seeddispersal can occur through wind and water, or viaanimals such as birds, mammals, etc. Operationalorestry experience and observations have shownthat seeds can be dispersed over surprisingly long

distances over relatively short time rames. Seedso light-seeded species, such as coniers, can travellong distances rom the nearest population cen-tres (Cwynar and MacDonald 1987). Coniers withsemi-serotinous cones, such as black spruce (Piceamariana), red pine (Pinus resinosa), and pitch pine(Pinus rigida), or example, seem particularly welladapted or such long-distance dispersal over hard-packed snow and ice. Ritchie and MacDonald (1986)have suggested that wind dispersal over snow mayalso explain the rapid post-glacial migration rateso coniers that have non-serotinous cones, such as

white spruce (Picea glauca). However, long-distanceseed dispersal o typically wind-dispersed conierscould also be explained through dispersal by birds

(Wilkinson 1997). Large or heavy-seeded species,such as those ound in mangroves (Geng et al. 2008),and especially those in highly ragmented environ-ments, may have greater diculty travelling acrosslandscapes (e.g., walnuts [Juglans spp.], hickories[Carya spp.]). Nevertheless, oaks (Quercus spp.)(Skellam 1951, Davis 1981) and American beech(Fagus grandiolia) (Bennett 1985) are capable orapid and widespread dispersal given the presence ocertain animal species.

Generally, by dispersing their seeds and pollen, or-est species can maintain their genetic diversity, andhence their long-term resilience to change overspace and time, by re-establishing themselves else-

where in avourable climates. However, anthropo-genic changes to landscapes and gene pools mayhave reduced this capacity, and population ragmen-tation has the potential to adversely aect the geneticand reproductive status o populations.

We are also concerned with the idea oin situ resil-ience, based on the potential or genetic adaptation,that is, the ability o a orest to maintain itselin situollowing a disturbance, and thereore we ocus morespecically on the role o genetic diversity as a actorin the capacity to adapt to a disturbance. Adapta-

tion in the genetic or evolutionary sense, wherebygene requencies are changed to promote successulgrowth and reproduction in a changed environment,has both short- and long-term components. It isimportant to understand the dierent rates at whichpopulations respond to environmental changes.rees are among the most genetically diverse o allorganisms (Hamrick and Godt 1990) and this diver-sity within natural populations provides the oun-dation or population stability in variable environ-ments (Gregorius 1996). Tis concept has been welldemonstrated with respect to adaptation to potentialpollutants (Pitelka 1988, Berrang et al. 1989, Scholtz

et al. 1989, Bazazz et al. 1995, Kull et al. 1996, Cantinet al. 1997), to pest populations (Burdon and Trall2001), and to various other physiological stresses.High levels o genetic diversity within a larger, lo-cal population or gene pool o a given tree species(e.g., typical boreal or temperate biome popula-tions) allows or a relatively rapid adaptive responseto an environmental challenge. Dierential survivalthrough natural selection pressures may result in anarrowing o the gene pool to promote those geno-types best able to survive disturbances, such as toxicchemicals, pest inestations or other types o inter-

specic competition, climate change, or soil waterand nutrient conditions. In this sense, these localpopulations may contain a subset o genotypes that

7/31/2019 cbd-ts-43-en

17/68

16


generation times approaching those o their preda-tors and parasites (e.g., willows, Salix spp.) - manyo which have generation times o less than a year.Understanding this point is crucial to understandinghow trees adapt and why maintaining natural levelso genetic diversity is so important.

Genetic changes to the gene pool based on the ac-tions o natural selection on the extant genetic di-versity o in situ gene pools can ollow a relativelyrapid population decline or collapse ollowing a dis-turbance, such as a major pest inestation. Tis pro-cess can then be reinorced by a longer-term process,whereby gene requencies change more slowly in thedirections orced by natural selection over many

generations o subsequent breeding and reproduc-tion. Individuals surviving a disturbance interbreedand propagate, avouring the gene requencies o thesurviving individuals. Over time, these gene re-quency changes are enhanced and rened to createa better-adapted population. However, species thathave inherently low levels o natural genetic diversitymay not be able to adapt to relatively sudden chal-lenges. For example, red pine is a tree species nativeto eastern North America that shows extremely lowlevels o detectable genetic diversity (Mosseler et al.1991, 1992, DeVerno and Mosseler 1997). Natural

populations o this species are vulnerable to pest in-estations and inections by ungal pathogen such asArmillaria spp. and Sclerroderris lagerbergii, whichcan eliminate entire populations (e.g., McLaughlin2001).

Diversity at the genetic level must also be comple-mented by diversity at the species level, particularlyby species groups such as pollinators (e.g., insects,bats, birds) and seed-dispersal organisms (e.g., manybirds and mammals) that may aect the long-termresilience o orest ecosystems. Without these asso-ciated species groups, tree species may be restricted

are pre-adapted (sensu Davis and Shaw 2001, Jumpand Penuelas 2005) to environmental changes. Us-ing experimental populations o yellow birch (Betulaalleghaniensis), Bazazz et al. (1995) demonstratedthe potential or populations to respond to varyinglevels o CO2, and the genetic complexity and mag-nitude o genetic responses to population actorssuch as density and competitive interactions. Suchexperiments demonstrate the overall capacity orresilience o orest tree populations to anticipatedincreases in CO2 or ozone (Berrang et al. 1989; e.g.,in aspen [Populus tremuloides]) or combinations othese gases (Kull et al. 1996) based on extant levelso genetic diversity within populations at any giventime. Tese kinds o experiments also indicate how

dicult it is to predict the way in which species willrespond to anthropogenically-caused changes (Ba-zazz et al. 1995), or to other environmental changesin the uture (DeHayes et al. 2000).

Concerns have been expressed that predicted cli-mate changes (IPCC 2007) may occur too quicklyor species to adapt (Huntley 1991, Davis and Shaw2001, Jump and Penuelas 2005), but genetically di-verse species are capable o rapid evolution (Geberand Dawson 1993). Many species have adapted torapid changes and have done so repeatedly over geo-

logical time through dispersal and genetic changesbased on the extant genetic diversity within local orregional gene pools, suggesting long-term genetic-based resilience to change. Tere is considerableevidence or adaptation in the geological and ossilrecord (Bernabo and Webb 1977, Webb 1981, Davis1983, Huntley and Birks 1983, and review by Geberand Dawson 1993). Such adaptation has been dem-onstrated by orest plants during or ollowing pastglacial and interglacial episodes, which were charac-terized by relatively rapid climate change (Huntleyand Webb 1988).

Nevertheless, a common misunderstanding persistsabout the nature o genetic adaptation in specieswith long generation times. Te general perceptionseems to be that, given the long-generation timeso many long-lived tree species, trees are at a severedisadvantage in terms o a suitably rapid responseto environmental challenges. However, trees are notentirely dependent on their generation time to dem-onstrate an adaptive or evolutionary response, butcan respond reasonably rapidly based on the inher-ently high levels o genetic diversity that character-ize most tree species. I evolution and adaptation in

species with long generation times were dependenton generation time, there would be no trees le onEarth with the possible exception o those that have

e u Bf n- Pk, ab, c

Credit:iStock000007432521

7/31/2019 cbd-ts-43-en

18/68

17


in their ability to adapt to change through seed dis-persal, pollination, and gene ow important pro-cesses or maintaining genetic diversity and repro-ductive success within populations. For example, acertain amount o gene ow among populations isrequired to minimize the adverse eects o inbreed-ing and inbreeding depression on growth, reproduc-tion, survival, and genetic diversity in small, isolatedpopulations o species in highly ragmented land-scapes. Small, isolated populations at the margins

o the geographic range may also be o special im-portance to the resilience o orests under climate-change scenarios because such population islandsoen serve as well-adapted seed sources or popu-lation migration under environmental change (Cw-ynar and MacDonald 1987). It can be assumed thatsuch populations at the geographic-range marginshave experienced some physiological stresses whileliving at the limits o their eco-physiological toler-ances. Such populations may have become adaptedthrough natural selection and some degree o geneticisolation (Garcia-Ramos and Kirkpatrick 1997) andcontain special adaptations that may enhance their

value as special genetic resources or adaptation andresilience to change.

3. The relationships among

biodiversity, productivity and

function, and resilience and stability

We review published information on the relationship

between biodiversity and productivity to provide

an understanding of the mechanisms that may be

important to function in forests systems. Through

this review, we suggest below that there is a funda-mental relationship among biodiversity, production,

and resilience and stability in forests and that this

relationship is important with respect to adaptive

management in forests under climate change. Here

we consider climate, weather conditions, soil par-

ent material as extrinsic (exogenous) physical inputs

to terrestrial ecosystems and the role of species as

intrinsic (endogenous) to ecosystem functioning.

There is considerable ongoing debate over the role

that biodiversity plays in ecosystem function and

stability owing to the highly complex nature of the

relationships among species and the synergistic roles

of extrinsic factors and intrinsic factors, including

genetic factors, in ecosystems (see e.g., Waide et

al. 1999, Kinzig et al. 2001, Loreau et al. 2002, for

summary discussions). Nevertheless, in the absence

of biodiversity there would be no ecosystems and no

functioning. Further, there is evidence that complexforest ecosystems are more productive than less di-

verse ones (under the same conditions) (e.g., Phil-

lips et al. 1994), and generally that forest systems

comprised of few species are highly prone to various

catastrophes including disease and invasion (Scher-

er-Lorenzen et al. 2005).

3.1 Teoretical background

Te relationship between diversity and productiv-ity is variable (Waide et al. 1999) and dependent

on the scale considered (Chase and Leibold 2002).Much o the work done to understand the relation-ship between species diversity, ecosystem processes,and production has necessarily been done in highlycontrolled low-diversity systems at small scales, es-pecially using grasses (e.g., ilman and Downing1994, ilman et al. 1996, Hector et al. 1999, Hector2002), or in other controlled systems (e.g., Naeem etal. 1995). Few studies have examined more connect-ed systems with multiple trophic levels and complexproduction webs, such as orests, nor have they con-sidered larger scales. While the work on simple tro-phic systems has, at best, limited applicability in or-

ests, it does present theoretical predictions or whatspecies do in ecosystems and so is briey discussedhere. In particular, two main competing hypotheseshave been identied to predict the relationship be-tween biodiversity and productivity in ecosystems:the niche complementarity hypothesis (ilman et al.1996, ilman and Lehman 2001) and the samplingeect hypothesis (Aarssen 1997, Doak et al. 1998).Under either hypothesis, a certain level o saturationis expected where no more eective use o resourcescan be achieved regardless o increased species rich-ness (Hooper et al. 2005).

Tese hypotheses are related to some earlier alter-nate constructs, including: the rivet hypothesis,

nu w az 18 u. rw

p bk pu

. P tb, Bz.

Credit:J.Pa

rrotta

7/31/2019 cbd-ts-43-en

19/68

18


where individual species are suggested to perormadditive roles (Ehrlich and Ehrlich 1961); the key-stone hypothesis, postulating that some species aresubstantially more important than others in control-ling productivity, and which is closely related to theredundancy hypothesis, which suggests that mostspecies live o excess energy in the system or playminor roles in production and so are largely insig-nicant in ecosystem unction.

Te niche complementarity (or niche dierentia-tion) hypothesis (see above) predicts that as speciesare added to a system, the productivity in the systemwill increase until vacant niches are lled because oeective partitioning o resources. Te coexistence

o species then is assured through interspecic di-erentiation as a direct response to competition orresources. I species are able to avoid competition byoccupying dierent niches, then production in thesystem will increase accordingly (e.g., ilman andLehman 2001, ilman et al. 2002). Niche dierentia-tion models also consider the concept o acilitation,where one or more species may enhance the capac-ity o another species to survive and reproduce (e.g.,ectomycorrhizal ungi on tree roots or legumes ingrasslands). However, ew keystone unctional rolesamong plants are known (e.g., C3 and C4 grasses,

nitrogen xers).

A competing model, the sampling (or selection) e-ect hypothesis, suggests that dominant competitors(sampled rom the regional species pool) will playthe greatest roles in ecosystem unctioning and asdiversity increases, unctioning in the system willbe controlled by these dominant species because otheir greater likelihood o being present in a diversesystem (e.g., Aarssen 1997, Huston 1997). Tis resultis achieved because the best competitors will alwayscontrol resources within a system. Niche dieren-tiation models predict coexistence among species,

while sampling eect models predict dominance byone or a ew species, especially or systems in equi-librium. Various studies suggest support or one orthe other o these models (e.g., Hooper and Vitousek1997, ilman at al. 2002, Hooper and Dukes 2004)or suggest that the capacity to conduct the experi-ments has been limited by almost intractable designproblems or analysis constraints (e.g., Huston 1997,Allison 1999, Schmid et al. 2002).

Tese two competing hypotheses will be aectedby scale o observation (Waide et al. 1999) and little

inormation is available at large scales such as or-est landscapes. Chase and Leibold (2002) workingwith production in pond systems ound productivity

F w p u

qu

declined with species richness at a local scale (uni-modal) but was monotonically increasing at regionalscales, but these patterns dier depending n the eco-system type (e.g., Waide et al. 1999). Measuremento orest production will be similarly inuenced bythe scale o measurement. Mechanisms or dier-ent responses at small and large scales might includeregional heterogeneity in environmental or edaphicconditions, dierent orest communities, or multiplestable states or the same orest system.3.2 Evidence o a diversity-productivity relation-ship in orests

esting the theories o the relationship between

diversity, productivity, and resilience in orests isdicult owing to the inability to control eitherextrinsic or intrinsic variables within these complexecosystems. Furthermore, niche partitioning is well-known in orests (e.g., Leigh et al. 2004, Pretzsch2005), with many uncomplicated examples such astap and diuse rooting systems, shade tolerant andcanopy species, and xeric and hydric species. Someconounding eects also aecting production inorests include successional stage, site dierences,and history o management (Vila et al. 2005). Speciesmixtures change with successional stage in orests,

rom those rapidly-growing species avouringopen canopy environments to those capable oreproducing and surviving in a more shaded canopyenvironment. Various plant species are adapted tosite types that are dened by soils, topography, andmoisture levels, but opportunistically may be oundacross a range o sites. Many orests, includingmost temperate orests, have undergone manydirect anthropogenic-related changes and so anunderstanding o community structure must be inthe context o the human history related to the stand.

Credit:A.nce

7/31/2019 cbd-ts-43-en

20/68

19


almost double the biomass. Pretzsch (2005) andJones et al. (2005) provided separate examples ocomplementarity between tree species in long-term,simple two species mixture experiments. Vila et al.

(2005) ound that overall production in Catalonianopen canopy orests was superior or mixed speciesstands than or pure stands, although individualproduction within the dominant species was nothigher, indicating an ecosystem rather than anindividual response. Schulze et al. (1996) oundno evidence that mixed species had a positive eecton production in European temperate stands andEnquist and Niklas (2001) reported no relationshipbetween plant diversity and total biomass intheir stands. Using experimental tropical treeplantations, Healy et al. (2008) used redundancy

analysis to suggest that diversity explained 23-30%o the variance in productivity (environmentalactors explained the rest). In boreal orests, jackpine (Pinus banksiana) was observed to have greaterdiameter when growing in mixedwood stands, asopposed to in pure stands on similar sites and at thesame ages (Longpr et al. 1994), suggesting a levelo complementarity. Wardle et al. (1997) ound arelationship between increasing plant unctionaldiversity and orest production (including biomassaccumulation) ollowing varied re requency, onisland systems in hemiboreal Sweden. Caspersen andPacala (2001) ound a positive relationship between

carbon storage and high tree species diversity,compared to lower carbon storage in stands with lowtree species diversity, across multiple types o orests.

Fw u p ub p, bk p b p . l P tb,

P s, Bz.

For example, long-term selection harvesting mayhave reduced relative abundances among tree speciesin a given stand, thereby altering the competitiveconditions and stand production. Developing aclear understanding o the species-productivityrelationship in orests must take these several actorsinto account, use a very broad sampling approach,and/or test the relationship experimentally to controlthe various actors.

Several orest studies have ound a positiverelationship between diversity and productionin stands, while ewer have not. O the 21 studiesconsidered in our review (excluding studies usingherbicides, thinning, ertilization, and N-xing

acilitation to eliminate conounding eects), 76%suggested a positive eect o mixed species (i.e.,number o species) on ecosystem production (table1). In plantations, the eects o mixing species canbe neutral owing to competition and so the resultso such experiments can be directly related to thespecies mixtures that were selected. On the otherhand, acilitation and additive eects on meanannual increment were seen in many studies (Kelty2006, Piotto 2008), especially in studies where annitrogen-xing species was included (Forrester et al.2006, Piotto 2008)

In Costa Rica, Ewel et al. (1991) experimentallydeveloped orest communities on burned plots.Tree treatments involved various successionalcommunities, while a ourth limited productionto a sequence o monocultures. Tey ound thatthe multi-species plots developed much higher soilertility over time than did monocultures, indicatingsuperior production and nutrient retention incomplex systems. Ewel et al. (1991) also noted muchgreater depletion o soil nutrients in short-livedmonocultures than in stands using perennial plants.

Also in Costa Rica, tree species richness wascorrelated to increased production in aorestationexperiments by Montagnini (2000), a result alsoreported by Erskine et al. (2006) or Australiantropical plantations o individual and mixed species.In one o the ew published studies not to reporta positive relationship between production anddiversity, Finn et al. (2007) ound that Australiantropical plantations that had been invaded byendemic species rom nearby natural orests did notresult in increased production, presumably becauseo inter-specic competition eects. Parrotta

(1999) was able to show acilitation eects in mixedplantings o tree species in experimental tropicalplantations, with mixed species plots producing

Credit:J.Parrotta

7/31/2019 cbd-ts-43-en

21/68

20


Table 1. Summary of published studies in forests that tested the relationship

between species richness and some measure of production (e.g., biomass

increment, soil C, etc.).

Studies testing eects o herbicides, thinning, ertilization, and nitrogen-xing plant acilitation wereexcluded. Observational reers to studies where data were gathered rom existing orest stands andexperimental reers to directed planting or removal experiments. See text or details o individual studies.

ef up p pu

au/ F pob

expP nu

Prokopev 1976 Boreal Expt X

Ewel et al. 1991 ropical Expt X

Longpr et al. 1994 Boreal Obs X

Schultze et al. 1996 emperate Obs X

Wardle et al. 1997 emperate Expt X

Parrotta 1999 ropical Expt X

Enquist and Niklaus 2001 emperate Obs X

Casparsen and Pacala 2001 emperate Obs X

Schroth et al. 2002 ropical Expt X

Petit and Montagnini 2004 ropical Expt X

Pretsch 2004 emperate Expt X

Jones et al. 2004 emperate Expt X

Vil et al. 2004 emperate Obs X

Erskine et al. 2006 ropical Expt X

Bristow et al. 2006 ropical Expt X

Finn et al. 2007 ropical Expt X

Kirby and Potvin 2007 ropical Obs X

Healy et al. 2008 ropical Expt X

Murphy et al. 2008 ropical Expt X

Piotto 2008Meta-analysiso 14 plantationstudies

Expt X

7/31/2019 cbd-ts-43-en

22/68

21


Tey concluded that orest managers should attemptto retain species diversity to increase productionand especially manage or species that maximizeunctions o interest, such as carbon storage.

Some o the above studies are within-site types andsome are across-site types. Across-site comparisonsprovide more variable results than do the within-sitecomparisons, as might be expected because largerscales include potentially conounding eects ohabitat variability, range boundaries, and dierentclimates. Depending on scale, these studies provideevidence that more diverse orests are generallymore productive than orests with low speciesdiversity. Further, many studies indicated that

carbon sequestration, a requently measured variableamong the studies, is enhanced by the presence omultiple complex levels o unctional groups inorests. Tis notion is urther supported by severalrecent studies showing that complex old-growthorests provide high-value carbon sinks and maycontinue to do so or centuries in all orest biomes,unless disturbed (Phillips et al. 1998, Baker et al.2004, Luyssaert et al. 2008, Lewis et al. 2009). In onlyone o these cases (table 1) was the direct additive orsynergistic relationship between number o species(or unctional species) and ecosystem productivity

quantiable, owing to the complexity o thesesystems. Te experimental data (table 1) all comerom two-or ew-species plantations, somewhatsimilar to the evidence rom highly controlledgrassland systems. Nevertheless, it is doubtul thatevidence o a biodiversity-productivity relationshipin orests can be derived experimentally in naturalorests through removal experiments, owing to thelarge number o uncontrollable variables, such as sitedierences and tree densities.

Mechanisms o complementarity eects observed inmixed species orest stands may be nutritional, as a

unction o improved soil condition (e.g., Ewel et al.1991, Brantberg et al. 2000, Hattenschwiler 2005), orrelated to improved partitioning o resources throughdierent rooting patterns and depths (Schmid andKazda 2001). While Scherer-Lorenzen et al. (2005)suggested that diversity matters less than expectedwith respect to its contribution to biogeochemicalcycles, Hooper et al. (2005) concluded that certaincombinations o species are indeed superior in termso soil nutrient retention and production. Clearlymore evidence is required to reduce the uncertaintyassociated with how complementarity operates

in orests. Arguably these various results maysupport either the niche dierentiation hypothesisor the sampling eect hypothesis and the evidence

supporting one over the other is sparse. However,the common theme rom most studies is that diverseorests are more productive than low-richness orestsand that unctional diversity within systems mattersconsiderably. Te evidence broadly supports theconcept that diverse orests provide more goods andservices than do orests with low species richness,especially planted orest monocultures (e.g., Pearceand Moran 1994, Srivasteva and Velland 2005, Diazet al. 2005, Dobson et al. 2006).

Many authors have advocated, and indeeddemonstrated, that it is not diversityper se thatinuences production and resource dynamics butrather it is the number o unctional species (or

unctional diversity) that is important (e.g., Phillipsand Gentry 1994). While studies have indicated alink between plant species richness and ecosystemproductivity (Phillips and Gentry 1994, Symstad et al.1998, Wardle et al. 1999, Schwartz et al. 2000, Schmidet al. 2002, ilman et al. 2002, Hector 2002), speciesrichness and unctional richness are not necessarilycorrelated (Diaz and Cabido 2001, Hooper et al.2005). Certainly, some species play much greaterunctional roles in systems than do others (Walker1994, Schlaper and Schmid 1999, Chapin et al.2000, Diaz and Cabido 2001), but species-specic

unctional roles may be idiosyncratic, with dierentkey species among similar ecosystems (Phillips et al.1994, Hooper et al. 2005). Nevertheless, most dataand almost all examples in the summary by Diaz andCabido (2001) come rom manipulated controlledsystems, especially relatively simple grasslands.Te concept o unctional diversity is compatiblewith either the niche complementarity or samplingeect hypotheses. Dierent unctional types couldcompete or the same resource or be sucientlydissimilar to occupy dierent niches within the samesystem.

3.2.1 d-pu p

Stone et al. (1996) concluded that more productiveecosystems are more resilient than less productiveones, and hence recover more rapidly romdisturbances. Functional diversity in orests isrelated to production in the ecosystem (Chapin etal. 1997, Diaz and Cabido 2003), and many speciesin orests appear to be redundant in terms o totalproduction (Pretzche 2005). Redundancy, whichis also reerred to as the insurance hypothesis

(Naeem 1998, Yachi and Loreau 1999), appears tobe a common and important trait in most orestsystems, contributing to their resilience ollowing

7/31/2019 cbd-ts-43-en

23/68

22


various disturbances, protecting against eectso species loss, or responding to environmentalchange. For example, several tree species havebeen lost, or substantially reduced in abundance,in temperate orest ecosystems, with little or noloss o productivity in that broad orest system(e.g., Pretzsch 2005), suggesting compensation byother species. Tereore, the redundancy providedresilience in terms o maintaining productivity inthe ace o species loss. Redundancy can also conersystem resilience and/or resistance in response tothe impact o disease and pests (see below: Jactelet al. 2005, Pautasso et al. 2005). Te resiliencethat redundancy provides in maintaining systemproductivity in response to species loss, disease and

pests, may not necessarily compensate or otherecosystem goods and services. For example, losso a particular species that had specic cultural oreconomic importance would mean a less valuableorest (e.g., Hooper et al. 2005). Furthermore, theremay not necessarily be redundancy or certainunctional species, such as nitrogen-xers, and theirloss would then have consequences or ecosystemprocesses (Brown et al. 2001).

While the evidence above supports the notion thatmixed-species orest ecosystems are more resilient

than monotypic stands, some natural monotypic, ornearly monotypic, orests do occur. For example, inthe boreal biome, natural stands o jack pine (Pinusbanksiana), Scots pine (P. sylvestris), lodgepole pine(P. contorta), and Dahurian larch (Larix gmelinii)are commonly dominated by single species. Tesestands sel-replace usually ollowing re over largelandscapes, with no change in production over time.Similarly, in wet boreal systems where re is absent,monotypic stands o a single species o r (Abiesspp.)occur and generally sel-replace ollowing insect-caused mortality. Generally, these monodominantboreal orest ecosystems tend to be relatively short-

lived and are prone to re or insect inestation, andso while not very resistant (relative to other oresttypes), they are highly resilient ecosystems despitetheir lack o unctional types and redundancy. Tehigh degree o seasonality in boreal orests maycontribute to the resilience among boreal monotypicstands, compared to in temperate and tropicalbiomes (Leigh et al. 2004), where orest speciesrichness is considerably higher (greater than anorder o magnitude) than in the boreal biome. Onlya ew types o monodominant canopy stands are alsoound in temperate orest types, such as pines and

eucalypts, or in tropical orests (e.g, Gilbertiodendron sp.).3.3 Diversity and stability

cu k a u lk t, c

For a system to have resilience, the state o interest(e.g., the mature orest type) must be stable over acertain time period. Considerable research hasexplored the concept that species diversity enhancesstability, dened as variation within dened bounds(time and space) or dynamic equilibrium, inecosystem processes in response to environmentalchange (e.g., Loreau 2000, Hughes et al. 2002).Te relationship between diversity and stability iscomplex and may resist generalization. Conusionover this issue stems rom debate over whetherstability reers to individual populations within

ecosystems or the stability o ecosystems and theirprocesses. For example, relatively recent work hassuggested that as diversity increases, stability withinindividual population declines (e.g., Moat 1996,ilman and Lehman 2001).

Ecosystems respond to environmental changethrough unctional compensation, or the dynamiccapacity o systems to maintain production, eventhough levels o output among individual speciesmay change (e.g., Loreau 2000). Tis concept isclosely linked to that o unctional redundancy indiverse ecosystems (Naeem 1998, Yachi and Loreau1999). Dynamic responses in diverse ecosystemsthat maintain stability to environmental change overtime may occur at genetic, species, or populationlevels. Tere appears to be low variability amongecosystem properties in response to change indiverse systems compared to those systems with lowdiversity, where higher variance is observed (Hooperet al. 1995, Ives et al. 1999, Lehman and ilman 2000,Hughes et al. 2002).

Loreau et al. (2002) noted the importance o regionalspecies richness that enables migration into systems

as a means to enhance ecosystem adaptability tochange over time. Immigration could enhanceboth genotypic and phenotypic responses to

Credit:J.Veilleux

7/31/2019 cbd-ts-43-en

24/68

23


environmental change enabling resilience in thesystem through compensation. Overall, the evidenceis consistent with the concept that diversity enhancesthe stability o ecosystem processes (Hooper et al.2005) and the ow o goods and services.

Ecosystems may exist in more than one stablestate (Holling 1973), a act supported by someexperimental evidence largely rom closelycontrolled experiments (Schroder et al. 2005).Drever et al. (2006) provided several examples oalternate stable states among the orest biomes. Itseems intuitive that orest ecosystems have multiplestable states that depend on the kinds o disturbancesthat orests regularly undergo (Marks and Bormann

1972, Mayer and Rietkerk 2004, Schroder et al.2005) and that many or all o these alternativestates may deliver similar goods and services. Forexample, regeneration trajectories ollowing wildredier in many orest types depending on previousdisturbances, intensity o the re, time since last re,whether or not a re occurs in a year with abundanttree seed, level o endemic insect inestation, ageo the trees, and many other actors (Payette 1992,Little et al. 1994, Hobbs 2003, Baeza et al. 2007).While the engineering resilience may be low, inthat the identical or similar species mix may not

result ollowing recovery rom the disturbance, theecological resilience is high because a orest ecosystemis restored. Lack o convergence to pre-disturbanceoristic composition does not necessarily imply alack o resilience with respect to other orest systemcharacteristics. Rather it implies that successionalpatterns dier depending on circumstances but thatthe system is ecologically resilient, even though thedominant canopy species composition has changedalong with certain ecological processes.

Te capacity o an ecosystem to stay within stablebounds is related to slow processes that can move

the system to another state, sometimes a state that isundesirable, rom a human perspective (Scheer andCarpenter 2003). Folke et al. (2004) suggested thatbiodiversity is one o those slow-changing variablesthat have consequences or ecosystem state, actingprimarily through species with strong unctionalroles. Te capacity o systems to maintain stabilityin the ace o environmental change is also relatedto the capacity o individuals within species to meetchallenges and to the possibility that other speciesmay increase their unctionality under changedregimes (biodiversity as insurance). A major

actor impeding the recovery and stability o orestecosystems is degradation and loss o unctionalspecies and reduced redundancy caused by land

use practices, including unsustainable harvesting.Degradation results in the ecosystem moving toan undesirable state that may have its own highresilience but be undesirable in terms o the reducedgoods and services that it provides.

3.3.1 d

Another measure o stability, and ultimately oresilience in the case o orest pests, is the capacityo an ecosystem to resist invasion by non-localspecies (i.e., community invasibility). Variousactors, both extrinsic and intrinsic to an ecosystem,such as availability o niches, system degradation,and ragmentation, may aect the capacity o

alien species to invade. Another actor which maypromot

cbd-ts-43-en

Documents