a comparison of the fate and effects of two pyrethroid insecticides (lambda-cyhalothrin and...

Ecotoxicology 4,219-244 (1995)

A comparison of the fate and effects of two pyrethroid insecticides (lambda-cyhalothrin and cypermethrin) in pond mesocosms D E B O R A H F A R M E R , I A N R. H I L L and S T E P H E N J. M A U N D *

Ecology and Soil Science Section, Zeneca Agrochemicals, Jealott's Hill Research Station, Bracknell, Berkshire, RG12 6EY, UK

Received 14 July 1994; accepted 30 November 1994

The fate and effects of two pyrethroid insecticides (lambda-cyhalothrin and cypermethrin) were investigated in replicated 25 m 3 pond mesocosms. Three pesticide treatments which simulated spray drift deposition were examined: 0.7 g a.i. ha -~ cypermethrin and 0.17 and 1.7 g a.i. ha -1 lambda-cyhalothrin. Based on the use rate and pesticidal activity of the chemicals, the cypermethrin and lower lambda-cyhalothrin rates were approximately equivalent. After applications, pyrethroid residues in the water column declined rapidly. Treatment-related effects were observed on some macroinvertebrate taxa, most notably the Asellidae and Gammaridae. Surface- dwelling insects also suffered initial knock-down, particularly in the 1.7 ga.i. ha-' lambda-cyhalothrin treatment, but there was recovery after the spray period. No adverse effects occurred on algae, macrophytes or zooplankton, but there were occasional enhancements (e.g. algal biomass and abundances of copepod nauplii and Rotifera) which may have been indirect effects. An overall comparison of the treatments indicated that the higher lambda-cyhalothrin rate had the greatest effects, whilst thd cypermethrin application had a somewhat greater impact than the lower lambda-cyhalothrin treatment rate (due to effects on peracarid crustaceans). The study indicated that should spray drift occur at the levels expected for either pyrethroid's normal use patterns, potential impacts on natural aquatic ecosystems would be minor and transient.

Keywords: pyrethroids; spray drift; mesocosm; residues; aquatic effects.

Introduction

Aquatic mesocosm studies are sometimes used in pesticide registration to generate data on environmental fate and ecological effects, where laboratory studies indicate there may be potential environmental concerns. Such studies have often been performed with lipophilic chemicals, such as pyrethroid insecticides, because their fate in the field differs substantially from that in standard laboratory toxicity tests. This is due to rapid and strong adsorption to organic matter and other surfaces, resulting in decreased exposure and a reduction in effects on organisms in nature (Hill 1985). Although such factors can often be predicted adequately from fate models and laboratory toxicity data, mesocosm studies have been used in the past to reduce uncertainty of such predictions by providing semi-natural conditions under which to measure the fate and effects of the chemical (see Hill et al. (1994a) for a review of previous aquatic field studies with pyrethroids).

*To whom correspondence should be addressed.

0963-9292 © 1995 Chapman & Hall

220 Farmer, Hill and Maund

This mesocosm study was performed with two pyrethroid insecticides, cypermethrin and lambda-cyhalothrin and was designed to assess and compare their fate and effects. Cypermethrin (trade names Cymbush, Cymperator and Demon) was first sold in the late 1970s and is now used throughout the world on a wide range of crops due to its cost-effective control of insects and low mammafian toxicity relative to other insecticides. Lambda-cyhalothrin (trade names Karate, Icon and Demand) is a newer pyrethroid, first marketed in 1985. It has similar properties and use patterns to cypermethrin but has better pesticidal efficacy at lower use rates. Lambda-cyhalothrin is also used for the control of public health pests and was the first insecticide to be passed by the World Health Organization Pesticide Evaluation Scheme (WHOPES) for the control of malaria and other tropical diseases.

Due to its greater pesticidal activity, lambda-cyhalothrin is applied at lower rates. For an equivalent use pattern, it is therefore expected to be found at lower concentrations in the environment. Chemical application rates for this study were selected to simulate potential spray drift rates within the range that would be expected from normal agricultural usage.

Methods

Test systems

The study site was located at Zeneca (formerly ICI) Agrochemicals, Jealott's Hill Research Station in Berkshire, UK. The eight mesocosms used in this study were constructed in 1984 and consisted of two groups of four steel-reinforced concrete tanks, each of 5 m × 5 m, with vertical walls of 1.2 m height (Fig. 1). The mesocosms were first filled with 15 cm of thoroughly mixed hydrosoil, which was obtained from local ponds and ditches. Mains water was then added to a depth of 1 m and organisms collected from nearby natural ponds were added to establish the biota. A number of macrophyte species colonized the mesocosms; some were initially present in the hydrosoil and others had colonized by the start of the pre-treatment year. The mesocosms were all interlinked by a water circulation system (Fig. 1) and from September to November 1985 (the pre-treatment year), water was circulated around the mesocosms at a flow rate of 2000 1 h -1 via a common mixing-chamber to promote uniformity of water quality and biota. Surveys were carried out in the pre-treatment year to establish baseline data for physicochemical and biological parameters. Prior to pesticide treatment, the water circulation system was stopped and sealed, leaving each mesocosm as an isolated experimental unit. Macrophyte growth was removed so that 60% of the basal area of each mesocosm was open water in order to ensure that the test chemicals reached the water column.

Pesticide application

Cypermethrin and lambda-cyhalothrin are broad-spectrum insecticides with similar structural and physicochemical properties (Table 1). During normal agricultural usage, it is possible that small amounts of either chemical may enter aquatic ecosystems via spray drift. The estimated percentage of the field application rate which may drift onto aquatic environments differs according to the crop which is being treated and the method of application. From aerial spraying of pyrethroids (commonly used in the USA for cotton) this has been estimated at 2% of the nominal field rate (Hill et al.

Fate and effects o f two pyrethroids in mesocosms 221

Pump chamber

Mains (city) water . . . . . ~ J , i =

I Inflow pipe to mesocosms .-...ID- I

I

,,

0.7 g ~01~

cypermethrin - I

1.7 g ai/ha lambda-cyhalot h rin

l 0.17 g a i / h a

lambda-cyhalothrin

I

I I P "

Control

5m i

- - I ~--- Overflow

Outflow pipe from mesocosms

0.17 g ai/na lambda-cyhalolhdn

0.7 g ai/ha cypermethrin

Control

1.7 g ai/ha lambda-cyhalothrin

Fig. 1. Plan of study mesocosms showing water circulation system and allocation to experimental treatments.


Table 1. Structural and physical properties of cypermethrin and lambda-cyhalothrin

Cypermethrin Lambda-cyhalothrin

Structure R = CI R = CF3

Chemical name O CN

( RS)-tr-cyano-3-phenoxybenzyl (1RS)-cis, trans-3-(2,2- dichloro-vinyl)-2,2- dimethylcyclopropane- carboxylate

Water solubility at pH 7 0.004 (mg1-1)

Octanol : water partition 6.6 coefficient (log P)

Organic carbon adsorption 160 000 (Ko¢ ml g-l)

Vapour pressure (ram Hg)

1:1 mixture of Z(1R,3R,o~S) and Z(1S,3S,otR), esters of a~-cyano-3-phenoxybenzyl 3- (2-chloro-3,3,3-trifluoroprop- 1-enyl)-2,2-dimethylcyclopropane-carboxylate 0.005

7.0

180 000

1.4 x 10 -9 1.5 × 10 -9

1994b). In Europe, application is usually performed by ground hydraulic or air-blast equipment on a wide range of crops including cereals, vines, hops and top fruit. Worst-case estimates of spray drift into surface waters for the crops and seasons in which pyrethroids are used range from 0.6% of the applied rate (ground hydraulic to field crops) to a maximum of 10% drift (air blast to top fruit) onto a water body which is 5 m downwind of application (Ganzelmeier et al. 1993). These drift rates decrease as the distance from a water body increases, such that a water body 20 m downwind from an application would be expected to receive only 0.1% and 1.5% from field crops and top fruit, respectively.

Typical application rates for lambda-cyhalothrin range from 7.5 g a.i. ha -1 for field crops to 30 g a.i. ha -1 for top fruit and vines. Cypermethrin is applied to similar crops at approximately four times these amounts because on a weight for weight basis, lambda-cyhalothrin is approximately 4-fold more active than cypermethrin. In order to compare realistically the fate and effects of the two chemicals on a basis of spray-drift deposition rates, it was therefore necessary to apply four times as much cypermethrin as lambda-cyhalothrin. Consequently, cypermethrin was applied to mesocosms at 0.7 ga.i. ha -x and lambda-cyhalothrin at 0.17 ga.i. ha -x. Both of these treatments are approximately equivalent to drift rates of 2% from field crops. An additional higher rate of lambda-cyhalothrin was applied at ten times the lower rate (1.7 g a.i. ha -x) against which any effects at the lower application rates could be compared. Each pesticide was applied on four occasions, with 2 week intervals between each application. Applications were made by spraying evenly over the surface of each mesocosm using a hand-held spray boom which spanned the width of a mesocosm and was carded by two people. The boom carried nine spray jets at 50 cm intervals.

Fate and effects of two pyrethroids in mesocosms 223

Sampling programme

During 1986 (the treatment year), sampling started 5 weeks prior to the first application and continued for 14 weeks after the final application. In general, physicochemical monitoring and biological sampling were carried out at various times in week-long sampling periods. The first 'sampling week' was week - 5 (sampling weeks were numbered relative to the first application in week 0). For the majority of ecological end-points, sampling was then performed subsequently at 2 week intervals from week - 1 to week + 13 and finally during weeks + 16 and + 19. However, zooplankton were additionally sampled every week from week - 1 to week + 19 and emerging insects were sampled every week from week - 1 to week +18. Hydrosoil and water residues were also collected at 2 week intervals, but additional water samples for residue analysis were collected 2 and 6 days after the first, second and fourth applications and 1 h, 24 h, 3 and 6 days after the third application.

Residue sample collection and analysis

Water residue samples were collected directly from the mesocosms using a hand-held vacuum pump which drew water through a cartridge containing an adsorptive matrix. Separate samples were taken from 15 cm below the water surface and 15 cm above the hydrosoil. Cartridges were fortified with an internal standard and extracted, The extracts were cleaned up and examined for pyrethroid residues by capillary gas-liquid chromatography (Hadfield et al. 1992). Limits of determination (lod) were 1 ng1-1 for each of the two enantiomer pairs of lambda-cyhalothrin and 2 ng 1-1 for each of the four isomer pairs of cypermethrin.

Hydrosoil cores were removed for residue analysis using 5 cm diameter thin-walled plastic tubes. The cores were frozen and cut into 2.5 cm depth sections prior to analysis. After fortifying with an internal standard, samples were extracted and the extracts cleaned up before analysing for pyrethroid residues by capillary gas-liquid chromatography. The lod was 0.2/~gkg -1 dry weight for each lambda-cyhalothrin enantiomer pair and cypermethrin isomer pair.

Physicochemical measurements

Water temperature, dissolved oxygen, pH and conductivity were measured in situ throughout the study using portable meters. In addition, samples were removed for measurements in the laboratory of turbidity, alkalinity, total organic carbon, nitrates and phosphates. At the start and end of the study, water and hydrosoil samples from each mesocosm were fully characterized. Water samples were screened for the presence of pesticides, other contaminants and physicochemical parameters including pH, alkalinity, hardness, conductivity, suspended solids, biological and chemical oxygen demand, nutrients and minerals. Hydrosoil was analysed for pesticides, pH, organic matter content, cation exchange capacity, textural class and microbial content.

~iological assessments

Plant communities were measured by a number of different methods. Phytoplanktonic algae were collected using a depth-integrating perspex tube sampler to remove water column samples from the entire 1 m depth profile. Samples were collected from three different locations in each mesocosm and combined. Periphytic algae were collected on 1 m long plastic substrates suspended at three sampling positions in the mesocosm.


Each substrate contained three strands stretching from the water surface to the hydrosoil. For both phytoplankton and periphyton, subsamples containing a minimum of 250 cells were removed in the laboratory and these samples were enumerated and identified to species.

Algal biomass was determined from cell biovolume measurements. Population biomass was then estimated assuming a mean density of 1.0 g m1-1. Primary production was estimated using a number of methods including light/dark bottle determination of photosynthesis and respiration (Gaarder and Gran 1927), chlorophyll and phaeophytin analysis by high-performance liquid chromatography (HPLC) and community metabolism using the three-point diel oxygen method (Lind 1979). Macrophyte and filamentous algal distributions were mapped as percentage surface cover and the major species present were identified.

Zooplankton were also sampled using a depth-integrating water column sampler. Water column samples were collected from three locations in each mesocosm and combined. After filtration, the organisms were identified and enumerated.

Macroinvertebrate populations were assessed using three methods: visual observations within quadrats, collection on artificial substrates and trapping of emerging adult insects. Visual observations were made in 2 m 2 quadrats in each mesocosm. Macro- invertebrates present in the quadrats over a 2 min period were recorded and their behaviour was categorized as normal, abnormal or dead. As well as regular observations throughout the study, additional observations were made after applications (1 h and 1 and 3 days). Artificial substrate samplers were constructed from plastic cylinders which are normally used as surface area enhancers for sewerage treatment plants (Anon 1979). These were placed in pairs, one at the pond surface and one resting on the hydrosoil at three locations in each mesocosm and were left to colonize for 2 weeks. After this time the samplers were removed from the mesocosms, the organisms washed off and then counted. Surface and bottom samples were processed separately. Emerging insects were collected in mesh-covered rectangular traps which were placed at the water surface. The traps covered a water surface area of 0.25 m 2 and two traps were placed in each mesocosm. Insects were collected from the traps twice a week throughout the study. Organisms collected from substrate samples and emergence traps were identified to family.

Experimental design and statistical analyses

One critical consideration in the experimental design of mesocosm studies is accounting for the variability between experimental units. This has two important implications for design of the experiment. Replication (to reduce experimental noise) should be sufficient to detect differences at the required level of sensitivity. Interspersion (to reduce experimental bias) should be adequate to avoid any potential impact of environmental gradients on study conclusions.

For this study, two replicates were selected on the basis of estimates of the variability of the key ecological end-points of interest. For macroinvertebrates and zooplankton (the most sensitive organisms in the laboratory) typical coefficients of variation (CV) are 50% or more. For these organisms, an ecologically significant impact should be designated as at least a 1-2-fold difference (Shaw et al. 1994). Smaller differences are probably irrelevant in natural ecosystems because of large seasonal variations, rapid generation times, recolonization and recovery. Calculation of differences detectable at


CVs of this order (Sokal and Rohlf 1981) indicated that with two replicates, differences of 1-2-fold would be detected for many end-points of interest in this study. After the completion of the study, measured CVs were used to calculate the sensitivity of the experiment for various end-points.

A systematic randomized design was selected as the most appropriate means of assigning experimental units to treatments. This type of design is often preferable to entirely randomized designs when it is suspected that differences between experimental units may affect the outcome of the study (Hurlbert 1984). It was predicted that for the present study, the physical positioning and aspect of mesocosms would affect certain end-points (through differences in exposure, ambient light conditions and temperature), so blocks of mesocosm were selected according to their location. The mesocosms were firstly split into two blocks of four (ponds 1-4 and 5-8, Fig. 1) to account for the largest potential environmental gradient (site length). Within each of these blocks, mesocosms were then assigned at random. However, if the randomized selection resulted in a poorly interspersed layout between pond blocks it was rejected and the procedure was repeated. This was thought to be of particular importance for pond exposure and, consequently, any layout that also placed both replicates of a treatment on the comers or centre of the site was also rejected. By this process a randomized design was achieved that ensured that the experimental units were adequately interspersed (Fig. 1). This approach is essential to eliminate bias in analysis of variance design field studies that have small numbers of replicates (Hurlbert 1984).

The null hypothesis of no difference between the treatment groups was tested by analysis of variance (ANOVA). Count data were first normalized by a log transformation and proportion data (%) were transformed using an arcsine transformation. If there was evidence from the F-test of a difference between treatment groups, then the pesticide treatments were individually tested against the controls with a two-tailed Students t-test using the estimate of residual variance from the ANOVA.

In one of the higher rate lambda-cyhalothrin-treated mesocosms, a population of stickleback (Gasterosteus aculatus) developed. In the treatment year, two adult fish were observed, presumably from accidental addition of eggs or fry during the initial stocking of the mesocosms. These adult fish had reproduced prolifically prior to the treatment period and many young fish were observed in this mesocosm throughout the year. Since fish were not present in any of the other mesocosms and there were obvious differences in the pond biology (probably due to fish predation and enhanced nutrient cycling), data from this mesocosm were not used. However, fish were observed during and following the pyrethroid treatment and no adverse effects were observed. For this rate a modified t-test was used to compare data points with the mean for the controls. This resulted in some loss of statistical sensitivity.

Results

Pyrethroid residues In mesocosms treated with 0.17 g a.i. ha -1 lambda-cyhlaothrin, measured water column residues were only slightly above the lod (2 ng 1-1) 1 h after application and were below the lod after 24 h. The lod was at 12% of the nominal applied (calculated for the time of application and assuming even mixing of total applied pyrethroid throughout the water column). A rapid decline in pyrethroid residues in water samples was also


observed for both the 1.7 g a.i. ha -1 lambda-cyhalothrin and 0.7 g a.i. ha -1 cypermeth- fin treatments. This is best demonstrated by the data for the higher lambda-cyhalothrin treatment since residues were above the lod on more occasions, as shown in Fig. 2. Following the third application, samples taken near the water surface 1 h after application contained pyrethroid residues of 94 ng1-1, approximately 50% of nominal values. Residues in samples taken at the same time from approximately 1 m depth were 23 ng1-1. The mean water residue value was 59 ng1-1 (assuming a linear distribution), which is 34% of the nominal. After 24 h, mixing of residues through the water column had occurred and concentrations had declined to a mean residue of 23% of the nominal. Following the third application in the 0.7 g a.i. ha -1 cypermethrin-treated mesocosms, the mean residue concentration in water samples taken near the water surface 1 h after application was 35 ng 1-1, again approximately 50% of the nominal. Residues in samples from approximately 1 m depth averaged 31 ng 1-1. After 24 h, this had declined to 13% of the nominal. In samples taken 2 days after application from both the 1.7 g a.i. ha -1 lambda-cyhalothrin and 0.7 g a.i. ha -1 cypermethrin treatments, residues averaged 9% and by 6 days the mean residue was equal to the lod of 2 ng1-1 (1% of the nominal).

Figure 2 also shows residues of lambda-cyhalothrin in hydrosoil samples from the higher lambda-cyhalothrin treatment rate. Measurements were made 8 days after each application and continued at 2-3 week intervals for 13 weeks after the final application.

1

i g o E 0

._=

F

I -

4

Week

W A T E R

14%

~ 3 i 6% 6%

II 9%

HYDROSOIL

Pyrethroid Spray Drift Applications leach 4.25 rng par mesocosm)

3

i

t34%

18%

Io1 ,1213

K e y :

• Concentration in water (mean of surface and bottom)

0 Less than Iod 11%)

0 - 2 . 5 cm hydrosoil depth

2 .5 - 5 . 0 crn hydrosoil depth

5 .0 - 7 , 5 crn hydrosoil depth

% % of amount applied Water as % of single application Hydrosoil as % of total applied

1 8 % 21%

22% 22%

Fig. 2. Measurements of pyrethroid residues in mesocosm water and hydrosoil following successive applications with the 1.7 g a.i. ha -1 lambda-cyhalothrin treatment.

t 13%


There was an increase in residues in the surface hydrosoil (0-2.5 cm depth) peaking at 7 #gkg -1 (17% of the total applied pyrethroid) after the third application. Following this, residues increased in lower depth fractions (2.5-5.0 cm and 5.0-7.5 cm) and decreased at the surface. By November, 13 weeks after the final application, only 13% of the total applied pyrethroid was recovered. At that time, residues in hydrosoil samples,from both the lower lambda-cyhalothrin and cypermethrin treatments were at or below the lod in the 0-2.5 cm depth fractions and absent from lower depths. Residues were below the lod (0.4/~gkg -1) at depths greater than 7.5 cm, throughout the study.

Physicochemistry

Characterization data collected before treatment indicated that the mesocosm water was moderately hard and was low in suspended solids and dissolved organics. Nitrogen : phosphorus ratios were less than 30 : 1 and phosphorus levels were between 0.02 and 0.2 mgl -~ indicating that the mesocosms were mesotrophic (Wetzel 1983). The mesocosm hydrosoil was classified as a sandy clay loam with 7.7-9.9% organic matter and pH 7.0-7.5. Before applications began, neither the water or hydrosoil contained any pesticides or contaminants.

Water temperature, dissolved oxygen and pH remained comparable in all mesocosms throughout the study. Temperature, measured at midday, increased from 13 °C in May to 21 °C in July and decreased again to 8 °C in November. Dissolved oxygen, measured at midday, followed a similar trend with highest levels of 15 mg1-1 recorded in the summer months. There were no apparent seasonal trends in pH, which averaged pH 9. There were, however, differences between treatment groups in conductivity, alkalinity and turbidity (Fig. 3). Conductivity measurements (Fig. 3a) in the 10% rate lambda-cyhalothrin treatment were higher than controls after the application period (one mesocosm data only). Turbidity was significantly higher in all treatments after the application period (Fig. 3b) and alkalinity was significantly higher in all treatments at the end of the study (Fig. 3c).

Plant communities

There were no adverse effects of pyrethroid applications on algal chlorophyll content, productivity or community metabolism. Abundance, biomass and chlorophyll content followed similar trends in all mesocosms. Community metabolism measurements indicated that a dynamic community of primary producers existed within the mesocosms, with photosynthesis and respiration in balance throughout the study and no treatment-related adverse effects. Phytoplankton biomass increased in the 10% rate lambda-cyhalothrin treatment (Fig. 4a) and periphyton biomass (Fig. 4b) also showed a treatment-related increase towards the end of the application period. In all treated mesocosms after the first application, there was an indication of a treatment-related enhancement in phytoplankton gross primary productivity, as measured by light/dark bottles (Fig. 4c).

Phytoplankton assemblages were dominated by Cryptophyta (the majority of these being Cryptomonas spp.), Cyanophyta and Chlorophyta. Bacillariophyta, Eugleno- phyta and Chrysophyta were also numerous and Pyrrhophyta were sampled occasionally. Seasonal succession of the different groups occurred, with periodic short-term blooms particularly of Cyanophyta (Fig. 5). There was no apparent effect of treatment

228

a) Conductivity b) Turbidity

Farmer, Hill and Maund

350

300 ®

E ._e o3

l l l i I\

\

2 5 0 ! | = = | | = i i !

-1 1 3 5 7 9 11 13 16 19

Week Number

1o i l l

8 I

6 z

• • • . . o ; . =

, ° °

. : / . \ " .

: : / / " , \ " . ,

/** "* : / ~ \ • - o ~.. ~,~ , ¢ ,

2 , , , , , , , , , ,

-1 1 3 5 7 9 11 13 16 19

Week Number

c)Alkalinity

80

70

.=

~ 6 0 0 t ~

o }

l l l i

/\ / \~.. ~-* --

o , • ; " % . I: /,¢,~,......

t% ..*. so '~'**/- -

I

- - I

i i i i i i l i i

-1 1 3 5 7 9 11 13 16 19

Week Number

Key:

l spray drift applications

control

m _ 1.7 g ai h a l lambda-cyhalothdn

. . . . . . 0.17 g ai ha ~ lambda-cyhalothrin

. . . . . . . . . . 0.7 g ai ha 1 cyperme~dn

Fig. 3. Physicochemical measurements from control and pyrethroid-treated mesocosms. Statistic- ally significant differences are denoted by * for p = 0.05 and ** for p = 0.01.

Fate and effects of two pyrethroids in mesocosms

a) Phytoplankton biomass b) Periphyton biomass

229

15

®

- - 1 0

I \ I I

0 , -5

fl 8

!

!

I .**

I I I I I I I I I I

-1 1 3 5 7 9 11 13 16 l g

Week Number

6

®

E o~ 4

0 I

-5

~-I ... . /..* \ ;

I I I I I I I I i I

-1 1 3 5 7 9 11 13 16 19

Week Number

c) Phytoplankton Gross Productivity

700

"7 600

O .c 500 o?, E

400

o o~ 300 E

200

100

I

/I / I /.I

- ; -1 1 3 5 7 9 11 13 16 1"9

Week Number

Key:

spray drift applications

- - control

- - - - - 1.7g ai ha "1 lamloda-cyhalothrin

. . . . . . 0.17g ai ha-1 lambda-cyhalothrin

. . . . . . . . . . 0.7 g ai ha 1 cypermethrin

Fig. 4. Phytoplankton and periphyton biomasses and gross primary productivity from control and pyrethroid-treated mesocosms. Statistically significant differences are denoted by * for p = 0.05 and ** for p = 0.01.

230

a) Control


b) 0.7 g ai/ha Cypermethrin

100

80

E 60

o

o_ 4 0

20

-5 -1 1 3 5 7 9 11 13 16 19 Week Number

O0

8O

60

4O

20

0 -5 -1 1 3 5 7 9 11 13 16 19

Week Number

c) 0.17 g ai/ha Lambda-cyhalothdn d) 1,7 g ai/ha Lambda-cyhalothrin

100 ,, ...... 100

~ _ 6 0

i i

-5 -1 1 3 5 7 9 11 13 16 19 -5 -1 1 3 Week Number

7 9 11 13 16 19 Week Number

Key:

m Cyanophyta [ ] Chlorophyta m Cryptophyta

D Euglenophyta D Chrysophyta D Baeillariophyta

Spray drift applications

Fig. 5. Relat ive abundances of phytoplankton orders collected from control and pyrethroid-treated mesocosms. Statistically significant differences are denoted by * for p = 0.05 and ** for p = 0.01.


on phytoplankton community structure. The most dominant groups amongst the periphyton were Cyanophyta (mostly Lyngbya keutzingii), Chrysophyta and Chloro- phyta. As with phytoplankton, the periphyton abundances, biomass and chlorophyll content followed a similar trend of general decrease, irrespective of the application rate, throughout the study period. Taxonomic richness for both phytoplanktonic and periphytic algal communities was good, with 35 and 56 species, respectively.

Prior to spray applications, much of the macrophyte biomass was cleared from the mesocosms to give at least 60% open water surface. There was, however, rapid regrowth and by the beginning of August, most mesocosms contained over 70% surface cover (mainly Elodea canadensis and Potamogeton crispus). Filamentous algal populations (dominated by Cladophera spp. and Chaetophora spp.) also increased during the summer. There were no effects of treatment on macrophyte communities.

Zooplankton

Of the three families of Cladocera present in the mesocosms, the Chydoridae and Daphniidae were the most abundant, whereas the Polyphemidae were only occasionally sampled. Both Copepod adults and nauplii (mainly Cyclopoida) occurred in large numbers throughout the year. Rotifera were most abundant in all mesocosms from mid-July to the end of the study in November. There were no adverse treatment- related effects on any individual groups (Fig. 6), but there appeared to be indications of treatment-related increases in populations of Chydoridae (Fig. 6b), copepod nauplii (Fig. 6c) and rotifers (Fig. 6d), though these were not statistically significant. The greatest increases were observed in the higher lambda-cyhalothrin treatment.

Macroinvertebrates Initial surveys of the mesocosms indicated that diverse macroinvertebrate communities had developed in all mesocosms prior to the application period (Table 2). During visual assessments made 1 h after spraying, abnormally behaving Notonectidae and Gyrrini- dae were observed. Pyrethroid effects were manifested as hyperactivity together with apparent loss of coordination and erratic movements. In subsequent assessments, dead hemipterans and coleopterans were also found and these were eliminated from the higher lambda-cyhalothrin treatment and significantly reduced in the cypermethrin treatment. There was, however, evidence of some recovery by reinvasion, mainly in mesocosms treated at the lower lambda-cyhalothrin rate.

The macroinvertebrate taxa which were most sensitive to the pyrethroid applications were the amphipod and isopod crustaceans (Fig. 7). After the first application, there were significant reductions in the abundance of Asellidae (Fig. 7a) in the higher lambda-cyhalothrin treatment. Effects from the cypermethrin treatment were only seen after the second application, but significantly fewer organisms were sampled thereafter until the end of October when some recovery was observed. The lower lambda-cyhalothrin treatments did not affect the Asellus population. The freshwater shrimp, Gammarus spp., was even more sensitive to the pyrethroid applications (Fig. 7b). After the first application, abundances on substrate samplers were substantially decreased in all treated mesocosms. Following subsequent applications, there continued to be significant decreases in the lower lambda-cyhalothrin treatments. However, in the higher lambda-cyhalothrin- and cypermethrin-treated mesocosms,

232

a) Daphniidae b) Chydoridae


180 '

160 '

1 4 0 '

120 '

"~ 100 '

80 '

E 60 ' Z

40"

20"

J i l l

.'. i !

, i' ', ,'1,' A / " , ~ ,~v/i

-1 5 10 15 Week Number

1000,

8 0 0 '

600 '

Q,.

.~ 400.

Z

200

l l , i l

II

Key:

I spray drift applications

control

_ m _ 1.7 g ai ha 1 lambda-cyhalothrin

. . . . 0.17 g ai ha "1 lambcla-cyhalothdn

. . . . . . . . 0.7 g ai ha I cypermethdn

° % , • . . . . . . . . . / \


c) Copepod naupUi d) Rotifera

1000

80O

~ o

4 0 0

Z

2O0

0

1111

/ / \

I..~ \ I / - \ I.: • \ ~/ ',....i /-.>

-1 .5 10 15

Week Number

700 jill I

500 I

ix/~ \


Fig. 6. Abundances of major zooplankton taxa collected from control and pyrethroid-treated mesocosms. Statistically significant differences are denoted by * for p = 0.05 and ** for p = 0.01.


Table 2. Macroinvertebrate taxa present in the mesoeosms before the pesticide applications were made

Superorder Phylum Class Order Superfamily

Subphylum Subclass Suborder or family

Platyhelminthes Turbellaria

Mollusca Gastropoda Pulmonata

Annelida Oligochaeta Hirudinea

Arthropoda Crustacea Malacostraca

Uniramia

Tricladida Planariidae Dendrocoelidae

Peracarida Amphipoda Isopoda

Ephemeroptera Schistonota

Odonata Zygoptera Anisoptera

Triehoptera Annulipalpia Integripalpia

Spicipalpia Hymenoptera

Hemiptera (Heteroptera) Nepomorpha

Coleoptera

Diptera

Lymnaeidae Planorbidae

Gammaridae AseUidae

Baetidae

Coenagfioniidae Aeschnidae

Hydropsychidae Leptoceridae Limnephilidae Phryganeidae Rhyacophlidae Braconidae Ichneumonidae Chalcidoidea

Corixidae Notonectidae Dytiscidae Gyrrinidae Hydrophilidae Tipulidae Chaoboridae Chironomidae

populations of Gammaridae were eliminated. There was no indication of recovery before the end of the study in November.

The dipteran family Chironomidae were collected in large numbers from all mesocosms, both as larvae (on substrate samplers) and emerging adults (in floating traps). After the first application, there were significantly fewer chironomids in surface

234

a) Asellidae b) Gammaridae


300

E 8 o g2oo ® E

(3.

81oo P ®

E Z

r,, \

I .

°" ,.~r o O ° ° ' * ° • .o " . . ~ t ~ r t . "

, , _ *,* -5 -1 1 3 5 7 9 11 13 16 19

Week Number

600

500 E 8 o o 400

E

200

E ~-100

/ ! i

-5 -1

'~.." - . ~ .- . . . . * _ * _ * *

i '= V , *,* *,* *,* *,* * * *,* 1 3 5 7 9 11 13 16 19

Week Number

Key:


control

- - ~ - 1.7 g ai ha-1 larnbda-cyhalothrin

. . . . . . 0.17g ai ha-1 lambda-cyhalothrin

. . . . . . . . . . 0.7 g ai ha -1 cypermethrin

Fig. 7. Abundances of Peracarida collected from artificial substrate samplers in control and pyrethroid-treated mesocosms. Statistically significant differences are denoted by * for p = 0.05 and ** for p = 0.01.

substrate samples in all treatment groups (Fig. 8a). However, after the third application numbers in treated mesocosms were similar to or greater than those in control mesocosms. Abundances of larvae on benthic substrate samplers appeared to increase during the application period and were generally somewhat higher in the pyrethroid- treated mesocosms. On the last two sampling dates, there were significantly more in both the lambda-cyhalothrin treatments (Fig. 8b). Compared to control abundances, emerging chironomids (Fig. 8c) appeared to increase in the lower lambda-cyhalothrin treatment from after the last application until the end of the study.


a) Chironomidae on surface substrates b) Chironomidae on bottom substrates

235

100"

8 o 80 ®

E

o -

~ 60

o

40 ® J~

z 20

Illl ,so

8

* * .. "~ .~ 5O

/ / ~ / , I

I f ,4 . . -x ,** | , • ,

-5 -1 1 3 5 7 9 11 13 16 19

Week Number

i . . . . / ".t ;* : ?-~,,,/ ,', ~v • I l

- ~ , ' x f , , , . Z " ~ * * **

I i I I I I I I I I I

-5 -1 1 3 5 7 9 11 13 16 19

Week Number

c) Chironomidae in emergence traps

60 i l l l

50

4O c i . m

3o ®

J~

Z 20 ¸

10

0 i

-5

I I I i i

I I I i I

f | I I

I , v,I 1 I i : ' 1

' , ' : i ' = I I:" I ~ ' , ' ' , l : : ' % ", , I I , : ! t

~. x I i I t I : • i \~., t la I I ~ : : =

E' . . ~ l , " ,,,~: : ".l

i i I i i i i i i i

-1 1 3 5 7 9 11 13 16 19

Week Number

Key:


- - control

- - - - - 1.7g ai ha 1 lambda-cyhalothdn

. . . . . . 0.17g ai ha ~ lambda-cyhalothdn

. . . . . . . . . . 0.7 g ai ha "1 cypermethdn

Fig. 8. Abundances of Chironomidae (artificial substrate sampler and emergence trap data) in control and pyrethroid-treated mesocosms. Statistically significant differences are denoted by * for p = 0.05 and ** for p = 0.01.

236

a) Baetidae on substrates

30

E 0 o 0

E 20

el

~ 1 0 '

E Z

Spray "ddtt" applications

I I

I I

\

..• * *

I L

\ I ' ",°

--." 1. / . . " - ' L "

I °" ~ * * * " ,"~ , ' . , ,---~

-5 -1 1 3 5 7 9 11 13 16 19

Week Number


b) Baetidae in emergence traps

Spray "ddtt" applications 20

15

1=

P 10 i i I !

E It t i l l tl l ' , , ~1 I I I I

• . . ' , , 0 I i ! ! ! ! i e i

-5 -1 1 3 5 7 9 11 13

Week Number

| !

16 19

Key:

spray drift applications control

m _ 1.7 g ai h a l lambda-cyhalothrin

. . . . . . 0.17g ai ha -1 lambda-cyhalothrin

. . . . . . . . . . 0.7 g ai ha "1 cyperrnethrin

Fig. 9. Abundances of Baetidae (artificial substrate sampler and emergence trap data) in control and pyrethroid-treated mesocosms. Statistically significant differences are denoted by * for p = 0.05 and ** for p = 0.01.

Nymphs and emerging adults of the mayfly family Baetidae were collected both on substrate samplers and in emergence traps (Fig. 9). After the last application, significantly more nymphs were collected on substrate samplers (Fig. 9a) in the lower lambda-cyhalothrin treatment, followed 4 weeks later by an increase in adults emerging from these mesocosms (Fig. 9b). After treatment, no adults were collected from emergence traps in the higher lambda-cyhalothrin treatment and abundances in cypermethrin-treated and control mesocosms were low.

Pyrethroid applications had no adverse effects on Turbellaria, Gastropoda and Annelida (Fig. 10). There was some evidence of treatment-related increases in Gastropoda (Fig. 10b) and Annelida (Fig. 10c) after treatments, but this was not statistically significant.


a) Turbellaria b) Gastropoda

237

120

E 100

E 80

O .

" O ®

60

6

.~ 40

z

20

J i l l II ~2~o

E 200

/ \" i '~ 150

° ° ° _ _

t "..,- ~ 100

. / 50

i I i I i | i i i i i

-5 -1 1 3 5 7 9 11 13 16 19

Week Number

o" : /

/\ .. ...... / I ! / I.:" \ / //

! i i i i i i i i i |

-5 -1 1 3 5 7 9 11 13 16 19

Week Number

c) Annelida

70

60

o

E

~- 40

~ ao

E 20

Z

10

;°°

~.."..I \ :I i. I '(

J " "~" " I " . . ' l I ..... , , ' . ~ "- ..'"-~'~"!'\ I • ° " , . . o .

-~ -; ; ~ ~ -~ ~1"11~ 1~ 1"9 Week Number

Key:

1 spray drift applications

control

- - ~ - 1.7 g ai h a l lambda-cyhalothrin

. . . . . . 0.17g ai ha -1 lambda-cyhalothdn

. . . . . . . . . . 0.7 g ai ha "1 cypermelhdn

Fig. 10. Abundances of Turbellaria, Gastropoda and Annelida (artificial substrate sampler data) in control and pyrethroid-treated mesocosms. Statistically significant differences are denoted by * for p = 0.05 and ** for p = 0.01.

238 Farmer, Hill and M a u n d

Discussion

Fate

The distribution and dissipation of both cypermethrin and lambda-cyhalothrin in the mesocosms were comparable, as would be expected for chemicals with such similar physicochemical characteristics. There was an initial rapid loss of both pyrethroids from the water column, with half-lives of approximately 1 day. Similar data are reported for other synthetic pyrethroids, even in differing study designs utilizing natural ponds, farm ponds, mesocosms and microcosms (reviewed by Hill et al. 1994a). Such behaviour was expected because the pyrethroids are known to be rapidly and strongly adsorbed to surfaces of both biotic and abiotic compartments of the aquatic ecosystem (Hill 1985, 1989; Lozano et al. 1989; Heimbach et al. 1992; Heinis and Knuth 1992).

Residues in the hydrosoil increased during the application period. This was also observed in similar studies with esfenvalerate (Lozano et al. 1989; Heinis and Knuth 1992) which also showed adsorption to plant and algal surfaces. After the application period, any residues adsorbed to particulate matter in the water column may have been deposited on the hydrosoil surface. Subsequent increases in residues further down the hydrosoil profile may have resulted from mixing brought about by the movements of benthic organisms. Data available from similar studies (Muir et al. 1985; Lozano et al. 1989; Heimbach et al. 1992; Hill et al. 1994a) indicate that the hydrosoil is the major sink for pyrethroid residues, but variability between samples and the length of post-treatment sampling does not allow conclusions to be drawn concerning the persistence of residues into the following year. However, laboratory studies using sediment-water systems (Hill 1985; Hamer et al. 1992) have shown that once pyrethroids become adsorbed they are much less bioavailable and so potential effects are substantially reduced.

Ecological effects

To provide a summary of the overall ecological effects of each treatment, the number of significant differences between control and treatments for all the end-points measured in the study (after applications began) were summed (Table 3). By combining these into broad classifications, it is clear that the most substantial effects were caused to macroinvertebrate communities at all treatment rates (Fig. 11). There is also a dose-response relationship between application rate and degree of effect. These data suggest that at similar application rates, overall effects of lambda-cyhalothrin on macroinvertebrates may be lower than those of cypermethrin. However, these data are biased by effects on the peracarid crustacea, which appear more adversely affected by cypermethrin. Effects on all other taxa are similar at comparable application rates for the two chemicals (Table 3).

Effects on macroinvertebrates followed a similar pattern in all of the pyrethroid treatments and were probably related to differences in exposure between the different types of organisms. Earliest effects were observed on semi-aquatic organisms at the water-air interface, which would probably have received substantial doses of chemical in the form of direct contact with the spray. For this reason, the quadrat observation method used was particularly useful for assessing-the immediate effects of the pesticide sprays. Symptoms of hyperactivity are commonly associated with this kind of pesticide (Miller and Salgado 1985). The effects on these organisms were only transient both due


Table 3. Total number of statistical comparisons be tween t reatments and control for each end-point where t rea tment values were significantly higher ( + ) or lower ( - ) than the controls

0.17 ga. i . ha -1 1.7 ga. i . ha -1 lambda- lambda- 0.7 g a.i. ha -I cyhalothrin cyhalothrin cypermethrin

Parameter + - + - +

Physicochemistry DO/temperature/pH/condition Alkalinity 1 1 1 Turbidity 3 2

Phytoplankton Gross photosynthesis Productivity 1 1 Total numbers Total biomass 3 Chlorophyll a

Periphyton Total numbers 2 Total biomass 1 2 Chlorophyll a

Filamentous algae

Macrophytes

Zooplankton Rotifera Copepod adults Copepod nauplii Daphniidae Chydoridae 1 1 Total numbers

Substrates Turbellaria 1 Gastropoda Oligochaeta Hirudinea Asellidae Gammaridae 7 Baetidae 7 Coleoptera 1 Chironomidae 3 1 5

Emergence traps Baetidae 5 Chaoboridae 1 Chironomidae 2

Ouadrat visual observation Planorbidae 1 Lymnaeidae Notonectidae Coleoptera 2

1 1 1 l 4

2 2

23 14 22 21

2 2 1 1 1 6 1

1 1 1 1

240

Physico- chemical Phytoplanklon Pedphyton

,~ + 20 r (Sl) (SS) (23)

£ +10

-10

|

I "6

E

u~ ¢b (n

-20 -

-30 - -

-40 -

- 5 0 "

-60 -

[ ] 0.17 g ai/ha lambda-cyhalothdn

I~1 0.7 g ai/ha cypennethdn

• 1.7 g ~ k,mMa-~'yh,~othan


Zooplankton Macroinvertebrates

(lOS) (s11)

Fig. 11. Comparison of number of detected significant differences (p = 0.05) between treatments and control. Total number of comparisons are shown in parentheses. On the y-axis, + represents the number of significant increases in comparison to controls and - represents the number of significant decreases.

to the further dissipation of the pesticide and also probable recoloniz~/tion from winged life stages.

Amphipod and isopod crustaceans (Peracarida) were the most adversely affected benthic organisms. The sensitivity of Gammaridae appeared to be somewhat greater than the Asellidae, with effects occurring more rapidly and to a greater extent in the former. In the laboratory the concentrations at which effects of cypermethrin occur on these two organisms are very similar (Stephenson 1982). The difference in response observed here between the two organisms is probably due to their exposure. Gammaridae principally inhabit the sediment surface or water column, whereas Asellidae are epibenthic, often burrowing into the hydrosoil. Consequently the exposure of Gammaridae to bioavailable pyrethroid may have been higher since when the chemicals reached the hydrosoil they will have been rapidly and strongly adsorbed.

Data from this study suggest little or no recovery of the peracarid crustacea from


these impacts. However, this may not be a true reflection of nature because the potential for Gammarus or Asellus populations (or similar fully aquatic organisms) to recover by reinvasion is minimized by the enclosed nature of the system, 'even' treatment (no unaffected areas) and the relatively small target such a system provides for immigration compared to natural ponds. In the cypermethrin treatment, although the Asellus population was initially affected, it was not eliminated and recovery was observed towards the end of the year. In natural aquatic systems it is unlikely that the whole body of water would receive pesticide spray drift as in this study, thus invasion from unaffected areas would occur. In addition, these organisms may also be found in rivers and streams where concentrations of pesticide are less due to the dilution factor of flowing water. Thus a risk assessment based on a study of this type without consideration of natural environmental factors could be over-severe because of the low potential for recovery of some organisms in experimental mesocosms.

Like the peracarids, the ephemeropteran family Baetidae has been shown to be very sensitive to pyrethroids in laboratory water-only toxicity tests, with a 72 h Ecs0 of cypermethrin for Cloeon dipteran of 12 ng 1-1 (Stephenson 1982). Although pyrethroid residues were found above this concentration in treated mesocosms, there were no decreases in numbers of mayfly nymphs. Again, the lack of effects on these organisms may be due to reduced exposure because the organisms live close to the substrate. Effects on the Chironomidae were complex, with populations decreasing during spraying, particularly reflected by impacts on surface substrate samplers. This was followed by an increase in the numbers of larvae in all treated mesocosms, particularly the lower lambda-cyhalothrin treatment, probably due to an increase in detrital food material from dead organisms, increases in algae and decreased predation by Hemiptera and Coleoptera. Increases in abundance of larvae were reflected by subsequent apparent increases in emergence from these mesocosms.

Zooplankton were far less severely impacted than has been reported elsewhere for pyrethroid insecticides (Hill et al. 1994a). Copepoda have previously been shown to be the most sensitive zooplankton to the pyrethroids (Hill 1989; Lozano et al. 1992). However, in this study they were unaffected and there was a dramatic increase in the numbers of copepod nauplii following spray application, which was most apparent in the higher lambda-cyhalothrin treatment. Increases in juveniles can be triggered by an increase in adult food supply and in this case coincides with increased rotifer populations, a major food source of the predatory cyclopoids. Such increases in rotifer populations are commonly observed after pyrethroid applications and can be due to reductions in some of their predators or competitors (Yasuno et al. 1988; Heimbach et al. 1992; Lozano et al. 1992). In this case, the enhancement may have been caused by reduction of the macroinvertebrate crustacea, removing competition for algal and detrital food sources. Also, substantial numbers of decaying organisms may have released plant nutrients, leading to the observed increases in primary productivity. This would also have enhanced rotifer populations. Increases in the abundance of Chydoridae in treated mesocosms may have also been due to reductions in competition for detritus with affected macroinvertebrates. Similar responses have been observed by Tooby et al. (1981) following the application of the pyrethroid deltamethrin to natural ponds.

The absence of direct toxic effects on the phytoplankton and periphyton populations was expected since laboratory tests indicate that the pyrethroids are virtually non-toxic


to these communities (Hill 1985). There were, however, initial increases in these populations which were probably due to a reduction in grazing pressure as a result of decreased populations of isopods and amphipods. The other populations of grazers and detritivores present in the mesocosms (Turbellaria, Gastropoda and Annelida) were unaffected by pyrethroid applications. They were however able to benefit from decreases in grazing competition and increases in detrital matter and populations of these organisms tended to increase (although not significantly).

Experimental design and sensitivity

One common criticism of field studies of this type (ANOVA design) is that the sensitivity to detect differences between treatments and controls is relatively low (Eberhardt 1978). Experimental sensitivity depends mainly on the variability of the end-point measured and the number of replicates that are utilized in the experimental design (Sokal and Rohlf 1981). The sensitivity of aquatic mesocosm studies is often low because end-points tend to be highly variable (due to both intrinsic and sampling variability) and replication is limited (due to logistical and cost constraints). However in mesocosm studies, which are attempting to simulate natural ecosystems, the objective is to detect effects which are of significance at the population, community and ecosystem levels of organization, rather than those on individual organisms (as is the case in the laboratory). For such effects to be ecologically important, it has been suggested that impacts need to be at least 0.5-1.5-fold differences (Shaw et al. 1994) due to the resilience, rapid recovery and functional redundancy of many aquatic ecosystems. For organisms with very rapid life-cycles, such as the zooplankton and phytoplankton, ecologically important effects probably will not occur until there are differences very much higher than this, perhaps in the region of at least 2-fold differences. Smaller differences are probably meaningless in comparison to natural fluctuations of these organisms. Consequently, it is possible to remain confident that there are no adverse impacts on the ecosystem without having relatively high experimental sensitivity (such as might be expected for a study at the organism or suborganism level).

In order to examine the sensitivity of this study, control coefficients of variation for the various end-points in the study were used to estimate the sensitivity of the experiment, using the iterative formulae described by Sokal and Rohlf (1981). Increases in the coefficient of variation lead to decreases in experimental sensitivity and this relationship between CVs for end-points in this study with two replicates is shown in Table 4. Coefficients of variation from control mesocosms were used as an estimate of the variability of certain study end-points. This allows an estimate of the sensitivity of the experiment to be made. From these data, it can be seen that many of the macroinvertebrate end-points were relatively sensitive. For organisms whose abundance tended to fluctuate over short periods, such as zooplankton and phytoplankton, experimental sensitivity was less, but nevertheless differences of around 2-3-fold would have been detected. To bring about a 1-fold increase in sensitivity would, in most cases, have required replication that would not have been logistically feasible (Table 4) in the present study where two pyrethroids were compared, one at two application rates. From this analysis it can be inferred that where ecologically significant differences occurred in this study as a result of pyrethroid treatment, they will have been detected in the majority of cases.


Table 4. Coefficients of variation for certain study end-points and smallest differences from control which would be detectable under various study designs

Difference Number of replicates required to Coefficient detected detected differences of: of variation (fold) with

Population (controls) (%) two replicates 1-fold 2-fold 3-fold

Primary producers Total phytoplankton 44 2.3 6 3 2 Total periphyton 60 1.7 4 2 2

Zooplankton Rotifera 69 2.6 8 3 2 Daphniidae 51 2.0 5 2 2 Chydoridae 56 2.2 6 2 2 Copepod nauplii 64 2.4 7 3 2 Copepod adults 71 2.7 9 3 2 Total zooplankton 59 2.3 6 3 2

Macroinvertebrates a Turbellaria (ESS) 24 0.9 2 2 2 Gastropoda (ESS) 25 1.0 2 2 2 Annelida (ESS) 37 1.4 3 2 2 Gammaridae (ESS) 42 1.6 4 2 2 Asellidae (ESS) 43 1.6 4 2 2 Chironomidae (ESS) 42 1.6 4 2 2 Chironomidae (ET) 57 2.2 6 2 2 Baetidae (ESS) 31 1.2 3 2 2 Baetidae (ET) 20 0.8 2 2 2

aESS, artificial substrate sampler; ET, emergence trap.

Acknowledgements

We would like to thank the following for their contributions to the study: Miss N. Allison, Mr B. Arbuckle, Mr P. Askew, Ms D. Castle, Mrs V. Ely, Mr E. Farrelly, Mr M. Hamer , Mrs S. Hill, Mrs D. Jackson, Mr W. Lucassen and Mrs J. Runnalls. We are also grateful to Mr E. Mclndoe and Mr J. Reeks for the statistical analysis of the data.

References

Anon (1979) Biological Methods for the Surveillance of River Water Quality. University of Aston Report, DOE. Contract No. DER/480/100.

Eberhardt, L.L. (1978) Appraising variability in population studies. J. Wildl. Manag. 42,207-38. Gaarder, T. and Gran, M.M. (1927) Investigations of the production of plankton in Oslo Fjord.

Rapp. Process-Verbaux. Reunions. Cons. Perma. Int. Explor. Mer. 42. Ganzelmeier, M., Koepp, H., Spangenberg, R. and Streloke, M. (1993) Wann Pflanzenschutz-

mittel Abstandsauflagen erhalten. Pflanzenschutz-Praxis 3/1993, 14-15. Hadfield, S.T., Sadler, J.K., Bolygo, E. and Hill, I.R. (1992) Development and validation of

residue methods for the determination of the pyrethroids lambda-cyhalothrin and cypermethrin in natural waters. Pestic. Sci. 34,207-13.


Hamer, M.J., Maund, S.J. and Hill, I.R. (1992) Laboratory methods for evaluating the impact of pesticides on water/sediment organisms. Brighton Crop Protect. Conf. - Pests Dis. 2, 487-96.

Heimbach, F., Pfltiger, W. and Ratte, H.-T. (1992) Use of small artificial ponds for assessments J . . .

of hazards to aquatic ecosystems. Environ. Toxtcol. Chem. 11, 27-34. Heinis, L.J. and Knuth, M.L. (1992) The mixing, distribution and persistence of esfenvalerate

within littoral enclosures. Environ. Toxicol. Chem. 11, 11-25. Hill, I.R. (1985) Effects on non-target organisms in terrestrial and aquatic environments. In

Leahey, J.P. ed. The pyrethroid insecticides, pp. 151-262. London: Taylor & Francis. Hill, I.R. (1989) Aquatic organisms and pyrethroids. Pestic. Sci. 27,429-65. Hill, I.R., Shaw, J.L. and Maund, S.J. (1994a) Review of aquatic field tests with pyrethroid

insecticides. In Hill, I.R., Heimbach, F., Leeuwangh, P. and Matthiessen, P. eds. Freshwater field tests for hazard assessment of chemicals, pp. 249-71. Michigan: Lewis Publishers.

Hill, I.R., Travis, K.Z. and Ekoniak, P. (1994b) Spray-drift and run-off of foliar applied pyrethroids to aquatic mesocosms: rates, frequencies and methods. In: Graney, R.L., Kennedy, J.H. and Rodgers, J.H. eds. Aquatic mesocosm studies in ecological risk assessment, pp. 201-40. Special Publication, Society of Environmental Toxicology and Chemistry. Michigan: Lewis Publishers.

Hurlbert, S.H. (1984) Pseudoreplication and the design of field experiments. Ecol. Monogr. 54, 187-211.

Lind, O.T. (1979) Handbook of Common Methods in Limnology. St Louis, Toronto, London: The C.V. Mosby Co.

Lozano, S.J., Brazner, J.C., Knuth, M.L., Heinis, L.J., Sargent, K.W., Tanner, D.K., Anderson, L.E., O'Halloran, S.L., Bertelsen, S.L., Jensen, D.A., Kline, E.R., Balcer, M.D., Stay, F.S. and Siefert, R.E. (1989) Effects, Persistence and Distribution of Esfenvelerate in Littoral Enclosures. US EPA Duluth and University of Wisconsin-Superior, Report DU E104/PPA 06/7592 A.

Lozano, S.J., O'Halloran, S.L., Sargent, K.W. and Brazner, J.C. (1992) Effects of esfenvalerate on aquatic organisms in littoral enclosures. Environ. Toxicol. Chem. 11, 35-47.

Miller, T.A. and Salgado, V.L. (1985) The mode of action of pyrethroids on insects. In Leahey, J.P. ed. The pyrethroid insecticides, pp. 151-262. London: Taylor & Francis.

Muir, D.C.G., Rawn, G.P. and Grift, N.P. (1985) Fate of the pyrethroid insecticide deltamethrin in small ponds: a mass balance study. Z Agricult. Food Chem. 33,603-9.

Shaw, J.L., Moore, M., Kennedy, J.H. and Hill, I.R. (1994) Design and statistical analysis of field aquatic mesocosm studies. In Graney, R.L., Kennedy, J.H. and Rodgers, J.H. eds. Aquatic mesocosm studies in ecological risk assessment, pp. 85-104. Special Publication, Society of Environmental Toxicology and Chemistry. Michigan: Lewis Publishers.

Sokal, R.R. and Rohlf, F.J. (1981) Biometry, 2nd edition. New York: W.H. Freeman. Stephenson, R.R. (1982) Aquatic toxicology of cypermethrin. 1. Acute toxicity to some

freshwater fish and invertebrates in laboratory tests. Aquatic Toxicol. 2, 175-85. Tooby, T.E., Thompson, A.N., Rycroft, R.J., Black, I.A. and Hewson, R.T. (1981) A Pond

Study to Investigate the Effects of Fish and Aquatic Invertebrates of Deltamethrin Applied Directly onto Water. UK MAFF Report PRD 1276.

Wetzel, R.G. (1983) Limnology, 2nd edition. Saunders College Publishing. Yasuno, M., Hanazato, T., Iwakuma, T., Takamura, K., Ueno, R. and Takamura, N. (1988)

Effects of permethrin on phytoplankton and zooplankton in an enclosure ecosystem in a pond. Hydrobiologia 159,247-58.

a comparison of the fate and effects of two pyrethroid insecticides (lambda-cyhalothrin and...

Documents