Fate of carbofuran and its effects on aquatic macroinvertebrates in Canadian prairie parkland ponds

Ecotoxicology 4, 169-189 (1995)

Fate of carbofuran and its effects on aquatic macroinvertebrates in Canadian prairie parkland ponds M A R K W A Y L A N D * and D A V I D A. B O A G *

Department of Zoology, Biosciences Building, University of Alberta, Edmonton, Alberta T6G 2E9 Canada

A multipond study was conducted in 1986 to determine the fate of carbofuran and its effects on aquatic macroinvertebrates in alkaline, Canadian prairie parkland ponds. Four ponds were sprayed with carbofuran in late July; four other ponds served as controls. Sixteen hours after spraying, concentrations of carbofuran in the water column ranged from 9-32 /~g1-1. After 124 h, concentrations had declined to 3-12/~g 1-1. Carbofuran partitioned into submersed aquatic plants, ranging from three to 46 times greater in the plants than in the surrounding water. Carbofuran was below detection limits in most sediment samples. Among the aquatic macroinvertebrates considered in this study, the crustacean Hyalella azteca and trichopteran larvae declined significantly in abundance following the application of carbofuran. Coenagrionidae and small chironomid larvae did not decline following carbofuran application. Hyalella azteca abundance remained relatively low in treatment ponds through to May 1987, while trichopteran larval abundance had recovered by August 1986. There were no readily apparent shifts in community structure in this study, although slight, disproportionate increases in Chironominae and Coenag- rionidae in the treatment ponds following spraying may have resulted from such factors as reduced competition or a change in the size or composition of the food base.

Keywords: carbofuran; insecticide; macroinvertebrates; prairie ponds.

Introduction

Because of the extensive use of insecticides and the increasing frequency of aerial application on the Canadian prairies, there is increasing concern that the innumerable small water bodies, commonly referred to as prairie pot-holes, may become contaminated directly by overflight spraying or indirectly through drift deposit (Grue et al., 1986; Mineau et al., 1987; Sheehan et al., 1987; Forsyth, 1989).

A large proportion of North America's duck populations rely on prairie pot-holes as a source of high-protein invertebrate food throughout the breeding and brood-rearing portions of their annual cycle (Swanson and Meyer, 1973; Bellrose, 1980). Contamina- tion could potentially reduce their invertebrate standing crops and, thus, decrease availability to waterfowl.

*Present Address: Canadian Wildlife Service, 115 Perimeter Road, Saskatoon, Saskatchewan, S7NOX4, Canada.

*Present Address: 6746 Amwell Drive, R.R. 1, Brentwood Bay, BC, VO5 1A0, Canada.

0963-9292 © 1995 Chapman & Hall

170 Fate of carbofuran and its effects on aquatic macroinvertebrates

In recent years, the carbamate insecticide carbofuran (2,3-dihydro-2-2 dimethyl-7- benzofuranyl methylcarbamate) has been one of the most widely used insecticides on the Canadian prairies. However, aside from the well-known relationship between pH and hydrolysis of carbamates (Aly and EI-Dib, 1972; Charnetski et al., 1977; Chapman and Cole, 1982), the work done on partitioning of carbofuran in a farm pond (Klaassen and Kadoum, 1979), the work done on the persistence of granular carbofuran in rice paddy water in the tropics (Seiber et al., 1978; Siddaramappa et al., 1978) and of technical-grade carbofuran in a model ecosystem (Yu et al., 1974), the fate of carbofuran in aquatic systems remains poorly understood (NRCC, 1979). A more thorough under- standing of the fate of carbofuran in prairie ponds is needed in order to assess its effects in such ecosystems.

Even less is known about the effects of carbofuran on aquatic macroinvertebrates than is known about its fate and persistence. The sum of our information on this topic appears limited to laboratory or field studies of its acute toxicity to Daphnia (Yu et al., 1974; Hartman and Martin, 1985; Johnson, 1986), midge larvae (Diptera: Chironomidae) (Mulla and Khasawinah, 1969; Karnak and Collins, 1974; Johnson, 1986) and mosquito larvae (Mulla et al., 1966b).

In this study, we examined the fate and persistence of carbofuran and its effects on selected macroinvertebrates in a series of alkaline ponds located in central Alberta, Canada. We hypothesized that a rate of application representing a hypothetical maximum concentration in the water that could occur after a direct deposit while spraying for grasshopper control would have a deleterious effect on macroinvertebrate populations. Because of the rapid degradation rate that we expected to see in these alkaline ponds, we were particularly interested in examining post-spray recovery rates of populations of affected taxa as well as post-spray shifts in macroinvertebrate assemblages through elimination of competitors, prey or predators by a pesticide. This was one aspect of a larger study designed to assess the effects of carbofuran on prairie pond ecosystems (Wayland, 1989). In addition to the results reported here, it included a study of the mortality rates of selected macroinvertebrate taxa (Garnmarus lacustris, Hyalella azteca, Chironomus tentans, Limnephilus and EnaUagma) confined in small cages in prairie ponds for several days and exposed to carbofuran (Wayland and Boag, 1990) and a mesocosm study examining the seasonal effects of carbofuran on pond macroinvertebrate assemblages (Wayland, 1991).

Methods

Study area

The study ponds are located in central Alberta approximately 90 km north of Edmonton near the village of Clyde (54014 ' N, 113°36 ' W). The eight ponds used in this study are dug-outs or borrow pits which had been dug in the late 1950s and early 1960s. All ponds are surrounded by agricultural land which was either fallowed or planted with grasses or cereal crops.

The ponds are steep-sided with maximum depths in the range 1.7-2.8 m. Emergent vegetation, mainly Typha latifolia is restricted to narrow strips around the perimeters of the ponds. Most ponds support dense growths of submersed vegetation, mainly Cera- tophyllum demersum, Myriophyllum sp. and Potamogeton to a depth of 1.0-1.5 m. All

Wayland and Boag

Table 1. Selected physicochemical characteristics of the study ponds

171

Pond pH ~ Conductivity b Turbidity u % sediment Dissolved O2 d (pmhos cm -1) (NTUs) organic matter c (rag 1-1)

T1 8.4 1686 5.8 6.4 + 1.3 7.5 + 1.4 T2 8.8 953 25.0 4.1 + 0.6 8.8 + 0.5 T3 8.9 404 6.6 3.5 +__ 1.5 7.4 + 1.0 T4 8.5 427 5.3 4.5 + 0.8 8.0 + 1.0 C1 10.1 341 1.3 - 10.8 + 0.1 C2 10.1 329 3.4 - 12.0 + 1.9 C3 8.9 744 6.9 - 8.6 + 0.4 C4 9.8 326 3.3 - 9.6 + 0.7

"Mean values based on means from three sampling dates between July and September 1986 (two to four samples taken during midday from each pond on each date). bValues based on one sample taken from each pond during August 1986. NTU, nephlometric turbidity units. CMean (+SE, n = 4 samples per pond) percent organic matter of 5 cm thick core samples. ~Mean +SE based on means of samples taken from near the pond bottom during midday on three dates between July and September (two to four samples per pond per date).

ponds are alkaline (pH in the range of 8.0-10.0). Some physicochemical parameters associated with the study ponds are shown in Table 1.

Fate and persistence Four ponds were randomly assigned for t rea tment with carbofuran. Surface areas for each of these rectangular ponds were measured. Average depths were est imated by recording depths at 3 m intervals along both axes (length and width) and then averaging these estimates. Measurements of surface area and mean depth were used to calculate the volume of each t rea tment pond on the day it was sprayed.

Two ponds were sprayed on 23 July 1986 and the other two on 30 July 1986 with an emulsifiable formulat ion (Furadan 480 Flowable ®) commonly used to control grasshoppers (Orthoptera) . The ponds were sprayed f rom a canoe using a backpack sprayer. Care was taken to ensure uniform application of the insecticide. The target initial concentrat ion in each pond was 14pg 1-1, a concentration that might be expected in a pond with a mean depth of 1 m following contamination f rom a direct overflight while spraying for grasshoppers at the recommended rate.

Sixteen hours after spraying, water, submersed vegetation and 5 cm thick sediment samples were sampled at two shallow sites (25-75 cm) and two deep sites (76-125 cm) within each of the t reatment ponds. Water samples were taken 15-30 cm below the surface and stored in 21 glass jars. HzSO4 was added immediately to lower the p H to 5.0. Sediment samples were obtained with a 5 cm diameter core sampler. Sediment and plant samples were stored in plastic bags. Water t empera ture and p H were recorded at each site at the t ime of sampling. All samples were stored on ice in plastic coolers and returned to the laboratory within 8 h. Water samples were refrigerated at 4°C. Analyses of water samples were completed within 24 h of sampling. Sediment and plant samples


were frozen and analyses done at a later date. Sampling procedures were repeated 124 h after spraying.

In the laboratory, residues were extracted from water samples with 100% dichlor- omethane (DCM) following addition of 200 g NaCI. Using rotary evaporator and nitrogen blow-down techniques, the extract was concentrated on 1 ml of the concentrate was exchanged to acetonitrile prior to analysis by reverse-phase high pressure liquid chromatograph (HPLC) with a UV detector. HPLC conditions were as follows: the column was a Serva Octadecyl-Si 100 Polyol, the eluent was a 1 : 1 mixture of acetonitrile and water and the detector wavelength was 280 nm.

Sediment samples were thawed, air-dried and then ground and sieved. Subsamples weighing 25 g were placed in a polytron homogenizer and residues extracted with methanol. The extract was filtered, exchanged with ethyl acetate and concentrated to 2 ml.

Plant samples were thawed and homogenized in a food processor. Twenty-five gramme subsamples along with 100ml ethyl acetate and 150g anhydrous sodium sulphate were then blended in a stainless steel blender for 5 min. The extract was then filtered and concentrated to 5 ml.

Final extracts of sediment and plant samples were analysed by gas chromatography/ mass spectrometry (GC/MS) in the selected ion mode of operation. For sediment samples, the column was 25 m x 0.20 mm internal diameter (ID), the initial temperature was 100°C and the final temperature was 300°C. For plant samples, the column was 30 m x 0.30 mm ID, the initial oven temperature was 100°C and the final temperature was 280°C. For both types of samples, the carrier gas was helium with a flow rate of 1 ml min -1.

Standards of carbofuran were analysed with the three types of samples so they could be quantified. For 10% of the samples, blanks, spikes and duplicates were analysed. Recoveries from water, soil and plants were 110 + 11, 68 + 9 and 108 + 12%, respectively (x + CV). Analysis of duplicates indicated reasonable precision for water and plant samples (mean CVS, 13 and 15%, respectively). Duplicates were not done for sediment samples because carbofuran residues were below the detection limit (0.02#g g- l ) for most samples. Detection limits for water samples and plants were 0.5/~g 1-1 and 0.02/~g g-l , respectively.

The rates of loss (Kob) of carbofuran from water and submersed plants were calculated using regression techniques. Half-lives (tl/2) of carbofuran in each pond were calculated according to Seiber et al. (1978) as follows:

Kob = (In(C0) - ln(C)/t (1) t m = 0.693/Kob (2)

where Co and C are concentrations at 16 and 124 h, respectively, t is the interval in hours, Kob is the degradation rate constant and tl/2 is the half-life.

Effects on macroinvertebrates

To determine carbofuran's immediate effects on pond macroinvertebrates, two control and two treatment ponds were sampled 2 days and 1 day before spraying and 3 and 4 days after spraying. Samples were taken from four permanently marked shallow sites (25-75 cm) and four deep sites (76-125 cm) which had been randomly selected at the beginning of the study. A 500 cm 2 stovepipe sampler (Rosenberg 1973) was used to

Wayland and Boag 173

collect invertebrates according to a technique described in detail by Wayland (1989). Samples were preserved in 70% ethanol and returned to the laboratory where they were sorted using a sugar flotation technique (Flannagan, 1973). Invertebrates were identified to species or genus wherever possible. Chironomid larvae were identified only to subfamily. Because of difficulties in identifying early instar damsel fly nymphs, this group was identified only to family (Coenagrionidae). Zooplankton was not considered in this study.

To evaluate carbofuran's seasonal effects on macroinvertebrates, four control and four treatment ponds were sampled approximately biweekly from mid-June until late Septem- ber. A final sampling was carried out in mid-May 1987. The sampling and sorting procedures followed the same protocol as outlined above.

In order to avoid pseudoreplication (Hurlbert 1984), mean abundances of the numerically-dominant groups were calculated for both shallow and deep sites in each pond on each sampling date. Data obtained from the biweekly sampling program were smoothed in a manner modified from Prepas (1984) by pairing successive sampling dates and calculating mean abundances for the deep and shallow sites for each pair of weeks. Thus, mean abundances from shallow and deep sites in each pond from 12 June were paired with their corresponding values from 26 June and new means calculated. Simi- larly, 10 July was paired with 22 July, 6 August with 20 August and 9 September with 24 September.

Because invertebrate abundance varies greatly with depth in ponds and lakes (Ali and Mulla, 1976; Mittelbach, 1981; Thorp and Diggins, 1982), the Wilcoxon signed rank test (Snedecor and Cochran, 1980) was used to determine whether any of the invertebrate groups considered in this study were distributed differently between depths. These analyses were based on the mean abundance of each taxon within each depth zone of the study ponds for each pair of weeks. However, data from the treatment ponds during the post-treatment phase of this study were not included in these analyses because it seemed possible that the insecticide might influence depth-distribution patterns of invertebrates in treated ponds. If the distribution of a particular taxon was found to differ between shallow and deep sites, data from shallow and deep zones were analysed separately when examining the effects of carbofuran. However, if depth did not influence invertebrate distributions, then data from shallow and deep sites were pooled, a mean obtained and subsequent analyses were based on mean abundances within each pond as a whole.

Data were transformed to stabilize variances (Downing, 1979) and homogeneity of variances was tested using the F-max test (Sokal and Rohlf, 1981). Within the text, back- transformed means have been reported (Sokal and Rohlf, 1981). Figures show transformed means and standard errors.

Repeated measures ANOVA (Winer, 1971) was used to evaluate the effects of carbofuran on the abundance of individual taxa. The main effect was treatment (control and carbofuran-treated) and the subplot effects were sampling date and treatment × sampling date interaction (T × D). When interaction or date effects approached significance (p <~ 0.10), planned mean comparisons between dates within treatment levels were done according to Winer (1971); the Dunn-Sidak method (Sokal and Rohlf, 1981) was used to adjust the experiment-wise error rate. For the experiment that measured immediate effects, only adjacent sampling dates were compared. For the experiment evaluating seasonal effects, comparisons were made among all post-


treatment sampling periods and between the two pre-treatment sampling periods as well as between the last pre-treatment and first post-treatment periods.

R e s u l t s

Fate and persistence

Water samples taken 16 h after carbofuran application suggested that the initial concentrations of carbofuran were higher than the target value of 14#g 1-1 in two of the four t rea tment ponds and approached the target concentration in the other two ponds (Table 2). Using degradat ion rates in water as shown in Table 3, we calculated according to Seiber et al. (1978) that the initial concentrations in the water were 9.8, 16.3, 45.7 and 36.8#g 1-1 in ponds T1, T2, T3 and T4, respectively. The higher initial concentrations

Table 2. Concentrations of carbofuran in the water column, submersed plants and sediments in the four treatment ponds at two times after spraying

Carbofuran concentration c

Pond Hours pH a Temperature Water Plants Sediment post-spray (oc)b ~ g 1-1) (~g g-l) (~g g-l)

T1

T2

T3

T4

16 9.3 19 9.2 0.30 0.04 (17-22) (9.1-9.3) (0.09-0.97) (0.03-0.05)

124 9.0 - 3.1 0.08 ND e (1.2-8.0) d (0.04-0.16)

16 8.9 19 14.4 0.08 ND (18-21) (11.8-17.6) (0.02-0.25)

124 9.0 - 4.1 0.07 e ND (2.4-6.9)

16 8.9 19 32.5 0.44 0.05 f (17-22) (12.9-81.9) (0.15-1.28)

124 8.9 - 2.8 0.10 ND (1.6-4.7) (0.05-0.21)

16 8.0 16 32.6 0.20 0.13 f (14-20) (24.6-43.2) (0.04-0.92)

124 8.9 - 11.9 0.04 d ND (8.0-17.7) (0.02-0.10)

aMean, n = 4.

bMean daily temperature (range), based on max-min temperatures for 24 h period before sampling. CMean (95% CL, n = 4), values back-transformed from loge(Y + 1) where Y is carbofuran concentration. an = 3. Carbofuran not detected in fourth sample. Detection limit = 0.5/~g 1-1 for water and 0.02 #g g-1 for plants and sediments. eCarbofuran not detected in any samples (n = 4). fCarbofuran detected in only one sample.

Wayland and Boag

Table 3. Rates of loss of carbofuran from water and submersed plants in four ponds sprayed with this insecticide

Pond Medium Intercept a Slope a R 2b /1/2 c (h)

T1 Water d 2.38 -0.0102 0.87 68 Plants 5.91 -0.0120 0.61 58

T2 Water 2.86 -0.0117 0.89 59 Plants e . . . .

T3 Water 3.85 - 0.0229 0.90 30 Plants 6.30 -0.0138 0.69 50

T4 Water 3.63 -0.0093 0.88 74 Plants 5.53 -0.0143 0.58 48

alntercept and slope values based on loge(Y + 1) where n = 8, Y is carbofuran in #g 1 -I of water or p.p.b, of plant based on samples taken at 16 and 124 h post-spray. bCoefficient of determination. tHai f-life. an=7. eCarbofuran was below detection limit in three of four samples at 124 h. No attempt was made to estimate slope or half-life.

175

in ponds T3 and T4 were unplanned and may have resulted from inaccurate measure- ment of the pond volume, of the mass of Furadan ® used in the sprayer or a combination of the two.

As expected carbofuran degraded rapidly in all t reatment ponds (Table 2). The calculated tl/2 of carbofuran in the water column ranged from 30 to 74 h (Table 3). A two-way A N O V A revealed a significant pond × sampling time interaction (p = 0.0004, 3,23 dr) with respect to loss rates from the water column, indicating that the Kob between 16 and 124 h was not the same across all ponds. Unplanned comparisons of slopes (Sokal and Rohlf, 1981) indicated that the KobS in the water column were similar among ponds T1, T2 and T4 but that pond T3 differed from the other three ponds (p < 0.05). This difference occurred despite the fact that pond T3 was similar to both ponds T1 and T2 in pH and water temperature (Table 2). Pond T4 had the longest tl/2 in the water column and the lowest pH and water temperature, which is consistent with earlier work demonstrating the important role that these variables play in the degradation of carbamate insecticides in water (Aly and E1-Dib, 1972; Chapman and Cole, 1982).

Mean concentrations of carbofuran in submersed aquatic plants ranged from 0.08 to 0.44/~gg -1 16h after spraying. Carbofuran bioconcentrated in plants. P l a n t : w a t e r ratios of carbofuran ranged from 3.4 in pond T4 124 h after spraying to 35.7 in pond T3 124 h after spraying (see Table 2).

Half-lives of carbofuran in plants ranged from 48 to 58 h (Table 3), loosely paralleling

the tl/2 in water.


With the exception of pond T1 and one sample from the shallowest site in each of ponds T3 and T4, carbofuran was not detected in bottom sediments (Table 2). Sediments in pond T1 contained 0.038 _+ 0.010t~g g-1 carbofuran 16 h after spraying. Carbofuran was not detected in any sediment samples taken at 124 h.


Sixty-seven taxa were recorded in samples taken from the study ponds. However, 58 occurred in <50% of the samples and 46 were never recorded in at least one of the study ponds. We subjectively determined that only five groups were sufficiently abundant to warrant analysis: the amphipod Hyalella azteca, larvae of the subfamily Chirono- minae (Chironomidae), nymphs of the mayfly genus Caenis, nymphs of the damselfly family Coenagrionidae and larval caddisflies (Trichoptera), representing nine genera and five families. These five taxa accounted for 50% of the total number of macroinvertebrates recorded, the remaining 62 taxa accounting for the other 50%.

Of these taxa, only H. azteca and Caenis showed different abundances between shallow and deep zones (Wilcoxon signed rank tests, p < 0.01). The three other taxa were not different in their depth distributions (p = 0.09, 0.10 and 0.38 for Chironominae larvae, coenagrionid nymphs and trichopteran larvae, respectively). Therefore, data from shallow and deep zones were combined for these three taxa in subsequent analyses.

To identify any immediate, deleterious effect of carbofuran on macroinvertebrates, it was necessary to determine whether changes in abundance between sampling dates were of either a different magnitude or direction in the treatment ponds relative to the control ponds. Moreover, it was necessary to demonstrate that before spraying with carbofuran, any changes in abundance of macroinvertebrates recorded in the treatment ponds paralleled those in the control ponds from one date to the next.

Within 2 days after the application of carbofuran, the abundance of H. azteca declined significantly in the shallow zones of the treated ponds in a manner which was not paralleled in the controls (T x D: F = 9.65, 3,6 df, p = 0.01) (Fig. 1); in the deep zones this pattern was not significant (T × D" F = 0.69, 3,6 df, p = 0.59), although there was a trend towards a decline in the carbofuran-treated ponds (Fig. 1). Although H. azteca did not decline in the deep zone of pond T2, a decline in the deep zone of pond T4 was strongly evident. In that pond, its abundance in the deep zone declined from pre-treatment levels of 44 _+ 5 and 77 _+ 20 to post-spray abundances of 9 +_ 2 and 6 +_ 3 (x + SE, n = 4, based on untransformed data). The only significant change noted when means of adjacent dates were compared, was in the shallow zones of treated ponds between the last pre-spray and first post-spray sampling dates (t' = 5.0, 6 df, k = 3 comparisons, p < 0.01); during this period their abundance declined by -91%.

The patterns of change in larval Chironominae between adjacent dates, although showing a post-spray decline, did not differ significantly (F = 1.3, 3,6 dr, p = 0.35) (Fig. 1). Thus, spraying with carbofuran did not produce any significant reduction in this taxon.

For coenagrionid nymphs, there was no significant change (F = 1.5, 3,6 df, p = 0.3) after exposure to carbofuran based on the experimental procedure used (Fig. 1). It is possible that low numbers of coenagrionids coupled with the high residual variance (CV = 80%) may have masked a reduction in damsel fly nymphs following carbofuran application. In pond T2 the mean (_+SE, n = 8) numbers of Coenagrionidae per sample were 4 +_ 2 and 6 + 3, 2 days and 1 day before spraying and only 0.5 +_ 0.3 and 0.3 _+

Wayland and Boag

H. aztoca

+

H. ~tz't~a

177

o - - - -

~1.4 o.. ~ 1.2

0 0.8

0.6

ChlmnomlnaQ Coenagdonldae 4~

0 2 _ +

o~

=o

-1 2 l n c l l o p t e r a DAYS o ~ - ~ 0 r ~ s r - s ~ . ~ y

-1 2

DAYS PRE- AND POST-SPRAY

-2

Fig. 1. Abundance (mean + SE) of selected groups of macroinvertebrates in two control ponds (C2 and C3) and two treatment ponds (T2 and T4). Samples taken 1 and 2 days before spraying and 2 and 3 days after spraying. The carbofuran spraying day (23 July) is indicated by the arrow.

0.0, 2 and 3 days after spraying. Coenagrionid nymphs were observed dying or in distress on the surface of pond T2 1 day after spraying. Similar declines were not noted in pond T4 despite the higher concentration of carbofuran in that pond (Table 2). Thus, it is possible that carbofuran reduced coenagrionid numbers in pond T2 but not in pond T4.

Four genera of trichopteran larvae (Triaenodes, Limnephilus, Molanna and Agrypnia) occurred in the ponds during late July. Their numbers were pooled to determine whether or not trichopteran larvae were adversely affected by exposure to carbofuran. The abundance of these larvae changed in the treated ponds in a manner that was not paralleled in the control ponds (F = 4.9, 3,6 dr, p = 0.048). The only significant change that occurred was in the treated ponds between the last pre-spray and first post-spray dates (t' = 3.7, 6 df, k = 3 comparisons, p = 0.04) (Fig. 1); between these two sampling dates, trichopteran larvae declined -92%.

Over the whole season, the abundance of H. azteca, in treatment ponds was reduced relative to that in control ponds (Fig. 2). Only in the shallow zone, however, did this decline approach significance (T × D: F = 2.4, 4,24 df, p = 0.08). Comparisons of means indicated no change in abundance in shallow zones of control ponds (all comparisons:


~__~ C O N T R O L

i TREATMENT H. az'teca H. a2"teca

ShaJh~ Deep

C o e n a g r i o n i d a e ~ C h i r o n o m i n a e

tl'liltl t t t 0 ~ ~ ~ 0 ~

JUNE JULY AUG. SE,c~T. MAy JUNE JULY AUG, SEPT. MAY

M O N T H M O N T H

Fig. 2. Abundance (mean + SE) of the indicated macroinvertebrate groups in four control and four carbofuran-treated ponds from June 1986 until May 1987. The week of carbofuran spraying (23-30 July) is indicated by the arrow.

24 df, k = 5 comparisons, p > 0.10). However in treatment ponds, a significant reduction occurred between the July pre-spray sample and the August post-spray sample (t' = 3.6, 24 dr, k = 5, p < 0.01); this reduction was approximately 7-fold from 23 to three individuals per sample (mean, back-transformed from log(x + 1)) (Fig. 2). This reduction was sustained through to May 1987 in the shallow zones of treated ponds (Fig. 2) which suggests that populations of H. azteca, had not recovered 10 months after exposure to carbofuran. In the deep zones of the treatment ponds a trend towards reduced numbers of this amphipod, following spraying, was similar to that in the shallow zone (Fig. 2), although low mean abundances and high residual variation prevented detection of a significant reduction (T × D, p = 0.23; date, p = 0.57). In the deep zones of the control ponds H. azteca abundance suggested a stable or slightly increasing trend throughout the study, whereas in the treated ponds there was an overall declining trend mainly because of the August mean value which decreased relative to that in the control ponds (Fig. 2). The overall decline in the treatment ponds reflects the drop in abundance recorded in ponds T3 and T4 where mean numbers (__+SE, n = 4) in the deep zone fell from 10 + 2 and 60 + 12 individuals per sample pre-spraying to 2 + 4 and 0.8 + 0.5


individuals post-spraying. Initial concentrations of carbofuran in these two ponds were higher than in pond T1 in which the abundance of this amphipod in the deep zone was relatively constant from July (5 +_ 2 individuals per sample) through to August (7 + 3 individuals per sample).

Throughout the study, the abundance of Coenagrionidae was relatively similar in treated and control ponds with one exception; in the treated ponds, the mean abundance of this taxon is increased significantly between August 1986 (x = 3 individuals per sample) and May 1987 (x = 9 individuals per sample) (t' = 3.6, 24 df, k = 5 comparisons, p < 0.01) (Fig. 2). This contrasts with the control ponds where there were no significant changes in mean abundance among any of the post-treatment months (x = 3 individuals in August, 5 in September and 4 the following May) (all comparison p > 0.10).

In neither larval Chironominae nor the mayfly Caenis was a differing pattern of seasonal abundance recorded in the treatment and control ponds. The absence of a significant T × D interaction for both taxa (Chironominae: F = 0.9, 4,24 df, p = 0.49; Caenis shallow zone: F = 0.59, 4,24 df, p = 0.67; Caenis deep zone: F = 0.1, 4,24 df, p = 0.99) suggested that neither of these taxa were numerically affected by exposure to carbofuran in the treated ponds. For Chironominae, mean larval abundance in control ponds increased most noticeably between August and September (Fig. 2) (t' = 2.7, 24 df, k = 5 comparisons, p = 0.07), whereas in treated ponds, this group increased continuously between August 1986 and May 1987 (t' = 3.6, 24 df, k = 5 comparisons, p < 0.01). Their increase in the treatment ponds was slightly greater than in the control ponds (Fig. 2).

Despite the decrease in Trichoptera demonstrated in the samples taken on the second and third days after spraying (Fig. 1), their seasonal abundance in the treated ponds was not significantly lowered below that in the control ponds as indicated by the absence of a T × D interaction (F = 1.5, 4,24 df, p = 0.24). Trichoptera reached seasonal lows in both treated and control ponds in late July just before application of the insecticide. Thereafter they increased and their numbers peaked in May 1987 in both treated and control ponds (Fig. 3). Throughout this study, nine genera representing five families were encountered. There were three leptocerids (Mystacides, Triaenodes and Oecetis), three limnephilids (Limnephilus, Nemotaulius and Anabolia) and one phryganeid (Agrypnia) in addition to Molanna (Molannidae) and Polycentropus (Polycentopodi- dae). Only Triaenodes, Limnephilus, Molanna and Agrypnia were present at the time the ponds were sprayed. The low abundance of larvae coupled with the absence of several genera from the ponds at the time they were sprayed made it unlikely that carbofuran would exert a strong seasonal effect on Trichoptera. Nevertheless, it is noteworthy that the mean abundance of trichopteran larvae was higher in the treated ponds than in the control ponds in June and July before spraying, whereas after spraying it was consistently lower (Fig. 3). This trend in the data can be attributed in part to the relative post-treatment scarcity in treated ponds of those genera that were present at the time of spraying. In the treated ponds, 94 larvae of the four genera that were in the ponds immediately before or after spraying were obtained in samples during the pre- treatment phase of the study examining seasonal effects. In the control ponds, 38 such larvae were collected during the same period. During the entire post-treatment phase of the study, only 47 such larvae were obtained from the treatment ponds, while 216 were collected in samples from control ponds. A 2 × 2 contingency table analysis taking into account total pre-spray and post-spray abundances of these four genera in treated


and control ponds indicated that their post-spray abundances were independent of their pre-spray abundances (Z 2 = 109.0, 1 df, p < 0.005). In contrast, when considering only those genera that were absent in samples taken immediately before and after spraying, pre-spray and post-spray abundances were not independent in treatment and control ponds (X 2 = 3.6, 1 df, p > 0.50). There were large increases in this group during the post-treatment phase of the study in both treatment and control ponds (from 23 to 169 and from 7 to 140 in treatment and control ponds, respectively). This is similar to the increase in the control ponds by the genera present in the ponds at the time of spraying

Caenis Caenis ShJlOW Deep

1.2

1 I 11 I O ~ M'¥

0.9 Triehoptera ~ C~ONT~O~

JUNE J U L Y AUG. SEPT. MAY MONTH

Fig. 3. Abundance (mean + SE) of the indicated macroinvertebrate groups in four control and four carbofuran-treated ponds from June 1986 until May 1987. The week of carbofuran spraying (23-30 July) is indicated by the arrow.

but is the opposite of the pattern of this group in the treated ponds. These data suggest that the reduction in trichopteran larval abundance in treated ponds relative to control ponds during the post-treatment phase of the study was due mainly to the deleterious effects of carbofuran on those genera present in the ponds at the time of spraying. Unfortunately, small numbers of larvae per sample and high among-pond variance prevent meaningful statistical analyses of each group of larvae separately according to the original experimental design using repeated measure ANOVA.


Discussion

Fate and persistence

The degradation of carbofuran and other carbamate insecticides have been shown to be log-linear functions of time (Aly and E1-Dib, 1972; Seiber et al., 1978; Gibbs et al., 1984). Assuming that degradation rates in the treatment ponds also followed a log-linear pattern over time, it was possible to estimate degradation rates for each pond despite the fact that they were sampled only twice following spraying. If our assumption of log-linear degradation rates was wrong, then our calculated degradation rates and half- lives may also be in error.

At 25°C, the estimated tl/2 of carbofuran through hydrolysis in aqueous solution was 5 days at pH 8.0 and 0.8 days at pH 9.0 (NRCC 1979). Chapman and Cole (1982) calculated the tl/2 from the hydrolysis rate constant and found it to be 7 days at pH 8.0 and 25°C. Estimates of tv2 in the study ponds which had temperatures in the range of 16-19°C and pH values ca. 9.0 except for T4 in which the pH was 8.0 at the time of spraying (Table 2), were in the range of 30-74 h, a time span that is bracketed by the half-lives reported above.

The more rapid degradation of carbofuran in pond T3 when compared to the other ponds coupled with the similarity in mean pH values for these ponds (Table 2) suggests that factors other than pH may also be important in determining carbofuran degradation rates in alkaline prairie ponds.

Rates of loss of carbofuran from aquatic plants in this study appeared to parallel its corresponding rates of loss from water, possibly because carbofuran degradation in the water column resulted in desorption from sediments or plant surfaces as has been suggested in other studies (Hughes et al., 1980; Huckins et al., 1986).

The relatively low recovery of carbofuran from pond sediments agrees with the earlier observation that carbofuran is unlikely to accumulate on pond bottoms (NRCC 1979) but, with the exception of pond T1, differs from the results of Klaassen and Kadoum (1979) who recorded carbofuran residues two to eight times higher in the sediments than in overlying water up to 3 days after application. We cannot explain why carbofuran did not accumulate in sediments in ponds T2-T4.

Carbofuran concentrations were three to 50 times higher in plants than in surrounding water. Studies dealing with other pesticides have shown similar trends (Mulla et al., 1966a; Hurlbert et al., 1970; Rawn et al., 1982; Huckins et al., 1986). The tendency of carbofuran to accumulate on aquatic plants may have implications for its bioavailability to aquatic invertebrates, many of which are phytophillous in pond ecosystems (Krull, 1970; Voights, 1976). It has been shown that, for hydrophilic pesticides such as carbofuran (Kenaga and Goring, 1980), bioconcentration by invertebrates appears to be unrelated to features of their microhabitat (Muir et al., 1983). In spite of this finding, the possibility exists that under field conditions promoting rapid degradation of a hydrophilic pesticide, epiphytic invertebrates may be subjected to high levels of exposure as a result of adsorption-desorption phenomena at the plant-water interface. This is consistent with the suggestion in the literature that desorption of pesticides from sediments may result in temporarily higher concentrations in pore water or water closely associated with the sediments than in overlying water, resulting in greater concentrations in sediment- associated invertebrates than in those in overlying water (Lynch and Johnson, 1982;


Muir, et al., 1985). Yet it contrasts with the phenomenon of reduced bioavailability of pesticides to biota in the presence of highly organic substrates as has been examined in detail by DiToro et al. (1991). The phenomenon of microhabitat-induced effects of pesticide bioavailability needs to be more fully examined, especially as it relates to aquatic plants and their epiphytic fauna.


The addition of carbofuran to the study ponds at concentrations in the range 9-32/xg 1-1 16 h after spraying, immediately reduced the abundance of H. azteca in the shallow zones (where it was most abundant) and possibly in the deep zones of treated ponds. The toxicity of carbofuran within this range of concentrations to H. azteca has been reported previously (Wayland and Boag, 1990; Wayland, 1991). Carbofuran at 5/xg 1-1 caused a 25% increase in mortality of caged H. azteca; at 25/xg1-1 the increase in mortality was 81% (Wayland and Boag, 1990). At 25 #g 1-1, the abundance of H. azteca in enclosures declined to 10% of pre-treatment levels (Wayland, 199!).

The initial reductions in 14. azteca abundance in the treated ponds persisted for the remainder of the summer and into the early part of the following spring indicating the absence of any mechanism that would enable the populations to recover after the ponds were sprayed. This result contrasts with the gradual post-treatment recovery of H. azteca reported by Wayland (1991) in another study. In that study, it was suggested that a low level of recruitment of 14. azteca probably continued throughout the summer. After the initial, carbofuran-induced kill, the low level of recruitment exceeded mid- and late- summer mortality of the few juveniles that survived the treatment, thus enabling recovery to occur. In contrast, juvenile mortality, which is normally high during mid- summer in temperate populations of H. azteca (Cooper, 1965), matched or exceeded mid-summer recruitment in the control enclosures, causing the population to remain stable during the post-treatment period. In this study, breeding H. azteca (in amplexus) were not observed during or after spraying. While this observation is inconsistent with evidence suggesting that H. azteca is capable of producing offspring into late summer in northern temperate lakes (Cooper, 1965; Mathias, 1971; de March, 1977; Wen, 1992), it may explain, nonetheless, the lack of any recovery by this species up to 10 months after spraying (Fig. 2).

While mortality rates are not apparently inversely related to density in H. azteca (Wilder, 1940), they do vary temporally according to the age structure of the population (Cooper, 1965). Thus, a sudden insecticide contamination of a pond could cause the premature removal of an age class that was destined to die at a later date anyhow. Through age-related natural mortality, populations in unaffected ponds might eventually decline relative to those in contaminated ponds which suffer large-scale, sudden and non-age-related mortality. There is some evidence of such a pattern in the shallow zones of the study ponds. However, in the deep zones, no such pattern emerged (Fig. 2).

The abundance of Trichoptera declined in the treated ponds immediately following spraying. This result corresponded well with our observation of many stressed and dying phryganeid and limnephilid larvae on the surface of ponds T1 and T4, 16 h after spraying. The decrease in Trichoptera is consistent with the high mortality of Limnephi- /us larvae that were held in cages and exposed to carbofuran at 25 fig l- 1 (Wayland and Boag, 1990).

Unlike H. azteca, Trichoptera increased again following the initial post-spray decline.


By August, their abundance in treated ponds was only slightly less than pre-spray levels and had surpassed pre-spray levels by September (Fig. 3). Overall, seasonal changes in trichopteran abundance paralleled those in the controls. Patterns of change in the study ponds were consistent with life-history information on caddisflies adapted to lentic, permanent waterbodies. This information indicates that such caddisflies are abundant as larvae during spring and early summer, decline as a result of pupation and emergence from late spring through mid-summer and increase again during the fall (Winterbourn, 1971a,b; Bert6 and Pritchard, 1986). Thus a mid-summer application of a short-lived insecticide such as carbofuran may not affect that portion of the trichopteran fauna of prairie ponds whose life cycles are characterized by spring or early summer emergence and late summer egg hatching.

Species adapted to temporary waterbodies could be more prone to long-term effects since they do not emerge until later in the summer (Bert6 and Pritchard, 1986). We did not determine the life cycles of the four genera that were present as larvae at the time that carbofuran was sprayed. Nevertheless, it is interesting that their abundance did not increase in the treatment ponds to the same extent as in the control ponds during the post-spray period. Finally, trichopteran larval populations can be quite high during the spring (Fig. 3). Thus the effects of an earlier date of insecticide application may be quite different from those observed in this study since recolonization would depend heavily on adults dispersing from neighbouring ponds that were not affected. This is an important consideration because carbofuran spraying for grasshopper control is normally done during May and June on the Canadian prairies. Our late July application date may complicate interpretation of the results with regards to the effects of carbofuran spraying for grasshopper control on pond macroinvertebrates because their emergence and larval development phenologies may influence their larval densities in a pond at the time it is sprayed as well as their sensitivity to carbofuran.

Within the range of concentrations used in this study, there was no detectable effect of carbofuran on larval Chironominae (Fig. 1). The toxicity of carbofuran to Chironomi- nae varies widely with species and/or test conditions. The 48 h ECs0 for Chironomus riparius was 56pg1-1 at a pH of 8.5-9.0 (Johnson, 1986). Chironomus tentans is apparently more sensitive with an LCs0 of 1.6pg 1-2 under unspecified test conditions (Karnak and Collins, 1974). Acute mortality in the field was 88% for Goeldichironomus holoprasinus and 90% for Chironomus stigmatarus at an application rate of approximately 94pg 1-2 (calculated from Mulla and Khasawinah, 1969). In this study, neither the species nor generic composition of the Chironominae were quantified. However, several larvae collected from ponds T2 and T4 during the days immediately prior to and following spraying were identified. Those from T2 were predominately Tanytarsus, while those from T4 were mainly Paratanytarsus, Glyptotendipes and Chironomus. It is possible that these groups were not as sensitive to carbofuran as C. tentans which suffered high mortality when held in small cages in the treatment ponds at the time of spraying (Wayland and Boag, 1990).

Over the whole study, the numbers of Chironominae larvae were not affected substantially by the mid-summer application of carbofuran. Nevertheless, a slightly greater increase in their abundance was recorded in the treated ponds than in the control ponds during the post-spray phase of the study (Fig. 2). Kennedy et al. (1970) observed a similar pattern in ponds treated with methoxychlor, wherein numbers of chironomid larvae had increased to a greater degree in treated ponds 28 days after spraying. They


suggested that a reduction in competitors or predators and/or the organic enrichment resulting from the large-scale die-off of invertebrates immediately after treatment could have accounted for the disproportionate increase. In this study, it is possible that the slight, disproportionate increase in the treated ponds relative to the control ponds resulted from the reduction in a potential competitor such as H. azteca, which, like many of the Chironominae, is a benthic herbivore and detritivore (Hargrave, 1970), feeding primarily on fine and coarse particulate organic matter which it scavenges from sediment or submerged plant material (Rooke, 1986). Nevertheless, since this difference was slight and non-significant, the possibility that it might have been due to sampling variability alone cannot be discounted.

There was no detectable reduction in Coenagrionidae following carbofuran spraying. This is consistent with the results of cage experiments with the coenagrionid nymph, Enallagma, wherein the application of carbofuran in these ponds failed to cause significant mortality (Wayland and Boag, 1990). Nevertheless, the lower mean survival of Enallagma reported in that study ( -67% in control ponds and 43% in treated ponds) coupled with the observation of stressed and dying individuals on the surface of pond T2 16 h after spraying, suggests that some carbofuran-induced mortality did occur.

During the post-treatment phase of the study, the abundance of coenagrionid nymphs in control ponds remained fairly stable, whereas in treated ponds it increased (Fig. 2). However, this difference was not uniform among ponds, nor did the magnitude of the difference increase with increasing carbofuran concentration. Therefore, it is unlikely that this was a secondary result (sensu Hurlbert, 1975) of the carbofuran treatment. Indeed, the larger post-treatment increase in the treated ponds relative to the controls could be attributed primarily to an increase in only one of the treatment ponds (T2). In this pond, Coenagrionidae abundance increased from 4 + 1 in August to 14 _+ 4 the following May (x _+ SE, n = 8). Increases in the three other treatment ponds were more modest: from 0.4 + 0.1 to 5 + 2 in T1, from 2 + 1 to 4 + 1 in T3 and from 6 + 2 to 10 + 3 in T4. During the same period their abundance increased marginally in two control ponds (from 8 + 2 to 12 + 2 and from 0.5 + 0.0 to 2 + 1) and decreased marginally in the other two. It is possible that some factor independent of carbofuran such as increased dispersal of adults to or increased deposition and hatching of eggs during the post-spray period in pond T2 and to a lesser extent in ponds T1 and T4, might have accounted for the difference between treatment and control ponds. It is noteworthy that a great deal of asynchrony in emergence phenologies and timing of egg deposition has been demonstrated in two cocnagrionid species that occur commonly in central Alberta (Baker and Clifford, 1981, 1982). If such asynchrony existed among coenagrionid populations in the study ponds, patterns of population change over time may have differed among ponds.

Caenis abundance declined through emergence and reached seasonal lows by late July around the time that carbofuran was sprayed (Fig. 3). Their abundance began to increase again in August after carbofuran had degraded. Thus, there was little opportunity for carbofuran to have a negative impact on this taxon in this study. On the contrary, since both Caenis and H. azteca are considered to be deposit feeders on epibenthic algae (Hargrave, 1970; Cummins, 1973), it seemed possible that the carbofuran-induced reduction in H. azteca might have reduced competition and/or increased food availability enough to enable Caenis to increase at a faster rate in the treated ponds. While previous studies have indicated that competitive interactions can affect rates of population change for aquatic invertebrates in benthic communities (Hall et al., 1970; Cantrell and


McLachlan, 1977) and that pesticide-induced changes in benthic communities can result in large increases in abundance of resistant groups (Sanders et al., 1981; Cuffney et al., 1984), we did not detect such a secondary effect on Caenis in this study (Fig. 3).

Notwithstanding the seasonal impact on H. azteca which had been one of the numerically dominant taxa in the ponds before spraying, it is possible that the benthos in the treated ponds was not sufficiently changed by carbofuran to allow detection of any secondary effects, with the possible exception of the Chironominae as noted above. One explanation for the lack of secondary effects is that differences between H. azteca and the other taxa in microhabitat use may have reduced competitive or predator-prey interactions. Hanson (1990) found that amphipods including H. azteca dominated rooted plant weedbeds whereas other invertebrates, including chironomids, dominated Chara stands in a lake in central Alberta. Segregation of different macroinvertebrates, including H. azteca and chironomids according to macrophyte stand composition was also noted in a small river in southern Ontario (Rooke, 1986) and in a series of lakes in southern Qu6bec (Cyr and Downing, 1988). As noted above, Caenis dwells on pond bottoms (Magdych, 1979), whereas H. azteca uses both sediments and macrophytes (Hargrave, 1970; Rooke, 1986). The trichopterans in this study occur mainly on macrophytes and occur only rarely on sediment (Merritt and Cummins, 1984). Perhaps the specific habitat use patterns of these taxa prevented them from taking full advantage of the carbofuran-induced reduction of H. azteca. Another possible explanation for the lack of secondary effects is that differences among taxa in feeding modes may have reduced competition for food. For example, a filter-feeding chironomid, Glyptotendipes paripes, which is very common in prairie ponds, was unaffected by changes in density of a collector-gatherer chironomid, C. riparius (Rasmussen 1985). Thus, changes in the density of H. azteca which exhibits a generalized feeding mode, including that of a collector-gatherer, shredder and possibly a scraper (Hargrave, 1970; Rooke, 1986), may not affect filter-feeding chironomids in prairie wetlands. However, such changes may affect invertebrate taxa using similar feeding modes, for example, collector-gatherer chironomids. It is possible that the relatively weak secondary response of the Chironomi- nae to the carbofuran-induced reduction of H. azteca in this study was primarily due to a response by collector-gatherer chironomids. Most of the trichopterans in this study are classified as collector-gatherers and shredders, similar to H. azteca. It is therefore puzzling that they did not respond positively to the reduction in H. azteca. Perhaps slight differences in microhabitat use, as noted above or an increase in predation pressure as predators shifted prey types to compensate for the reduction in H. azteca as a food source may have prevented such a response.

Although, we did not detect strong evidence of secondary effects, it would not be appropriate to conclude that such effects did not occur, since 50% of the total number of invertebrates recorded in this study were not considered in the analysis.

Acknowledgements

We thank the landowners who graciously permitted use of their ponds for this study: M. Allison, A. Champagne, R. Hoogland, A. Lazarz, R. Petryshyn, J. Roelofs, A. Sustrik and J. Yaremko. T. Tessier, G. Hutchison, L. LeClair, R. Pattenden, D. Sheloff, R. Wayland, D. Wrubleski and Shi Yufang provided capable assistance during various phases of this study. Residue analysis was done by Enviro-Test Labs, Edmonton. This

186 Fate o f carbofuran and its effects on aquatic macroinvertebrates

study was funded by The World Wildlife Toxicology Fund (Canada), the Delta Water- fowl and Wetlands Research Station, a Natural Science and Engineering Research Council (NSERC) grant to D.A.B. and an NSERC scholarship to M.W.

References

All, A. and Mulla, M.S. (1976) Chironomid larval density at various depths in a southern California water percolation reservoir. Environ. Entomol. 5, 1071-4.

Aly, O.M. and E1-Dib, M.A. (1972) Studies on the persistence of some carbamate insecticides in the aquatic environment. In Gould, R.F. ed. Fate of organic pesticides in the aquatic environment, pp. 210-43. Washington DC: American Chemical Society.

Baker, R.L. and Clifford, H.F. (1981) Life cycles and food of Coenagrion resolutum (Coenagrioni- dae: Odonata) populations and Lestes disjunctus (Lestidae: Odonata) populations from the boreal forest of Alberta, Canada. Aquat. Insects 3, 179-91.

Baker, R.L. and Clifford, H.F. (1982) Life cycle of an Enallagma boreale (Selys) population from the boreal forest of Alberta, Canada (Zygoptera: Coenagrionidae). Odonatologica 11,317- 22.

Bert6, S.B. and Pritchard, G. (1986) The life histories of Limnephilus externus (Hagen), Anabolia bimaculata (Walker) and Nemotaulius hostilis (Hagen) (Trichoptera: Limnephilidae) in a pond in southern Alberta, Canada. Can. J. ZooL 64, 2348-56,

Cantrell, M.A. and McLachlan, A.J. (1977) Competition and chironomid distribution patterns in a newly flooded lake. Oikos 29, 429-33.

Chapman, R.A. and Cole, C.M. (1982) Observations on the influence of water and soil pH on the persistence of insecticides. J. Environ. Sci. Health B17,487-504.

Charnetski, W.A., Erickson, D., Pullishy, G.M. and Kortsh, W.E. (1977) Carbofuran Degrada- tion in Southern Alberta Pond and Lake Water. Edmonton, Alberta, Canada: Environmental Protection Service, Alberta Environment.

Cooper, W.E. (1965) Dynamics and production of a natural population of a freshwater amphipod, Hyalella azteca. Ecol. Monogr. 35, 377-94.

Cuffney, T.F., Wallace, J.B. and Webster, J.R. (1984) Pesticide manipulation of a headwater stream: invertebrate responses and their significance for ecosystem processes. Freshwater Invertebr. Biol. 3, 153-70.

Cummins, K.W. (1973) Trophic relations of aquatic insects. Ann. Rev. Entomol. 18, 183-206. Cyr, H. and Downing, J.A. (1988) The abundance of phytophilous invertebrates on different

species of submerged macrophytes. Freshwater Biol. 20, 365-74. de March, B.G.E. (1977) The effects of photoperiod and temperature on the induction and

termination of reproductive resting stage in the freshwater amphipod, Hyalella azteca (Saussure). Can. J. Zool. 55, 1595-600.

DiToro, D.M., Zarba, C.S., Hansen, D.J., Berry, W.J., Swartz, R.C., Cowan, C.E., Pavlou, S.P., Allen, I-I.E., Thomas, N.E. and Paquin, P.R. (1991) Technical basis for establishing sediment quality criteria for nonionic organic chemicals using equilibrium partitioning. Environ. Toxicol. Chem. 10, 1541-83.

Downing, J.A. (1979) Aggregation, transformation and design of benthos sampling programs. J. Fish. Res. Board Can. 36, 1454-63.

Flannagan, J.F. (1973) Sorting Benthos Using Flotation Media. Fisheries Research Board of Canada Technical Report 354.

Forsyth, D.J. (1989) Agricultural chemicals and prairie pothole wetlands: meeting the needs of the resource and the farmer - Canadian perspective. Trans. N. Am. Wildlife Nat. Res. Conf. 54, 59-66.

Gibbs, K.E., Mingo, T.M. and Courtmanch, D.L. (1984) Persistence of carbaryl (Sevin-4-oil) in


woodland ponds and its effects on pond macroinvertebrates following forest spraying. Can. Entomol. 116, 203-13.

Grue, C.E., DeWeese, L.R., Mineau, P., Swanson, G.A., Foster, J.R., Arnold, P.M., Huckins, J.N., Sheehan, P.J., Marshall, W.K. and Ludden, A.P. (1986) Potential impacts of agricultural chemicals on waterfowl and other wildlife inhabiting prairie wetlands: an evaluation of research needs and approaches. Trans. N. Am. Wildlife Nat. Res. Conf. 51, 357-83.

Hall, D.J., Cooper., W.E. and Werner, E.E. (1970) An experimental approach to the production dynamics and structure of freshwater animal communities. Limnol. Oceanogr. 15, 839-928.

Hanson, J.M. (1990) Macroinvertebrate size-distributions of two contrasting freshwater macroinvertebrate communities. Freshwater Biol. 24, 481-91.

Hargrave, B.T. (1970) Distribution, growth and seasonal abundance of Hyalella azteca (Amphi- poda) in relation to sediment microflora. J. Fish. Res. Board Can. 27, 685-99.

Hartman, W.A. and Martin, D.B. (1985) Effects of 4 agricultural pesticides on Daphnia pulex, Lemna minor and Potamogeton pectinatus. Bull. Environ. Contain. Toxicol. 35, 646-51.

Huckins, J.N., Petty, J.D. and England, D.C. (1986) Distribution and impact of trifluralin, atrazine and fonofos residues in microcosms simulating a northern prairie wetland. Chemos- phere 15, 563-88.

Hughes, D.N., Boyer, M.G., Papst, M.H., Fowle, C.D., Rees, G.A. and Baula, P. (1980) Persistence of three organophosphorus insecticides in artificial ponds and some biological implications. Arch. Environ. Contain. Toxicol. 9, 269-79.

Huribert, S.H. (1975) Secondary effects of pesticides on aquatic ecosystems. Residue Rev. 57, 81-148.

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

Hurlbert, S.H., Mulla, M.S., Keith, J.O., Westlake, W.E. and Dirsch, M.E. (1970) Biological effects and persistence of Dursban in freshwater ponds. J. Econ. Entomol. 63, 43-52.

Johnson, B.T. (1986) Potential impact of selected chemical contaminants on a northern prairie wetland: a microcosm evaluation. Environ. Toxicol. Chem. 5, 473-85.

Karnak, R.E. and Collins, W.J. (1974) The susceptibility to selected insecticides and acetylcholi- nesterase activity in a laboratory colony of midge larvae, Chironomus tentans (Diptera: Chironomidae). Bull. Environ. Contain. Toxicol. 12, 62-9.

Kenaga, E.E. and Goring, C.A.I. (1980) Relationship between water solubility, soil sorption, octanol-water partitioning and concentration of chemicals in biota. In Eaton, J.G., Parrish, R.R. and Hendricks, A.C. eds. Aquatic toxicology and hazard assessment, pp. 78-115. Philadelphia: American Society for Testing and Materials.

Kennedy, H.D., Eller, L.L. and Walsh, D.M. (1970) Chronic Effects of Methoxychlor on Bluegills and Aquatic Invertebrates. Washington DC: United States Fish and Wildlife Service, Techni- cal Paper of the Bureau of Sports Fisheries and Wildlife, No. 53.

Klaassen, H.E. and Kadoum, A.M. (1979) Distribution and retention of atrazine and carbofuran in farm pond ecosystems. Arch. Environ. Contain. Toxicol. 8, 345-53.

Krull, J.N. (1970) Aquatic plant-macroinvertebrate associations and waterfowl. J. Wildlife Manage. 34, 707-18.

Lynch, T.R. and Johnson, H.E. (1982) Availability of a hexachlorobiphenyl isomer to benthic amphipods from experimentally contaminated natural substrate. In Pearson, J.G., Foster, R. B. and Bishop, W.E. eds. Aquatic toxicology and hazard assessment, pp. 273-87. Philadel- phia: American Society for Testing and Materials.

Magdych, W.P. (1979) The microdistribution of mayflies (Ephemeroptera) in Myriophyllum beds in Pennington Creek, Johnston County, Oklahoma. Hydrobiologia 66, 161-75.

Mathias, J.A. (1971) Energy flow and secondary production of the amphipods Hyalella azteca and Crangonyx richmondensis occidentalis in Marion Lake, British Columbia. J. Fish. Res. Board Can. 28, 711-26.

188 Fate o f carbofuran and its effects on aquatic macroinvertebrates

Merritt, R.W. and Cummins, K.W. (1984) An Introduction to the Aquatic Insects of North America, 2nd edn. Dubuque, IA: Kendall/Hunt.

Mineau, P., Sheehan, P.J. and Baril, A. (1987) Pesticides and waterfowl on the Canadian prairies: a pressing need for research and monitoring. In Diamond, A.W. and Filion, F.L. eds. The value of birds, pp. 133-47. Princeton: Cambridge and Princeton Press.

Mittelbach, G.G. (1981) Patterns of invertebrate size and abundance in aquatic habitats. Can. J. Fish. Aquat. Sci. 38, 485-94.

Muir, D.C.G., Rawn, G.P., Townsend, B.E., Lockhart, W.L. and Greenhalgh, R. (1985) Bioconcentration of cypermethrin, deltamethrin, fenvalerate and permethrin by Chironomus tentans larvae in sediment and water. Environ. Toxicol. Chem. 4, 51-61.

Muir, D.C.G., Townsend, B.E. and Lockhart, W.L. (1983) Bioavailability of six organic chemicals to Chironomus tentans larvae in sediment. Environ. Toxicol. Chem. 2, 269-81.

Mulla, M.S. and Khasawinah, A.M. (1969) Laboratory and field evaluation of larvicides against chironomid midges. J. Econ. Entomol. 62, 37-41.

Mulla, M.S., Keith, J.O. and Gunther, F.A. (1966a) Persistence and biological effects of parathion residues in waterfowl habitats. J. Econ. Entomol. 59, 1085-90.

Mulla, M.S., Metcalf, R.L. and Geib, E.F. (1966b) Laboratory and field evaluations of new mosquito larvicides. Mosquito News 26, 236-42.

NRCC (1979) Carbofuran: Criteria for Interpreting the Effects of its Use on Environmental Quality. Ottawa: National Research Council of Canada, Associate Committee on Scientific Criteria for Environmental Quality, NRCC 16740.

Prepas, E.E. (1984) Some statistical methods for the design of experiments and analysis of samples. In Downing, J.A. and Rigler, F.H. eds. A manual on methods for the assessment of secondary productivity in fresh waters, pp. 266-335. London: Blackwell Scientific Publications.

Rasmussen, J.B. (1985) Effects of density and microdetritus enrichment on the growth of chironomid larvae in a small pond. Can. J. Fish. Aquat. Sci. 42, 1418-22.

Rawn, G.P., Webster, G.R.B. and Muir, D.C.G. (1982) Fate of permethrin in model outdoor ponds. J. Environ. Sci. Health B17, 463-86.

Rooke, B. (1986) Macroinvertebrates associated with macrophytes and plastic imitations in the Eramosa River, Ontario, Canada. Arch. Hydrobiol. 106, 307-25.

Rosenberg, D.M. (1973) Effects of dieldrin on diversity of macroinvertebrates in a slough in central Alberta. PhD thesis, University of Alberta, Edmonton.

Sanders, H.O., Walsh, D.F. and Campbell, R.S. (1981) Abate: Effects of Organophosphorus Insecticide on Bluegills and Invertebrates in Ponds. Washington, DC: United States Fish and Wildlife Service Technical Paper No. 104.

Seiber, J.N., Catahan, M.P. and Barril, C.R. (1978) Loss of carbofuran from rice paddy water: chemical and physical factors. J. Environ. Sci. Health, BI3, 131-48.

Sheehan, P.J., Baril, M., Mineau, P., Smith, D.K., Harfenist, A. and Marshall, W.K. (1987) The Impact of Pesticides on the Ecology of Prairie Nesting Ducks. Ottawa: Canadian Wildlife Service, Technical Report Series No. 19.

Siddaramappa, R., Tirol, A.C., Seiber, J.N., Heinrichs, E.A. and Watanabe, I. (1978) The degradation of carbofuran in paddy water and flooded soil of untreated and retreated rice fields. J. Environ. Sci. Health, BI3, 369-80.

Snedecor, G.W. and Cochran, W.G. (1980) Statistical Methods, 7th edn. Ames, IA: The Iowa University Press.

Sokal, R.R. and Rohlf, F.J. (1981) Biometry, 2nd edn. San Francisco: W.H. Freeman and Co. Swanson, G.A. and Meyer, M.I. (1973) The role of invertebrates in the feeding ecology of

Anatinae during the breeding season. In Waterfowl habitat management symposium, pp. 143-85. Moncton, NB, Canada: Atlantic Waterfowl Council.

Thorp, J.H. and Diggins, M.R. (1982) Factors affecting depth distribution of dragonflies and other benthic insects in a thermally destabilized reservoir. Hydrobiologia 82, 33-44.


Voights, D.K. (1976) Aquatic invertebrate abundance in relation to changing marsh vegetation. Am. Midl. Nat. 95,313-22.

Wayland, M. (1989) Fate of carbofuran and its effects on macroinvertebrates in prairie ponds. MSc thesis, University of Alberta, Edmonton.

Wayland, M. (1991) Effect of carbofuran on selected macroinvertebrates in a prairie parkland pond: an enclosure approach. Arch. Environ. Contain. Toxicol. 21,270-80.

Wayland, M. and Boag, D.A. (1990) Toxicity of carbofuran to selected macroinvertebrates in prairie ponds. Bull. Environ. Contam. Toxicol. 45, 74-81.

Wen, Y.H. (1992) Life history and production of Hyalella azteca (Crustacea: Amphipoda) in a hypereutrophic prairie pond in southern Alberta, Can. J. Zool. 70, 1417-24.

Wilder, J. (1940) The effects of population density upon growth, reproduction and survival of Hyalella azteca. Physiol. Zool. 13, 439-59.

Winer, B.J. (1971) Statistical Principles in Experimental Design, 2nd edn. New York: McGraw- Hill.

Winterbourn, M.J. (1971a) An ecological study of Banksiola crotchi (Trichoptera: Phryganeidae) in Marion Lake, British Columbia. Can. J. Zool. 49, 637-45.

Winterbourn, M.J. (1971b) The life histories and trophic relationships of the Trichoptera of Marion Lake, British Columbia. Can. J. Zool. 49, 623-35.

Yu, C.C., Booth, G.M., Hansen, D.J. and Larsen, J.R. (1974) Fate of carbofuran in a model ecosystem. J. Agricult. Food Chem. 22, 431-4.

Fate of carbofuran and its effects on aquatic macroinvertebrates in Canadian prairie parkland ponds

Documents

Transcript of Fate of carbofuran and its effects on aquatic macroinvertebrates in Canadian prairie parkland ponds