Population Consequences of Fipronil and Degradates to Copepods at

The predominant data used in ecological risk assessment today are individual-based rather than population-based; yet environmental policies are usuall...
0 downloads 0 Views 466KB Size
Environ. Sci. Technol. 2004, 38, 6407-6414

Population Consequences of Fipronil and Degradates to Copepods at Field Concentrations: An Integration of Life Cycle Testing with Leslie Matrix Population Modeling G . T H O M A S C H A N D L E R , * ,† TAWNYA L. CARY,† ADRIANA C. BEJARANO,† JACK PENDER,‡ AND JOHN L. FERRY‡ Department of Environmental Health Sciences, Norman J. Arnold School of Public Health, and Department of Chemistry, University of South Carolina, Columbia, South Carolina 29208

The predominant data used in ecological risk assessment today are individual-based rather than population-based; yet environmental policies are usually designed to protect populations of threatened species or communities. Most current methods in ecotoxicology are limited by largely logistic/ technology-driven requirements that yield data for a relatively small number of test species and end points that focus on acute lethality or sublethal nonproductionbased parameters (e.g., biomarkers, mutagenesis, genetic change, physiological condition). A contrasting example is presented here showing the predictive ability of meiobenthos-based full life cycle toxicity testing to extrapolate multi-generational effects of chemicals on variables of import to population growth and maintenance. Less than 24-h-old larvae of a meiobenthic copepod were reared individually in 96-well microplate exposures to parent and degradates of the phenylpyrazole insecticide fipronil. Survival, development rates, sex ratio change, fertility, fecundity, and hatching success were tracked daily for 32 d through mating and production of three broods in spiked seawater. These data were then inserted in a Leslie (Lefkovitch) matrix stage-based population growth model to predict relative rates of population increase (λ) and changes in net population growth with time and toxicant concentration. Field-reported test concentrations produced strong reproductive (52-88%) and net production (40-80%) depressions for parent (at 0.25 and 0.5 µg/L), desthionyl (0.25 and 0.5 µg/L), and sulfide (0.15 µg/L) moieties as compared to controls. Spiked sediment exposures of 65-300 ng of fipronil/g of dry sediment yielded significantly reduced production rates per female that were 67-50% of control production. The consistent reproductively linked impacts of fipronil and its degradation products at the population maintenance levels suggest risks to sediment-dwelling crustaceans at concentrations well below noneffects for * Corresponding author telephone: (803)777-0091; fax: (803)7773391; e-mail: [email protected]. † Department of Environmental Health Sciences. ‡ Department of Chemistry. 10.1021/es049654o CCC: $27.50 Published on Web 08/31/2004

 2004 American Chemical Society

most aquatic test species based on risk assessment data from primarily acute and sub-life cycle toxicity tests.

Introduction For ecotoxicology to have the highest informative value to risk assessment, test methods must yield sufficient biologically meaningful information to extrapolate chemical effects on individuals up to the population (and higher) levels of biological organization. Unlike human environmental risk assessment, which is charged with determining an individual’s risk of acquiring disease or harm from environmental hazards, ecological risk assessment operates to protect ecological resources at the species population level. Thus, toxicological information that provides, for example, a population’s risk of extinction or loss to some threshold density is of highest value to management. As articulated in two recent papers in this Journal (1, 2), most traditional ecotoxicological methods are strongly focused on determining acute lethality to individualssor measuring individualbased biochemical to tissue-organ system sub-life cycle effects. These approaches have little predictive strength in extrapolating the population-level significance of subtle chronic sublethal effects manifested at the individuals’ genetic > biochemical > organ systems levels or the ability of a threatened population to reproduce/recruit and maintain a stable population density and age structure under acute or chronic stress. Population (and higher order) risk assessment is best served by species that allow exposure and full life cycle measurements through the reproductive and recruitment time windows in the lab or field. Unfortunately, much of the data pool available to predict higher order contaminant risks to aquatic systems is from only a handful of logistically amenable species such as water fleas, aquatic insects, amphipods, and annelid worms. Thus ecotoxicology through much of its young history has been technologylimited with regard to population-based risk assessment since the preponderance of our traditional test methods are so heavily standardized on species with reproductive life cycles of months to years and exposure concentrations/durations that are typically sub-life cycle. Microinvertebrate-based toxicity testing can ameliorate some logistical shortcomings of long-lived vertebrate and invertebrate species since many microinvertebrates are key in ecological processes and function and most have rapid life cycles. One particular group of mostly short-lived microinvertebrates, the meiobenthos (i.e., sediment-dwelling metazoans and foraminifera 106 m-2, standing biomass of ∼1 g m-2, and carbon production rates of ∼20 g of C m-2 yr-1 (6, 7). This community consists primarily of nematodes, copepods, and foraminifera with >95% residing obligately in oxic sediment horizons (8). Meiobenthic copepods serve as a predominant food source for juvenile fishes, shrimps, and crabs (9, 10) and are often the most acutely sensitive meiobenthic taxon to pollution (3). Many meiobenthic copepods have rapid life cycles (15-25 d; 3) that may allow quick recovery and genetic adaptation to polluted sediments (11, 12). In this study, we use a meiobenthic copepod-based full life cycle bioassay (13) as one example of how microplatebased culturing technologies can be adapted to provide VOL. 38, NO. 23, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6407

FIGURE 1. Example life cycle graph of copepod life stage development times showing, for the control microplates, the proportions of each life stage progressing (or failing to progress) to the next life stage through 24 d (i.e., the time required for controls to produce two hatched broods of nauplii). All proportions are exclusive of mortality (i.e., dead individuals were not scored as developmental failures). empirically based population-level projections of sublethal toxicant impacts on end points of import to population maintenance and risk management (e.g., stage-specific survival, development rates, and reproductive rates). We focus on exact measurement and modeling of life-stage-specific sublethal effects of a unique receptor-specific phenylpyrazole insecticide (fipronil) commonly used near estuarine environments for rice production, turf grass management, termite treatment, and residential insect control (13). To demonstrate the utility of this approach for predicting population and multi-generational effects, empirical life cycle data were modeled for each compound using a life stage classified Leslie matrix approach (14-16). Toxicological effects data even at the population (and higher) levels would be of little value to risk assessment without considering their context within the environmental fate of fipronil. Thus, we tested water- and sediment-borne fipronil parent compound plus two of its predominant degradation products (water only) at a range of concentrations reported from a large coastal watershed in Louisiana.

Experimental Methods Test Organism. Amphiascus tenuiremis cf. Mielke (1974) is an infaunal sediment-dwelling harpacticoid copepod that is well-suited for evaluating sublethal reproductive and developmental toxicity of sediment-associated contaminants due to its moderate acute sensitivity, high chronic sensitivity, ease of culturing in sediments or seawater, 18 d life cycle at 25 °C, and small size (0.4 µm) (3). A. tenuiremis passes through six naupliar molts in 8-9 d and five copepodite molts in 7-8 d at 25 °C. Figure 1 is a life cycle graph of developmental stages, times, and fecundity (through two broods) for this species. Collection of Test Organisms. Gravid A. tenuiremis were captured directly from monoculture sediments in the laboratory (17) and pipetted to a 12-well plate containing seawater and 75 µm mesh cup inserts. The inserts allowed hatching nauplii to fall ∼3 mm to the well bottom over a 24 h period while retaining larger gravids. Captured nauplii were transferred individually to microwells of 96-well ultralow attachment polystyrene microplates (Corning Costar, Corning, NY). 6408

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 23, 2004

Preparation of Nominal Fipronil Treatment Solutions. A 5 mg/mL working stock in pesticide-grade acetone was made for each compound and used to spike nominal treatment solutions. The parent fipronil (FP) stock was made using technical grade FP (5-amino-1-[2,6-dichloro-4-(trifluoromethyl)phenyl]-4-[(trifluoromethyl)sulfinyl]-1H-pyrazole-3-carbonitrile; 98.0% pure) obtained from ChemService, Inc. (West Chester, PA). The desthionyl fipronil (DSF) stock was made using crystals, verified by GC-MS, and purified (99%) from UV-exposed technical-grade fipronil obtained from ChemService, Inc. The fipronil sulfide (FS) stock was diluted from a 100 µg of FS/mL of acetone stock obtained from AccuStandard, Inc. (New Haven, CT; 98% pure). An acetone carrier control (CTL ) 0.10 µL of acetone/L) and the following nominal treatment concentrations were made by adding the appropriate amount of stock solution to 100 mL of pre-aerated, 0.22 µm filtered synthetic seawater (30 ppt, Instant Ocean): 0.25 and 0.50 µg of FP/L, 0.25 and 0.50 µg of DSF/L, and 0.075 and 0.15 µg of FS/L. Test concentrations were selected to be sublethal and within environmentally reported field concentrations (18). All spiked solutions were stirred for 1 h in the dark at 20 °C prior to addition to test microwells. Rearing of Nauplii/Copepodites. Rearing/testing of individual nauplii in microwells followed methods described in Chandler et al. (13). Briefly, 21 (3 per treatment) 96-well low-binding microplates were hydrated for 1 h with deionized water, emptied, allowed to dry, and then re-loaded with one nauplius per testing microwell (n g 40/microplate). Excess transfer seawater was ∼98% removed from each well using a gastight Hamilton microsyringe and reloaded with 200 µL of control or treatment solution. Microplates were placed in an incubator at 25 °C with a photoperiod of 12 h:12 h. Each nauplius was fed 2 µL of a fresh algae mixture (∼107 cells of 1:1:1 Isochrysis galbana:Dunaliella tertiolecta:Phaeodactylum tricornutum) on day 0 of the bioassay and every 6 d thereafter throughout the experiment. Every 3 d, ∼90% of the overlying treatment solution was removed and replaced with 200 µL of fresh treatment solution to ensure proper water quality and a consistent exposure. Isolated nauplii and copepodites were individually monitored within test microwells for up to

TABLE 1. Mean Proportional Life Stage Transitions ((1 SD) and Fecundities Observed over the Complete Life Cycle Exposure of Amphiascus tenuiremis to Fipronil and Its Degradates in 96-Well Microplatesa treatments

Pee

Pen

Pnc

Pnn

Pcf

Pcc

Pfg

Pff

Fen

0.25 FP 0.5 FP 0.25 DSF 0.5 DSF 0.075 FPS 0.15 FPS

0.09 ( 0.11 0.04 ( 0.06 0.09 ( 0.15 0.12 ( 0.16 0.08 ( 0.10 0.06 ( 0.08

0.91 ( 0.11 0.96 ( 0.06 0.91 ( 0.15 0.88 ( 0.16 0.92 ( 0.10 0.94 ( 0.08

0.84 ( 0.04 0.70 ( 0.06 0.85 ( 0.06 0.85 ( 0.06 0.84 ( 0.05 0.77 ( 0.12

0.16 ( 0.04 0.30 ( 0.05 0.15 ( 0.05 0.15 ( 0.05 0.16 ( 0.04 0.23 ( 0.10

0.49 ( 0.05 0.56 ( 0.12 0.48 ( 0.05 0.56 ( 0.09 0.59 ( 0.07 0.43 ( 0.06

0.02 ( 0.03 0.01 ( 0.02 0.05 ( 0.04 0.02 ( 0.01 0.03 ( 0.02 0.04 ( 0.03

0.40 ( 0.14 0.10 ( 0.09 0.71 ( 0.08 0.41 ( 0.15 0.56 ( 0.03 0.42 ( 0.05

0.60 ( 0.14 0.90 ( 0.09 0.29 ( 0.08 0.59 ( 0.15 0.44 ( 0.03 0.58 ( 0.05

13 ( 4 16 ( 2 12 ( 4 14 ( 4 14 ( 4 14 ( 4

a

See life cycle graph (Figure 1) for definitions of life-cycle proportions (Pxx values).

28 d as they developed to the copepodite and adult life stages, respectively. Nauplii and copepodites remained in their original testing microwells throughout development to virgin adult. Survival and developmental progress were recorded daily for each individual nauplius and subsequent copepodite. Dissolved oxygen, salinity, and pH were measured in fresh treatment solutions every 3 d to ensure suitable water quality (19). Individualized Mating of Virgin Adults. On day 18 of the bioassay, >80% of individual surviving copepodites in the CTL, 0.25 µg of FP/L, 0.25 µg of DSF/L, 0.50 µg of DSF/L, and 0.075 µg of FS/L treatments had molted into reproductively mature adults and were collected by sex within treatment for pairwise mating. Due to strong developmental depression, the 0.50 µg of FP/L and 0.075 µg of FS/L treatments matured later than the other treatments and therefore were mated on day 21 of the bioassay. For each treatment, adult virgin males and females were removed from testing microplates and transferred within sex to 50 mL crystallizing dishes containing their respective treatment solutions. A single male or female copepod was removed from crystallizing dishes and placed pairwise individually by treatment into previously unused microwells (n ) 28-41 mating pairs across three replicate microplates/treatment). Mating wells were loaded with 200 µL of fresh CTL or treatment solution, and each mating pair was fed 2 µL of algae mixture as described above. Using an inverted stereomicroscope, each mating well was visually monitored over 12 d at the same time each day for the following end points: survival, abnormal behavior (e.g., not feeding, depressed movement), reproductive success (percent of females producing two viable clutches), frequency of aborted brood sacs, post-mating days to female brood sac extrusion (through three clutches of eggs), days from brood sac extrusion to naupliar hatch (through two broods), first and second brood sizes (number of hatched nauplii/female), and percent hatching success over two clutches (no. of hatched nauplii/no. of hatched nauplii + no. of unhatched eggs). Only end points showing significant differences (p < 0.05) from the CTL are reported here. Stage-Structured Population Growth Modeling. Multigenerational population-level effects of FP, DSF, and FS were estimated using empirical microplate life cycle data fitted to a matriarchal stage-structured Leslie matrix model (RAMAS EcoLab 2.0; Applied Biomathematics, Setauket, NY) (14, 15, 20, 21). A five-stage (embryo to nauplius to copepodite to virgin female to gravid female) matrix model was used to project offspring production through three generations based on (a) stage-specific survival and life stage transition rates, (b) the proportion of copepodites developing into virgin females (thereby capturing sex ratio shifts), (c) the proportion of females producing two viable clutches, and (d) fecundity (i.e., hatched nauplii/female) through two clutches. Figure 1 and Table 1 show the mean proportions of individuals, by treatment, able to develop from nauplius through reproductive maturity in microplates. Since this model is matriarchal (i.e., focused on female production to capture gender shifts

induced by toxicants), final population projections should be considered relative to control mating pair projections rather than interpreted as absolute field abundance projections under FP, DSF, and FS exposures. Empirical life cycle toxicity test data were inserted in the Leslie matrix to derive instantaneous rates of population increase (λ) to predict population growth for each mating exposure in each triplicate microplate. A total of 10 simulations was run for each microplate in each mating exposure, with λ values averaged across microplates. Population growth was modeled through three filial generations beginning with 120 F0 nauplii. As a tool for risk assessment, we were most interested in using microplate data to produce consistent population growth predictions for subsequent comparisons across treatments and controls. Thus, a simple, unconstrained Lefkovitch adaptation (15) of the Leslie matrix model was employed (i.e., life stage structured, simple exponential growth, no density dependence, and no carrying capacities invoked) (22). It should be realized, however, that in estuarine systems competition, predation, and environmental carrying capacities would certainly influence population outcomes for these low trophic-level organisms. FP-Spiked Sediment Bioassay. Surface sediments were collected from pristine Bread and Butter Creek (North Inlet, SC; 33°20′ N, 79°10′ W) and processed per Chandler (23). A FP working stock of 100 µg/mL was made using the same technical-grade FP (above) from ChemService, Inc. Stage I copepodites (n ) 50/test chamber) were collected on 70 µm mesh from flow-through sediment monocultures of A. tenuiremis and reared for 16 d in 10 mL aliquots of 0, 65, 105, 180, and 300 ng/g of FP-spiked sediments under continuous flow (20 °C, 30 s, 12 h:12 h photoperiod; 23-25). Each test chamber (50-mL volume; 4 chambers/trt) received 2 mL of concentrated algae every 3 d (1.4 × 107 cells/mL; Isochrysis galbana, Dunaliella tertiolecta, Phaeodactylum tricornutum; 1:1:1). After 16 d, the contents of each chamber were collected on a 45 µm sieve and preserved/stained with a 5% formalin and 1% Rose Bengal solution. Replicates were counted for stage-specific survivors, sex ratios, clutch sizes, percent gravid females, and total F1 offspring production. Statistical Analyses of Fipronil Effects. Data were arcsine square root or square root normalized and analyzed using a General Linear Model (GLM) mixed procedure with treatments, microplate replicates, and sex (where applicable) as fixed effects (R ) 0.05). Statistical interactions among fixed effects were also tested. The 0.50 µg of FP/L treatment yielded only three reproductive data pointssas a result of strong fipronil reproductive suppressionsand was therefore removed from statistical analysis. The following dependent variables were tested for treatment effects: stage-specific survival, development rates, percent of females becoming gravid, days to brood sac extrusions, days from extrusion to naupliar hatch, brood size through two clutches, and hatching success. Leslie matrix projections of λ were compared across replicate treatment microplates using ANOVA and Tukey’s test of additivity. Sediment-bioassay survival, percent gravid, VOL. 38, NO. 23, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6409

FIGURE 2. Larval nauplius to juvenile copepodite cumulative development curves, by day, for controls and those treatments producing significant delays relative to the control. and offspring production were analyzed using one-way ANOVA and Dunnett’s multiple comparison tests. Statistical significance was set at p < 0.05. Chemical Analysis of Fipronil Analytes. Nominal spike concentrations were measured using a liquid:liquid methyl tert-butyl ether (MTBE) extraction technique. Briefly, fresh samples were collected in triplicate and placed (3 mL) into 20 mL amber vials at test initiation and every 3 d immediately prior to solution renewal. Due to the low test concentrations required in this chronic bioassay, samples were analyzed using a liquid 2:1 MTBE extraction. For each replicate sample, an internal standard (0.25 µL/L of 4-bromoanisole) was added to the sample:MTBE mixture. Samples were vortexed for 1.5 min and sonicated for 5 min at 23 °C. The MTBE layers containing extractable analytes were transferred to 15 mm i.d. amber crimp-top GC vials and analyzed for fipronil analytes using a Hewlett-Packard 5890 series II gas chromatograph equipped with a 63Ni electron capture detector (GC-ECD) (Hewlett-Packard, Palo Alto, CA) operated in splitless mode. The carrier gas was helium with injector and detector temperatures at 230 and 310 °C, respectively. A DB-5 30 m × 0.25 mm i.d., 0.25 µm film thickness column (Agilent Technologies, Palo Alto, CA) was operated at an initial temperature of 100 °C for 1 min with a 10°C/min increase to 270 °C and a 5 min hold. Retention times for FP, DSF, and FPS under these conditions were 14.4, 12.8, and 14.2 min, respectively. GC-measured concentrations of all exposure treatments consistently ranged from 82 to 108% of nominal values; therefore, all concentrations are reported hereafter as the nominal a priori targeted values.

Results Survival Rates. Naupliar to adult survival rates in the control, 0.25 FP, both DSF and the 0.075 FS microplates averaged >80% over the 32 d assay period. The 0.5 FP and 0.15 FS exposures were slightly more lethal yielding 69.2 and 73.6% overall survival rates, respectively. However, survival rates were sufficiently variable across microplates such that only the 0.5 FP exposure yielded significantly depressed survival relative to the CTL (p ) 0.018). Development Rates. Nauplius to copepodite developmental rates (e.g., days to copepodite; Figure 2) were 6410

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 23, 2004

FIGURE 3. Percent of females able to produce one, two, or three broods of viable offspring over 11-14 d of mating. modeled/compared across treatments, using logistic regression, to determine time required for 80% of the microplate populations to reach the copepodite stage. The number of days for 80% of all nauplii to reach the copepodite stage in the CTL was 11.3 d (Figure 2A). Similarily, in the 0.25 and 0.5 DSF and the 0.075 FS exposures, nauplii reached 80% copepodites on days 11.3, 11.0, and 11.0, respectively (Figure 2A). In contrast, nauplii reared in 0.15 FS required 19 d to reach the copepodite stage (Figure 2D; p < 0.02). At 0.25 and 0.5 FP, nauplii required 14.3 and 16.5 d, respectively, to become 80% copepodites (Figure 2B,C; p < 0.02). For the copepodite to adult development window, delays of only 0.1-1.3 d were seen over all treatments relative to the CTL, and none were significant. Reproductive Effects. The number of adult mating pairs able to produce viable offspring through g12 d of mating was greatly reduced in the 0.5 DSF, the 0.15 FS, and both FP treatments (all p e 0.02; Figure 3). Reproductive success was defined as percent females able to produce at least two viable clutches. The 0.5 DSF and 0.15 FS treatments showed 52 and 54% decreases in percent success, respectively, while both FP treatments showed decreases of 56 and 88% as compared to controls (CTL > 85% success; Figure 3). For those females able to reproduce, the time required to grow and extrude their first and second brood sacs was also delayed significantly by 3.0 and 2.3 d, respectively, for the 0.5 DSF treatment relative to the CTL (Figure 4). Additionally, both FP treatments produced delays of 2.5 and 2 d to first and second brood sac extrusion; however, data were variable and n-sizes were low

FIGURE 4. Time to egg-clutch extrusion for the first, second, and third naupliar broods by treatment. due to poor reproductive success under FP exposure (Figure 3). Stage-Based Leslie Matrix Population Growth Models. The microplate culturing approach allows full life cycle tracking of individuals’ survival, development, and reproductive output; thus, it is well-suited for robust prediction of population-level responses through time via a stage-based adaptation of the Leslie matrix (15, 16). Model-predicted instantaneous rates of natural increase (λ) for CTL and FPexposed copepods ranged from 1.39 for CTL down to 1.18 for 0.5 FP under optimal culture conditions of abundant food, no disease, and no competition or predation. However, if they could be measured, field rates would almost certainly be lower. λ was never observed to fall below 1 (i.e., the population replacement level) because the low test concentrations yielded correspondingly low lethality. FS at 0.075 µg/L produced a slightly higher λ of 1.43. When λ values were computed at the appropriate statistical level of individual microplates within treatments, then all treatments except for 0.075 FS and 0.25 DSF had empirical λ values significantly lower than controls. When the matrix model (and λ) was used to project three generations of production from an F0 starting population of 120 first-stage nauplii (Figure 5A), first generation abundances were strongly depressed only in the 0.15 FS and both FP treatments. By the third generation, however, all treatments except for the 0.075 FS showed strong depressions relative to the control. FP parent at these test concentrations had the strongest multi-generational impact yielding F3 production depressions of 3.6-5-fold relative to controls. To determine which life stage elements had the largest impact on predicted population growth under the influence of the six fipronil treatments, the matrix model was subjected to a simple “sensitivity” analysis. Various end point values (e.g., stage-specific survival rates, fecundity, etc.) from the treatment Leslie matrixes were substituted one at a time into the fipronil-free control matrix and computed again over three generations to determine relative influence of each on offspring production. The only empirically measured life stage element (value) able to induce “treatment” responses from the control matrix (Figure 5B) similar to original treatment matrix responses (Figure 5A) was “proportion of females successfully reproducing”. Overall, the third generation production projections from these sensitivity substitutions (5B) were very similar to nonmanipulated treatment matrix results with the exception of stronger projected depressions in the 0.5 FP and 0.075 FS treatments. Stronger depressions for these two treatments in the control matrix sensitivity analysis were caused by the lack of an empirical “mitigating” effect of slightly more abundant females observed in the original 0.5 FP (44:56% male:female) and 0.075 FS (41:59%

FIGURE 5. (A) Three-generation Leslie (Lefkovich) matrix projections, by treatment, of naupliar production by an initial F0 cohort of 120 nauplii. (B) Sensitivity of the control response matrix to substitution of treatment reproductive failure rates (i.e., proportions of fipronilor degradate-reared copepodites successfully mating and producing two broods). male:female) matrixes. This consistency of within treatment responses between Figure 5, panels A and B, is a straightforward demonstration of the population-level influence of fipronil’s (and its degradates’) strong negative impact on reproductive success (Figures 3 and 4), excepting the lowest DSF and FS treatment concentrations. Sediment Bioassay. Survival rates of copepodites to the adult stage over treatments and CTL ranged from 80 to 92% and were not statistically different. Sex ratios were similarly unaffected by FP. The total number of offspring recovered after 16 d of sediment culture, however, was significantly depressed (>33%) in all FP treatments relative to the CTL (p ) 0.004; Figure 6A). When offspring number was normalized to the number of surviving females per treatment (i.e., yielding production per female), net production was significantly and sharply depressed in all FP treatments (p < 0.002; Figure 6B). On average, the 300 ng of FP/g of sediment treatment yielded 200% of the control at only 0.15 FPS), was more chronically toxic to nauplii at lower concentrations than FP or DSF, and yielded an almost equal reproductive failure rate (61% failed mating pairs at 0.15 FS) as parent FP at 0.25 µg/L. Clearly, given FS’s strong depressive effect on development and reproductive success, it should be judged as hazardous to copepod growth and reproduction as the FP parent; plus the sulfide is much more persistent in sediments (33) where these animals normally live. Similar to trends seen in previous copepod studies with parent FP alone (13, 34), FP significantly delayed naupliar growth and molting to the first sub-adult life stage (the copepodite) in dose-response fashion. This delay in growth was much less pronounced in the copepodite to adult development window (only up to 1.3 d longer than the normal 7-8 d). For this crustacean, which has a median lifespan of ∼47 d (35), a 5-7 d total delay in growth to reproductive maturity would translate into 2-3 fewer clutches of eggs per lifetime (and significant negative impact on population growth) even if lifespan and clutch extrusion rates were not affected by fipronil at these concentrations. Delays in extrusion rates were common however, which would further diminish population growth potential with time (Figure 4). As this copepod is intertidal in habitat, such delays could present problems with timing of reproduction to meet spring or neap tides or to avoid the presence of migrating meiobenthic predators such as salmonid and sciaenid fishes (e.g., refs 36 and 37). In previous studies that focused only on the copepodite to adult development windows, FP at 0.220.42 µg/L induced stronger male-specific reproductive failure rates (e.g., 71% vs 56% of CTL failure here) (13, 34). The additional exposure time in the naupliar to copepodite window seems to confer moderate resistance to FP reproductive damage in those copepods that survive. However, mortality rates were significantly higher in fipronil treatments of this nauplius to adult assay (