Populations of desert pupfish (Cyprinodon macularius; pupfish), a federally-listed endangered species, inhabit irrigation drains in the Imperial Valley agricultural area of southern California. These drains have varying degrees of selenium (Se) contamination of water, sediment, and aquatic biota. Published Se toxicity studies suggest that these levels of Se contamination may pose risk of chronic toxicity to Se-sensitive fish, but until recently there have been no studies of the chronic toxicity of Se to desert pupfish.

A life-cycle Se exposure with pupfish was conducted to estimate dietary and tissue thresholds for toxic effects of Se on all life stages. The dietary exposure was based on live oligochaete worms (Lumbriculus variegatus) dosed with Se by a laboratory food chain based on selenized yeast. Oligochaetes readily accumulated Se from mixtures of selenized and control yeasts. The protocol for dosing oligochaetes for pupfish feeding studies included long-term (at least 28 days) feeding of a low-ration of yeast mixtures to large batches of oligochaetes. Oligochaetes were dosed at five Se levels in a 50-percent dilution series. Pupfish were simultaneously fed Se-dosed oligochaetes and exposed to a series of Se concentrations in water (consisting of 85 percent selenate and 15 percent selenite) to produce exposures that were consistent with Se concentrations and speciation in pupfish habitats. The nutritional characteristics of oligochaete diets were consistent across the range of oligochaete Se concentrations tested.

The life-cycle exposure started with laboratory-cultured juvenile pupfish that were exposed to Se through sexual maturation and reproduction (150 days; F0 exposure). The Se exposure continued with eggs, larvae, and juveniles produced by Se-exposed parents (79 days; F1 exposure). Selenium exposure (water and diets), Se bioaccumulation (whole-body and eggs), and toxicity endpoints (juvenile and adult survival and growth; egg production and hatching success, larval survival and deformities) were documented throughout the life-cycle study.

Selenium concentrations in water (as much as 52 micrograms per liter [μg/L]) and diets (as much as 53 micrograms per gram [μg/g], on a dry weight basis) bracketed concentrations reported in pupfish habitats. Juvenile F0 pupfish rapidly accumulated Se and bioaccumulation models indicated that pupfish had reached more than 97 percent of maximum whole-body Se concentrations by the time they reached reproductive maturity. Adult pupfish accumulated whole-body Se concentrations that averaged about 40 percent of those in the oligochaete diets. Selenium concentrations in eggs and F1 juveniles were similar to or slightly greater than Se concentrations in F0 adults. Juvenile F0 pupfish contained selenomethionine fractions (62–71 percent of whole-body Se) greater than the average reported for wild pupfish from the Imperial Valley (53 percent).

Selenium exposure had minimal effects on survival or growth of juvenile and adult pupfish. There was evidence of toxic effects on pupfish in the highest Se treatment (Se–5), including reduced growth of F0 and F1 juvenile pupfish (17–21 percent less than controls) on some sampling dates. These growth reductions did not persist to subsequent sampling dates, but reduced growth of F1 pupfish in the Se–5 treatment was associated with reduced survival (12 percent less than controls).

Egg production was greatest in the controls and decreased with increasing Se exposure, reaching a minimum (51 percent less than controls) in the Se–4 treatment, but egg production was reduced by only 24 percent in the Se–5 treatment, a lesser reduction than in other Se treatments except Se–1. There was no statistically significant overall effect of Se treatment on mean pupfish egg production, reflecting large variation among replicates and among sampling dates. However, comparisons of daily mean egg production for 23 sampling dates indicated that egg production in each of 5 Se treatments was significantly less than controls on multiple (3–7) sampling dates, but no mean for any Se treatment was significantly greater than controls on any date. Significant reductions in daily egg production occurred mainly during the middle of the study and egg production increased in several Se treatments during the final 2 weeks of the study. These results suggest that pupfish egg production, although a highly variable endpoint, was adversely affected by elevated Se exposure.

Neither egg hatching success nor survival of F1 larvae indicated clear evidence of Se toxicity. Egg hatching success did not differ significantly among treatments, with means ranging from 84–91 percent. The frequency of morphological deformities (primarily spinal deformities) was greater in larvae 10 days post-fertilization (dpf) from a preliminary reproduction study than in older larvae (14 dpf) from the main reproduction study. The frequency of larval deformities was generally greater in Se treatments than controls, but mean frequencies did not differ significantly among treatments. Survival of F1 larvae to 21 dpf was not reduced significantly by parental Se exposure, but the Se–5 treatment had the lowest larval survival (84 percent), and lowest combined egg hatching and larval survival (76 percent).

Results of the Se treatments indicate that pupfish were insensitive to Se toxicity through most of their life cycle. Consistent toxic effects on survival and growth of juvenile and adult pupfish (defined as at least 10 percent reduction compared to controls) occurred only in treatment Se–5, which had a mean dietary Se concentration of 52 μg/g and a mean pupfish whole-body Se concentration of 27 μg/g. These apparent toxicity thresholds for growth and survival rank among the least sensitive chronic Se toxicity values reported for nonreproductive endpoints in freshwater fish. Comparisons of these thresholds to surveys of Se concentrations in the Imperial Valley suggest that risks of Se toxicity are low in pupfish habitats. The dietary threshold was about twice as high as the greatest mean Se concentrations reported in midge larvae from seven sites in the Imperial Valley. Whole-body thresholds were greater than mean whole-body Se concentrations reported for field-collected pupfish from three sites and for the sailfin molly (Poecilia latipinna), a potential bioaccumulation surrogate for pupfish, from seven sites.

Reduced egg production, although highly variable, was the most sensitive response of pupfish to Se exposure. Toxic effects on egg production (reductions of 24–51 percent relative to controls) occurred in the four highest Se treatments, corresponding to reproductive toxicity thresholds of 7.3 μg/g for Se in diet, 3.4 μg/g in pupfish (whole body), and 4.4 μg/g in pupfish eggs. These thresholds are substantially lower than published Se toxicity values for reproductive effects in other freshwater fish (for example, 17–24 μg/g in eggs). Reduced egg production has not been reported as a sensitive endpoint in Se toxicity studies, although abnormal ovarian development has been reported in Se-exposed fish, and reduced egg production has been reported as a sensitive response of other Cyprinodon pupfish to other environmental stressors.

Selenium concentrations in tissues of pupfish, mollies, and diet items from Imperial Valley sites frequently exceeded concentrations associated with reduced pupfish egg production in the laboratory study. Reduced egg production may limit the ability of pupfish populations to persist and recover in Se-contaminated habitats in the Imperial Valley and elsewhere in their limited range. However, these apparent risks of Se toxicity are not supported by recent surveys of desert pupfish populations in the Imperial Valley. These surveys indicated that desert pupfish made up a small, but variable, component of fish communities in Imperial Valley habitats, including sites with increased levels of Se exposure, and that pupfish distribution and population density indicated no clear relationships with Se concentrations in diets or fish tissues. Additional studies could determine the role of egg production in the maintenance and recovery of desert pupfish populations in Se-contaminated habitats.