Effects of temperature on hatching and growth performance of embryos and yolk-sac larvae of a threatened estuarine fish: Longfin smelt (Spirinchus thaleichthys)

Abstract The abundance of longfin smelt (Spirinchus thaleichthys) in the San Francisco Estuary (SFE) has declined to less than 1% of historic numbers, contributing to their listing as “threatened” under the California Endangered Species Act. To prevent local extirpation, a captive culture program is being developed for the longfin smelt at the Fish Conservation Culture Laboratory at UC Davis. To begin systematically addressing current issues with larviculture methods, we tested the effects of temperature on longfin smelt embryos from 7 clutches with different parentages. We reared embryos and newly hatched non-feeding yolk-sac larvae in freshwater at 9, 12, or 15 °C until the age when these fish underwent mass mortality (>90% mortality) during the experiment (3, 4, or 5 days post-hatch (dph), depending on rearing temperature). We measured hatch success, hatch morphometrics (notochord length, dry mass, yolk volume, and oil volume), length and dry mass growth rates, yolk and oil resorption rates, routine oxygen consumption rate, and locomotor activity (percent of time spent moving, total distance travelled, and average swimming velocity). Embryos and larvae reared at 15 °C performed poorly, experiencing reduced hatch success, hatch notochord lengths, growth rates, and earlier mass mortality. Embryos reared at 9 and 12 °C had similar hatch success but at hatch, larvae were longer at 9 compared to 12 °C. However, larvae grew faster in 12 compared to 9 °C resulting in both temperature groups having similar lengths prior to mass mortality (4 dph for 12 °C and 5 dph for 9 °C). Temperature had no effect on routine oxygen consumption rate among temperature groups but locomotor activity was elevated at 9 and 15 °C groups compared to 12 °C, suggesting that locomotor activity consumed a higher proportion of the energy budget, leaving a smaller fraction for growth. Finally, larval swimming speeds, one day prior to each temperature's respective day of mass mortality, were higher at 12 °C (3 dph) compared to 9 (4 dph) and 15 °C (2 dph), which could lead to higher foraging success. Interclutch variation was high, however, suggesting that parentage effects may result in brood-specific variation in the responses of larvae to water temperature. Taken together, these data suggest that rearing longfin smelt at 15 °C would have detrimental effects and we suggest that cooler temperatures such as 9 or 12 °C may improve the culturing of the earliest life stages of longfin smelt.