Assessing the relative importance of temperature, discharge, and day length on the reproduction of an anadromous fish (Alosa alosa)

Abstract

Abstract Climate change threatens anadromous fishes such as the allis shad (Alosa alosa) populations of which have declined since the 20th century in Europe. Sensitivity to climate change could be quantified by determining the fish’s spawning behaviour, defined as the timing of reproduction (i.e. spawning events) as a function of temporally variable environmental factors. The cues that fish use to time reproduction could determine their response to climate change. A machine learning technique (boosted regression tree) was calibrated using a 14‐year dataset composed of daily measures of environmental factors and fish occurrences during reproduction. The boosted regression tree provides complete insight into the complex relationship between the spawning probability, i.e. the probability for a fish to reproduce, and environmental factors that might evolve with climate change. The spawning probability was positively related to day length (44.6%) and water temperature (34.7%) and negatively related to river discharge (20.7%). Optimal reproductive conditions corresponded to a difference in day length between 0 and 0.04 hr, a water temperature between 15 and 26°C and a river discharge between 55 and 665 m3/s; these conditions are currently being utilised by allis shad populations in the Garonne and Dordogne rivers, France. This study highlights the relative influence of environmental factors on the observed spawning period as well as the evolution of habitat suitability during the 14‐year period. The novelty of this study stems from assessing population process data, i.e. the occurrence of fish reproduction, rather than mere occurrence data in an ecological niche model study. Climate change may lead to a shift in spawning phenology, as the water temperature and river discharge will also change. Therefore, conservation plans need to integrate these effects on spawning grounds.