Strong evidence for changing fish reproductive phenology under climate warming on the Tibetan Plateau.

Phenological responses to climate change have been widely observed and have profound and lasting effects on ecosystems and biodiversity. However, compared to terrestrial ecosystems, the long-term effects of climate change on species' phenology are poorly understood in aquatic ecosystems. Understanding the long-term changes in fish reproductive phenology is essential for predicting population dynamics and for informing management strategies, but is currently hampered by the requirement for intensive field observations and larval identification. In this study, a very low-frequency sampling of juveniles and adults combined with otolith measurements (long axis length of the first annulus; LAFA) of an endemic Tibetan Plateau fish (Gymnocypris selincuoensis) was used to examine changes in reproductive phenology associated with climate changes from the 1970s to 2000s. Assigning individual fish to their appropriate calendar year class was assisted by dendrochronological methods (crossdating). The results demonstrated that LAFA was significantly and positively associated with temperature and growing season length. To separate the effects of temperature and the growing season length on LAFA growth, measurements of larval otoliths from different sites were conducted and revealed that daily increment additions were the main contributor (46.3%), while temperature contributed less (12.0%). Using constructed water-air temperature relationships and historical air temperature records, we found that the reproductive phenology of G. selincuoensis was strongly advanced in the spring during the 1970s and 1990s, while the increased growing season length in the 2000s was mainly due to a delayed onset of winter. The reproductive phenology of G. selincuoensis advanced 2.9 days per decade on average from the 1970s to 2000s, and may have effects on recruitment success and population dynamics of this species and other biota in the ecosystem via the food web. The methods used in this study are applicable for studying reproductive phenological changes across a wide range of species and ecosystems.