Using post-settlement demography to estimate larval survivorship: a coral reef fish example.

Many species have multi-stage life cycles in which the youngest stages (e.g., larvae) are small, dispersive, and abundant, whereas later stages are sessile or sedentary. Quantifying survival throughout such early stages is critical for understanding dispersal, population dynamics, and life history evolution. However, dispersive stages can be very difficult to sample in situ, and estimates of survival through the entire duration of these stages are typically poor. Here we describe how demographic information from juveniles and adults can be used to estimate survival throughout a dispersive larval stage that was not sampled directly. Using field measurements of demography, we show that detailed information on post-settlement growth, survival, and reproduction can be used to estimate average larval survivorship under the assumption that a typical individual replaces itself over its lifetime. Applying this approach to a common coral reef fish (bicolor damselfish, Stegastes partitus), we estimated average larval survivorship to be 0.108% (95% CI 0.025-0.484). We next compared this demography-based estimate to an expected value derived from published estimates of larval mortality rates. Our estimate of larval survivorship for bicolor damselfish was approximately two orders of magnitude greater than what would be expected if larval mortality of this species followed the average, size-dependent pattern of mortality inferred from a published sample of marine fishes. Our results highlight the importance of understanding mortality during the earliest phases of larval life, which are typically not sampled, as well as the need to understand the details of how larval mortality scales with body size.