Individuals matter more than replicates: distribution of sampling effort in isotopic niche estimation

The number of individuals to be sampled is a key element in the sampling design of any study as it directly affects the estimations and inferences made. Additionally, in cases where several replicates per individual can be taken, it is important to define how the sampling effort will be distributed between the intraindividual and interindividual components (within and between individuals, respectively). Determining how samples should be distributed among these components can help optimize the available resources and reduce bias in the estimations. To study population trophic diversity, the total niche width (TNW) is usually estimated, which is an approximation of resource diversity at the population level. TNW is the sum of the resource diversity consumed at the intraindividual (replicates) and interindividual (individuals) level. In this study, the effect of prioritizing the number of individuals or the number of replicates on the accuracy and precision of TNW estimations was tested. Multiple isotopic (δ13C and δ15N) values per individual in populations with different degree of individual specialization were simulated. Then, isotopic data from natural populations within the same species (available published studies) were used to assess the results obtained with simulated data. It was found that TNW estimations were more accurate and precise when prioritizing the number of individuals over the replicates, along the entire individual gradient of trophic specialization. Therefore, it is advisable to prioritizing the number of individuals. This methodological contribution should be considered in future studies that use repeated measures of isotopic data to estimate TNW.