Untested assumptions about within-species sample size and missing data in interspecific studies

Abstract

Phylogenetic comparative studies rely on speciesspecific data that often contain missing values and/or differ in sample size among species. These phenomena may violate statistical assumptions about the non-random variance component in sampling effort. A major reason why this assumption is often not fulfilled is because the probability of being sampled (i.e., being captured or observed) may depend on species-specific characteristics. Here, we test this assumption by using information on within-species sample sizes and missing data from five independent comparative datasets of European birds. First, we show that the two estimates of data availability (missing values and withinspecies sample size) are positively correlated and are associated with research effort in general (the number of papers published). Second, we demonstrate biologically meaningful relationships between data availability and phenotypic traits. For example, population size, risk-taking, and habitat specialization independently predicted within-species sample size. The key determinants of missing data were population size and distribution range. However, data availability was not structured by phylogenetic relationships. These results indicate that the accuracy of sampling is repeatable and distributed non-randomly among species, as several species-specific attributes determined the probability of observation. Therefore, data availability seems to be a speciesspecific trait that can be shaped by ecology, life history, and behavior. Such relationships raise issues about non-random sampling, which requires attention in comparative studies.

Range geometry and socio‐economics dominate species‐level biases in occurrence information

“…These results support the notion that while large ranges are bound to overlap with more sampling locations (compare Garamszegi & Møller, 2012), large, irregularly shaped ranges are severely constrained in the detail with which a given number of records could cover them. Range size was the single most important predictor, with a strong positive effect on record count and a strong negative effect on range coverage.…”

Measuring avian specialization

“…An index of specialization is only as reliable as the underlying data. The quality of information about traits varies from species to species, might be incomplete or inaccurate in some cases, depending on the quality and number of studies conducted on each species (Ducatez & Lefebvre, ; Garamszegi & Møller, ; McKenzie & Robertson, ). Furthermore, the type of variable used to fill out the trait‐features can also influence the index.…”

Escape from predators and genetic variance in birds

“…Most statistical analyses assume that single observations provide equally precise information about the deterministic part of total process variation; that is, the standard deviation of the error term is constant over all values of the predictor variable (Sokal & Rohlf, ). Garamszegi & Møller (, , ) have shown that bias due to variation in sample size among observations is a major problem in comparative analyses and equally significant as the biases caused by treating species‐specific estimates as statistically independent. If this assumption of similarity in sampling effort is violated, weighting each observation by sampling effort allows the use of all data, giving each datum a weight that reflects its degree of precision due to sampling effort (Draper & Smith, ; Sokal & Rohlf, ).…”

“…If this assumption of similarity in sampling effort is violated, weighting each observation by sampling effort allows the use of all data, giving each datum a weight that reflects its degree of precision due to sampling effort (Draper & Smith, ; Sokal & Rohlf, ). This procedure also allows both rare and common species to be included and hence avoids any bias in sampling due to rarity (Garamszegi & Møller, ). Therefore, we weighted statistical models by sample size.…”