A comparison of abundance estimators for small mammal populations

Abstract

ABSTRACT. A major difficulty in the application of probabilistic models to estimations of mammal abundance is obtaining a data set that meets all of the assumptions of the model. In this paper, we evaluated the concordance correlation among three population size estimators, the minimum number alive (MNA), jackknife and the model suggested by the selection algorithm in CAPTURE (the best-fit model), using long-term data on three Brazilian small mammal species obtained from three different studies. The concordance correlation coefficients between the abundance estimates indicated that the probabilistic and enumeration estimators were highly correlated, giving concordant population estimates, except for one species in one of the studies. The results indicate the adequacy of using enumeration estimates as indexes for population size when scarce data do not allow for the use of probabilistic methods. Differences observed in the behavior of enumeration and probabilistic methods among species and studies can be related to the exclusive sampling design of each area, species-specific movement characteristics and whether a significant portion of the population could be sampled.

Optimising sampling methods for small mammal communities in Neotropical rainforests

“…, Pacheco et al. ). Nonetheless, grid‐based studies may have some advantages, including providing the ability to estimate the movements easily (e.g.…”

“…Combining these two trap configurations may be the best approach when the goal is to maximise rodent captures. Indeed, grids and line transects are both widely used to quantify the abundance of small mammals (D'Andrea et al 1999, Pardini et al 2005, Pacheco et al 2013. Nonetheless, grid-based studies may have some advantages, including providing the ability to estimate the movements easily (e.g.…”

Climate‐driven range shifts are stochastic processes at a local level: two flying squirrel species in Maine

“…While this index underestimates the true population size, enumeration indices of small mammal populations are comparable to probabilistic estimates (McKelvey and Pearson , Pacheco et al. ). The study design allowed for high replication in an area with natural geographic barriers, both of which enhance the accuracy of enumeration indices (McKelvey and Pearson , Pacheco et al.…”

A 3-YEAR POPULATION STUDY OF Guerlinguetus brasiliensis ingrami (RODENTIA, SCIURIDAE) AT THE SERRA DA BOCAINA NATIONAL PARK, RIO DE JANEIRO STATE, BRAZIL

“…Guerlinguetus b. ingrami had a low capture probability, and the result was a different pattern between POPAN model and MNKA ( Figure 2). This occurred because capture probability and the similarity between MNKA and probabilistic population sizes are positively correlated as demonstrated by Pacheco et al (2013). Also, population sizes of G. b. ingrami showed a slight tendency to decrease throughout the study, except for the last month ( Figure 2).…”

“…MNKA is an enumeration technique that sums the number of individuals found in a trap session, as well as the number of individuals found before and after that session (Krebs 1966). Although MNKA may be more biased than the probabilistic models as it may underestimate population sizes, MNKA is still used in small mammal studies, especially those with relatively small sample sizes (< 50 individuals; Pacheco et al 2013). Therefore, we used MNKA as a reference, to assess whether JS estimates are indeed above this method, and to provide estimates that can be used for comparison by future studies.…”