Multi‐species occupancy models as robust estimators of community richness

Abstract

1. Understanding patterns of diversity is central to ecology and conservation, yet estimates of diversity are often biased by imperfect detection. In recent years, multi-species occupancy models (MSOM) have been developed as a statistical tool to account for species-specific heterogeneity in detection while estimating true measures of diversity. Although the power of these models has been tested in various ways, their ability to estimate gamma diversity-or true community size, N is a largely unrecognized feature that needs rigorous evaluation.2. We use both simulations and an empirical dataset to evaluate the bias, precision, accuracy and coverage of estimates of N from MSOM compared to the widely applied iChao2 non-parametric estimator. We simulated 5,600 datasets across seven scenarios of varying average occupancy and detectability covariates, as well as varying numbers of sites, replicates and true community size. Additionally, we use a real dataset of surveys over 9 years (where species accumulation reached an asymptote, indicating true N), to estimate N from each annual survey.3. Simulations showed that both MSOM and iChao2 estimators are generally accurate (i.e. unbiased and precise) except under unideal scenarios where mean species occupancy is low. In such scenarios, MSOM frequently overestimated N. Across all scenarios, MSOM estimates were less certain than iChao2, but this led to over-confident iChao2 estimates that showed poor coverage. Results from the real dataset largely confirmed the simulation findings, with MSOM estimates showing greater accuracy and coverage than iChao2. Community ecologists have a wide choice of analytical methods, and both iChao2and MSOM estimates of N are substantially preferable to raw species counts. The simplicity of non-parametric estimators has obvious advantages, but our results show that in many cases, MSOM may provide superior estimates that also account more accurately for uncertainty. Both methods can show strong bias when average occupancy is very low, and practitioners should show caution when using estimates derived from either method under such conditions. K E Y W O R D SChao estimator, community richness, data augmentation, gamma diversity, imperfect detection, multi-species occupancy models, occupancy modelling S U PP O RTI N G I N FO R M ATI O NAdditional supporting information may be found online in the Supporting Information section.

Estimates of Species Richness and Composition Depend on Detection Method in Assemblages of Terrestrial Mammals

“…Given the considerable attention paid to the mammals of the study region, we are confident that actual richness is known (13 species), in agreement with atlas cumulative data [ 37 ]. Species richness calculated with the non-parametric index iChao2 clearly overestimated true richness in contrast with recent analyses indicating that iChao2 tends to underestimate richness both in virtual and real communities with a larger number of species (20–125 species; [ 59 ]). Although Tingley et al [ 59 ] concluded that iChao2 was an accurate estimator, they also showed that accuracy notably decreases when mean species probability of occupancy is low (0.1) and highly variable (SD = 1.0), regardless of variation in the average species probability of detection.…”

“…Species richness calculated with the non-parametric index iChao2 clearly overestimated true richness in contrast with recent analyses indicating that iChao2 tends to underestimate richness both in virtual and real communities with a larger number of species (20–125 species; [ 59 ]). Although Tingley et al [ 59 ] concluded that iChao2 was an accurate estimator, they also showed that accuracy notably decreases when mean species probability of occupancy is low (0.1) and highly variable (SD = 1.0), regardless of variation in the average species probability of detection. However, we found considerable overestimation of species richness even with high probabilities of occupancy (mean ± SD = 0.77 ± 0.37; n = 13 species) which are closer to scenarios where iChao2 performed better [ 59 ].…”

“…Observed species richness is usually thought to underestimate species richness, as increasing sampling effort increases the number of species observed until reaching an asymptote [ 60 ]. Our results show that, although estimating richness has been recommended in most situations [ 59 , 61 ], in some cases the observed species richness may approach true richness more closely than estimated richness, even using a moderate number of temporal replicates (four sessions in our study).…”

“…Although Tingley et al [ 59 ] concluded that iChao2 was an accurate estimator, they also showed that accuracy notably decreases when mean species probability of occupancy is low (0.1) and highly variable (SD = 1.0), regardless of variation in the average species probability of detection. However, we found considerable overestimation of species richness even with high probabilities of occupancy (mean ± SD = 0.77 ± 0.37; n = 13 species) which are closer to scenarios where iChao2 performed better [ 59 ]. Observed species richness is usually thought to underestimate species richness, as increasing sampling effort increases the number of species observed until reaching an asymptote [ 60 ].…”

Effects of seasonality and land use on the diversity, relative abundance, and distribution of mosquitoes on St. Kitts, West Indies

“…While we were able to detect the majority of species predicted to be present on St. Kitts, our actual ability to detect a given mosquito species was low on average (9% average detection rate). That being said, multi-species occupancy models are robust to low detection probabilities as long as mean site-level occupancy is relatively high, which it was in this study (30%) [ 44 ]. Using the site-level species richness estimated from our multi-species occupancy model, we noted some potential correlation between percentage of local land cover and mosquito species richness.…”