Data from active and passive sampling methods provide useful information on spatially dynamic populations and assemblages, such as recolonization patterns following disturbance and invasive spread. However, combining information from different sampling approaches can introduce errors when aggregating data. The goal of this study was to reconcile the information content of these two types of sampling strategies by assessing the population sampled by each and the precision of their measurements of fish distribution and abundance.
We simultaneously collected fish with two types of passive samplers and one active-sampler at 20-35 fixed plots encompassing different regions of the Florida Everglades over a six-month period. Passive sampling techniques consisted of 24-hour sets of unbaited traps, either with or without drift-fences to direct fish movement. The active method consisted of a 1-m2 throw-trap, an enclosure-type sampler. Occupancy and detection probabilities were modelled for nonnative species and for locations differing in recovery time from seasonal drying. Information from both types of sampling method were integrated into an encounter model to estimate activity patterns relative to local density and the relationship of these patterns to discrepancies between standardized abundance estimates.
Results/Conclusions
Passive sampling methods are essentially activity samplers that continuously sample over a longer temporal window than enclosure samplers, and these activity samplers provided the largest occupancy estimates for fishes found at low densities. Contexts where species occurred at low densities included habitats newly colonized post-disturbance or by invasive spread of a non-native species. Passive samplers are potentially vulnerable to over-estimation of local abundance if a species is highly active or attracted to traps, because these behavioral factors can inflate encounter rates relative to less active species with similar abundances. Activity levels estimated from the predator-encounter model increased with degree of disagreement between standardized estimates of abundance from passive and active sampling methods.
Choice of sampling method affected spatial estimates of nonnative fish prevalence and abundance as well as native fishes used as ecological indicators for Everglades restoration. Our enhanced understanding of how information derived from active and passive sampling methodologies relate to one another can be used to improve impact assessment, management of non-native fishes, and generate a more accurate probabilistic framework for predicting patterns of colonization by native and non-native fishes.