The risks of learning: confounding detection anddemographic trend when using count-based indices forpopulation monitoring

Abstract

Theory recognizes that a treatment of the detection process is required to avoid producing biased estimates of population rate of change. Still, one of three monitoring programmes on animal or plant populations is focused on simply counting individuals or other fixed visible structures, such as natal dens, nests, tree cavities. This type of monitoring design poses concerns about the possibility to respect the assumption of constant detection, as the information acquired in a given year about the spatial distribution of reproductive sites can provide a higher chance to detect the species in subsequent years. We developed an individual- based simulation model, which evaluates how the accumulation of knowledge about the spatial distribution of a population process can affect the accuracy of population growth rate estimates, when using simple count-based indices. Then, we assessed the relative importance of each parameter in affecting monitoring performance. We also present the case of wolverines (Gulo gulo) in southern Scandinavia as an example of a monitoring system with an intrinsic tendency to accumulate knowledge and increase detectability. When the occupation of a nest or den is temporally autocorrelated, the monitoring system is prone to increase its knowledge with time. This happens also when there is no intensification in monitoring effort and no change in the monitoring conditions. Such accumulated knowledge is likely to increase detection probability with time and can produce severe bias in the estimation of the rate and direction of population change over time. We recommend that a systematic sampling of the population process under study and an explicit treatment of the underlying detection process should be implemented whenever economic and logistical constraints permit, as failure to include detection probability in the estimation of population growth rate can lead to serious bias and severe consequences for management and conservation. Autocorrelation, citizen science, demographic monitoring, den, detection probability, learning period, nest, population size, population trend, reproduction.