An integrated state-space model to assess Italian alpine galliforms status from count and bag data

Abstract

The assessment of wildlife population sizes and their trends is one of\nthe most important research fields in conservation biology, as it is\nused to identify vulnerability soon enough to implement measures in\nthreatened species, or to set up sustainable harvesting rates in\nexploited populations. Yet, because field work is expensive, may be\ndifficult in terms of logistics and because some populations of the same\nspecies may be monitored by different stakeholders, population status\noften rely on fragmented and heterogenous information on sub-populations\ncollected through various monitoring programs. In this context, data\nintegration, i.e. the simultaneous analysis of different datasets in a\nsingle modelling framework allows to get unbiased and more precise trend\nestimates than separated analysis that in turn may lead to more adequate\nmanagement policies. In this study we developed an integrated\nstate-space model to jointly model populations growth rates from\nindividual counts and hunting bags data for three hunted species of\nmountain Galliformes in Italy. We examined population trends at various\nspatial scales and disentangled the potential effect of game management\nplans from biological factors. The integration of counts and bags\nsucceeded in improving growth rate parameter precision and in reducing\nproxy-specific bias by increasing the sample size and extending data\nseries length. On a 19-year basis, all three species exhibited negative\nmean growth rates. We did not find strong regional patterns for Rock\nptarmigan and Rock partridge, as a likely consequence of prevailing\neffects of local environmental conditions on population growth rate.\nBlack grouse eastern populations exhibited lower growth rate than\nwestern populations. Our paper demonstrates that an integrated model of\ndifferent index of population size of game species can provide more\naccurate values than separate analysis, we advocate to consider such an\napproach for other wildlife monitoring cases for which data is scarce.