Estimation of deer population dynamics using a bayesian state-space model with multiple abundance indices

Abstract

Deer population control is important in wildlife management, because overabundance of deer is a problem worldwide. For practical deer population control, deer population dynamics and the factors that influence them need to be evaluated in low‐cost and time‐efficient ways. However, in traditional methods of estimation, such as cohort analysis, large numbers of deer need to be caught for many years, and the ages of the deer must be determined. We estimated deer population dynamics by using a Bayesian state‐space model with multiple deer abundance indices (seen deer per unit effort, pellet group count, and block count) and numbers of deer hunted and culled in Yamanashi Prefecture, central Japan. In the state process of our state‐space model, latent deer abundance at year t in location m (Dt,m), with m being each cell of a grid mesh covering Yamanashi Prefecture, was assumed to decrease annually through hunting and culling, to increase with the population growth rate of each mesh (rm; which was determined from the percentages of forest, evergreen forest, and artificial grassland), and to fluctuate stochastically. In the observation process, Dt,m was assumed to be correlated with the deer abundance indices and a Gaussian white noise in the deer abundance indices. The estimated Dt,m was correlated with each deer abundance index, but the correlation coefficient was the greatest for pellet group density. The percentage of hunted and culled deer needed to reach 30% to reduce the annual growth rate (Dt,m/Dt−1,m). Increasing the percentage of artificial grassland increased rm. Our results showed that 1) deer abundance could be estimated by using only deer abundance indices in addition to population growth rate and the percentage of hunted and culled deer; and 2) preventing the intrusion of deer onto artificial grassland and intensive culling on artificial grassland were important to decrease deer abundance. © 2013 The Wildlife Society.

Appropriate vegetation indices for measuring the impacts of deer on forest ecosystems

“…Deer abundance is increasing in the Americas, Europe, and Asia (Stewart and Burrows, 1989;Fuller and Gill, 2001;Rooney, 2001;Iijima et al, 2013). Accordingly, serious impacts of deer on forest ecosystems, including tree debarking (Akashi and Nakashizuka, 1999;Verheyden et al, 2006;Vospernik, 2006;Kiffner et al, 2008;Takeuchi et al, 2011), reduction of sapling density (Horsley et al, 2003;Kumar et al, 2006;Gill and Morgan, 2010;Randall and Walters, 2011;Nuttle et al, 2014), browsing on saplings and understory vegetation (Takatsuki, 1989;McLaren et al, 2000;Morellet et al, 2001;Rooney and Waller, 2003;Heuze et al, 2005;Akashi et al, 2011), changes in species diversity and/or composition of understory vegetation (Stewart and Burrows, 1989;Horsley et al, 2003;Beguin et al, 2011;Suzuki et al, 2013;Filazzola et al, 2014), and soil disturbance (Beguin et al, 2011) have been documented.…”

“…However, constructing multiple deer enclosures with variable deer densities is challenging. Recently, absolute deer density has been estimated by Bayesian modeling (Yamamura et al, 2008;Iijima et al, 2013); in these cases, estimated absolute deer density could be used to clarify the effect of deer density on vegetation indices. We sought to evaluate the relative effect of deer density and other factors on vegetation indices and to clarify which vegetation indices are most appropriate for measuring deer impacts on forest ecosystems by considering the effects of tree size, snow depth, light condition, and understory vegetation type.…”

Susceptible conditions for debarking by deer in subalpine coniferous forests in central Japan

“…In recent years, the population densities of large ungulates, especially deer species, have increased worldwide (Stewart and Burrows 1989;Fuller and Gill 2001;Rooney 2001;Apollonio et al 2010;Iijima et al 2013). An increase of deer density has been shown to result in more prevalent debarking of trees (Akashi and Nakashizuka 1999;Nagaike and Hayashi 2003;Iijima and Nagaike 2015) and browsing of saplings and understory vegetation (Gill and Beardall 2001;Beguin et al 2009;Suzuki et al 2013) in forests.…”

“…In addition to these factors, spatial variation of deer impact was observed in lower elevation forests: the higher the deer density, the higher the proportion of debarked trees (Iijima and Nagaike 2015) and browsed saplings (Akashi et al 2011) and the lower the sapling density (Beguin et al 2009). The spatial variation of deer density across low-elevation areas was explained by the presence of attractive habitat (e.g., artificial grassland; Kamei et al 2010;Iijima et al 2013). However, the attractive habitat in subalpine forests, which are located on steeper slopes and have more complex topography than lower elevation areas, has not been well studied.…”

Vegetation Recovery in Response to the Exclusion of Grazing by Sika Deer (Cervus nippon) in Seminatural Grassland on Mt. Kushigata, Japan

“…Recently, the population of sika deer (Cervus nippon) in Japan has increased sharply because of the declining numbers of hunters and their aging, extinction of predator, and so on [7,8]. The effects of deer on natural grassland vegetation have been reported worldwide (e.g., [9]).…”