Relationships among behavior, habitat, and population density in a cyclic population of ruffed grouse

Abstract

Understanding how patterns of habitat use by animals vary in relation to population density is of major interest to ecologists and wildlife managers. For decades, biologists have linked high populations of ruffed grouse (<italic>Bonasa umbellus</italic>) with aspen (<italic>Populus</italic> spp.) forests in the northern part of their range. However, male ruffed grouse in northern populations also select territories in conifer forests, even when apparently suitable aspen forests nearby remain vacant. This selection of cover types that are presumed to be of lower quality presents a conundrum to biologists: why would males select inferior cover types when better cover types are available? To address this conundrum, I investigated the behavior, habitat, and population dynamics of a population of male ruffed grouse to evaluate relationships among these factors and elucidate mechanisms of their cover type selections. Little is known about the mechanisms underlying habitat selection in most species. Studying the behavior of individuals within a population is one technique used to evaluate these mechanisms. Ruffed grouse are an ideal species in which to study behavior, habitat, and population dynamics because the males are easily counted, they are territorial, their territories are relatively easy to locate, and many populations are cyclic. I conducted my study at the Cloquet Forestry Center in northern Minnesota and used data collected over the course of one complete 10-year ruffed grouse population cycle (2002-2011) during which time the population reached both a high and low point in the numbers of territorial males counted each year. The grouse population was declining when the study begin in 2002 and did not begin its cyclic increase until 2006. I used repeated auditory surveys to locate the territories of drumming males and estimate the population density of territorial male grouse. I assessed behavior questions by using automated video systems to record the activities of a stratified (by aspen and conifer cover types) random sample of 23 male ruffed grouse from all of the males detected on my surveys. The number of breeding males on my 1,419-ha study area varied from a low of 47 in 2005 to a high of 134 in 2010. I assumed that rates of male ruffed grouse display were indicative of their interactions with conspecifics. I developed a priori models that described the relationships among male display rates and environmental attributes, structured these models as generalized linear models, and used information theoretic model selection to evaluate these models. The null model (i.e., intercept only model) was the top-ranked model (received the lowest AIC_c score). I used generalized linear models with random effects and information theoretic model selection to evaluate a priori models that described relationships among male ruffed grouse densities per cover type and environmental variables associated with their territory locations. Aspen-dominated cover types contained the highest densities of male ruffed grouse during all years (0.09 to 0.24 males/ha during 2002-11) except 2006 and 2011, when density was highest in northern hardwoods-dominated cover types (0.12 and 0.13 males/ha in 2006 and 2011, respectively) followed by aspen-dominated cover types (0.10 and 0.12 males/ha in 2006 and 2011, respectively). The density of males in conifer-dominated cover types was always less than the density of males in aspen-dominated cover types (0.02 to 0.08 males/ha during 2002-11). In addition to consistently higher densities of males in aspen- versus conifer-dominated cover types, densities of males in aspen-dominated cover types exhibited a higher intrinsic rate of increase than in conifer-dominated cover types during all years. However, the consistent occupancy of conifer-dominated cover types by males throughout the study also suggested that males perceived conifer-dominated cover types as suitable sites for territories and display behavior. At the cover type scale, the top-ranked model related to male grouse population density included an interaction of cover type and year variables. The mean density of male ruffed grouse within landscapes during all years ranged from 0.04 to 0.10 males/ha with minimum and maximum densities of 0 and 0.16 males/ha, respectively, during the entire study. At this landscape scale, an interaction of the estimate of the relative evenness of cover types (Shannon's Evenness Index) within landscapes and year comprised the top-ranked model explaining male ruffed grouse densities. Based on my study, if there is a difference in quality of conifer and aspen cover types, it manifests itself either in the specific stand structure used for male breeding displays or some life stage or seasonal habitat need of the ruffed grouse other than male breeding display cover. My results show that grouse populations achieve their highest densities in heterogeneous landscapes when aspen-dominated cover types, composed of multiple age classes, are the key cover types on landscapes. Landscapes that are less heterogeneous and where the majority of cover types are not dominated by aspen also harbor grouse, but at lower densities than heterogeneous aspen-dominated landscapes. Traditionally, managers have focused on the distribution and abundance of aspen-dominated cover types to enhance ruffed grouse habitat. However, my findings suggest that managers should focus on managing the spatial complexity of cover types on landscapes, even when aspen-dominated stands are in low abundance.