At the turn of the 20th century, North American deer populations, particularly in the East, were reduced to relict populations that totalled less than 300 000 individuals. Their steady recovery started in the 1930s, was well under way by the middle of the 20th century, and today deer abundance is estimated to reach 30 million individuals (Côté et al., 2004). The causes of this change in deer populations are multiple. They include the eradication of predators, favourable hunting regulations, increased edge habitat as a result of forest fragmentation, land use changes with access to rich forage in fields and, more recently, climate change. As a result of this continent-wide increase, deer populations now reach historic heights in many places (McCabe & McCabe, 1997; Côté et al., 2004) and can be considered both a conservation success and a conservation concern because of negative effects on understory vegetation, forest regeneration and public safety and health (Côté et al., 2004).
Leopold et al. (1947) warned early about the threats expanding deer herds posed to forest ‘health’, and recent local studies have demonstrated cascades of negative effects triggered by high deer populations on plants and animals (Côté et al., 2004; Martin et al., 2010). The impact on North American forest understory had already become so prevalent at the end of the 20th century that Diamond (1992) wondered about the need to ‘shoot deer to save nature’. Many local studies (deCalesta, 1994; McShea & Rappole, 2000; Allombert et al., 2005; Gill & Fuller, 2007; Martin et al., 2010, 2011; Cardinal et al., 2012) and a recent data-based modelling exercise (Newson et al., 2012) have demonstrated links between expanding deer populations and declines of numerous understory-dependent songbirds. These results led authors (Allombert et al., 2005; Martin et al., 2011) to suggest that continent-scale increases in deer populations should have continent-wide negative effects on forest birds that depend, for feeding or nesting, on habitat features affected by deer such as understory vegetation and abundance of deciduous trees.
Here, we show for the first time, a continent-wide link between increased deer populations and declines in understory-dependent forest–songbirds. We anticipate our essay will stimulate increased attention to potential continent-wide ecological consequences of increasing deer populations and ultimately to the indirect consequences of human-induced depletion of large carnivores.
Seven local North American studies on deer and songbirds (Appendix S1 and S2) have analysed how 73 forest–songbirds responded to deer. These studies were conducted in boreal, temperate and deciduous forests severely subjected to browsing by overabundant introduced or native deer (Appendix S2). In these studies, deer negatively affected 37 songbird species, of which 33 (89%) could be identified a priori as potentially deer-sensitive (i.e. they depend on the understory for feeding and nesting). Among the remaining 36 species that are not, a priori, deer-sensitive, 29 were unaffected by the presence of deer and 7 increased in response to deer.
We used Breeding Bird Survey (Sauer et al., 2011) data to assess population trends for these 73 songbird species between 1966 and 2009, a period of marked continent-wide increases in white-tailed (Odocoileus virginianus) and mule or black-tailed (Odocoileus hemionus) deer (Côté et al., 2004). Of the 73 species, 29 declined at the continent-scale during this period. Of those, 19 (65%) had been identified as deer-sensitive by the local studies (see Appendix S1 for details).
We quantified recent spatial variation in deer abundance with an index of deer biomass/km² calculated at the scale of each of the 63 North American states and provinces (Crête, 1999). The index varied from fewer than 100 to more than 300 kg km−². We retained 49 states and provinces (77%), in which local population trends could be estimated for at least 11 deer-sensitive songbirds. We analysed how the proportion of declining species in the deer-sensitive and non-deer-sensitive categories varied with deer biomass per km². In the non-deer-sensitive category, the proportion of declining species did not vary significantly across deer biomass categories and ranged between 15 and 19%. In the deer-sensitive category, the proportion of declining species always exceeded 20% and significantly increased to 46% with increasing deer biomass (Fig. 1). The 17 states or provinces with highest biomass were in the eastern United States. Those with the lowest biomass were in the North and West, areas where deer live with their predators, grey wolves (Canis lupus) and/or cougar (Felis concolor) (Crête, 1999). Among the 17 states with the highest deer biomass (>301 kg km−²), the proportion of deer-sensitive songbirds that declined exceeded 35% in all but two (Michigan & Mississippi). Overall, the proportion of deer-sensitive species that declined was highest in Alabama (67%; 10 of its 15 deer-sensitive species declined) and lowest in Utah (0%; 12 deer-sensitive species analysed).
Figure 1. Ratio of number of bird species declining and number of species analysed in 49 North American States or Provinces for deer-sensitive (grey) and nonsensitive (black) species in relation to deer biomass (<100 kg km−²: 11 states, 101–301 kg km−²: 21 states, >301 kg km−²: 17 states). We tested differences with a post hoc Tukey's HSD test. Different letters refer to statistically significant differences (P < 0.05). The error bars correspond to standard error.
Download figure to PowerPoint
Our results indicate that bird species for which local studies have shown an impact of deer decline more strongly in states where deer populations are higher. On the other hand, bird species that were not previously found to be deer-sensitive did not show any trend correlated with ungulate density. In the absence of predators, hunting is recognized as an important factor controlling deer populations (Côté et al., 2004). However, for various reasons, the number of hunters is declining in the northern hemisphere, and we suspect that this trend will continue with the propagation of chronic wasting disease (CWD). In consequence, deer populations will continue to increase in North America and Europe in the years to come.
Of the 310 forest-breeding species in the United States 22% are declining. Over the past 40 years declines occurred mainly in eastern forests (http://www.stateofthebirds.org/2009/). Here, we have shown that deer herd expansion could be an important player in this decline. Our essay is based on correlation between deer abundance and the decline of forest birds since the 1960s. We are aware of the limits of this simple approach, particularly because it does not include the other factors generally recognized as related to bird population declines, such as fragmentation of breeding habitat, anthropogenic habitat change, resource use and invasive species (Robbins et al., 1989; Robinson et al., 1995). In addition, in our study many of the deer-sensitive species are neotropical migrants that are also exposed to the destruction of their wintering habitats (Robbins et al., 1989; Terborgh, 1989). However, the contrast in response of our two a priori established species groups (sensitive or not to deer) is striking enough to suggest a role of deer in continent-scale songbird distributions and to encourage further attention to the question. Deer alone may not threaten any species with extinction but could locally be the straw breaking the camel's back for many songbird populations when acting in concert with other stresses. With respect to species diversity in forest bird communities, another concern is the extent to which deer have already caused continent-wide shifts in forest bird community structures through a reduction in the prevalence of understory species.
- Top of page
- Supporting Information
Simon Chollet is a PhD student at the CEFE (Centre d'Ecologie Fonctionnelle et Evolutive) in Montpellier, France. His research focuses on the impacts of overabundant ungulates on plants and bird communities.
Jean-Louis Martin is a senior scientist at CNRS and is based at CEFE. His research focuses on how species and communities adjust to change, in particular change induced by humans (species introduction, land-use abandonment or intensification) and the lessons for conservation. He is a founder of RGIS (http://rgis.cefe.cnrs.fr), a research consortium that has worked for over fifteen years on the ecological consequences of the current increase in deer populations. Most of the work by RGIS takes place on Haida Gwaii in British Columbia, Canada.
Author contributions: S.C. compiled the data and performed the analysis. Both authors contributed equally to initial design and ideas and to writing of the paper.