Population density and plant availability interplay to shape browsing intensity by roe deer in a deciduous forest

Highlights

• The functional response relating browsing intensity to plant availability was strongly non-linear and density-dependent.

• When plant availability was low, browsing increased with plant availability with an increasing rate with roe deer density.

• When plant availability was above 12.5% browsing intensity increased with both plant availability and roe deer density in an additive way.

• Forest susceptibility to browsing increases with increasing competition for food, especially when plant availability is low.

• Considering functional responses improves the accuracy of management tools when monitoring the herbivore-plant system.

Abstract

Browsing damage in forests relies on a complex interaction between herbivore density and forest understory composition and relative availability. Although variation in the amount of browsed twigs is sometimes used to assess abundance of large herbivores, the potential confounding effect of resource availability on this relationship has not been investigated yet. To fill the gap, we measured how browsing intensity of the woody plants varied in response to changes in both roe deer (Capreolus capreolus) abundance and vegetation availability from an intensive long-term monitoring. We estimated plant availability and consumption by roe deer from a modified Aldous method throughout a 14 year-long period during which we experimentally manipulated population density. The functional response was strongly non-linear and density-dependent. When plant availability was low (<12.5%), browsing intensity strongly increased with plant availability with an increasing rate of roe deer density, whereas beyond this threshold, browsing intensity slightly increased with both plant availability and population density in an additive way. Thus, forest susceptibility to browsing increases with increasing competition for food, especially when plant availability is low. Moreover, the interplay between browsing intensity and population density at low plant availability prevents the use of browsing intensity to monitor roe deer abundance when plant availability is low. Our findings provide clear evidence that relying on key ecological concepts such as functional responses improves the accuracy of management tools when monitoring changes of the herbivore-plant system over time.

Introduction

Populations of large herbivores have strongly increased in range and abundance over the last decades in temperate areas (Côté et al., 2004, Apollonio et al., 2010). This general and continuous increase in the abundance of large herbivores has led to major changes in forest composition (Ramirez et al. 2019). By consuming plant tissues, herbivores affect the growth, survival and reproduction of plants (Hidding et al., 2012, Sabo, 2019). Heavy browsing generally favours the abundance of species resistant to browsing, at the expense of plant species preferred by herbivores (Tremblay et al., 2006, Hidding et al., 2013, Perea et al., 2014). In addition, the high browsing pressure exerted by abundant herbivores has strongly affected the forest landscape, mainly by reducing availability of understory (Fuller, 2001, Rooney and Waller, 2003). Overall, by affecting tree recruitment and forest regeneration, browsing by ungulates is currently considered as one of the main threats to forest (Schuck and Requardt, 2008, Perea et al., 2014, Ramirez et al., 2019).

The ecological consequences of browsing on plants depend on deer selectivity and on the tolerance of plants to browsing (Long et al., 2007, Perea et al., 2014, Sabo, 2019). Availability of understory vegetation is known to be one of the key factors affecting the sensibility of forest regeneration to browsing damage (Gill, 1992, Vospernik and Reimoser, 2008). The amount of understory vegetation available to large herbivores is closely linked to the dynamics of overstory vegetation (Gill et al., 1996, Vospernik and Reimoser, 2008). In most exploited forests, the removal of canopy increases the amount of understory vegetation available to herbivores following man-made or natural clearing (Gill et al., 1996, Olesen and Madsen, 2008, Vospernik and Reimoser, 2008). As the forest stand grows up, canopy closure lets the amount of understory vegetation available for herbivores decrease drastically (Gill et al., 1996, Olesen and Madsen, 2008, Vospernik and Reimoser, 2008). To understand better the forest-large herbivore system and take the appropriate management rules, we need to decipher how variation in population abundance and plant availability shapes browsing intensity (Côté et al., 2004, Tremblay et al., 2006). Relying on the concept of functional response (Holling, 1959), plant consumption should increase non-proportionally with plant availability (Abrams, 1982, Spalinger and Hobbs, 1992), leading the intensity of browsing to increase faster at low plant availability and to saturate at high plant availability (type II functional response). Moreover, for a given plant availability, browsing intensity should obviously increase with population abundance. While the latter relationship has been intensively investigated for a long time (see e.g. Aldous, 1944, Morellet et al., 2001, Chevrier et al., 2012), the expected confounding effect of functional response has been mostly overlooked in studies of plant-herbivore systems at very fine spatial scales (i.e. at the scale of plant consumption, 4th order selection : Johnson, 1980).

We aim to fill this knowledge gap by measuring browsing intensity in relation to both abundance of large herbivores and plant availability. We took benefit from an experimental manipulation of roe deer (Capreolus capreolus) abundance in the intensively monitored forest population of Trois-Fontaines, France, in which browsing intensity and plant cover were recorded over 14 years following a slightly modified version of the method developed by Aldous (1944) and adapted by Ballon et al., (1992). We expected (1) browsing intensity to increase with roe deer density (Morellet et al., 2001, Tremblay et al., 2006, Chevrier et al., 2012), (2) a functional response to occur so that browsing intensity should not increase proportionally with plant availability (Illius et al., 2002, Van Beest et al., 2016), and (3) the functional response to be density-dependent, with increasing roe deer density leading to increase intraspecific competition for food (Gill et al., 1996).