Relative influence of inter- and intraspecific competition in an ungulate assemblage modified by introduced species

Abstract

Interspecific competition from introduced and naturally colonizing species has potential to affect resident populations, but demographic consequences for vertebrates have rarely been tested. We tested hypotheses of interspecific and intraspecific competition for density, body mass, and fertility of adult female Roe Deer (Capreolus capreolus) across a heterogeneous forest landscape occupied by two introduced deer species: Mediterranean Fallow Deer (Dama dama); and subtropical Reeve's Muntjac (Muntiacus reevesi). Species-specific deer densities in buffers around culling locations of 492 adult female Roe Deer, sampled over seven years, were extracted from spatially explicit models calibrated through annual nocturnal distance sampling. Roe Deer fertility and body mass were related to species-specific deer densities and extent of arable lands using piecewise structural equation models. Reeve's Muntjac density was lower at higher Fallow Deer densities, suggesting interspecific avoidance via interference competition, but greater when buffers included more arable land. Roe Deer body mass was marginally greater when buffers included more arable land and was independent of deer densities. However, Roe Deer fertility was unrelated to female body mass, suggesting that fertility benefits exceeded an asymptotic threshold of body condition in this low-density population. However, Roe Deer fertility was slightly greater rather than reduced in areas with greater local Roe Deer density, suggesting negligible intraspecific competition. In contrast, Roe Deer was less fertile in areas with greater Reeve's Muntjac densities; thus, interspecific exceeded intraspecific competition in this assemblage. In contrast, we found no support for any effects of Fallow Deer density on Roe Deer density, body mass, or fertility. Complex networks of interspecific competition operating in this deer assemblage include: interspecific interference from Fallow Deer exceeded habitat effects for Reeve's Muntjac; and interspecific competition from introduced, smaller sedentary Reeve's Muntjac reduced fertility, unlike intraspecific, or potential competition with larger, more mobile, Fallow Deer for native Roe Deer. Mechanisms driving Roe Deer fertility may include interspecific behavioral interference or stress-resource depletion is considered less likely because Roe Deer fertility was independent of body mass. Findings emphasize the importance of ensuring appropriate management strategies for controlling invasive species.

Keywords: feral deer; interference competition; interspecific competition; intraspecific competition; invasion biology; invasive species.

Fig. 1.

Piecewise structural equation models (SEMs) testing hypotheses of interspecific competition among three deer species, and intraspecific competition of Roe Deer. Observational model representing all hypothesized directional causal effects (arrows) between variables (rectangular boxes, see Table 1 for details) that were tested by SEMs.

Fig. 2.

Densities of Reeve’s Muntjac, Roe Deer, and Fallow Deer across Thetford Forest (2011–2017), showing density surfaces (individuals/km2) interpolated at a 100 × 100 m grid from annual density surface models (DSMs) including latitude and longitude. For Fallow Deer, surfaces show the 3-year moving average, due to lower encounter rate, larger group size, and thus greater sampling variance.

Fig. 3.

Relation of Roe Deer fertility (A, B), Roe Deer body mass (C), and Reeve’s Muntjac density (D, E), to species-specific local deer densities (A, B, D) and local extent of arable lands (C, E), supported by model averaging across piecewise structural equation models (SEMs). Fertility, body mass, and Reeve’s Muntjac density were predicted from models including variables measured at the Akaike-weighted mean buffer radius (497 m) following model averaging of SEMs across incremental buffer radii (from 400 to 600 m). For fertility, boxplots represent the distribution of individual Roe Deer with either two embryos versus zero or one embryos (combined reference level).

Fig. 4.

Results of piecewise structural equation models (SEMs) testing hypotheses of interspecific competition among three deer species, and intraspecific competition within Roe Deer, showing supported causal effects (arrows) relating Roe Deer body mass and fertility to intra- and interspecific deer densities and local extent of arable lands, showing model-averaged coefficients (across incremental buffer radii, 400–600 m, corresponding to 59–113 ha) and their 95% confidence interval. Models incorporate random effects of cull year and forest block; dotted line represents the correlated error structure for which the correlation coefficient is also reported.