Insect-mediated apparent competition between mammals in a boreal food web

Abstract

While the important role of animal-mediated interactions in the top-down restructuring of plant communities is well documented, less is known of their ensuing repercussions at higher trophic levels. We demonstrate how typically decoupled ecological interactions may become intertwined such that the impact of an insect pest on forest structure and composition alters predator-prey interactions among large mammals. Specifically, we show how irruptions in a common, cyclic insect pest of the boreal forest, the spruce budworm (Choristoneura fumiferana), modulated an indirect trophic interaction by initiating a flush in deciduous vegetation that benefited moose (Alces alces), in turn strengthening apparent competition between moose and threatened boreal caribou (Rangifer tarandus caribou) via wolf (Canis lupus) predation. Critically, predation on caribou postoutbreak was exacerbated by human activity (salvage logging). We believe our observations of significant, large-scale reverberating consumer-producer-consumer interactions are likely to be common in nature.

Keywords: apparent competition; habitat selection; insect outbreaks; species conservation; species interactions.

Fig. 1.

Simplified schematic of a boreal caribou–wolf–moose system illustrating the indirect effects of a SBW outbreak on trophic interactions. Links L1 to L5 are described by arrows. The blue arrows indicate effects for an early or less-impacted stage of SBW outbreak (e.g., 3 y of defoliation), and green arrows indicate effects at a later or more severe outbreak stage (e.g., 10 y of defoliation). The yellow arrows represent effects of salvage logging inclusive of the associated road network introduced by logging.

Fig. 2.

Observed percentage cover of available deciduous vegetation in SBW-infested forest stands in relation to an index of budworm cumulative severity in 2018 (Côte-Nord region, Québec, Canada). The points represent raw values, and colored ribbons indicate 95% CIs.

Fig. 3.

Relationship between moose density and percentage cover of salvage-logged forest stands (Côte-Nord region, Québec, Canada). We compared moose density before and after SBW outbreak (2006 versus 2018) within areas that were subject to salvage logging, with a spatial reference where salvage logging was low (<20%), conforming to a before-after-control-impact design. The boxplots describe moose density in 2006 and in 2018 in relation to the percent cover of areas infested by SBW for at least 1 y and then cut (SBW_cut) within plots (each 60 km²) characterized by a low (<20%) or a high (>20%) percent cover of SBW_cut in 2018. The black points correspond to the mean density of moose, the center value is the median, edges of the box are 25th and 75th percentiles, and whiskers represent ± 1.5 the interquartile range.

Fig. 4.

Study area in the province of Québec, Canada. The green boxes represent maps of the distribution and defoliation severity based on aerial survey data of the SBW outbreak (68) in 2011 (Top) and 2018 (Bottom) within the study area. The blue dots represent the distribution of GPS radiocollared caribou in 2006 to 2011 and 2012 to 2018 used in the analyses. The red line represents the northern limit for forest management activities. The hatched area defines Hunting Area 18 in Côte-Nord region within the study area.