Introduction
Reproduction is vital to population persistence and distribution dynamics. Reproductive success is tightly linked to the quality and spatial distribution of available suitable habitat (Pulliam & Danielson 1991; Kurki et al. 2000) and so anthropogenic landscape change can markedly alter a species’ spatial distribution. These effects are typically negative, through fragmentation and habitat loss (Fahrig 1997; Fahrig 2002; Fahrig 2003) but are positive for some species, facilitating range expansions or invasions (Ewers & Didham 2006; Didham et al. 2007). Linking spatial variability in reproductive success with landscape change (or disturbance) is key to understanding mechanisms of invasion and range shifts, an increasingly important endeavor under climate change (Lawler et al. 2008; Lawler et al. 2009).
Quantifying spatial variation in reproductive success has been mostly limited to taxa with stationary offspring such as plants (Muñoz & Arroyo 2006) and nesting birds (Rosenberg, Swindle & Anthony 2003; León-Ortega et al. 2017). Mammals are much harder to quantify due to their large size, widespread ranges, and vagile young. Camera trapping (Burtonet al. 2015; Steenweg et al. 2016) can bridge this data gap, generating data on mammalian distribution and density. Many mammal species keep young at heel during early maternal care and this state can can be likewise observed with camera traps. Applied to camera data for grizzly bears (Fisher, Wheatley & Mackenzie 2014) and European brown bears (Burton et al.2018), we showed how spatial variation in reproductive success can be modelled to identify landscape mechanisms affecting success. Though further elaborated since (MacKenzieet al. 2017) the diverse opportunities of this approach have yet to be widely realized. Here, we illustrate how camera trap data can help infer mechanisms of species invasion and range expansion, using an example from the Nearctic boreal forest.
Boreal landscapes have been markedly changed by widespread and economically important resource extraction (Schindler & Lee 2010; Venier et al. 2014). The epicenter of change are Canada’s oil sands, the third largest global oil deposit and a driver of global economies (Bayoumi & Mhleisen 2006). Petroleum exploration and extraction create an altered landscape without analogs (Schneider, Dyer & Parks 2006; Pickell, Andison & Coops 2013; Pickell et al. 2015). Landscape change affects the entire boreal forest mammal community (Fisher & Burton 2018), but most notably manifest in woodland caribou declines (Rangifer tarandus ) (Hervieux et al. 2013; Hebblewhite 2017). Wolf predation is a primary cause (Boutin et al. 2012), with wolf populations bolstered by high-density invading white-tailed deer (deer; Odocoileus virginianus ) (Latham et al. 2011; Latham et al. 2013).
White-tailed deer range expansion is a pan-continental phenomenon (Laliberte & Ripple 2004; Heffelfinger 2011) impacting entire ecosystems (Côté et al. 2004). Research on deer expansion south of the boreal has focused on population biology (DeYoung 2011), movement (Beier & McCullough 1990), and predation (Ballard et al. 2001). Boreal deer invasion has been linked to landscape and climate change (Dawe, Bayne & Boutin 2014; Fisher & Burton 2018; Fisher et al. 2020) but the mechanisms remain unidentified. We sought to examine whether anthropogenic landscape change is linked to spatial patterns of deer reproductive success, as a possible mechanism of boreal forest invasion.
Deer balance energy intake from early-seral deciduous forage (Ditchkoff 2011) with metabolic demands markedly increased by cold temperatures and deep snow, historically limiting white-tailed deer range (Parker, Barboza & Gillingham 2009; Hewitt 2011). In the boreal, climate change has produced warmer winters (Karl & Trenberth 2003); concurrently, landscape change has generated more abundant early-successional vegetation (Finnegan, MacNearney & Pigeon 2018; Finnegan, Pigeon & MacNearney 2019; MacDonald et al. 2020) that is strongly spatially linked to deer abundance and persistence (Fisher et al. 2020). Deer mortality risk is greatest in the first year of life (Lesage et al. 2001), decreasing markedly for 1-2 year-olds (Delgiudiceet al. 2006). Fawn growth and survival is largely based on maternal body condition, governed by food availability (Therrien et al. 2008), so examining how spatial resource availability contributes to breeding success within the first year helps us understand how landscape change contributes to boreal deer expansion.
We hypothesized that anthropogenic landscape change in the northern boreal forest is providing resource subsidies that bolster reproductive success for invading white-tailed deer. If true, we predicted that anthropogenic features representing conversion of mature forest to early seral vegetation would explain variability in the spatial distribution of deer reproductive success. We define reproductive success as a deer occurrence with at least one attendant fawn in the summer months. This requires that a female achieve oestrus, breed, produce offspring, and maintain that offspring into the summer months, thus drawing close to recruitment–and is a measure that can be consistently applied to all mammal species with attendant young at heel.