This study on the movements of invasive wild pigs and their regional connectivity potential using GPS satellite collaring is the first ever in Canada and represents the only connectivity research on wild pigs in the country to date. Our results indicate a high degree of potential landscape connectivity throughout much of the prairie regions of mid-western North America, with a matrix dominated by a deciduous forest, crop, waterbodies, and wetlands mixture being the easiest areas for wild pigs to move through. These findings are mirrored by a large sample of independently collected sighting and wild pig GPS data indicating that our models likely reflected wild pig selection tendencies and therefore their likelihood of occurrence in various landcover types. Differences in the landscape matrix created two pinch points along the U.S.-Canadian border in southern Saskatchewan and southwestern Manitoba. These two areas could readily facilitate transboundary movements and should continue to be monitored via baited trail cameras (given that some monitoring efforts are already in place, Aschim and Brook 2019, MacDonald and Brook 2023) for potential southern spread of this invasive species across international borders.
Landscape connectivity is one of the key drivers of species dispersal and expansion and is generally defined by the availability of habitat that can meet various life history needs within a landscape matrix (Dickson et al. 2013; Gantchoff and Belant 2017). Despite deciduous forest having the greatest conductive value in this study, it rarely played a role in the most conductive regions due to its fragmented nature amongst crops and wetlands. As a result, despite wild pig predilection for this landcover type, they occurred in landscapes dominated by crops with interspersed interconnected bodies of water and wetlands amidst deciduous forest and grassland along the periphery of crop fields. While agricultural crops were a dominant landscape feature within the northern prairies that potentially restricts wild pig movement given their dependence on cover for thermoregulation (VerCauteren et al. 2019), previous work has shown that taller crops (e.g., corn), along with grasslands, and interspersed waterbodies may act as the thermoregulatory patches wild pigs require (Santos et al. 2004; Paolini et al. 2018; Kramer et al. 2022). Furthermore, it has been shown that (i) the conglomeration of crops often leads to an increase in wild pig range (Keuling et al. 2008), (ii) wild pigs have higher movement rates within crops (Kay et al. 2017), (iii) wild pigs have higher population growth rates and probability of establishment in landscapes with access to agricultural crops (Tabak et al. 2018), and (iv) pig expansion in the U.S. prior to 2009 had been closely associated with agriculture (Snow et al. 2017). Thus, while further research is required to determine the specific rate at which expansion can occur through an agriculturally dominated matrix; given that the northern prairies have all the components for wild pig persistence (i.e., food, cover, and thermoregulation) it is very likely this crop matrix will lead to southern expansion without aggressive management intervention and that at least some of this expansion has already occurred.
Landscapes are naturally heterogeneous, however, large-scale landscape homogenization has occurred in many areas due to agricultural practices whereby monoculture crops are planted over large areas. Thus, while there is clear evidence that crops positively facilitate wild pig expansion, this homogenization greatly reduces regional connectivity (e.g., see Iowa; Fig. 3). This also holds true for native habitats which are seemingly important to connectivity. For example, northern Minnesota is predominately wetlands, and has very low connectivity (Fig. 3). Although homogeneous landscapes typically have negative impacts on the biodiversity of flora and fauna (Olden et al. 2006), they can also make it difficult for some species to adequately acquire the necessary resources for expansion and establishment. For example, generalist brown bears (Ursus arctos) traveled farther in homogenous environments to gain adequate dietary diversity (Mangipane et al. 2018). While such homogenous areas may have apparently negative impacts on wild pig expansion and colonization, it is important to understand wild pigs are generalists (Ballari and Barrios-García 2014), and they may eventually move into, and take advantage of, sub-optimal areas, especially for dispersal (as seen in other species; Hemmingmoore et al. 2020). Indeed, wild pigs may be able to thrive (though perhaps at lower densities) in areas with land cover types less suited to their needs (Baber and Coblentz 1987). Additionally, given the prevalence of crops in highly conductive regions, expansion may also have some temporal component, as crops are highly ephemeral with fields being essentially barren with hard packed snow for half of the year. Barren habitats are avoided and are areas that have lower potential connectivity, likely a result of their lack of cover greatly increasing the risk of being hunted or preyed upon, despite the fact that they are easily traversable by this species (Morelle et al. 2015). It is also important to note that intense human presence limits wild pig selection and activity (e.g., wild pigs select for fewer and safer habitats and are less active with human presence; Ohashi et al. 2013), which is reflected in the decreased connectivity around cities and towns. However, roads and highways do not appear to limit connectivity to the same degree as cities and towns, nor do they appear to limit pig movements (i.e., do not create pinch points), likely because they contain enough of a suitable matrix around them to negate their effects. However, roads with higher volume traffic and multiple lanes may increase collisions and mortalities (Beasley et al. 2014), which could lead to a reduction in dispersal success.
Wild pigs have been expanding across the Canadian prairies at an unprecedented rate (40,936 km2/yr; Aschim and Brook 2019), initially due to the widespread distribution of domestic wild boar farms starting in the 1980s (Michel et al. 2017), but additional supplementation now occurs through other exotic pig farms or released pets (MacDonald and Brook 2023). These farms were the original source that generated free-ranging populations, and wild pigs have been rapidly dispersing from these source areas (Michel et al. 2017, Aschim and Brook 2019, MacDonald and Brook 2023). Although there is literature that points to potential dispersal and establishment rates (e.g., 2.5 km/year at the colonization front in Belgium and a maximum dispersal distance of 4.5 km from the natal home range by females in Sweden; Truvé and Lemel 2003; Morelle et al. 2016), giving a potential baseline for this species. However, given that the movement capabilities of wild pigs in Canada are greater than seen elsewhere (e.g., home ranges for our collared pigs were, on average, 29-fold larger than the global average; Kramer and Brook unpublished data, Garza et al. 2018), it is very likely that dispersal distances in Canada are larger than anywhere else on earth. These unprecedented broad movements could be a function of lower human density or a lack of resources to meet their needs at smaller spatial scales. It is thus imperative that aggressive science-based management actions be enacted to hinder their spread. Ideally, prevention of initial invasions is the first and most cost-effective course of action, but once populations are entrenched, resource managers can try to contain a species at their source to prevent secondary spread (Drury and Rothlisberger 2008). It is clear that complete eradication of wild pigs in Canada is no longer feasible given their vast established range and very limited control efforts. We do know that wild pigs at northern latitudes can make extensive use of tall and densely planted agricultural crops (Kramer et al. 2022) and access to even small quantities of agricultural crops can dramatically increase population growth rates (Tabak et al. 2018), highlighting these landcover types as a key focus of management efforts. For example, changing farming practices which result in a matrix configuration dominated by shorter crops may help in mitigating wild pig spread (Kramer et al. 2022). In contrast, control efforts could be focused on preventing wild pigs from accessing areas that facilitate expansion, particularly during summer months when crops are likely to play a larger role as both high-quality food and hiding cover (Paolini et al. 2018, Kramer et al. 2022). This is especially pertinent to our study extent, as the regions Canadian pigs would most likely expand south into exhibit the highest potential wild pig population densities of anywhere in the northern U.S. (Lewis et al. 2017).
We acknowledge that this study does present some limitations. First, our sample size of GPS collared pigs was relatively small with collar malfunction, high hunting pressure within Saskatchewan, and physiological constraints of collaring wild pigs making it difficult to collect data for the life of the collar (e.g., collars slipping off prematurely). Despite this, statistically, our sample size was likely more than adequate to describe wild pig preferences (Street et al. 2021). Furthermore, our use of two independent validation datasets comprised of both an extensive amount of wild pig sightings, along with use of an independent wild pig GPS information, both showed that sightings and independent GPS data fell within high connectivity areas, adding confidence to our results. A second drawback was that, given that the U.S. portions of our extent are mostly, and have historically, been pig free, validation data was simply not available. As a result, our model should be verified and modified, if necessary, at a later time using information on wild pigs at monitoring sites along the Canadian-U.S. border as well as data on wild pig occurrences if they ever become established within our U.S. states of interest. At a minimum, our work provides guidance on where transboundary monitoring efforts should be concentrated, although a clear effort is being made by U.S. federal and state agencies to remain vigilant to wild pig reports across this expanse (as seen by the National Feral Swine Task Force).
As wild pigs continue to expand their range across the northern prairies management efforts will ultimately have a considerable effect on wild pig range expansion and establishment success. Economic and environmental damage by wild pigs (Pimental 2007; Barrios-Garcia and Ballari 2012; U.S. Department of Agriculture 2014) and their ability to transmit harmful diseases to livestock (e.g., pseudorabies, Müller et al. 2011; African Swine Fever, Blome et al. 2013) is well documented, and a desire to avoid these impacts (particularly since this area contains half of domestic swine production in both countries) highlights the need for enhanced local, national, and international collaboration. Our results show the potential for rapid and uncontrolled expansion of wild pigs in the northern prairies of North America and helps guide efforts towards prudent monitoring as this species continues to expand its range.