Introduction
Context dependence, where relationships differ in different situations, is widespread in ecology and invasion ecology (Catford et al.2022; Pyšek et al. 2020). For example, invasion can have a positive, negative, neutral or unimodal relationship with disturbance (Moles et al. 2012). Such variation can make it difficult to generalise across studies, systems and species, can prompt questions about the validity of hypotheses, and can hamper invasive species management. Like other subfields of ecology (Rillig et al. 2019), most studies in invasion ecology consider only one or two factors (Jeschke et al. 2020; Moles et al. 2012; Zhang et al. 2022), and most invasion hypotheses only consider one or two processes (Latombe et al. 2021). However, multiple factors can limit species’ growth, survival and reproduction in heterogenous environments (Farrior et al. 2013; Harpole et al. 2016; Harpole & Tilman 2007; Tilman 1982), and unless the primary limiting factor in a given situation is alleviated, a response to increases in other factors will be slight or not occur (Cade et al. 1999; Tilman 1990). For example, in a resource-limited system, increasing species’ seed dose (or propagule pressure) will not result in increased invasion if resource availability, rather than seed availability, is the main factor constraining species’ abundance (Fig. 1) (Kaiser et al. 1994). The concept of limiting factors (or environmental constraints (Tilman 1990) or Sprengel-Liebig’s Law of the Minimum (van der Ploeg et al. 1999)) is central to thinking in community and restoration ecology (Farrior et al. 2013; Greet et al.2022; Seabloom 2011; Tilman 1982) (but see Danger et al. 2008), as well as agronomy (van der Ploeg et al. 1999), but invasion ecology typically concentrates on drivers or correlates of invasion rather than on the factors that limit it (Catford et al. 2009; Taylor et al. 2016). In this paper, we argue that a greater focus on ecological mechanisms rather than ecological phenomena in invasion ecology, and a greater focus on limiting factors specifically, will help us better explain and anticipate context dependence and enable more accurate predictions of invasion and its impact on native diversity. Because species coexistence relies on species being able to invade communities from which they are absent (Chesson 2000), increased mechanistic understanding of invasion is central to community ecology too.
Invasion by plant species requires seeds, resources, suitable abiotic conditions and biotic interactions, and sufficient time and space for populations to establish and grow; any of these factors could limit invasion when in short supply (Catford et al. 2009; Pearsonet al. 2018). However, their relative scarcity – and thus importance – may vary from situation to situation depending on the species-site mixes in question. Variation in species’ life histories means that what may limit one species may not limit another (Catfordet al. 2019; Tilman 1990). If a study considers and manipulates multiple factors in parallel, effectively alleviating multiple limiting factors, then the effect of these factors should become visible (Fig. 1). However, if invasion is co-limited by two essential factors but only one limiting factor is alleviated, then otherwise positive or negative relationships may appear neutral. Because of this, studies that investigate multiple factors and hypotheses in combination may be more likely to find evidence for underlying mechanisms than studies that manipulate one factor at a time. Considering factors that are likely to limit invasion in a given situation can help us to understand and predict context dependence. A species-site matrix provides a structure for considering the hierarchy and interplay of limiting factors and for making such predictions (Fig. 1).
Imagine we have two groups of species, “fast invaders” characterised by resource acquisitive traits, fast growth rates, high seed production and poor competitive ability, and “slow invaders” with conservative traits, slow growth rates, low seed production and high competitive ability (Craven et al. 2018; Reich 2014). These species might invade early, mid and late successional sites (Fig. 1a). Succession theory tells us that early succession should primarily be seed-limited and late succession resource-limited (Catford et al. 2012a; Huston & Smith 1987). However, the extent of limitation should differ between the two groups of species reflecting their life histories (and associated tradeoffs, Tilman 1990), and these differences may be most pronounced in mid-succession where seed- and resource-limitation are less extreme. We might thus expect that fast invaders will be resource-limited in mid succession, whereas slower-growing and less resource-demanding slow invaders will – in the initial years of invasion – primarily be limited by their intrinsic growth rates and seed availability. Once we have a species-site matrix portraying the hierarchy of limiting factors, we can then imagine how seed addition or disturbance might modulate these conditions, revealing underlying relationships (Fig. 1b). Where seed-limited, invader abundance should primarily be determined by seed dose, whereas species’ competitive ability for limiting resources and disturbance (or other actions that increase resource availability) should largely determine the extent of invasion when resources are limiting. Alleviating primary limiting factors can allow effects of other, sequentially limiting factors to be observed, thus revealing the interplay of limiting factors.
By considering factors that are likely to limit resident plant diversity and how invasion may affect those limiting factors, we can also use the concept of limiting factors to predict the conditions under which invasion may reduce resident diversity (cf community diversity made up of both residents and invaders). For example, in the very early stages of succession, resident diversity would likely be limited by resident seed availability and to a lesser extent by resident population growth rates. Resident diversity in mid succession will likely be limited by a combination of resources, seeds and growth rates, moving to resource-limitation in late succession (Fig. 1). In the first few years of invasion, especially when invasion is small-scale, invading plants are unlikely to alter the seed availability or population growth rates of resident plants, so invaders will primarily affect resident diversity through resource competition. By consuming resources, like space, soil nitrogen or water, plant invaders can increase the extent of resource limitation, and thus potentially reduce the occupancy and abundance of resident species. The diversity impacts of invasion will depend on context though, such that invasion may have little impact on resident species when invaders are poor competitors or when factors other than resources primarily constrain resident diversity (i.e. plant invaders will only reduce diversity if they effectively lower the capacity of Liebig’s barrel; Fig. 1a).
In this study, we use a seed addition experiment in Minnesotan grasslands to test whether the concept of limiting factors can help predict and explain the cover abundance of two groups of invading species and their impact on resident diversity in communities that vary in successional stage. Our three-year experiment involves four treatments in a partially factorial design: i) successional stage of recipient community; ii) disturbance; iii) invader type; and iv) invader seed dose. Invader type and community successional stage provide a matrix of conditions where we expected the relative importance of seed-, resource- and growth rate-limitation on invader abundance and resident diversity to vary (Fig. 1). Invader seed dose, disturbance (which elevates resource availability) and invader type (fast vs slow invaders) enable us to effectively modulate invader seed-, resource- and invader growth rate-limitation respectively and to test our predictions. Variation in invader type and abundance in the invaded communities enabled us to test how invasion can reduce resident diversity via its effect on resource availability.
Through simultaneous invasion of up to 14 grassland species (10/14 native but all species formerly absent from the experimental sites), we also examine the importance of interactions among invaders, specifically whether resource acquisitive “fast” invaders constrain the abundance of more conservative and competitive “slow” invaders via niche pre-emption, or whether slow invaders constrain fast invaders via resource competition for soil N and soil water (Catford et al.2012a). We use measures of resource availability (light at ground level, soil moisture, soil nitrate) and, where relevant, resident species richness to help understand the trends observed. We use “invasion” to describe the colonisation and establishment of previously absent (“invading”) plant species in this paper, using the term invasion in a broad mechanistic sense (Chesson 2000; Tilman 2004). The processes that we examine and discuss may thus relate to native species added for restoration, range-shifting and range-expanding species, as well as human-mediated alien species invasion and community ecology research (Seabloom et al. 2003). To avoid confusion, we use the term “factor” to refer to limiting factors only, and “treatment” to refer to the four factors manipulated in the experiment.