Introduction

With worldwide trade intensifying, opportunities for species introduction outside of their native range have increased (Hulme 2009; Levine and D’Antonio 2003). In spite of ongoing debates on the impacts of exotic plant invasion on native biodiversity (Gilbert and Levine 2013; Thomas and Palmer 2015), there remain many situations where invasive plants are considered a nuisance and have to be managed (Simberloff 2005). When diverse native plant communities give ways to dense stands of invasive plant species, habitat quality and biogeochemical processes are often compromised (Blossey 1999; Zedler and Kercher 2004). Invasive plant species can therefore greatly undermine conservation or restoration goals where biodiversity is valued, for instance in nature reserves or ecologically significant habitats.

Exotic invasive plants can be controlled by herbicides, mowing, burning or labour-intensive practices like slashing or hand-felling. If control is not selective, the resident plant community is disturbed or has disappeared, and if the invader is still part of the local species pool, the risk of reinvasion can be high. For seed dispersed invasive species, the success of seedlings in the colonization/establishment phase (Fig. 1) will determine the potential to (re)establish a self-sustaining population (Catford et al. 2009; Dietz and Edwards 2006; Van Kleunen and Johnson 2007). This usually represents the most vulnerable stage for many invasive plants (Fraser and Karnezis 2005) and therefore seedling establishment provides a critical window of opportunity for plant invasion management. Restoring a competitive plant cover that can interfere with invasive plant establishment can be effective, but synthesis of biodiversity theory, ecological mechanisms, and experimental evidence relevant to biodiversity restoration is lacking, possibly delaying applications (Esler et al. 2010).

Fig. 1
figure 1

The biological invasion process in which key determinants such as propagule pressure, abiotic constraints, and biotic resistance determine the invasion outcome

Full size image

The recruitment of invaders depends on three factors that together determine the outcome of invasion—biotic resistance, abiotic constraints, and propagule pressure (Catford et al. 2009; Miller et al. 2014). Biotic resistance refers to the ability of species in a resident community to resist or limit invasion by plants (Catford et al. 2009; Levine et al. 2004). Various components of a community, including competitors, consumers and/or pathogens, can contribute to biotic resistance to invasion (Levine et al. 2004). The role of biotic interactions has been well-documented in community ecology (Byun et al. 2013; Fox 1987; Fridley et al. 2007; Levine and D’Antonio 1999; Pokorny et al. 2005; Prieur-Richard et al. 2000), and is central to our understanding of how communities at risk of invasion assemble after disturbances (Clark and Johnston 2011).

Abiotic constraints refer to environmental conditions, stressors, or filters that can suppress the recruitment of invaders and/or resident competitors depending on their tolerance (Melbourne et al. 2007; Weiher and Keddy 1995). Limited nutrients, low light level, or unsuitable climatic conditions, for example, constitute abiotic constraints (Davis et al. 2000; Davis and Pelsor 2001). Disturbances may release invaders from abiotic constraints, providing invasion opportunity (Hobbs and Huenneke 1992).

Propagule pressure has also been identified as a key element in determining invasion outcome (Lockwood et al. 2005). Pressure depends on propagule size—the number of individuals arriving at any one time, and frequency of arrivals. Colonization pressure is also used to refer to the number of species released into a single location, a variant of propagule pressure (Lockwood et al. 2009). An obviously important and initially underestimated factor in the invasion process, propagule pressure interacts with other factors (biotic resistance or abiotic constraints) in ways that need to be better understood.

In this paper, we review current knowledge on these three key determinants of invasion success, focusing on a common field situation in which biodiversity needs to be restored after disturbances and plant invaders are part of the species pool. We elaborate on the ecological mechanisms at play for each determinant and emphasize their relevance for ecological restoration and exotic invasive plant management. This synthesis allows us to identify gaps in knowledge to inform future studies as well as opportunities and constraints for field applications aiming at the restoration of invasion-resistant habitats.

Mechanisms of invasion resistance and related ecological concepts

Several mechanisms have been proposed to explain how biotic resistance, abiotic constraints, and propagule pressure each acts during invasion (Fridley et al. 2007; Funk et al. 2008; MacDougall et al. 2009; Sax et al. 2007; Shea and Chesson 2002). These mechanisms are not necessarily mutually exclusive, since several processes may work synergistically or in alternation, depending on context and scale (Fridley et al. 2007; Pauchard and Shea 2006). Moreover, concepts and mechanisms are often related through scale, from individual species to community to ecosystems.

Biotic resistance

Niche and fitness difference

Each species has acquired through natural selection a set of traits (canopy height, specific leaf area, type of reproduction, timing of germination, seed size, etc.) that determines its fitness in a particular environment and the level of resources (nutrient, light, moisture, etc.) that it needs or can tolerate. The niche of a species describes the habitat requirements and ‘behaviors’ (for a plant, examples could be dormancy or foraging through deep roots) that may allow for persistence and reproduction. A functional group framework associates species in terms of their traits, and consequently of their niche and fitness similarity (Drenovsky et al. 2012; Eisenhauer et al. 2013; Funk et al. 2008). Simply put, the way in which a species interacts with another species depends on their respective traits and niche requirements and how similar or dissimilar these traits or niches are.

Biotic resistance can be explained by mechanisms of niche difference and fitness difference between resident and invading species. ‘Limiting similarity’, from classical competition theory (Macarthur and Levins 1967; Weltzin et al. 2003), postulates that there is a limit to the similarity in niche overlapping or resource use between resident species and invading species. According to this theory, an invading species will not establish where a resident species occupies a similar niche or has similar traits. A functional group interpretation of the concept of limiting similarity can also be considered: the related ‘Fox’s assembly rule’ hypothesizes that the lack of a certain functional group in a resident community will make that community more susceptible to invasion by species from that particular functional group (Fox 1987; Von Holle and Simberloff 2004) until niches become saturated.

When niches overlap, the fitness difference between the resident species and the invader determines which one will be competitively excluded (MacDougall et al. 2009). Performance traits such as plant height (Gaudet and Keddy 1988), biomass (Gaudet and Keddy 1988; Lulow 2006; Rinella et al. 2007), plant cover (Gerhardt and Collinge 2003) or plant size (Schamp and Aarssen 2010) can be indicative of potential for biotic resistance. Biomass of resident communities, in particular, has been identified as a good indicator of competitive ability (Gaudet and Keddy 1988) and level of biotic resistance (Lulow 2006). High biomass of resident species could reflect the way in which available resources are utilized and, consequently, whether resources are made (or not) available for potential invaders. Moreover, difference in the timing of life events can contribute to differential plant fitness and temporal partitioning of resources. Early-season emergence, for instance, can contribute to fitness (Verdú and Traveset 2005) and biotic resistance at the critical establishment stage (Firn et al. 2010). Species that establish early and grow fast benefit from a so-called ‘priority effect’, eventually inhibiting a subsequent invader from establishing (Mwangi et al. 2007). Priority effect is expressed as ‘first come, first served’ in terms of available resources (Byun et al. 2013; Fukami et al. 2005; Stuble and Souza 2016b; Young 2001).

In the context of manipulating biotic resistance for restoration purposes, there is experimental evidence that establishing native species functionally similar to potential invaders, or establishing fast growing species early, may be successful in limiting exotic plant invasion. Fargione et al. (2003), introducing species (both native and exotic) into 147 experimental prairie-grasslands plots at Cedar Creek Natural History, showed that established species most strongly inhibited introduced species from the same functional group. In a prairie restoration experiment, Larson et al. (2013) found that early robust establishment of native species increased biotic resistance to Cirsium arvense (more so than functional similarity). Leffler et al. (2014) examined the ability of monocultures and assemblages of prairie species (native or non-invasive exotic) to resist invasion by two exotic invasive species, Bromus tectorum (cheatgrass; an annual grass) and Isatis tinctoria (dyer’s woad, a biennal forb). They found that monocultures of a single growth form were in general more resistant to invasion by invasive plants of the same growth form. It must be noted that applying the limiting similarity concept to ecological restoration assumes that the identity of the potential invader is known. To increase biological resistance to several known potential invaders or to undetermine invaders, the concept must be extended to include a diversity of established native species, as we discuss below.

Diversity effect

At community level, the diversity-resistance hypothesis predicts a positive relationship between species diversity and biotic resistance (Levine and D’Antonio 1999). Based on the mechanisms previously mentioned, diverse communities can have fewer unused niche spaces and more efficient resource use through niche partitioning than species-poor communities, thus resisting invasion (Funk et al. 2008; MacDougall et al. 2009). Additionally, the more species in a resident community, the more likely a resident species’ niche will overlap with that of invaders, increasing competition through limiting similarity. Niche partitioning can be demonstrated empirically by increasing the number of neighboring plants (Kennedy et al. 2002), canopy complexity (Lindig-Cisneros and Zedler 2002) or resource uptake partitioning in a community (e.g., soil nitrogen forms) (Ashton et al. 2010; Booth et al. 2003; Frankow-Lindberg 2012). When resident species are functionally very different from each other, functional diversity, through trait complementarity, can help maintain both community stability and resistance (Fargione and Tilman 2005; Funk et al. 2008).

The related hypothesis of ‘insurance effect’ states that species diversity increases community-wide stability under a fluctuating or heterogeneous environment (Ives et al. 2000; Loreau et al. 2001; Tilman et al. 2006), with a potential scale effect in the field (Balvanera et al. 2006; Fridley et al. 2007). Trait or functional diversity allows species with suitable traits to fill available niches when conditions change, preventing invaders from taking advantage of fluctuating resource availability (Davis et al. 2000). In this regard, functional group richness (or functional diversity based on trait complementarity) could be a more direct indicator of invasion resistance than species diversity alone (Symstad 2000; Pokorny et al. 2005; but see also Prieur-Richard et al. 2000). All types of functional groups can be of equal importance (Pokorny et al. 2005; Rinella et al. 2007), but others predict that functional group composition (i.e., ratio of each functional group) can be equally important (Prieur-Richard et al. 2000; Tilman et al. 1997).

For any particular community, diversity effect can be divided into ‘selection effect’ and ‘complementarity effect’ (Loreau and Hector 2001). The effect in question can be measured in terms of any process of interest such as productivity, resistance to invasion, etc. A selection effect is significant if certain dominant species in a community mainly influence diversity effects (Emery and Gross 2007; van Ruijven et al. 2003). Complementarity effect is significant when resources partitioning or positive interaction between species contribute most to diversity effect (Kennedy et al. 2002; Levine et al. 2004; Levine and D’Antonio 1999). Empirical evidence indicates that both complementarity and selection effects contribute to a net diversity effect on invasion resistance (Fargione and Tilman 2005; Frankow-Lindberg et al. 2009), although their relative contribution is not always quantified.

For restoration purposes, ecological theory and empirical evidence suggest establishing high native plant functional diversity after disturbance can increase resistance to invasion by exotic invasive species. For example, Phragmites australis seed germination is sensitive to light level and competition, and seedling recruitment through seed dispersal contributes to invasion of disturbed habitats (Albert et al. 2015). To test mechanisms of biotic resistance to P. australis seed invasion in freshwater wetland (Byun et al. 2013, 2015) first classified 35 wetland plants from the regional pool of native species within four functional groups based on trait similarity. An additive competition design was conducted with P. australis and different mixtures of native plants and invasion success (P. australis seedlings) was measured. The functional group characterised by fast growing annual plants was most resistant to early invasion by P. australis, supporting priority effect, followed by the functional group of tall long lived perennials with rhizomes. Mixtures of four species were more resistant than monoculture due to complementarity diversity effect. Comparing resistance to invasion from three exotic invasive species of North American grasslands (Cirsium vulgare, Melilotus officinalis, Bromus inermis), in plots seeded with low and high richness seed mixtures (15 and 97 species respectively), high richness plots were more efficient at resisting invasion (Nemec et al. 2013). Sheley and Half (2006) found that forb mixtures were significantly more successful against the invasive Centaurea maculosa than forb monocultures. A garden experiment by Ammondt and Litton (2012) showed that combination of two or three Hawaiian native species had a greater negative impact than monocultures on biomass and fecundity of the invasive grass Megathysus maximus, also suggesting that increased functional diversity may improve restoration success. Experimental studies conducted in the context of restoration, however, do not always show diversity effect. In an experiment with different planting methods and seed species richness using seed mixtures commonly used by practitioners, Larson et al. (2011) found that greater species richness did not translate into greater resistance to exotic invasion. Similarly, Leffler et al. (2014) found that a diverse mixture of grasses was no less, but also not more resistant to invasion than a monoculture, In such cases, higher plant functional diversity may still be advisable if it brings other benefits (e.g., enhance habitat quality) than biotic resistance alone, or other possibly abiotic constraints on resistance must be evaluated.

Enemy release hypothesis

Pathogens or herbivory by native organisms may facilitate exotic plant invasion. For example, the overabundance of deer in North America facilitates the success of invasive Alliaria petiolata and Microstegus vimineum (Knight et al. 2009). When pathogens or herbivores are viewed as important determinants of plant invasion however, it is more often within the framework of the Enemy release hypothesis (ERH). ERH stipulates that species invading new regions gain competitive advantage over resident species as they leave behind their herbivores, pathogens, parasites or predators (Keane and Crawley 2002). Support for this hypothesis comes mostly from (1) biogeographical studies showing that damages from enemies are fewer on invasive plant species in the invaded range than on plants from the same species in the native range (Colautti et al. 2004), (2) from fewer damages on invasive compared to native species within the same area (Cappuccino and Carpenter 2005), and (3) successful biological control programs (Mitchell and Power 2003). While ERH may be determinant for some exotic introduction, it is not universal and several studies failed to find patterns consistent with this hypothesis (Parker and Gilbert 2007).

ERH frequently serves as justification for biological control of exotic species using herbivore insects or pathogenic fungi specific to their plant host and originating from its native area. While biological control rarely eliminates invasive plants, it may reduce their population size, productivity or fertility, thereby increasing biotic resistance of restored native plant communities (Van Driesche et al. 2010). One invasive plant species that has attracted much attention regarding the effect of biological control using imported exotic insect herbivores is Lythrum salicaria (purple loosestrife) and its biological agents Galerucella calmariensis and Galerucella pusilla (leaf beetles) (Yeates et al. 2012). In a survey of 36 Lythrum stands in Central New York 10 years after exposures to the biological control agents, Grevstad (2006) found that plants were shorter and showed various signs of beetle damage, but that stem density and coverage remained unaffected, a pattern consistent with other surveys. Biological control for the invasive Ageratina riparia (mist flower) in New Zealand, on the other hand, was very successful. Within five years of the introduction of the Entyloma ageratinae (white smut fungus) and Procecidochares alani (gall fly) as biological agents, mean percent cover of mist flower in heavy infested sites declined from 81 to 1.5% (Barton et al. 2007). Biological control of invasive species can be a valuable tool because it may be potentially self-sustaining, non-polluting and relatively inexpensive (Culliney 2005). While this field of research is actively increasing, biological control agents, however, exist for only a few species within a specific area, and development for additional species is slowed by the risk of damages to non-target native species. Consequently, it is still very rarely applied in restoration projects.

Abiotic constraints

Abiotic filtering and fluctuating resources

Studies describing the variation in the levels of invasion among habitats in different regions and on different continents have found similar patterns of invasion for similar habitats. The most and the least invaded habitats remain the same, despite being in different regions (Chytrý et al. 2008; Pysek and Chytry 2014), emphasizing the importance of abiotic conditions (Kalusová et al. 2015). Harsh abiotic constraints suppress intolerant species, including invaders (Melbourne et al. 2007; Parepa et al. 2013; Weiher and Keddy 1995). Invasion levels therefore tend to be low in harsh climatic conditions and nutrient-poor habitats (Chytrý et al. 2008). Field studies have shown that abiotic constraints play a significant role in determining invasibility in relation to flooding (Collinge et al. 2011; Gerhardt and Collinge 2003), sediment salinity (Dethier and Hacker 2005), soil nutrients (Goldstein and Suding 2013) overstory tree composition (Von Holle 2005), or extreme climatic events (Collinge et al. 2011; Goldstein and Suding 2013). A strong environmental filtering process may lead to trait underdispersion (traits are more similar to each other than what would be expected by chance) and phylogenetic clustering during community reassembly (Adler et al. 2013; Procheş et al. 2008).

The ‘fluctuating resource availability’ hypothesis states that plant communities become more vulnerable to invasion when the amount of unused community resources (i.e., resource availability) increases (Davis et al. 2000). A combination of abiotic and biotic factors determines resource availability. When the supply of resources (abiotic factor) is controlled, or if resource acquisition by resident communities (biotic factor) increases, invasion will occur less frequently. The notion that competition is less important in recently disturbed environments where resident plants are not sequestering all available resources underlies this hypothesis (Davis et al. 1998; Grime 1998), which has been the subject of much debate and is supported by some empirical evidence (Davis and Pelsor 2001; Frankow-Lindberg 2012; Iannone and Galatowitsch 2008). Any disturbance event that changes abiotic constraints to increase resource availability can trigger an invasion.

One of the most common application of a change in abiotic conditions in the context of restoration to manage plant invasion concerns the lowering of nitrogen availability. N enrichment was found to stimulate exotic invasive plant species in many environments. Addition of carbon in the form of sawdust or sucrose reduces N availability, thus presumably favouring native species at the expense of fast-growing exotic species. An example where this approach was successful is reported by Prober et al. (2005), who showed that repeated sucrose applications in grassy woodlands in a temperate region of Australia significantly increased the abundance of native perennial grasses at the expense of exotic annuals. Creating a crop plant cover on bare soil (thereby lowering light availability) controlled 89% of invasion by P. arundinacea, but also suppressed desired species (restoration target communities) by 57% (Iannone and Galatowitsch 2008). In the same experiment, applying sawdust (high C:N ratio) lowered available nitrogen in soil and decreased invasion by 59%,while not affecting the desired target communities. Alternatively, sawdust addition brought no significant benefits to native plants of California grassland over a 2-year period (Corbin and D’Antonio 2004). Despite the mixed results from the literature, carbon addition or other modification of the abiotic environment may still hold promise as a tool for restoration, depending on habitats, but requires a good understanding of the response of the invading species involved.

Propagule pressure

The role of propagule pressure in the invasion process is relatively straightforward: ‘the more you introduce, the more you get’ (Lockwood et al. 2009). For example, propagule pressure alone explained 56% of the variance in exotic species richness in a study (Lonsdale 1999). In a restoration experiment, Phalaris arundinacea (reed canary grass) was sown in wetland mesocosms at different density (0, 10, 50, 100, or 500 seeds/m2) with a mixture of native species (Reinhardt Adams and Galatowitsch 2008). Increasing the seed density of the invader increased invasion success (more shoots, more biomass of P. arundinacea). There is, however, considerable interest in the relationship between propagule numbers and invasion success (the dose–response curve) (Lockwood et al. 2005). It is uncertain whether there are always increasing invasion benefits from increasing propagule pressure, or whether saturation occurs beyond a certain threshold (Byun et al. 2015; Lockwood et al. 2005). If so, it is not entirely clear what ultimately determines that threshold. For Hieracium lepidulum, an invasive perennial herb in New Zealand, recruitment varied not only with propagule pressure, but also with the biotic and abiotic characteristics of habitats including resident plant cover and plant richness (Miller et al. 2014). This suggests taking into account complex interactions when evaluating the role of propagule pressure and that the effects of propagule pressure are likely to be exacerbated in areas displaying poor biotic resistance.

Managing the propagule pools of invasive species early in the restoration process could significantly reduce invasion risk but also poses challenges. Eliminating on-site seed bank often requires destructive or costly approaches (chemical control, tarping). Moreover, managing invasive species populations from the surrounding matrix that most contribute to seed rain in the restored site (closest populations, those facing predominant wind, etc.) is not always possible, and so nearby propagule sources are commonly ignored in restoration projects. Evidence suggests, however, that restoring a plant cover will reduce the effect of propagule pressure on invasion success.

Interplay between determinants of biological invasion

Depending on their focus, studies are not always consistent regarding the importance of each of the main determinants of biological invasion, but it is also unlikely that a single mechanism will govern the entire invasion process in all types of habitats and for all invaders. Rather, biotic resistance, abiotic constraint, and propagule pressure interact with each other in ways that depend on habitat conditions and the species involved (Berg et al. 2016; Perelman et al. 2007; Warren et al. 2012). The intensity of abiotic constraints will determine the relative importance of abiotic versus biotic factors in invasion resistance (Gerhardt and Collinge 2003). In stressful and harsh environments, abiotic constraints may entirely determine the fate of invaders (Chytrý et al. 2008; Dethier and Hacker 2005; Wang et al. 2006b). In benign or intermediate conditions, in contrast, biotic resistance can become as, or more important than abiotic constraints (Gerhardt and Collinge 2003; Naeem et al. 2000; Perelman et al. 2007; Thomsen et al. 2006a, b). Abiotic constraints can modulate biotic resistance in direct as well as indirect manners (Byun et al. 2015; Collinge et al. 2011; Perelman et al. 2007). Similarly, environmental heterogeneity and biotic resistance can modify invasion outcome from propagule pressure alone (Eschtruth and Battles 2011; Miller et al. 2014). Understanding and manipulating invasion resistance therefore requires a synthetic approach: evaluating not only the individual effect of each factor, but also disentangling their interaction effects (Table 1).

Table 1 Hypotheses about determinants of invasion
Full size table

The importance of considering simultaneously different mechanisms of biological invasion for management can be illustrated using a well-known macrophyte invader of wetlands of North America, Phragmites australis. Experiments showed that abiotic constraints can modulate biotic resistance through direct and indirect effects (Byun et al. 2015). High water level always directly affected invasion by P. australis by reducing recruitment success, but also indirectly affected invader recruitment through the resident plant cover, reducing or increasing their ability to resist invasion. For example, Typha latifolia increased its biotic resistance to invasion with increased flooding, whereas Lolium multiflorum suffered from flooding, decreasing its biotic resistance effect. The presence of a dense resident plant cover also lowered the threshold at which invasion success occurred even when propagule supply increased and biotic resistance was most effective under low propagule pressure of the invader. These results mirror those of Miller et al. (2014) for Hieracium lepidulum. We know from the manipulation of species and functional group richness and composition in resident communities that they all have a significant impact on biotic resistance (Byun et al. 2013). Increasing biotic resistance through the restoration of a resident plant cover, however, will be most efficient when considering abiotic constraints and propagule pressure.

Enhancing ecological resistance

Restoring a native plant cover after control can be an effective measure to prevent or delay invasion. A plant cover can be self-regenerative, so repeated control interventions become less necessary although monitoring is recommended. Table 2 summarizes the results of research projects focusing on plant cover to control invasive plants. This approach, however, raises many practical issues, including how to select and combine species and how to determine the appropriate conditions where ecological resistance will be enhanced. Invasive plant control programs usually do not evaluate the role of native revegetation following removal (Kettenring and Adams 2011) and, to date, empirical tests have shown mixed results in terms of the effectiveness of plant restoration for controlling invasion. This may be because these tests are more often based on available species than on the community assembly principles that determine species associations. The selection and combination of plants is crucial as just randomly adding species may increase diversity but not necessarily enhance resistance, or worse may result in imbalanced competition amongst certain plant types, which may actually increase the susceptibility to invasion. Failure to detect the effect of a plant cover on resistance may also result from the fact that other modulating factors such as the ones explored in this study (abiotic constraints, propagule pressure) are not considered, while in some cases they may override biotic resistance effect. In any case, it is unlikely that biotic resistance alone will completely suppress invasion without complementary preventive measures such as reducing propagule pressure from nearby sources or pathways and address, when possible, site conditions for abiotic filtering.

Table 2 Research relevant to plant cover restoration to control invasion
Full size table

As the ultimate test of ecological theories (Ewel 1987), the control of invasion requires a thorough and operational understanding of invasion mechanisms that could apply to a range of environments and taxa. The complex nature of the invasion process where abiotic and biotic factors interact in ways that are not yet fully understood makes not only field applications, but also research experiments rather daunting. Since rarely a single mechanism (or factor) governs the entire invasion process, multiple drivers must be addressed simultaneously. Even though we know more about resistance mechanisms and functional traits of plant species, most restoration studies have tested the effect of specific local species on an invader instead of specific resistance mechanisms. The latter are often inferred a posteriori. Yet, testing the interaction between species traits (both of the resident species and the invaders), abiotic condition, and propagule pressure could lead to useful generalisation. While we need control and restoration case studies with local species at different sites, the development of a coherent, dynamic and adaptive framework around biotic/ecological resistance will require going beyond the idiosyncrasy of the many species and systems being tested. Developing such a coherent framework would be beneficial as many current managerial frameworks quickly become obsolete or are applied incorrectly or out of context. This will require identifying and sharing emerging principles at the crossroad between community ecology, invasion ecology, and ecological restoration. Given the evidence uncovered, emphasizing the functional diversity of the restored community seems a promising approach when facing potentially multiple invaders and/or fluctuating abiotic conditions and could inform species selection for restoration.