Results
The standardized protocol for recording plant species communities along mountain roads has been thoroughly tested in the field on all continents except Antarctica (Figure 4). The first survey was carried out in eight regions in 2007 and has been repeated every five years since, resulting in one baseline historical survey (2007) and up to two resurveys (2012 and 2017). The number of regions has increased since 2007, with 18 regions performing the survey by 2018 (Figure 4). The global database currently includes circa 2,700 plots and >100,000 observations of >5,000 vascular plant species.
One of the most striking findings of the global MIREN surveys to date has been to document the importance of roads in facilitating mountain invasions. Specifically, we have found that non-native species richness in roadsides decreases with increasing elevation, but generally peaks in the lower third of the elevation gradient (Alexander et al., 2011). Moreover, the vast majority of non-native species found at high elevation are also present at low elevation, with only 2.3% of high-elevation non-native species being specifically adapted to alpine environments (Alexander et al., 2016). These findings indicate that non-native species are first introduced and become established at low elevation sites, following this they spread to higher elevations (Alexander et al., 2011). At higher elevation sites, non-native species generally become increasingly filtered out by environmental pressures, so fewer warm-adapted perennials reach higher elevations (McDougall et al., 2018). We have also revealed that the number of non-native species declines with increasing distance from the road (Seipel et al., 2012; Haider et al., 2018), indicating that the native plant community serves as a second environmental filter that selects for more shade- and moisture-tolerant perennials (McDougall et al., 2018). In addition to non-native species, the MIREN surveys have shown that native species also use roads as corridors (Lembrechts et al., 2017). Interestingly, Lembrechts et al. (2017) found that occurrence optima are higher in roadside habitats than faraway habitats, and moreover that some alpine species have shifted their ranges downwards due to altered abiotic conditions and competitive release in roadside habitats (see also e.g. Lenoir et al., 2010).
Globally, the MIREN surveys have demonstrated that native plant species richness does not follow a consistent pattern in non-roadside (semi-)natural habitat along elevation gradients, suggesting the existence of additional region-specific mechanisms, such as biome, vegetation type and human activity. These mechanisms are now the subject of further study. In contrast, a clearer elevation signal is present on roadside plots, with total species richness peaking at mid-elevations in most regions (Haider et al., 2018). Further, we have observed a reduction in community dissimilarity (beta-diversity) along roadsides relative to more distant plots, which is amplified by the arrival of non-native species along mountain roadsides homogenizing plant community composition (Haider et al., 2018). The MIREN surveys have also provided insight into the vulnerability of habitats regionally (Pollnac et al., 2012), the genetic background of successful invasions (Haider et al., 2012) and the impact and management of local invasions (McDougall et al., 2011a). For example, in the Greater Yellowstone Ecosystem in the United States, we found that non-native species emergence varies with elevation and habitat type, which provided land managers valuable information for mitigating biological invasions (Pollnac et al., 2012). Moreover, in the dry Mediterranean Andes in Argentina, which are characterized by treeless vegetation, the survey demonstrated how non-native plant species can successfully spread from the roadside into natural vegetation at low and intermediate elevations, thus highlighting the susceptibility of these types of ecosystems to invasion (Aschero et al., 2017). By contrast, the alpine vegetation of northern Norway has been shown to be more vulnerable to invasion than its low elevation counterpart, indicating that vegetation structure plays an important role in community invasibility (Lembrechts et al., 2014). Finally, the MIREN surveys have already generated information about regional floras. An excellent example is the discovery of a new species of Poaceae during MIREN monitoring in Kosciuszko National Park, Australia – this species was named after the network: Poa mireniana (Walsh & McDougall, 2018).