Temporal patterns
Native and non-native ruderal species temporal dynamics deviated surprisingly much from our hypotheses. Indeed, we expected to find an increase in ruderal species over time as a result of the joint impact of climate and land use change on the vegetation composition along the Rallarvägen, since both have been shown to be advantageous for ruderal species (Chiuffo et al. 2018; Guo et al. 2018; Thuiller et al. 2005). Following this hypothesis, we can indeed assume that already shortly after the building of the railroad at the beginning of the 20th century substantial numbers of ruderal species were already present along the Rallarvägen trail - both native and non-native to the region, as the bare ground that remained accommodated up to 82 ruderal species, of which 44 were non-native, according to the first survey in 1903 (Fig. 2). Contrary to expectations, however, ruderal species (both the total and the non-native ruderals separately) were most numerous along the trail during the time of the railroad building and settlement expansion (1903-1913). The building of the E10 road (completed in 1983) did not result in an additional influx of ruderal species, nor did the substantial climate warming since the late 1980s. In 2021, total ruderal species richness was just below the levels of 1913, while non-native ruderal richness was only half of the numbers in 1903 and 1913. The drop in 1983 must be at least partially a methodological anomaly, but even then, we must consider the total number of ruderal species to have remained roughly constant through time, and the number of non-native ruderals to be decreasing (see ‘limitations’ below).
What is more, patterns in the community temperature index (expressed as the Ecological
Indicator Values for Temperature; EIV-T) revealed that most warm-adapted ruderal species were already present in the early 20thcentury, after the building of the railroad, with little evidence of an influx of warm-adapted species due to climate change in recent years (Fig. 3). In fact, we even observed a steady decline in mean EIV-T of the non-native ruderals up until 2021. Our data showed that 36 (mostly non-native) ruderal species with high EIV-Ts (≥ 4) in the historical years, had disappeared by 1983, and did not reappear in 2021, while species observed for the first time more recently had on average lower EIV-T values. It is likely that those warm-adapted ruderal species were introduced along the Rallarvägen with its initial construction (and likely the addition of non-native soil introduced into gardens in settlement areas). After this initial introduction, however, these species did not manage to sustain a stable population, potentially due to a combination of harsh winters and the lack of continuous disturbance after 1913 (Lembrechts et al. 2016b; Niittynen & Luoto, 2017; Rendeková et al. 2019). Interestingly, the lack of additional significant disturbance events along the trail in recent years, seems also to have prevented these species from re-establishing now that the climate has warmed to an average 1.0°C (ANS, 2019; Lembrechts et al. 2016b). In time, many of the ruderal species along the trail were thus again outcompeted by the native species that thrive in subarctic mountain environments – demonstrating community resilience. Nevertheless, as climate change progresses, we expect that warmer temperatures might still tip the scale, allowing subsequently introduced warm-adapted non-native species to gradually establish successfully in areas where disturbance is prominent (Dainese et al. 2017).