Long-term monitoring shows the upward spread of non-native plant species along mountain roads around the world

The number of non-native plant species in mountain regions around the globe has increased in the last 5-10 years, and already establishing species are rapidly spreading to higher elevations.
Published in Ecology & Evolution
Long-term monitoring shows the upward spread of non-native plant species along mountain roads around the world

Compared to lowland ecosystems, mountains have so far been less invaded by non-native plant species. However, an increasing number of studies report the presence of non-native plants in mountain ecosystems around the world. Climate warming and increasing human pressure at high elevation are expected to further facilitate the establishment and spread of non-native species in the future.  In 2005, the Mountain Invasion Research Network (MIREN) was founded to understand patterns, processes and impacts of plant invasions in mountain regions around the world. To support these goals, researchers established a long-term standardized protocol to survey non-native (and native) plants in mountains regionally and globally. We had to be patient, but now, finally, we gathered enough data to look at long-term trends for the first time.

The MIREN protocol in action

In 2007, the first MIREN roadside surveys were conducted in Hawaii, New South Wales (Australia), Oregon and Montana (USA), Central and South Chile, Switzerland and Tenerife (in 2008), with a further three regions (India, Norway and Victoria (Australia) in 2012. In each region, every five years the vegetation is monitored along several roads, all of them open to traffic at least at some point during the year. Each road is divided by elevation into ideally 20 transects and at each transect, three 2m x 50m plots are laid-out in a T-shape in order to distinguish between disturbed habitats directly next to the road and more semi-natural habitats away from the road (Figure 1). Data collection consists of both site (geolocation, elevation, road type) and vegetation measurements (name, abundance and cover of each vascular plant species present). The number of regions conducting the survey is growing, and today the data set consists of 18 regions around the world, with the most recent re-survey completed in 2022.

Figure 1: Layout of the MIREN design. a) Placing of the 20 sampling transects along an example road. b) T-shaped structure of plots 1-3 at each sampling transect. (Figure adapted from Haider et al. (2022))

As more than 15 years have passed since the start of the project, we could now finally explore how the number and distribution of non-native plants in mountain ecosystems has been changing over time, and if these changes are consistent between mountain regions across the globe. In our recent paper in Nature Ecology and Evolution, we offer a first overview of the change in non-native species richness and spread dynamics in eleven regions (Fig. 2) over the last five to ten years.

Figure 2: Location of the 11 MIREN regions with temporal data used in the paper.

Non-native species are moving up and new species are moving in

Overall, we found a significant increase in non-native plant richness of on average approximately 16% over ten years. Considering that ecological processes such as the establishment of new species and range expansion often take place over long periods of time, this increase in the number of detected non-native species within a decade is surprising. However, the trends varied greatly among the eleven regions and the increase in species richness could only be detected by pooling observations from multiple independent regions, highlighting the importance of globally replicated studies to detect temporal trends within short time periods.

Figure 3: MIREN vegetation survey in Switzerland 2022. Left: Camille Brioschi identifying one of the approximately 600 plant species recorded in Switzerland in 2022 alongside one of the three Swiss roads. Right: One of the highest transects in Switzerland (top) versus one of the lowest transects, showing the great variation in vegetation (picture credits: Camille Brioschi, Fiona Schwaller).

Our results suggest plant invasions into mountain ecosystems are increasing as a result of  both a growing  number of new non-native species and an upward movement of species already present when the survey started. Moving to higher elevations and latitudes to follow their preferred temperature is a well-known escape strategy for plants in times of global climate warming, but while upward movements of native species in mountains are well recognized and relatively well documented, long-term studies of non-native species have so far been rare. We found that the upper elevation range limit of non-native plants expanded upwards in 10 out of the 11 surveyed regions. Given that our survey covered a period of only 5-10 years, these changes in distribution are occurring very rapidly. However, as non-native species are often initially introduced in the lowlands, from where they find their way into the mountains, an upward spread can be expected even without the effect of climate warming. This corresponds with the finding that in some of the surveyed regions, upward range limit shifts were mainly observed for species occurring in the lower or middle part of the surveyed elevational gradient, as might be expected if species are initially introduced to the bottom of the elevation gradient.

Figure 4: The European invader Verbascum thapsus enjoying its spot in a Chilean mountain roadside.

The need for statistical rigour

While exploring variation among species in the magnitude of upper range limit shifts, we found consistent negative relationships between species’ range limit shifts and their initial range limits, in some regions even resulting in slight average downslope shifts at high elevations. Similar patterns have been previously observed in temporal studies of native species distributions, and various ecological mechanisms have been proposed as possible explanations. However, due to two properties of finite environmental gradients like ours, this pattern of range limit shifts can be expected by chance alone. Firstly, the statistical phenomenon of regression toward the mean (see Mazalla & Diekmann (2022)) describes the fact that unusually large or small values will on average be followed by measurements closer to the mean, leading to an expected negative correlation between species’ initial range limits and their change over time. Secondly, the geometric constraint imposed by the finite elevational gradient leads to restrictions in observable range shifts, for example making it impossible to observe large upward range limit shifts for species already recorded near the upper limit of the survey. Consequently, we formulated a null model to evaluate to what extent our observed patterns deviated from the changes expected by chance alone. With time-series data on species range shifts becoming more and more available, null models are a powerful and necessary tool for robust interpretation of range shift dynamics.

Figure 5: Diverse landscapes and beautiful weather during the 2022 re-survey of the elevational transects in Tenerife. (picture credits: Meike Buhaly)

With this study, we show that in the current era of anthropogenic change, non-native species are continuing to advance into high elevation ecosystems. These changes are so fast that we can observe them over less than a decade, highlighting the speed at which our vegetation is changing. These findings highlight the urgency to extend monitoring schemes and enact management plans to prevent any negative consequences for high elevation ecosystems. We should act now, as our mountains are changing while we’re watching.


  1. Haider, S. et al. Think globally, measure locally: The MIREN standardized protocol for monitoring plant species distributions along elevation gradients. Ecology and Evolution 12, e8590 (2022).
  2. Mazalla, L. & Diekmann, M. Regression to the mean in vegetation science. Journal of Vegetation Science 33, e13117 (2022).

Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in