Introduction

Biological invasions are one of the main factors contributing to the loss of biodiversity worldwide, affecting ecosystems in all biomes (Pyšek et al. 2020). Although biological invasions by plants are currently largely concentrated in lowlands, climate change and human activities are increasingly funneling species upwards towards higher elevations (Alexander et al. 2016; Lembrechts et al. 2016; Dainese et al. 2017). Similarly, non-native species are moving toward higher latitudes, even into subpolar and polar areas (Pauchard et al. 2016; Lembrechts et al. 2017). The last two decades have seen a rapid increase in the number of studies of invasions into Arctic and alpine areas, starting to fill critical gaps in our understanding of plant invasions into mountain environments (Alexander et al. 2016; Fuentes-Lillo and Pauchard 2019). We now know that the patterns of invasion in mountain ecosystems are mainly conditioned by four key factors: propagule transport, abiotic conditions, biotic interactions and disturbance processes (Colautti et al. 2006; Pauchard et al. 2009, 2016). Unfortunately, many of these studies of invasive plant species in mountain areas focus on Europe and North America, with only a minority of studies from other continents such as South America (Alexander et al. 2016). Indeed, the latter continent has seen fewer than 200 publications on the topic during the last two decades, with 50% of these articles stemming from the mountain ecosystems of Chile and Argentina alone (Fuentes-Lillo and Pauchard 2019), leaving a considerable knowledge gap on mountain invasions in the páramo and puna (Barros and Pickering 2015; Fuentes-Lillo and Pauchard 2019).

The Andes is the main mountain range in South America (~ 500 to ~ 7,000 m.a.s.l). It has a geographical extent of more than 8,000 km, starting in the north of South America in Venezuela, at 8° N and extending all the way south to 69° S (Arroyo and Cavieres 2013). It comprises about 13% of the mountainous areas worldwide with a total area of ​3,000,000 km², with its highest peak reaching 6,962 m a.s.l (Insel et al. 2010). It hosts ecosystems with a high biodiversity of both flora and fauna due to the compression of climatic zones through elevational and latitudinal gradients (Pérez-Escobar et al. 2022). Historically, the Andes have been divided into three biogeographic zones. Firstly, the páramo, which includes the Andes from Venezuela (11°N) to the northern region of Peru (8°S), found in tropical latitudes. Secondly, the puna, which extends from the northern region of Peru (8°S) to the northern part of Chile (27°S), encompassing tropical to mid-temperate latitudes. Finally, the southern Andean steppe, located in temperate latitudes from northern Chile (27°S) to Patagonia (55°S) (Arroyo and Cavieres 2013). It also has a high level of endemism resulting from the speciation and migration that occurred during the last glacial period (Pallardy 2002). Indeed, with 6.7% of worldwide plant diversity and 23% of species considered endemic, the Andes Mountain range is one of the main hotspots of biodiversity worldwide (Barros and Pickering 2015; Barros et al. 2015). Nevertheless, these vulnerable ecosystems have experienced an increase in plant invasions, with more than 100 non-native plant species recently reported in mountain ecosystems in the Southern Andes steppe (Alexander et al. 2016; Fuentes-Lillo and Pauchard 2019).

On average, the increase in temperature pushes plant species upwards on mountains at a rate of 1.1 m yr− 1, although species do not necessarily move unidirectionally (Chen et al. 2011; Lenoir and Svenning 2015). Upward expansion of non-native plants is however likely much faster, especially when facilitated by anthropogenic disturbances (Marini et al. 2011; Lembrechts et al. 2016). Anthropogenic disturbances, including through industrial development, agriculture, livestock, tourism or mining (Barros et al. 2013, 2014), is one of the main drivers of the arrival and establishment of non-native plants in the Andes (Pauchard et al. 2016). Over the next few years, land use change in the Andes is expected to increase further, with direct implications for the establishment and dispersal of non-native plants (Pauchard et al. 2016; Bramer et al. 2018). Greater anthropogenic disturbance of these vulnerable ecosystems, combined with the impacts of climate change predicted for the coming decades, will likely jointly contribute to an accelerated upward migration of non-native plants in the Andes (Hernández-Lambraño et al. 2017; Fuentes-Lillo and Pauchard 2019). However, few studies have explored the distribution patterns of non-native plants along elevational and latitudinal gradients under different climate change scenarios, especially in Andean ecosystems (Lenoir et al. 2017; Fuentes-Lillo and Pauchard 2019).

The unique environmental and topographic characteristics of the Andes as a continent-spanning mountain range allow us to investigate how the factors of disturbance and climate change could favor the arrival, establishment and expansion of non-native plants at local, regional and continent-wide levels (Pauchard et al. 2009; De Frenne et al. 2013). This would allow not only a better understanding of the dynamics of these invasions, but also the development of multiscale protocols for the prevention and management of the growing diversity of non-native plants that threaten the biodiversity of these ecosystems. This is particularly relevant in higher elevation areas with higher endemism and where the level of protection (percentage of protected areas) is still not adequate (Pauchard et al. 2009; Alexander et al. 2016; Elsen et al. 2018).

Almost 15 years ago, a comprehensive research agenda was proposed to study biological invasions in mountain ecosystems (Pauchard et al. 2009). Until now, however, a unified evaluation for the Andes is lacking. Thus, we aim to identify the research gaps, summarize existing knowledge, and guide future efforts in the management of non-native plants in this highly vulnerable region. We use a systematic review to summarize current trends in plant invasions in different biogeographic areas of the Andes, taking advantage of the long latitudinal and steep elevational gradients of the mountain range as a natural experiment to advance the understanding of this global phenomenon in a rapidly changing world. Furthermore, we investigate the relative role of climate and anthropogenic disturbance as drivers of these distributional patterns, hypothesize about the future of plant invasions under a changing climate in the Andes, and summarize the limited evidence on the impacts of plant invasions in the system. Finally, we formulate a research agenda to inform protocols for the prevention and management of non-native plants in the region.

Methods

To determine non-native plant species richness patterns along both elevational and latitudinal gradients in the Andes, species richness data were extracted from both published and unpublished sources (the latter mostly data from the database from the Mountain Invasion Research Network (MIREN, http://www.mountaininvasions.org) and the biological invasions laboratory from the University of Concepcion, Chile (LIB, http://www.lib.udec.cl/). To obtain published articles, a bibliographic search was carried out on Web of Science (WoS) and Google Scholar using the following keywords (and Spanish translations thereof): ((invasiv * OR non * native OR introduced * OR alien * OR exotic) NEAR plant), (‘’ high elevation ' ‘OR’ ‘high-elevation’ ‘), (“Colombia”, “Venezuela”, “Ecuador”, “Peru”, “Bolivia”, “Argentina” and “Chile”) and (“Andean”, “mountains”, “Andes mountain range”) (Alexander et al. 2016; Fuentes-Lillo and Pauchard 2019). This search identified a total of 185 scientific publications, of which 25 publications were selected based on the criterion that they presented (a) a record of non-native plants associated with an elevational gradient, (b) a focus on community studies and not on a single non-native species, and (c) sampling sites along an elevation gradient. Occurrence data of non-native plants associated with each elevation were extracted from these publications. A total of 34 plots from nine studies were available for the páramo, 20 plots from six studies for the puna, and 120 plots from ten studies for the southern Andean steppe. The data obtained for both the páramo (Venezuela and Ecuador) and the southern Andean steppe (Argentina, central Chile, and southern Chile) were collected using the MIREN sampling methodology, which primarily assesses non-native plant communities near mountain roads (Fig. 1) (Haider et al. 2022).This resulted in 101 non-native species richness records, from seven distinct Andean regions covering a latitudinal range from 8°N to 37°S and incorporating an elevational range from 300 to 4,700 m.a.s.l. Additionally, 369 non-native plant species richness records were added from unpublished data (LIB database) concentrated in the southern Andes (36° to 53°S), across an elevational range from 337 to 1,665 m.a.s.l.

Fig. 1
figure 1

Non-native plants that grow associated with anthropogenic disturbances (presence of a road) in the highest elevation areas of the Southern Andean Steppe. a) Acacia dealbata growing on the roadside in the Southern Andes (38°41 × 05 S; 71°47 × 56´´E) at an elevation of 1,479 m.a.s.l, above the tree line of Araucaria araucana forest; (b) Verbascum virgatum growing on the roadside in the South and Central Andes in Chile (37°37´,| 35´´S; 71°35´,71´´E) at an elevation of 1,401 m.a.s.l, above the tree line, in volcanic soils; (c) Lupinus polyphyllus growing by the roadside in the subantarctic Andes (46°03´,  50´´S; 72°00´, 51´´E) at an elevation of 1,014 m.a.s.l; (d) Echium vulgare growing on the roadside in the South and Central Andes in Chile (36°91´,25´´S; 71°42´, 51´´E) at an elevation of 1,527 m.a.s.l

Full size image

We divide the Andes mountain range into three biogeographic zones based on the classification proposed by Arroyo and Cavieres (2013). We analyze elevational patterns in the species richness of non-native plants within each zone. A generalized linear mixed-effects model (GLMM) was used to test whether non-native plants richness changed as a function of elevation (linear and quadratic) and latitude (linear and quadratic) and their interactions. The model was adjusted to a Poisson distribution error with logit link; the biogeographic zone was used as a random effect. This model was built with the function “lmer” using the R statistical package “lmer4” (v1.1-31; Bates et al. 2015), The predicted values from the model were extracted using the package “ggeffect” (v1.2.0; Lüdecke 2018). To model the nonlinear changes in the richness of non-native plants along latitudinal gradients in both the southern and northern hemispheres simultaneously, we used a cubic b-spline polynomial with a single knot at 0 degrees latitude. This approach allowed us to determine distinct distribution patterns of non-native plants in both the northern and southern hemispheres (Guo et al. 2021). To create the cubic b-splines matrix, we employed the “bspline” function from the “splines2” package (v0.5.0; Wang and Yan 2021).

To analyze the similarities between the non-native plant species communities in each Andean country, a Jaccard similarity index was calculated using the function “vegdist” using the beta diversity obtained with the function “betadiver” (both from the “vegan” package). The significance level of the dendrogram was calculated based on the 95% of the pseudo similarity values obtained using bootstrapping with 10,000 iterations (v2.6-4Oksanen 2009).

All graphs were created with the package “ggplot2” (v3.4.1; Wickham 2008).

Results

The literature review indicated that for the Páramo, only Venezuela (8°N) and Ecuador (0°S) had records of non-native plants along elevation gradients. For the Puna, records were found only in Bolivia (16°S). In contrast, for the Southern Andean Steppe, the majority of records were concentrated in Chile, particularly in central Chile (37°S) and southern Chile (36–39°S) and subantarctic Chile (53°S).

Our survey of non-native plants indicated that the Páramo Andes host 42 non-native plants of which 29% are from the Poaceae family, 12% from the Asteraceae family and the remaining 59% from 16 other families (Table S1, Fig S1).

In the Puna, a total of 60 non-native plants were recorded, 25% from the Poaceae family, 18% from the Fabaceae family and 8% from the Rosaceae family; the remaining 49% came from 22 families (Table S1, Fig S1).

Finally, for the Southern Andean steppe, 99 non-native plants were recorded, of which 21% are from the Asteraceae family, 19% from the Poaceae family and 11% from the Fabaceae family, with the remaining 49% from 22 families (Table S1, Fig S1). Only four of the non-native plants were present in all three biogeographic zones: Dactylis glomerata (Poaceae), Plantago lanceolata (Plantaginaceae), Taraxacum officinale (Asteraceae) and Rumex acetosella (Polygonaceae) (Table 1). Twenty species were present in at least two of the biogeographic zones (Table 1). Of the 20 most common species, 35% belonged to the Poaceae family and 15% were from the Fabaceae family, and 75% have an European origin (Table 1). In the páramo, 62% of species have a forb growth form, 28% are graminoids and 4% are shrubs (Table S1). In the puna, 41% are forbs, 25% are graminoids, 18% are trees and 15% are shrubs (Table S1). Finally, for the Southern Andean steppe, 68% are forbs, 19% are graminoids, 7% are shrubs and 6% are trees (Table S1).

Table 1 The 20 most common non-native plant species (in Species name alphabetical order) present in the Andes mountain range, their growth form, biogeographic region of origin and countries of occurrence. For the complete list of species present in the Andes mountains, see Table S1
Full size table

Jaccard’s similarity analysis identified three groups of non-native plant communities: the first consisting of Venezuela and Bolivia, the second of Ecuador, Argentina and Central Chile, and the third of South and subantarctic Chile (Fig. 2).

Fig. 2
figure 2

Grouping of Jaccard similarity measures of the non-native plant communities among the main countries and regions along the Andes mountain range. Páramo = Ecuador and Venezuela; Puna = Bolivia, Southern Andean Steppe = Argentina, Central Chile, South Chile and Subantarctic Chile

Full size image

Within each of the three biogeographic zones of the Andes, we found a reduction in the richness of non-native plant species with increasing elevation, however with a strong latitudinal interaction. Non-native plant species richness peaked in the mid-latitudes, at around 39°S (driven by high species richness in Chile and Argentina in the northern half of the Southern Andean steppe, Fig. 3A, C; Table 2). Both in the páramoand the southern Andes steppe (mainly in south Chile), we observed a greater richness of non-native species growing above the treeline, than in the puna (Fig. 3A). In the páramo, a weak decrease in the richness of non-native plants at low elevations (< 1,500 m.a.s. l) was also observed (Fig. 3A).

Fig. 3
figure 3

 A, B, C. (A) Patterns in non-native plant species richness along latitudinal and elevational gradients in the Andes Mountains, separately by biogeographic zone: Páramo (top, Venezuela and Ecuador), Puna (middle, Bolivia), Southern Andean Steppe (bottom, Argentina, central Chile, South Andes, subantarctic Chile). The vertical lines represent the elevation range of the tree line in each of the studied sites. In the páramo, the treeline is situated between 3,400-3,600 m a.s.l. In the Puna, it ranges from 3,700 to 3,900 m a.s.l. Lastly, in the southern andean steppe, the treeline occurs at elevations between 1,600–1900 m a.s.l. The vertical lines correspond to the elevational range of the tree line for each of the studied sites,. (B) The location of each region for which data have been included, colored as in A) presented on a background map of South America showing elevation (m a.s.l.). (C) Latitudinal pattern of non-native plants richness in Andean ecosystems 8°N to -55°S. For model coefficients of the associated generalized linear mixed models, see Table 2

Full size image
Table 2 Estimates, standard errors and significance (P-values) of the effect of elevation, latitude, quadratic equations of elevation (Quad Elev) and latitude (Quad Lati) and their interactions on the richness of non-native plants, based on a Generalized Linear Mixed Model. All estimates are significant (P < 0.05)
Full size table
Table 3 Examples of reported impacts of non-native plant species on mountain ecosystems in the Andes, based on a literature review
Full size table

Discussion

Plant invasions in the andean region mirrors global mountain patterns

The overarching patterns of invasion found in the Andes resemble those observed across the globe. Indeed, the Poaceae, Asteraceae and Fabaceae families were identified as the most common non-native families globally, while more than 63.9% of non-native species globally have an origin in Europe compared to 70% found in our study in the Andes (Alexander et al. 2016). The results of the Jaccard´s similarity index are similar to the results obtained by Seipel et al. (2012), showing that plant community similarities between regions may be driven by similarities in climatic conditions. The grouping formed by Bolivia and Venezuela may result from the bioclimatic similarities associated with an orotropical biome, which could determine the presence of similar non-native plants (Rivas-Martínez 2010; Arroyo and Cavieres 2013). The same occurs in the grouping of Ecuador, Argentina, and central Chile, where the predominant dry season may function as an environmental filter, selecting the most drought-tolerant non-native plants (Rivas-Martínez 2010; Deil et al. 2011; Sandoya et al. 2017). Finally, the grouping observed between southern Chile and Subantarctic Chile is likely explained by the climatic similarities in terms of temperature (with annual averages of 0–5 °C) and precipitation, ranging from 717 mm in southern Chile to 1350 mm in Subantarctic Chile. Similar to other areas in the Andes, this can function as an environmental filter that favors non-native plants adapted to these low temperatures (Rivas-Martínez 2010; Arroyo and Cavieres 2013).

The observed declines along the elevational gradient in the Andes is in agreement with the observed global patterns of non-native plant species richness, where the highest richness of non-native plants is often found at low elevations (Pauchard et al. 2009; Alexander et al. 2011; McDougall et al. 2018). These patterns have been explained by the increasingly extreme climatic conditions and decreasing anthropogenic disturbance towards higher elevations (Pauchard et al. 2009; Alexander et al. 2011; Marini et al. 2011; Lembrechts et al. 2016). An exception often observed to the decline of non-native richness with elevation occurs in tropical regions, where non-native plant richness usually peaks at intermediate elevations due to high temperatures and drought in low elevation zones (Arévalo et al. 2005; Seipel et al. 2012).This pattern was also observed in this study in the páramo, where a (weak) peak in abundance occurs at intermediate altitudes (Fig. 3A).

The decrease in non-native plant richness at higher altitudes can be attributed to the extreme climatic conditions in that region (Rew et al. 2020). On the other hand, in lower altitude zones, the high diversity of native plants may act as a biotic filter, preventing the establishment of non-native species, which explains the low richness in these areas. Meanwhile, the high number of non-native plants observed in mid-altitude zones (2500–3000 m above sea level) is associated with both the presence of human settlements, which facilitates their dispersal and establishment as well as moderate climatic conditions that resemble temperate conditions of most European generalist non-native species (Sandoya et al. 2017).

The higher non-native plant species richness at intermediate latitudes observed in this study coincides with the global pattern (Guo et al. 2021) where a greater richness of non-native plant species is found at mid latitudes 40°S (Guo et al. 2021). This pattern of peak in non-native plant species richness at intermediate latitudes is likely associated with the higher density of human populations and anthropogenic activities (concentration of agriculture, tourism, mining etc.), which results in a higher propagule pressure in these ecosystems (Pauchard and Alaback 2004; Barros et al. 2015; Guo et al. 2021). On the other hand, the decline in richness of non-native plants in the páramo can be explained by biotic resilience resulting from the high native species richness per unit area (Guo et al. 2021); rapid ecosystem recovery after disturbances due to high diversity (Guo et al. 2021); lack of functional traits associated with shade tolerance (Fine 2002); and low colonization rate (low propagule pressure) (Fine 2002; New et al. 2007). In the Southern Andes (above 50°S), the lower richness may be associated with extreme climatic conditions, low population density and fewer anthropogenic activities (Pauchard et al. 2016; Guo et al. 2021).

Both the elevational and latitudinal distribution patterns can depend substantially on sampling effort, which is lower in the páramo and puna than in the South Andean steppe. Therefore, the observed distribution patterns of non-native plants, primarily between latitudes 11°N and 27°S, should be treated with caution, as there is a significant information gap regarding the richness of non-native plants along elevation gradients, particularly in countries like Colombia and Peru. Increasing the sampling intensity and coverage of elevation gradients in the páramo and puna could reveal a higher richness of non-native plants in these regions, potentially altering the current distribution patterns of non-native plants in the Andes.

Factors determining invasion: the importance of the human footprint in the Andes

In general, most invasions reported from the Andean region are highly associated with anthropogenic factors, such as forest-agricultural areas in the páramo and areas used for tourism and livestock grazing in the Southern Andes steppe (e.g., Barros et al. 2015). As in any other biome, human disturbance in mountains can affect the three other major constraints for plant invasions: abiotic conditions, biotic interactions and propagule pressure (Colautti et al. 2006; Pauchard et al. 2009), with multiple types of disturbance often favoring generalist species (Barros and Pickering 2014; Vásquez et al. 2015; Sandoya et al. 2017).

Recent decades have seen a progressive increase in the intensity of human disturbances in mountainous areas of South America resulting in a loss of pristine areas (Barros et al. 2014). Consequently, pristine areas are increasingly restricted to protected areas, which are often surrounded by agriculture and urban development (Fuentes-Lillo et al. 2021). Protected areas in the Ecuadorian and Colombian Andes, for example, are mainly located around population centers (Bax and Francesconi 2019). Furthermore, the anthropogenic development of mountainous areas in the Andes is associated with an increase in tourist activities within protected areas and their surroundings, including hotels, ski resorts and vacation homes (Barros and Pickering 2014; Barros et al. 2015).

Roads act as one of the main drivers of invasion into the Andes (Pauchard et al. 2009; Seipel et al. 2012). Roads connect agriculture, urban areas and tourism activities, turning them into key drivers for the introduction of non-native plants to these ecosystems and into protected areas (Pauchard and Alaback 2004; Barros and Pickering 2014). Dispersal mechanisms include transporting seeds attached to car tires, clothing and domestic livestock (Ansong and Pickering 2013). Roads also cause altered soil properties, including higher nutrient levels, changes in soil pH, increased drainage and more extreme microclimatic conditions (Forman and Alexander 1998; Müllerová et al. 2011; Alexander et al. 2016). Finally, roads can reduce competition from native species and thus favor non-native plant establishment through niche expansion (Bolnick et al. 2010; Lembrechts et al. 2014, 2016). The facilitative effects of roads towards plant invasions have been documented for several roads in South America, including in Ecuador (Sandoya et al. 2017), Bolivia (Fernández-Murillo et al. 2015), the Souther Andean steppe (Argentina and Chile) (Haider et al. 2018), the South-central Chile (Pauchard and Alaback 2004; McDougall et al. 2011; Haider et al. 2018) and Argentina (Haider et al. 2018; McDougall et al. 2018). A similar, yet smaller, impact has been observed for hiking trails in Chile, which to a some extent also increase propagule pressure, alter abiotic conditions and reduce competitive interactions (Liedtke et al. 2020).

In the central Andes of Peru, an example of the role of disturbance is the establishment of the Pennisetum clandestinum, a non-native African grass, associated with the elimination of vegetation above 3800 m.a.s.l caused by the introduction of cattle (Urbina and Benavides 2015). The introduction of cattle in the Peruvian Andes also facilitated the establishment of Pinus patula and Pinus radiata species at 4000 m.a.s.l (Raboin and Posner 2012). In the Bolivian Andes, the establishment of the non-native tree Eucalyptus globulus and the grasses Sorghun halepense and Cynodon lemfuensis has been favored by fire disturbance (Thomas et al. 2010). In the Central Andes of Chile and Argentina, the presence of non-native herbs Taraxacum officinale, Lactuca sativa, Rumex acetosella and Convolvulus arvensis are associated with the disturbance caused by tourism (formation of informal trails) and cattle (Muñoz and Cavieres 2008; Barros and Pickering 2014). In the southern Andes, the establishment of the evergreen shrub species Ulex europaeus is associated with fragmentation of the Andean forests (Altamirano et al. 2016). Similarly, the invasive pine Pinus contorta was favored by anthropogenic disturbance, including fragmentation and anthropogenic land use (Franzese et al. 2017).

In the Andes, most of the protected areas are located at intermediate elevation (~ 1000–2000 m a.s.l) (Elsen et al. 2018), while anthropogenic activities (agricultural, livestock, tourism and mining activities) are increasingly affecting higher elevation areas, generating higher propagule pressure, changes in abiotic conditions and reduction of biotic resistance across the whole elevation gradient (Barros et al. 2015; Pauchard et al. 2016; Fuentes-Lillo and Pauchard 2019). It is thus necessary to generate biosecurity protocols to avoid the arrival of new species and develop plans and mechanisms to increase the extent of protected areas at higher elevations, especially given that these areas have a great taxonomic diversity and high levels of endemism (Ansong and Pickering 2013).

Factors determining invasion: is climate a limiting factor?

Mountains are ecosystems with steep gradients of abiotic factors such as temperature, precipitation, UV radiation, nutrient availability and growing season length, which have functioned as an abiotic filter preventing the establishment of non-native plants in higher elevations (Pauchard et al. 2009). The invasion process at high elevations and latitudes is thus limited to non-native plants that are pre-adapted to extreme climatic conditions or show very flexible climate niches (Alexander et al. 2011; Lembrechts et al. 2016). These climatic limitations supposedly largely drive the characteristic decline in non-native plants with elevation observed in mountain systems worldwide (Alexander et al. 2011; Seipel et al. 2012; McDougall et al. 2018). However, in the Andes we observed that a large number of non-native plants managed to grow above the treeline (Fig. 3A). We speculate that these occurences could be explained by the presence of anthropogenic disturbances (i.e., roads, human settlements) that modify abiotic conditions (Lembrechts et al. 2016; Fuentes-Lillo et al. 2021). Some examples of non-native plants that grow above the tree line in the Andes include Acacia dealbata, Lupinus polyphyllus, Verbascum virgatum and Echium vulgare, which grow in the highest elevation areas above the tree limit (Fig. 1).

It is important to note that these non-native plants, in addition to growing above the treeline, are subjected to low temperatures, high radiation, extreme drought, and nutrient-poor, underdeveloped soils (primarily volcanic) (Fuentes-Lillo et al. 2021). Therefore, the presence of anthropogenic activities which modify soil conditions (nutrient addition) and increase water availability may explain these patterns. For example, a correlation has been found between the nutrient content (nitrogen, phosphorus and potassium) and the presence of T. officinale in the Central Andes, where nutrient-poor soils would be key to limiting the establishment of this non-native plants (Quiroz et al. 2011). In the Peruvian Andes, water availability was identified as the main limitation for the establishment of non-native plants above 3,700 m.a.s.l (Thomas et al. 2010). In the Colombian Andes, low water availability and low nutrient levels were key to limiting the establishment of the non-native species Anthoxanthum odoratum, Hypochaeris raditaca and Holcus lanatus above 3,500 m.a.s.l (Valencia et al. 2013). Other abiotic conditions might also be affected by disturbance (Lembrechts et al. 2016).

Climate change: predicted effects on the patterns of non-native plants in the Andes

The effects of climate change are less well documented in the Andes than in other mountain areas of the world (Pauchard et al. 2016; Fuentes-Lillo and Pauchard 2019). Nevertheless, current climatic models for the Andes indicate a warming rate of approximately 0.34 °C per decade, similar to what is forecasted for both the Arctic and the Antarctic, and higher than the global average (Bozkurt et al. 2017). These higher temperatures are likely to be accompanied by increased precipitation in the páramo and by complex changes in precipitation in the puna and southern andean steppe, for both rain and snow (Bozkurt et al. 2017b). These changes in temperature and precipitation will directly affect species redistribution as they track their climatic niche (Pauchard et al. 2016; Petitpierre et al. 2016; Dainese et al. 2017). For the páramo, several studies have determined that increases in temperature can move species distribution limits upwards in the mountains. For example, it is expected that every 3 °C temperature increase will result in upward shift of 600 m in species range limits (Arzac et al. 2011; Bramer et al. 2018; Rew et al. 2020). In addition, climate change will have major indirect impacts on Andean ecosystems. For example, predicted increases in annual precipitation levels will result in increased erosion processes and land displacement that may favor the establishment of non-native plants over the native community (Ruiz et al. 2008; Ansong and Pickering 2013). For the southern andean steppe, it has been predicted that climate change may increase drought events, which would favor the establishment of non-native species adapted to drought conditions (Molina-Montenegro et al. 2011).

The warming climate will directly affectthe upward expansion of non-native plants (Petitpierre et al. 2016). This change in the elevational patterns is mainly associated with an improvement in abiotic conditions as well as a significant improvement in dispersal patterns, associated with higher seed production, which significantly increases the probability of establishment of non-native plants (Teller et al. 2016). In the páramo Andes (e.g., Colombian and Peruvian Andes), the expected increased precipitation might also remove one of the main limitations for the establishment of non-native plants in the region (Thomas et al. 2010; Valencia et al. 2013).

Forthe ecosystems of the Andes, we are aware of only one study that evaluates the future distribution of non-native plants along elevational and latitudinal gradients. This study models the potential niche of Taraxacum officinale and Ulex europeus under a climate change scenario in 2050, predicting an expansion of the potential niche, generating the redistribution of these species from the valleys towards higher elevation areas, especially for the Andes of Chile and Argentina (Hernández-Lambraño et al. 2017).

Considering that the redistribution of non-native plants could be directed towards highlands and high latitudes due to the projected increase in temperatures and precipitation in the Andes, generating more experimental studies (e.g., use of open top chambers (OTC) coupled with modeling (Species Distribution Modeling and Ecological Niche Modeling) for different non-native plants is necessary to fully understand how future climate scenarios will affect the patterns of distribution of non-native plants in the Andes (Ebeling et al. 2008; Bellard et al. 2013). The results of these experimental studies and modeling will allow for the identification of future areas occupied by non-native plants, which will help to generate biosecurity protocols to manage, control and prevent future expansion of non-native species (Faulkner et al. 2020). In this context, generating hybrid (i.e., mechanistic and correlative) models that include microclimatic (high-resolution data) and anthropogenic variables could provide a more realistic approximation of the future distribution of non-native plants (Lenoir et al. 2017; Lembrechts et al. 2020).

Impacts

The impact of non-native plants on the ecosystems of the Andes has been poorly studied in comparison to other alpine ecosystems (Alexander et al. 2016) and other regions in the southern hemisphere, such as South Africa (van Wilgen 2012), Australia (Alexander et al. 2016) and New Zealand (Tomiolo et al. 2016).

Most of the studies in the Andes pertain to non-native trees that invade large areas (e.g., Pinus radiata, P. patula, P. contorta) (Table 3). These studies are often focused on evaluating the effect of non-native species on the reduction of biodiversity, due to changes in the richness and abundance of native species, effects on pollination regimes or biotic homogenization (Table 3). It is important to note that during the last 10 years, only 10 publications reported on the impact of non-native plants in the Andes, with the majority belonging to the páramo (Fuentes-Lillo and Pauchard 2019).

Prevention and management of invasive plants in the Andes: do we need a common policy framework?

Although studies have shown that, invasive non-native plants are a major threat to biodiversity and ecosystem services in South America (Table 3), the concern of the state and citizens regarding invasive plants is lower than in other continents such as Oceania, Europe, or North America (Speziale et al. 2012; Pyšek et al. 2020). The magnitude of other threats to biodiversity in South America (IPBES, 2018) may contribute to the underestimation of the role of invasive species and particularly plants in the biodiversity crisis. More than two decades ago, most South American countries agreed to create an international committee to address climate change and ratified the United National Framework Convention on Climate Change (Dimitrov 2010); Comisión Europea, 2019). In addition, all countries ratified the Convention on Biological Diversity (CBD), which includes an explicit mandate to prevent, control, and eradicate non-native plants that pose a threat to local biodiversity. Although most countries where the Andes occur are signatories to various biodiversity conventions and sustainable development strategies, public policies in South American countries related to the prevention and control of invasive species is poor, and reflects the low political and social understanding and interest. Research on invasive species and most efforts to control invasive plants are directed toward species that are economic pests or have socioeconomic impacts (Speziale et al. 2012; Aizen et al. 2019; Fuentes-Lillo and Pauchard 2019).

There are, however, important differencess among South America countries in the policies to tackle invasive species. For example, Ecuador has a legal basis that supports the management of non-native species, mainly in the Convention on Biological Diversity (CBD) article 8-h that promotes the prevention, control and eradication of non-native species. Additionally, Ecuador presents a strategic plan that seeks to identify pathways of introduction, prioritize non-native species and reduce the impacts of non-native species in both terrestrial and aquatic ecosystems by 2019 (MAATE 2019). During the last decades, Colombia has made significant progress in protecting ecosystems from invasive species, moving from a national code of renewable natural resources (CNRN law 2811 of 1974) that allowed the introduction of non-native species to the country with a special permit from the state, to a management plan for the prevention and management of introduced, transplanted and invasive species that aims to prevent, identify routes of introduction and reduce the impacts of non-native species (MADS, 2011). Both Peru and Bolivia have national action plans on non-native species led by the Ministry of Environment (MINAM) in Peru and the Ministry of Environment and Water in Bolivia (MMAYA). These plans seek to coordinate technical groups to evaluate the risk of non-native species, identify routes of introduction and reduce the impact of these species on ecosystem services, and have goals to be met by 2035.

In Argentina, the Ministry of Environment and Sustainable Development has a national strategy on invasive non-native species (ENEEI) that aims to minimize the impact on national resources, biodiversity, ecosystem services, the economy, public health and culture. This national strategy seeks to develop a framework to strengthen governance and effective protection of biodiversity, enhance the socioeconomic benefits that include natural resources and ecosystem services, and promote research on the impacts of invasive species to generate effective public policies to control their impact.

Finally, Chile has a national biodiversity strategy with a vision for the year 2030 that involves the objectives of the CBD. Within this national strategy, the aim is to promote good practices in the agricultural and forestry sector sector associated with the use of non-native species, in order to prevent the introduction, release and dispersal of invasive non-native species and/or potentially invasive non-native species to the natural environment. The strategic objectives also include promoting basic and applied research on invasive species, mainly on prevention, control and/or eradication and restoration mechanisms, increasing awareness and information to the public on invasive species and developing environmental education strategies related to biological invasions and their effects on national biodiversity. It is expected that by the 2030s that efficient biosecurity protocols will be implemented, with prioritization of non-native species, identification of introduction routes and finally at least ten restoration plans in areas that have been invaded by non-native species (MMA 2017).

Unfortunately, aside from variation in national public policies and regulations there is additional variation in the enforcement of these instruments across regions and taxa (Aizen et al. 2019; Comisión Europea, 2019). Most public policies, national strategies and regulatory frameworks do not apply specifically to the Andes Mountains, but these ecosystems are equally represented by each of these regulatory systems associated with each country that includes the Andes mountains (Aizen et al. 2019). Because these ecosystems have a high level of endemism, are important biodiversity hotspots, and that protected areas protect an inadequate proportion of the Andes (Castillo et al. 2020), it is necessary to generate specific public policies for these ecosystems. Nonetheless, for most environmental agencies, mountain environments are low priority in terms of prevention and control of invasive species, including because of the perception that the rugged terrain and inaccessibility of the Andes mountains makes them seen as less prone to invasions.

This review has highlighted that it is necessary to move towards a regulatory framework or regional strategy, including all the countries that make up the Andes, to develop, standardize and prioritize policies to identify introduction routes, prioritize non-native species for their control, and reduce the drivers causing non-native species introduction and spread. A common multinational framework across the Andes could reduce the regional impact of non-native speciesin the region Generating bilateral and multinational strategies to prevent the spread and enhance the management of invasive non-native plants in the Andes is critical. These strategies should move towards implementing joint risk assessments before introducing any non-native species, as European countries have adopted (Aizen et al. 2019). An example in the region includes the coordinated efforts between Chile and Argentina to create a bilateral strategy for vertebrates such as for the control of the beaver (Castor canadensis) in Tierra del Fuego in southern Patagonia (Sanguinetti et al. 2014; Aizen et al. 2019).

No less important for the prevention and management of invasive plants in the Andes is the extremely limited “on the ground” expertise and experience about control methods for these species. Although some agencies have conducted experimental initiatives, these efforts have been limited and usually confined to specific taxa. For example, for invasive tree species a few studies have been conducted for Pinus spp. control in Chile and Argentina (Pauchard et al. 2016), and for Ligustrum lucidum (glossy privet) in Argentina (Valfré-Giorello et al. 2019), but the amount of research and scope of operational scale actions are considerably lower than countries such as New Zealand (Edwards et al. 2021) and Australia (Pyšek et al. 2020). Forshrub species the situation is similar, with some effort on a few species receiving most of the attention as with Ulex europaeus control in Colombia (Gómez-Ruiz et al. 2013) and Chile (Norambuena et al. 2000), while for herb species there is no reported experience in control.

Conclusions

Our review challenges the paradigm that extreme climatic conditions are the main limitation for the redistribution of non-native plants in the Andes, instead shifting the focus to the role of anthropogenic factors on the redistribution of non-native plants, both at latitudinal and elevational extremes. An improved understanding of the effect of anthropogenic factors on invasions in the region will assist to reduce the arrival of non-native plants in the higher elevation areas.

Given that climate change may have implications for altitudinal and latitudinal redistribution (i.e., shifting towards higher latitudes), it is crucial to enhance the study of the climate change effect on the patterns of redistribution of non-native plants. This will provide valuable information for developing biosecurity protocols and efficient management plans to mitigate future impacts of non-native plants on mountain ecosystems and their native biodiversity.

Increasing the understanding of the impacts of non-native plants on mountain ecosystems and ecosystem services is of utmost importance, considering that it is one of the aspects of plant invasions in mountains that has received insufficient attention and the importance of the Andes for biodiversity conservation and as a source of important ecosystem services (e.g., water) for the people who inhabit or depend on these areas.

Unfortunately, even though invasive plant species have been recorded as an important threat to biodiversity in the Andes, especially under climate change scenarios, countries have not responded accordingly in terms of prevention and management policies. Future efforts must address the multinational nature of invasion processes. Coordinated policies across neighbouring countries will increase the chances for successful intervention. In this regard, research gaps on prevention and management tools must urgently be addressed. In addition, and to minimize the impacts from anthropogenic use that favors plant invasions, strengthening protected areas systems and improving management for biodiversity conservation is critical in the Andes.

Finally, we urge the scientific community to unify methodological studies that allow us to have a better understanding of the invasion process both along the elevational and latitudinal gradients of the Andean ecosystems, considering that the Andes functions as a natural laboratory to evaluate ecological processes related to biological invasions along the gradients of climate and anthropogenic disturbance inherently present in the system.