Effects of climate change on Andean biodiversity: a synthesis of studies published until 2015

Understanding how GCC is affecting the biological diversity of the Andes is of utmost importance and timeliness given the relevance of the region for biological conservation. Our research questions were (1) what is the spatial (country-level) and temporal distribution of the scientific research exploring links between GCC and biodiversity in the Andean region? (2) What are the methodological approximations, areas of research and subjects of study most commonly considered? And (3) What are the trends in biodiversity responses most commonly found under different GCC stressors? We found that the first paper on GCC and biodiversity in the Andes was published in 2001. Since then the annual rate of publications, as well as the variety of areas of research, has risen steeply. The 65 published articles we found are likely to represent 1% of the scientific literature dealing with tropical ecology. Of those, more than half of the studies were conducted in a single country, used mostly observation rather than modelling or experimental methodological approaches, and focused mainly on plants. Studies dealing with birds, mammals and reptiles were notoriously underrepresented. The high number of GCC stressors and the great variety of responses found in this synthesis makes it difficult to draw general conclusions. However, we found that observational, modelling and experimental studies report negative GCC impacts on the biological diversity of the region. Most generally, observation and modelling studies report contractions of the distribution ranges of Andean species, and negative effects on species population densities and individual performance. We conclude our review suggesting that networking, recovering historic field data and conducting large-scale ecosystem experimental studies are critical to improve our knowledge on the effects of GCC on Andean biodiversity.


Introduction
Global climate change (GCC) comprises environmental and ecological changes that affect ecological systems, including human populations worldwide. [1] Although global warming is by far the most widely recognized GCC stressor, increases of the atmospheric concentrations of CO 2 , nitrogen deposition and land-use land-cover change (LULCC), also have strong effects on biological diversity and ecosystem functioning at global scales. [1] GCC stressors do not act in isolation, but interactively, making their overall impacts on biological diversity and ecosystem functioning difficult to disentangle.
The Andes mountain chain is critical for biodiversity conservation at the global scale. [2] Indeed, the Tropical Andes are considered the most important of all biodiversity hotspots, as it holds the highest variety of species of plants, amphibians, birds and mammals. [2] The great diversity of plants and animals supply medicines, food, fibres, construction materials and decorative and cultural crafts to local human populations. [3][4][5][6] It is estimated that people in an Andean country use at least 25% of the plant species recorded for the country. [7] The natural vegetation of the Andes is home to the wild relatives of plants that are global staple foods, including potatoes, tomatoes and quinoa. [8] This region also delivers key ecosystem services (i.e., water provision, carbon sequestration) to more than 50 million people living in or near the Andean mountain chain. [9] Thus, understanding how ongoing GCC is affecting the biological diversity of the Andes is of the utmost importance and timeliness. [10] Across the globe, mountain regions are particularly vulnerable to GCC. Mountain habitats are reduced in size, experience a high geographic isolation and their environmental conditions (i.e., temperature, humidity) change considerable across small distances due to steep slopes. Evidence of GCC impacts on mountain systems has started to accumulate. For example, plant communities across European mountains have shown signs of thermal migration that include the decline of populations of cold-adapted species, and the increase of warm-adapted ones. [11,12] Andean ecosystems are also responding to GCC. For example, the Northern Andes, a critical area for the conservation of global biodiversity, is expected to lose as *Corresponding author. Email: selenebae@gmail.com many as 2000 endemic species as a consequence of GCC. [13] Species of trees in the Andes are shrinking their areas of distribution, [14] migrating upwards, [15] or experiencing higher rates of mortality and possibly of biomass loss. [16] Thus, GCC will dramatically impact the biological diversity and ecosystem functioning of the whole region. Moreover, Andean ecosystems have experienced strong reductions of their original area due to historic and present human activities. GCC is shifting the environmental conditions of the Andes at an unprecedented rate. The North and Central regions of the Andean range will experience climatic departure (i.e. strong divergence from historic climatic conditions) as soon as 2020. [17] In this part of the Andes, near-surface temperatures are expected to increase up to 5°C by the end of this century, and will generate spatially heterogeneous changes in precipitation. [18][19][20] Monitoring the volume and geographic distribution of scientific production in the field of ecology is important to evaluate the efficiency of the funds allocated by government and non-government institutions interested in promoting research and conservation of tropical areas. [21] Indeed, previous research of this type indicates that the amount of biological research in the Andean region is extremely limited compared to that conducted in the Amazonia and Central America. [21] Thus, providing a general view of the type of research, geographical distribution, and general responses of biological diversity to GCC is helpful to identify general trends of biological research. Identifying gaps on the patterns and rates of scientific production in the Andes may help to guide the interest of scientists and funding agencies in the region. In this study, we summarize the temporal trends of published research investigating the effects of GCC on the biodiversity of the Andean region up to 2015. We know of no previous synthesis of published information on this topic for the region. As such we focus our review in basic patterns presented by the scientific literature. Our research questions are, (1) what is the spatial (countrylevel) and temporal distribution of the scientific research exploring links between GCC and biodiversity in the Andean region? (2) What are the methodological approximations, areas of research and subjects of study most commonly considered? And (3) what are the trends in biodiversity responses most commonly found under different GCC stressors? We provide a review of published literature on the effects of GCC in the region, and hope to motivate scientists to accelerate research on this topic.

Methods
We began our synthesis by first searching in the Science Citation Index Expanded (Web of Science) for the period between April 1975 (the earliest date in Web of Science we had access to) and December 2015 using the following quotation: (climate change OR global warming OR global change) AND Ande* AND (species OR species richness OR species diversity OR biodiversity). We then selected articles that specifically addressed GCC effects on species diversity or organismal performance; however, we did not consider studies dealing with paleo-biology. We extracted the following information from each article: geographic scope, year of publication, type of study according to the main method used to assess the effects of GCC on species biodiversity (i.e. observation, experimental or modelling), area of research (e.g. biodiversity, extinction, physiology), type of organism studied (e.g. plants, animals, diversity in general) and the GCC stressor(s) (e.g. temperature increase, land-use cover change, etc.) driving the biological response investigated. For each study, the biological responses were categorized as increase, decrease or no change.
We used this information to analyze geographic and temporal trends of research production. We also discuss what type of organisms was more frequently investigated. Finally, we discuss general patterns of biodiversity responses to GCC by type of study: observation, experimental or modelling.

Geographic and temporal distribution of studies
We found 65 articles that investigated the effects of GCC on organismal performance or species diversity in the Andes (Table 1). Thirty-nine of these studies were conducted in a single country, mainly in Ecuador, Peru and Chile, whereas 26 studies spanned two or more countries ( Figure 1). The countries with the highest total number of studies were Ecuador (n = 31) and Peru (n = 30), whereas Argentina, Venezuela and Bolivia had about half the studies (n = 13, 13 and 15, respectively). Only three articles focused in the Andean region as a whole.
The first study to investigate GCC in the Andes was

Methods, areas of research and subjects of study
Observation studies were by far the most commonly used approach to investigate the effects of GCC in species diversity. Most of the studies (n = 37) used methods based on field observations. About a quarter of the articles (n = 19) used modelling techniques, and the rest (n = 10) used an experimental approach (Figure 2). We identified five main areas of research; the areas of 'Biodiversity' and 'Spatial distributions of species' were the focus of most studies, with 15 and 25 papers, respectively ( Table 2). It is noteworthy that the representation of the areas of research has changed through time (Figure 1). The early research on GCC and diversity in the Andes focused on general aspects of species distribution, biodiversity conservation and species extinction, mainly amphibian's species loss due to fungal infections (Figures 2 and 3, Table 2). Later on, the areas of study became more diverse, and had an increased representation of studies exploring physiology and forest dynamics.
With regard to the focus of the studies on different types of organisms, we found that most frequently research investigated the effects of GCC in species of plants (32 studies) compared to studies that considered species of animals or both types of organisms simultaneously (Figure 3(a)). Amphibians represented the most studied group of vertebrates, followed by insects and aquatic invertebrates (Figure 3(b)), whereas other groups of vertebrates, including birds, mammals and reptiles, were virtually excluded.

Study types and trends of biodiversity response to GCC stressors
Observation studies explored the variation of 10 GCC stressors on more than 50 types of responses of biological diversity (Figure 4). Most of the papers explored how biodiversity or organismal biology was affected by higher environmental temperature (n = 16) and GCC variables considered simultaneously (n = 13). The GCC stressors studied generally affected biological diversity or organismal performance in one way or another, very few studies reported no effects. The studies often reported changes in the spatial distribution of the species, decreases of the population growth rates and altered rates of plant growth and seedling establishment ( Figure 5).
Of the 19 modelling studies, 18 addressed the effects of all GCC stressors simultaneously on biodiversity (Figure 5). These studies focused on predicting how GCC will affect the spatial distribution of plant species, and the species diversity in general. Most of the studies predicted upward shifts and range contractions of species distributions, and decreased spatial ranges of species occupancy. Further, various studies predicted decreased species diversity under GCC.
The experimental studies largely focused on the effects of higher environmental temperatures ( Figure 6). These studies explored the effects of higher environmental temperature or higher altitudinal distribution of species on plants or plant-animal interactions, and on plant physiological responses. It is interesting to note that these studies found decreased plant performance under experimental evaluations of GCC, but some studies reported no responses or changes in physiological activity or plant performance.

Discussion
Our findings indicate an uneven distribution of studies among Andean countries, where Ecuador and Peru (and to a lesser extent Chile) have a high volume of publications compared to other countries that remain virtually unexplored, mainly Bolivia ( Figure 1). Indeed, other studies point out that much biological research is being conducted in Ecuador and Colombia compared to other countries, and that Ecuador occupies the best studied country of South America, including the Andean region, once research production is standardized by area. [21] Thus, the scientific production on the area of GCC biology in the Andean region appears to respond in the same way that general ecological research does. It is important to note that most of the ecological research of these countries is produced mainly by foreign research groups working in focal areas of the Andes, and that it does not reflect the scientific production of local researchers. [21] A study exploring the extent of ecological research in the Andes and Amazon between 1995 and 2008 in two prominent ecological journals found that only 77 of 373 studies focused in the Andean region during this period, and estimated that the total biological publications for both regions were of 1540 scientific articles. [21] According to our search, 13 studies were produced between 2001 and 2008 ( Figure 2); therefore, roughly 1% of the ecological research conducted during that time period explored the effects of GCC and Table 2. Breakdown of effects of GCC according to the areas of study.
The high number of GCC stressors and the great variety of responses found in this synthesis makes it difficult to draw general conclusions (Figures 4-6). However, we found that observational, modelling and experimental studies report negative GCC impacts on the biological diversity of the region. Most generally, observation and modelling studies report contractions of the distribution ranges of Andean species, and negative effects on species population densities and individual performance (Figures 5 and 6). Further, interspecific interactions (e.g. disease) altered by GCC have already resulted in widespread extinction of amphibians in the Andes. However, it is worth noting that the responses of biodiversity were highly variable, particularly when environmental factors were isolated in experimental studies ( Figure 6).
Andean species are responding to GCC by migrating to higher altitudinal ranges. [14,15,22] Differences on the rates of species upslope migration are likely to affect species diversity, interspecific interactions and the ecosystem persistence per se. Recent research indicates that even highly mobile Andean species have narrow environmental envelopes; [23] thus even moderate changes in environmental conditions may threaten the persistence of less mobile high-altitude species. Differences in the tempo of migration may also disrupt interspecific interactions, which may threaten the persistence of the species (e.g. if a species lacks a critical mutualism in its reproductive cycle), or alter their relative densities in the community (e.g. if a species experiences higher or lower predation rates). [24,25] Thus, differences in migration rates and their resulting effects on species interactions are likely to result in hybrid and novel ecosystems in the Andes. [26] A critical aspect of research will be to determine the areas where these changes will take place, and what actions can be taken to enhance species migration considering landscape permeability that include urban development, agriculture, land abandonment and ecosystem restoration, LULUCC.

Directions for future research
Pitman et al. [21] present a list of insightful recommendations to improve the scientific production in the Andes and Amazon. In this section, we focus on specific approaches that can be adopted to improve substantially   Table 1; letters indicate organisms (a = animals, p = plants, f = fungi). Colours unify factors and responses. and over a short-to mid-period of time our knowledge about the effects of GCC on Andean biodiversity. All the proposed directions of research and collaboration would certainly be more helpful if accompanied by strong efforts on human and institutional capacity building.

Networking
Recent networking initiatives have proven very powerful to move forward the field of ecology in the context of GCC globally. [27][28][29][30] In the Andean region, the Andean forest network [31] has brought together more than 40 scientists working in the region. Most researchers have contributed their invaluable field data to build a large data-set that is currently being used to conduct studies that explore Andean forests responses to GCC (e.g. [14,16]). In the same way, scientists associated to the GLORIA network are using a core methodology to monitor biodiversity and climate change in high Andean summits. [32,33] Hence, support to ecological networks in the Andes can greatly improve the scientific production of the region.

Recovering historic field data
The general lack of studies of GCC in the Andes may be explained in part by the young history of science in the region where long-term studies are mostly lacking. Thus, historic ecological data can serve as the base to conduct ecological surveys that may give us insights about the types and rates of change of Andean biodiversity and ecosystems. For example, a recent study examines patterns of species, vegetation and LULCC in the Chimborazo volcano using historic data collected by Humboldt in 1802. [22] More recent data, including biodiversity surveys [34] and permanent plots established in the Andes in the last decades can also be very valuable to assess midterm trends of change of the plant community, at least in terms of species dominance and stand structure.

Conducting large-scale ecosystem experimental studies
It is urgent to explore the interactions between nutrient availability and temperature on biodiversity in Andean ecosystems. The recent incorporation of stoichiometry concepts to the study of thermal adaptation [35] suggests complex responses to GCC, where species physiological responses to increases in temperature will also depend on the biogeography of nutrient limitation. We found no articles connecting biodiversity with ecosystem scale effects of GCC, with regard to higher atmospheric CO 2 concentrations and/or nutrient availability. Hence, our capacity to predict how these critical aspects of GCC will affect the diversity of the region is very limited; some evidence suggests that Andean forests can be highly responsive to higher availability of soil N and P. [24,36] Further, it is uncertain how soil nutrient limitation will change in highly heterogeneous tropical montane forests under GCC. [37] Experimental and modelling studies could provide some answers to these issues. Although database literature searches are far from providing a complete view of the existing literature on a subject, we think that this study provides a general overview of the effects of GCC on the biological diversity of the Andes.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This work was supported by the Universidad Técnica Particular de Loja [PROY-CCNN-1034] and the Swiss Agency for Development and Cooperation (SDC).