A global perspective on tropical montane rivers

See allHide authors and affiliations

Science  13 Sep 2019:
Vol. 365, Issue 6458, pp. 1124-1129
DOI: 10.1126/science.aax1682


Tropical montane rivers (TMR) are born in tropical mountains, descend through montane forests, and feed major rivers, floodplains, and oceans. They are characterized by rapid temperature clines and varied flow disturbance regimes, both of which promote habitat heterogeneity, high biological diversity and endemism, and distinct organisms’ life-history adaptations. Production, transport, and processing of sediments, nutrients, and carbon are key ecosystem processes connecting high-elevation streams with lowland floodplains, in turn influencing soil fertility and biotic productivity downstream. TMR provide key ecosystem services to hundreds of millions of people in tropical nations. In light of existing human-induced disturbances, including climate change, TMR can be used as natural model systems to examine the effects of rapid changes in abiotic drivers and their influence on biodiversity and ecosystem function.

Tropical mountains have been objects of fascination for renowned naturalists including Alexander von Humboldt (1), Edward Whymper (2), and many others (3). Their work and travels along tropical elevational gradients laid the foundation for a growing body of knowledge documenting high biodiversity and endemism and a vast array of ecosystem functions and services. At the center of tropical montane ecosystems are complex networks of rivers, which support diverse biological communities, shape landscapes, connect mountaintops with lowland plains, and export extraordinary amounts of energy, nutrients, and sediment. Tropical montane rivers provide water, food, and energy that fuels regional economies and sustains hundreds of millions of people (4), yet they are some of Earth’s most inadequately studied ecosystems.

We define tropical montane rivers (TMR) as lotic freshwater ecosystems along elevational continua in the tropical regions of the world (Fig. 1A). Here, we compare tropical rivers and stream ecosystems along latitudinal and elevational gradients to depict how key abiotic variables change, resulting in the distinctive biological and ecological patterns found in TMR. Current understanding of TMR has been limited by two main factors. First, a terrestrial bias in tropical research has hindered recognition of TMR as distinct ecosystems rather than merely conduits of water from mountains to lowlands. Second, existing research has focused on the endpoints of TMR: small streams flowing through high-elevation zones (5) or rivers meandering through lowland plains (6). Consequently, research efforts have largely overlooked the rich biodiversity and the dynamic functioning of rivers running through mid-elevation ranges of tropical mountains, thus offering an incomplete picture of the nature and importance of these systems.

Fig. 1 Global distribution of mountains.

(A) Tropical, subtropical, and temperate mountains of the world (66). Comparisons between tropical, subtropical, and temperate mountains at 2.5-min resolution: (B) Annual mean temperature (67), (C) annual mean precipitation (67), and (D) rainfall erosivity (16).

In this Review, we take a global perspective on TMR (Fig. 1A). By highlighting mountains as structural and dynamic landscape elements (7), we identify the distinctive characteristics shaping them, provide selected global examples, examine how TMR are changing, and offer suggestions for future research. We emphasize the interconnectedness and continuous character of TMR, from their high-elevation origins to their discharge into lowland floodplains (and ultimately oceans), and their far-reaching influence on distant ecosystems and the humans that inhabit them.

What shapes a tropical montane river?

TMR occur at the intersection of tropical latitudes and elevational gradients, resulting in specific patterns of environmental factors, including temperature, growth period, precipitation, and flow disturbance regimes, which shape their biodiversity, ecological processes, and distinctive ecosystem services.

Temperature changes in predictable ways along latitudinal and elevational gradients (1). Warmer temperatures, associated with equatorial latitudes and lower elevations, usually promote higher diversity for most biological groups, although exceptions have been reported. In line with this pattern, TMR show increasing fish and invertebrate richness toward lower elevations (810).

What are the mechanisms behind these diversity patterns? The climate variability hypothesis (CVH), expanded by Janzen (3), states that organisms evolving under the steady temperatures of tropical zones should have weaker dispersal capabilities and thermal physiological adaptations than organisms in temperate zones that are exposed to large intra-annual temperature changes. Accordingly, we would expect higher species turnover rates along elevational gradients in tropical versus temperate regions. Recent studies (11) have corroborated the CVH (3) by showing that aquatic invertebrates along TMR have narrower thermal tolerances (11); reduced dispersal rates; and higher speciation (12), cryptic invertebrate diversity, and species turnover rates (13) than their temperate counterparts. This integrative, mechanistic, and comparative approach stresses the importance of studying mountain gradients—and specifically temperature gradients—as natural model systems to understand diversity across latitude and elevation.

Another feature of TMR are their particular disturbance regimes, defined by the frequency, intensity, and extent of extreme hydrologic events (14), landslides (15), and rainfall erosion (16). These forms of disturbance are more common in tropical regions as a result of the interaction of many environmental factors, including the amount, frequency, and seasonality of precipitation, geologic history, orographic relief, soil type, slope orientation, and riparian landcover (17). In the tropics, precipitation patterns are strongly influenced by the annual and inter-annual variations of the Intertropical Convergence Zone (ITCZ) and the resulting cyclones and monsoons (6). Although precipitation in these areas is highly variable (Fig. 1C), the percentage of high-intensity rainfall events (>25 mm/hour) is much greater in the humid tropics (>30%) than in most temperate regions (<5%) (18). This intense hydrological cycling results in complex flow patterns with frequent, short-duration floods, which modify streambeds by dislodging large boulders and transporting huge volumes of sediment and woody debris downstream.

The high frequency of landslides (15) and erosion (16) in TMR (Fig. 1D) is tied to geological activity, topography, and precipitation. The combination of a mountainous geological setting and intense precipitation, even in drier and more seasonal tropical regions (e.g., some western Andes TMR), results in higher frequencies of mass wasting events (15), which greatly contribute to channel modification and sediment, nutrient, and carbon transport downstream (18). The highest rates of sediment transport in TMR have been reported in tectonically active tropical mountains in New Guinea, where sediment loads can be as high as 10,000 tons per square kilometer per year (19). TMR are also responsible for the transport of large amounts of terrestrially derived carbon, contributing significantly to the global carbon cycle (17).


The movement and transport of materials along mountain gradients highlight the importance of TMR as longitudinal and lateral connectors from the highlands to lowlands and floodplains. For example, in the Amazon River system, ∼93% of all sediment and sediment-bound nutrients come from the Andean Cordillera (20), and a large portion of these are trapped and stored in the extensive floodplains of the basin (21). This transport and storage strongly influence soil fertility, maintenance of riparian habitats, aquatic primary production, fish yield (22), and protein provisioning to human populations (23). Similarly, seasonal flood pulses caused by Andean-Amazon longitudinal-lateral connections play fundamental roles in maintaining the fertility, heterogeneity, diversity, and productivity of lowland floodplain environments (24).

Biological diversity

Most TMR are located in Earth’s biodiversity hotspots (Table 1) (25) and support extraordinary biological diversity and endemism (26). Nevertheless, a comprehensive revision of most freshwater taxa in TMR is nonexistent. Freshwater fishes are the best-known taxa, exhibiting extremely high diversity and endemicity in many tropical mountains (2628). In Ecuador, for example, 45% of fish species from the western Andean slopes are reportedly endemic (29). Amphibian diversity in these regions is also particularly high (30), and aquatic invertebrates show high cryptic diversity along montane gradients (13). Nevertheless, these diversity patterns have been extensively studied in only a few specific locations, and the patterns could well be different elsewhere.

Table 1 Comparison of selected characteristics across tropical mountain systems.

Principal mountain systems where tropical mountain rivers occur and selected social, geological, and biological characteristics. The human population was estimated by masking the Gridded Population of the World 2015 V4 database [publicly available (68)] with mountain polygons from (66). Ga, billion years; n.a., not available.

View this table:

León River (1800 m)—a tropical montane river in the Andean-Amazon region—running through pristine montane forest after heavy rains, transporting significant amounts of sediment into the Pastaza Basin in Ecuador.


Defining features of TMR might also affect species’ life-history traits. Lotic organisms have rheophilic adaptations, but some of these might be enhanced in species from TMR. For instance, fishes from different tropical mountains exhibit convergent and parallel evolution in their morphological adaptations (e.g., reinforced skeletons, hyperossification, sucker mouths, fusiform shape, and attachment organs), evidencing evolutionary responses to harsh flow and frequent disturbance events (31). We hypothesize that intense levels of disturbance and short temporal windows of opportunity between frequent disturbance events would also affect behavioral and developmental life-history traits. Organisms along TMR should experience faster growth and shorter development periods, which promote multiple life cycles per year as elevation decreases (32), in contrast with organisms in temperate rivers. These traits should in turn favor smaller clutch and body size at lower elevations in tropical rivers. It has been shown that stoneflies (order Plecoptera) display larger body size at higher elevations in the Orinoco basin (33), supporting this window-of-opportunity hypothesis. Nevertheless, the importance of disturbance in driving organisms’ life-history patterns along TMR gradients has yet to be tested conclusively.

Where are tropical montane rivers located?

TMR are globally distributed between latitudes 23.5° north and south of the equator (Fig. 2). Spanning a large region, TMR are heterogeneous in terms of geological origins and biotic characteristics (Table 1), making them challenging to describe. As examples of their importance and distinctiveness, we highlight the main characteristics of three TMR systems that occur at global biodiversity hotspots (Fig. 2A) and that support numerous ecosystem services for large human populations (Table 1).

Fig. 2 Tropical mountain rivers in three selected regions of the world.

(A) World map showing selected regions (66): Tropical Andes of South America (B), Eastern Arc Mountains of Africa (C), and Western Ghats of India (D). Red dots indicate major cities (with human population > 250,000).

Tropical Andes

The tropical Andes (Fig. 2B) extend from Venezuela to Bolivia along western South America and comprise the headwater regions of some of Earth’s largest rivers, including the Magdalena, Orinoco, and Amazon. The uplift of the Andes over the past 50 million years (Ma), and particularly during the late-middle Miocene (~12 Ma) and early Pliocene (~4.5 Ma), affected the regional climate of South America and reconfigured the continent’s drainage patterns, giving rise to distinct fluvial systems (34), effectively creating the Amazon rainforest. Elevational gradients in the present-day tropical Andes extend upwards of 6000 m. Massive amounts of water, sediment, and nutrients are exported from the Andes along montane river corridors, influencing the ecology and biogeochemistry of the Pacific and Atlantic coasts of South America, the Amazonian floodplains, and the Caribbean lowlands of Colombia (35).

TMR in the Andes support extraordinary biological diversity, growing human populations (Table 1), and rich cultural traditions. The tropical Andes region is a global biodiversity hotspot for multiple taxonomic groups (25), but only recently was this diversity recognized for obligate freshwater taxa like fish (8), aquatic invertebrates (13, 36), and other freshwater biota (29). Within just the Andean headwaters of the Amazon, 671 freshwater fish (~40% endemic) have been described, which is equivalent to ~5% of global freshwater fish diversity. This number is likely to increase by 200 to 300 species in the next decade given current trends in taxonomic description (9). Similar trends are expected for other taxa, such as invertebrates, in which cryptic diversity is common (13). An estimated 100 million people inhabit the tropical Andes region, and TMR are the primary source of fish protein and water for domestic, electricity-generation, and agricultural uses (Table 1).

Eastern Arc Mountains

The Eastern Arc (Fig. 2C) is a chain of ancient crystalline mountains that were uplifted at least 30 Ma ago and exist as 13 distinct blocks extending from the Taita Hills in southern Kenya to the Udzungwa Mountains in south-central Tanzania (37). The Indian Ocean influences the climate of the Eastern Arc and the hydrology of rainfall-driven fluvial systems that drain the mountain blocks. East African TMR that originate in the Eastern Arc include the Pangani, Wami-Ruvu, and Rufiji, draining mostly to the Indian Ocean. The headwaters of these rivers supply water for riparian human populations and several Tanzanian cities, including the growing urban metropolis of Dar es Salaam, Tanzania (with a population of ~6 million), which gets upwards of 80% of its water from the Ruvu river basin, draining the Uluguru block (38).

The Eastern Arc is globally recognized for its biological importance, and numerous studies have examined its endemic plants and its terrestrial invertebrate and vertebrate groups (37); however, similar studies have not been carried out for freshwater species. This region’s extraordinary species richness and endemism is explained by mountain building processes (e.g., relief and soil erosion) resulting in high landscape heterogeneity and barriers for dispersal, which increase isolation and species evolution (7). These same theories could be applied to the regional patterns in fish (39), amphibian, and aquatic invertebrate diversity. At least three endemic Odonata (dragonflies and damselflies) reproduce exclusively in rivers of the Eastern Arc (40). Much of the freshwater diversity of this montane region remains to be discovered. For example, an ichthyofaunal survey of Eastern Arc rivers in 2007 documented 75 fish species, of which 11 were newly described (41).

Western Ghats

Spanning an area of 160,000 km2, the Western Ghats (Fig. 2D) range runs along the western coast of peninsular India (42). The range’s uplift along the faulted edge of fractured Gondwanaland ~65 Ma ago defined the region’s climate and drainage patterns (43). The range intercepts the southwestern monsoon winds, creating strong longitudinal and latitudinal precipitation gradients (42) that, in turn, influence biophysical and ecological processes across peninsular India. With an annual rainfall of 2000 to 8000 mm, the Western Ghats constitute the catchment of >30 west-flowing rivers and the headwaters of several large east-flowing rivers such as the Godavari, Krishna, and Cauvery, which provide water for an estimated 400 million people across peninsular India (44).

The Western Ghats is a distinct freshwater ecoregion (26) and the world’s most densely populated global biodiversity hotspot (25), where, on average, at least 300 people per square kilometer live alongside distinctive and endangered wildlife. Despite limited research, 320 species of freshwater fish have been identified in the Western Ghats, and this number is expected to increase with further investigation (27). Twelve genera and 212 species of freshwater fish (66%) are endemic to the region, of which about half are threatened (44). Additionally, 60 species of freshwater mollusks, including Gondwanaland relict species, and 174 species of odonates (40% endemic) have been documented along the rivers of the Western Ghats, with greater endemism and diversity in high-elevation hill streams and waterfall spray zones (44). Numerous local communities also rely directly on these rivers for their livelihood and sustenance; it is estimated that ~56% and ~18% of fish and mollusk species, respectively, are harvested for human consumption (44).

How are tropical montane rivers changing?

TMR drain climate-sensitive areas, harbor species-rich highly endemic faunas, and integrate human influences along riverine and elevational gradients. Consequently, they may be sentinels of regional and global change. Several contemporary anthropogenic disruptions—hydropower development, water pollution, water withdrawals for human consumption, climate change, and biodiversity loss—disproportionately affect TMR.

Hydropower development

An explosion of new hydropower development is currently altering global rivers at unprecedented rates, much of it focused on TMR. In the Andean Amazon alone, an estimated 142 hydropower projects already existed or were under construction as of 2017, and an additional 161 had been proposed (45). Two major dams on the Madeira River constructed in the past decade have already disrupted downstream flow of sediment (21) and upstream movement of large migratory fish, with implications for regional Amazonian fisheries (46). Small dams (<20 MW) tend to be more common on TMR, often operating as water diversion schemes that result in kilometers of dewatered rivers (47). Proliferation of small hydropower development in the Western Ghats—including >440 projects in various stages of implementation just in Karnataka State—has altered river geometry, water chemistry, and fish assemblage composition, especially in dewatered reaches (48). Changes in the availability and quality of freshwater ecosystem services for human populations also accompany the ecological effects of hydropower development on TMR.

Water pollution and withdrawals

Many tropical countries’ primary urban and industrial centers are located at high elevations (>1000 meters above sea level), where air temperatures are cooler and susceptibility to debilitating tropical diseases such as malaria or dengue may be reduced compared with lowland areas. Examples include Mexico City, Bogotá, and Nairobi (population ~21 million, ~10 million, and ~6.5 million, respectively). Likewise, much of these countries’ agricultural production occurs in their tropical montane regions, like mid-elevation piedmont areas, which tend to have nutrient-rich soils. Consequently, TMR often drain areas where human activities are intense and where there usually is little to no wastewater treatment (49). Colombia’s Magdalena River system is a prime example: beginning at 3685 meters above sea level and flowing 1612 km between the eastern and central cordilleras of the northern Andes, the Magdalena’s basin covers one-quarter of Colombia but concentrates 80% of the country’s human population (~46 million). Moreover, most of this population is concentrated in montane areas, causing significant impacts on water quality and riverine biota (35). These effects are further exacerbated by deforestation and mineral mining, which increase sediment and heavy metal loads (50). Water withdrawals for irrigation and human urban consumption are also significant and occur at volumes that permanently alter the functioning of many TMR ecosystems, including an increasing number of perennial rivers that are becoming intermittent (51).

Climate change

High-elevation tropical areas and their biota are among the most sensitive to the documented and projected effects of climate change, particularly rising temperatures and glacier melt (52). Several studies have documented upslope shifts in terrestrial biota as a consequence of increased temperatures along tropical montane gradients (53), but few have examined implications of climate change for TMR (5) and the human populations that depend on them (Table 1). For example, the population of Dar es Salaam, Tanzania (6 million), which depends on nearby Eastern Arc montane rivers for water, is projected to exceed 13 million by 2035, placing it among the top 10 fastest growing cities in the world. Altered precipitation patterns or increased frequency and magnitude of droughts, both likely under future climate scenarios, could increase Dar es Salaam’s vulnerability with respect to water security (38). In terms of tropical montane biota, freshwater fish are expected to suffer severe range contractions in future climates given physiological and dispersal constraints. Additionally, tropical aquatic invertebrates will likely be vulnerable to small increases in temperature due to narrow physiological tolerance and low dispersal rates (12).

Biodiversity loss

TMR are susceptible to biodiversity loss as a consequence of the aforementioned factors, which alter the quality and quantity of habitat for river-dependent biota. The severity of these factors depends on local and regional conditions, but tropical mountain species are predicted to be particularly vulnerable to their synergistic effects (54). Stream-breeding amphibians in neotropical montane areas have been particularly susceptible to catastrophic declines related to the spread of chytridiomycosis panzootic (30) and habitat degradation (55). These declines resulted in subsequent changes in algal primary production, organic matter dynamics, food web structure, and aquatic-terrestrial linkages (56). Similarly, introduction of exotic species such as rainbow trout (Oncorhynchus mykiss) and brown trout (Salmo trutta) are very common in TMR globally (57). In the tropical Andes, trout have been shown to predate on native fish in the Andean genus Astroblepus, while also competing with them for habitat and food and impacting invertebrate populations (58). Overfishing is a pervasive concern in TMR (59), which can lead to tropical fish extinctions and has enormous consequences on nutrient cycling (60), carbon flow (61), and food security (23).

Future directions for research on tropical montane rivers

Biodiversity and ecosystem processes

As illustrated above, fish and other obligate aquatic vertebrates exhibit high levels of diversity and endemism in TMR (26). However, many montane areas remain inadequately studied, and assessments of their biodiversity are based on limited information. Large fish have been studied more extensively mostly because fisheries are a critical resource for human populations of tropical regions (23). In contrast, small noncommercial fish are poorly known, with presumably hundreds of undescribed species in many tropical basins around the world (9). For other freshwater obligate aquatic groups—invertebrates, algae, microbes—even the most basic information on their diversity, phylogeny, life-history adaptations, and ecological roles is scarce. For example, the ratio of autochthonous (algae) to allochthonous (litter) resources decreases along the Andean-Amazon gradient, resulting in changes in invertebrate stoichiometry, growth, and community composition (62). However, it is still not clear how individual species, or communities, can affect large-scale ecosystem processes, and how sensitive these processes are to climate change and land-use transformations. Research on other ecological processes like migration, biological interactions, food web dynamics, and sensitivity to altered flow regimes is also essential to better understand and manage these ecosystems.

Longitudinal, lateral, and vertical connectivity

Although we understand the critical role of TMR for the transport of nutrients and carbon-laden sediments that fuel lowland floodplain biological productivity, we still lack comprehensive information on the magnitude and extent of these connections. There is little information across mountain ranges on how far longitudinally and laterally we can track the chemical signatures and ecosystem processes through the landscape, and how important these connections are for productivity and ecosystem services of the lower basins (21) and of the oceans (63). This is especially important given the proliferation of hydroelectric projects in many tropical regions (21), which drastically alters connectivity and transport along montane rivers (45). Flow and channel connectivity are also ecologically important for the many fish species that migrate toward or along TMR (64). For example, goliath catfishes of the Amazon (Brachyplatystoma spp.) migrate thousands of kilometers to Andean TMR to spawn (65). Fish migratory patterns and their ecosystem-level roles remain poorly understood for most tropical basins around the world.

Human-induced disturbances

Human influence is increasing rapidly in tropical regions. Hundreds of millions of people from the tropics are using resources and ecosystem services provided by TMR (Table 1), changing water quality and quantity, overexploiting species, and altering channel connectivity. Although some information already exists on individual impacts of pollution, water withdrawals, species loss, and climate change, these and other threats can act synergistically. For example, hydropower dam construction on TMR that drain urban and agricultural areas can lead to rapid accumulation of sediments, which need to be frequently released to protect infrastructure. The sudden release of polluted water and sediments from dams results in large hydrological peaks, contamination of riverine habitats, and massive faunal mortality (21). It is critical to analyze not only the threats to TMR, considering their integrity and interconnectedness along elevational gradients, but also the interactive effects of multiple stressors. This integrative view will be essential for exploring mitigation and adaptation strategies in accordance with the functioning of these dynamic ecosystems.


TMR have received insufficient attention despite their ecological and societal importance. Although existing research attests to the high biological diversity and endemism of TMR throughout the world, most regions remain inadequately studied. Studies of TMR are usually focused on individual elevational bands, thus ignoring the unifying and dynamic nature of these rivers as they traverse elevation gradients. As the structure and functioning of TMR integrate the effects of elevation and temperature and the varying influence of geological and geographical setting on disturbance regimes, we propose that these systems could be used as natural models for exploring the relationships between these key factors and their influence on biodiversity patterns, life-history adaptations, ecological processes, and ecoservices at different spatial and temporal scales. Moreover, in light of the many pressing human-induced disturbances in TMR, they can serve as templates to document, model, and explore solutions for management and adaptation strategies for rivers in a changing world.

References and Notes

Acknowledgments: We thank C. Hoorn, J. E. Suárez, and two anonymous reviewers for their many helpful comments to improve this paper. Funding: We acknowledge funding from the U.S. National Science Foundation (DEB-9615349, DEB-1046408), USAID-funded Partnerships for Enhanced Engagement in Research (PEER) Program managed by the National Academy of Sciences, The MacArthur Foundation (16-1607-151053-CSD), and a collaboration grant from Universidad San Francisco de Quito. Competing interests: The authors declare no competing interests.
View Abstract

Navigate This Article