Understanding diversity patterns along environmental gradients lies at the heart of community ecology and conservation. Previous studies have found variation in bird diversity and density along “natural” elevational gradients in the Tropical Andes Hotspot. However, there is still a lack of knowledge about how bird communities respond to traditional land-use patterns, in association with other multiple drivers, along elevations. In the present study, we investigated biotic, abiotic and anthropogenic sources of variation associated with bird species diversity, density and turnover along a 3000-m elevational gradient, in southern limit of the Tropical Andes Hotspot, northern Chile. Over four seasons, we conducted 472 bird point count surveys and established 118 plots distributed across the Desert, Pre-Puna, Puna and High-Andean belts, where biotic, abiotic and anthropogenic factors were measured. We used mixed-effects models to estimate alpha diversity and multinomial Poisson mixture models to estimate species density, accounting for detectability. Species diversity and density increased until 3300 masl and then declined. This type of elevational pattern is characteristic of dry-based mountains, where environmental conditions are suitable at mid-elevations. Habitats shaped by traditional Aymara indigenous agriculture, associated with relatively high vegetation heterogeneity, hosted the highest values of bird diversity and density. Species turnover was structured by habitat type, while elevational ranges of most species were restricted to three relatively discrete assemblages that replaced each other along the gradient. Our study revealed a hump-shaped relationship between elevation and bird diversity and density in the Dry Tropical Andes Biodiversity Hotspot, supporting a diversity pattern characteristic of dry-based mountains of the world. Traditional Aymara agriculture may have constructed ecological niches for biodiversity at mid-elevations, enhancing vegetation heterogeneity, thus providing resources for resident and rare species. Increasing loss of traditional land-use may present a threat to the bird community in the Tropical Andes Hotspot.
Citation: Araneda P, Sielfeld W, Bonacic C, Ibarra JT (2018) Bird diversity along elevational gradients in the Dry Tropical Andes of northern Chile: The potential role of Aymara indigenous traditional agriculture. PLoS ONE 13(12): e0207544. https://doi.org/10.1371/journal.pone.0207544
Editor: Tim A. Mousseau, University of South Carolina, UNITED STATES
Received: July 26, 2018; Accepted: November 1, 2018; Published: December 5, 2018
Copyright: © 2018 Araneda et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: This work has been supported by CONICYT/Fondecyt Iniciación 11160932, the Centro de Investigación en Medio Ambiente (CENIMA) of the Universidad Arturo Prat, and the Centre for Intercultural and Indigenous Research (CIIR) (GRANT: CONICYT/FONDAP/15110006). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Montane zones are typically rugged landscapes uplifted to such a level that local climate is affected . Mountains host exceptional biodiversity due to the elevational gradient that results from the variation of climate and topography over distances of only a few kilometers . These ecological transition areas (ecotones) are often characterized by high species turnover rates or beta diversity . Furthermore, mountains have been subject to human land-use practices for millennia and, as a result, harbor the largest number of distinct ethnic groups, varied remnants of biocultural traditions, and human-habitat adaptations through agriculture [4,5]. General diversity patterns along elevational gradients are the result of the combined effects of complex, often nonlinear, processes that show covariation with elevation . Conventionally, species richness has been considered to decrease monotonically with increasing elevation, while the elevational ranges of some species are greater at high elevations than at low elevations. This is the so-called “Stevens’s rule” . However, Rahbek [8,9] rejected Stevens’s rule by showing that many elevational gradients have mid-elevation peaks in diversity.
Climatic variation can influence the composition of bird communities along elevational gradients [10,11]. In temperate regions, birds are sensitive to seasonality due to both resource bottlenecks for food and water availability, and temperature regulation requirements across seasons [12,13]. In arid mountain ecosystems, maximum bird diversity can occur in wetter and cooler climatic conditions, generally at mid-elevations [14,15]. Vegetation heterogeneity (e.g. vegetation strata with dense foliage) is often correlated with bird species richness at various geographical scales [6,16,17]. For example, mid-elevation habitats with high vegetation heterogeneity in the Eastern Himalaya influence peaks in species diversity and density along the elevational gradient . Furthermore, relatively high habitat diversity along elevational gradients may harbor many co-existing species within habitat types, resulting in high species turnover between different habitats [19,20].
Through various land-use practices, humans have shaped almost every corner of the Earth, and thus influence the diversity and structure of ecological communities . Within mainstream ecological literature, humans are generally treated as exogenous drivers of change . However, relatively recently, researchers have expanded this mainstream notion to identify humans as multidirectional participants in coupled social-ecological systems . For example, through traditional agricultural practices, such as terracing and ridged crop systems in the Andes, humans have for millennia constructed ecological niches for biodiversity in mountain areas [5,24–26]. However, the role of traditional agriculture, in association with multiple other drivers (e.g. climate and vegetation) along elevational gradients, has not been subject to detailed empirical assessment in bird community ecology studies [6,27].
The Tropical Andes Hotspot is the most diverse hotspot on Earth, with higher numbers of species and rates of endemism than any other [28,29]. The hotspot contains a high variety of habitat types resulting from steep altitudinal gradients and climatic factors caused by the interaction of complex topography, trade winds, oceanic influences  and, potentially, indigenous use of Andean slopes for agriculture over the course of millennia [31,32]. This variety of habitat types may host bird species with narrow environmental tolerances, resulting in limited species distributions along elevational gradients [1,7]. It may be expected that this pattern will relate to rapid turnover–or beta diversity–along the elevational gradient, particularly when local diversity–or alpha diversity–is a small fraction of the total landscape diversity .
Studies conducted on birds along elevational gradients in the Wet Tropical Andes (Peru, Colombia and Bolivia) have shown a decline in species richness with elevation, due to a decrease in temperature and vegetation cover, and proximity to human settlements [15,33–36]. By contrast, Kessler et al , found a unimodal pattern relationship between richness and elevation in the forests of the Bolivian Andes, with a peak in bird species richness associated with the presence of old-growth forests at intermediate elevations (2700–3150 masl). The Dry Tropical Andes Region, which includes northern Argentina and Chile, is still relatively unexplored, and most of the available literature is descriptive [38–42]. This is especially true of the southern limit of the Tropical Andes Hotspot, which includes a complex of mountain chains and valleys, bordered to the south by the extremely arid Atacama Desert [30,43].
In the present study, we examined bird diversity and density patterns along an elevational gradient in the Dry Tropical Andes of northern Chile. We then evaluated a suite of biotic and abiotic factors that may be correlated with bird diversity and density, focusing on climatic conditions, seasonality, elevation, vegetation heterogeneity and habitat type, and giving special attention to the association between indigenous traditional agriculture and bird communities along the elevational gradient. Finally, we assessed the turnover (beta diversity) and range patterns of each bird species along the elevational gradient. We predicted that (1) vegetation heterogeneity drives non-linear associations between elevation and diversity and density, and (2) habitat diversity along the gradient can host different bird assemblages, resulting in high species turnover and distinct communities. To test these predictions, we estimated alpha diversity, density and beta diversity along a gradient of 3000 meters of elevation. This gradient ranged from Desert belt through Pre-Puna–with semi-arid vegetation and indigenous Aymara agriculture–to Puna and the High Andean belt above 4000 meters of elevation.
The study was conducted in the Aroma/Chiapa basin, located on the western limit of the Volcán Isluga National Park, in the Dry Tropical Andes of northern Chile (Fig 1). Aroma/Chiapa is an east-west drainage basin with a length of 70 km, encompassing part of the Desert, Pre-Puna, Puna and High-Andean belts [43,44]. The area receives rainwater during the summer season (45.56 ± 35.32 mm), mainly between January and March when the Inter-tropical Convergence Zone (ITCZ) moves to the south [43,45]. The basin has almost continuous surface runoff, and its tributaries are of an ephemeral regime . The surveyed points ranged from 1200, where basin starts, (19°47’53”S 69°40’36”W) to 4120 m above sea level, where basin ends (masl; 19°38’2”S 69°9’19”W).
Circles show the 118 surveyed points, and the striped polygon indicates the western boundary of the Volcán Isluga National Park. Source: Base map from Natural Earth.
Vegetation cover varies with elevation, and vegetation formations have been described as distinct belts associated with different elevations . Luebert and Pliscoff  and Trivelli and Valdivia  proposed the following classifications. (1) Desert belt (<2500 masl), where it is possible to find the Inner Tropical Desert Formation, with sparse vegetation and dominance of Tessaria absinthioides and Ditichlis spicata. The Inner Tropical Low Desert Scrub Formation–an open and xeromorphic scrubland dominated by Adesmia atacamensis and Cistanthe celosioides–is also present in the Desert belt. In addition, it is possible to find intrazonal vegetation associated with streams, such as Pleocarphus revolutus, T. absinthioides and Cortaderia atacamensis. (2) Pre-Puna belt (2500–3200 masl), characterized by an Andean Tropical Low Desert Scrub Formation: a very open scrub, with or without succulents, generally dominated by Atriplex imbricata and Acantolipia deserticola. The Pre-Puna belt also comprises an Andean Tropical Spiny Forest Formation, which is a sparse, extremely xeromorphic forest, dominated by Cactaceae Browningia candelaris and Corryocactus brevistylus in the upper canopy, with a low shrub stratum of succulents. Riparian vegetation is represented chiefly by C. atacamensis and some Fabaceae trees. In this belt, it is also very common to find patches of traditional Aymara agriculture . Aymara agriculture chiefly comprises subsistence farming of a diverse set of traditional crops, such as maize, alfalfa, potatoes, oregano and banana passionfruit (Passiflora mollissima). This traditional land-use includes a water irrigation mechanism in the form of crop terraces [24,48]. Aymara agriculture has likely been conducted in the area for around 1100 years, and remains a system of communal ownership and management of the streams that descend from the Tata’Jachura volcano [49–51]. (3) The Puna belt (3200–4000 masl) is characterized by an Andean Tropical Low Scrub Formation of zonal vegetation, comprising a dense thicket dominated by Fabiana ramulosa, Diplostephium meyenii, Lophopappus tarapacanus and Baccharis boliviensis in the woody strata, which can reach 1m in height. Some Cactaceae such as C. brevistylus are also present in this belt. (4) The High-Andean belt (>4000 masl) is represented by the Andean Tropical Low Shrub Formation of zonal vegetation dominated by Parastrephia lepidophylla and P. quadrangularis, where abundant Festuca orthophylla and Tetraglochin cristatum can also be found. No intrazonal vegetation was found in the Puna or High-Andean belts.
We randomly established 118-point surveys, each at a distance of at least 150m from adjacent points. These points were grouped into 19 elevational intervals, generated by grouping three to seven survey points (6.21 ± 0.34) for every 150m of elevation. We surveyed each point twice during both the wet (February and April) and dry (November and May) seasons of 2016 and 2017 (n = 472). Each point survey lasted 6 minutes, during which every bird seen or heard within a 50m radius was recorded. The distance to each bird was estimated and grouped into two distance intervals (0–25 and 26-50m) for further analysis. Birds were recorded during the four hours of peak activity immediately after dawn. In each point survey we recorded temperature (°C), humidity (%) and wind speed (m/s) using a handheld weather monitor (WM-300 WindMate, Speedtech Instruments, USA; Table 1) [52,53].
We used previous studies of bird-habitat relationships along mountain elevational gradients to identify potential structural habitat attributes (hereafter covariates) that may influence distribution patterns of birds in the Andes [33,54–56]. We located habitat plots (50m radius; 0.79 ha; n = 118 plots) at the center of the previously described point survey. Each plot (point survey) was assigned to a habitat type according to the seven criteria described in Table 1. We then characterized the habitat structure at each plot [41,47,57] and estimated vegetation heterogeneity for six vertical strata based on a six-point scale (Table 1). Our definition of vegetation heterogeneity states that heterogeneous plots have many vegetation strata with dense foliage coverage [58,59]. Therefore, vegetation heterogeneity was estimated by summing the coverage index of each vegetation stratum (Table 1).
Bird diversity and composition.
We estimated alpha diversity using Generalized Linear Mixed-Effect (GLME) models with a Poisson type error , using lme4  and AICcmodavg packages  in R . GLMEs describes the relationship between a response variable and several explanatory covariates (fixed effects) collected in aggregated units at different levels (random effects). We tested the fixed effect of habitat type, elevation, heterogeneity and seasonality on bird richness. Elevational interval, year, elevational interval-by-heterogeneity and seasonality-by-heterogeneity interaction were used as random effects (Table 1). The strength of evidence of fifteen models was evaluated by calculating model weights (wi) and the AIC value [60,64]. Models with AIC < 2 were considered to be just supported by the data .
Species richness was defined as the total number of recorded species by point survey. While, alpha diversity was the estimated number of species for each point survey by the GLME. We calculated beta diversity by the dissimilarity in presence and absence of species composition comparing neighboring intervals, utilizing the Sørensen’s Index of Dissimilarity Sør = 2a/(2a + b + c), where a is the number of species common between two points, b the number of species unique to first point, and c the number of species unique to the second point . Sørensen’s index is dependent on variation in the matching component a, or the level of continuity in species composition between two points .
Bird detectability and density.
Bird density estimates will vary according to species detectability, which may be influenced by the distance of the recorded bird from the observer and other survey-specific covariates, including temporal factors and weather conditions [68,69]. The Multiple-Covariate Distance Sampling (MCDS) framework uses the observer distance distribution, y, and one or more additional covariates represented by the vector z, to model the detection function. Therefore, the probability of detection is denoted as g(y, z) .
Using MCDS we analyzed avian point surveys utilizing Multinomial Poisson Mixture Models . To estimate detection (p) and density (D) for each species across points, we used maximum-likelihood methods in the R-Unmarked [71,72] program from . To model D, we first assessed collinearity to reduce the number of covariates. Collinear covariates were removed (r > 0.7), maintaining only the ones predicted to be more biologically influential for each species [41,54,57]. The half-normal key function for the detection function was selected using Akaike's Information Criterion (AIC) [64,70]. Detectability (p) was estimated using eight covariates potentially affecting the scale parameter of the detection function: season, time, date, noise, temperature, humidity and wind speed (Table 1). Important covariates for each species were identified by AIC .
After correcting for p of each species, we estimated density (individuals per hectare) by taking habitat type (HAB), vegetation heterogeneity (HET) and elevation (ELE) as covariates. To obtain the best model for each species, we computed model weights (wi) and the AIC value, following the same rationale described above. Once a preferred approach was selected, we averaged the density estimates from competing models (ΔAIC <2) . For analyses of detectability and density, we used a subset of the most abundant species. For both alpha diversity and density, a Kruskal-Wallis with Wilcoxon post-hoc multiple pairwise tests was used to examine differences between habitat types. We also performed linear and nonlinear regression to fit models of bird diversity and density on vegetation heterogeneity and elevation, and ranked them according to Akaike’s information criterion adjusted for small samples (AICc) .
We obtained research permits from the Chilean National Park administration authority (CONAF number 194195) to work in the Volcán Isluga National Park, and consent from the “Tata’Jachura” Chiapa Aymara Indigenous Community.
Species composition and elevational range size
We recorded 49 bird species across 118 survey points along the elevational radient from 1200 to 4120 masl (See S1 Appendix). Seven bird orders were recorded, which included 14 families. Tyrannidae and Thraupidae were the most represented families, each with nine species. Furnariidae followed these families with eight species, then Columbidae with five and Trochilidae with four.
Elevational ranges showed that three relatively discrete assemblages of birds replaced one another along the elevational gradient (Fig 2). Only three species–Metropelia aymara, Falco femoralis and Asthenes modesta–were recorded across a broad range of elevations. Mid-elevation habitats in the Pre-Puna belt showed the largest number of species (2500–3300 masl). Here, 61.2% (30) of total bird species were found, being mainly Passeriformes, Apodiformes and Columbiformes. A total of 18 species were exclusive to the Pre-Puna belt, such as Patagioenas maculosa and Colibri coruscans, whereas 12 species had a broader elevational range and were also present in other belts, such as Lepthastenura aegitaloides and Muscisaxicola maculirostris. A total of 10 species (20.4%) were registered exclusively in the lowlands of the Desert belt (<2500 masl), such as Aeronautes andecolus and Xenospingus concolor. Finally, Passeriformes and Falconiformes accounted for the majority (both with four species) of the nine species (18.4%) found in the Puna and High-Andean belts.
Lines indicate the maximum and minimum elevational limits of each species range, and species are ordered along the abscissa by ranked elevational midpoints (range average). See “S1 Appendix” for codes of bird species.
Elevational ranges varied among species. Then, 19 (38.8%) of the species showed elevational ranges of 500 m, and 30 (61.2%) had a range of more than 1000 m. Only one species (L. aegitaloides) showed an elevational range of 3000 m, covering the entire study area. Six species were detected only at a single elevation point.
The medians of alpha diversity between elevational intervals were significantly different (Kruskal-Wallis test; H = 415.3, d.f. = 18, p<0.05). For GLME analyses, the model including habitat type (HAB), heterogeneity (HET), elevation (ELE) and seasonality (SEA) as fixed effects performed better than all the other models (Table 2).
The models are ranked in descending order according to AICc.
The pattern of alpha diversity along the elevational gradient was explained by a polynomial regression (p<0.05, r2 = 0.46, y = -8.9x3-8.8x2+14.6x+1.6; Fig 3A), in which increasing diversity was observed with increasing elevation up to 3500 masl (peak values of diversity), and then decreased. Alpha diversity varied from season to season, showing an increment in the wet season (0.20 [SE ± 0.007]; p<0.05). Vegetation heterogeneity also showed a positive association with alpha diversity (0.10 [SE ± 0.02]; p<0.05, r2 = 0.57, Y = 2.1x3 + 3.1 x2 + 20.9x + 1.6; Fig 3B). Finally, we found a positive association between alpha diversity and agricultural habitat (0.43 [SE ± 0.13]; p<0.05). Alpha diversity varied among habitat types (Kruskal-Wallis test; H = 234.8, d.f. = 5, p<0.05; Fig 4A), with values being relatively higher in agricultural and highland steppe habitats. By contrast, desert, riparian and arboreal shrublands showed relatively low values of alpha diversity (Wilcoxon post-hoc test p<0.05).
Relationship between (a) elevation and alpha diversity, (b) heterogeneity and alpha diversity, (c) elevation and density (individuals/ha), and (d) heterogeneity and density (individuals/ha) along an altitudinal gradient in the Dry Tropical Andes of northern Chile.
(a) Alpha diversity by habitat type and (b) bird density (individuals/ha) by habitat type, along an elevational gradient in the Dry Tropical Andes of northern Chile. Agr: Agricultural habitat; Arb: Arboreal shrubland habitat; Col: Columnar cactus habitat; Des: Desert habitat; HiS: Highland steppe habitat; Rip: Riparian habitat. Dotted lines indicate mean alpha diversity and mean density, respectively. Significant codes: <0.0001 ‘****’; <0.001 ‘***’; <0.01 ‘**’.
An increase in alpha diversity by elevational interval was observed in lowlands (<2500 masl). The lowest values of alpha diversity were found in the desert habitat, between 1800 and 1950 masl, with 0.09 (SE ± 0.004) species per interval (Table 3). In the interval between 1950 and 2100 masl–the riparian and arboreal shrubland habitat types–alpha diversity increased, reaching a value of 0.90 (SE ± 0.07) species per interval.
Alpha diversity increased from 1.08 (SE ± 0.04) species per interval at the midlands Pre-Puna belt (>2500 masl) to 4.64 (SE ± 0.15) species at the 2850–3000 masl interval. These peak values were recorded for riparian habitats and for areas with permanent presence of indigenous Aymara agriculture. Alpha diversity began to decrease with elevation, from 2.02 (SE ± 0.09) species per interval in columnar cactus habitats to 1.34 (SE ± 0.07) species per interval in highland steppe habitats in the Puna belt. In contrast, the High-Andean belt showed a higher alpha diversity than the Puna belt, reaching 1.94 (SE ± 0.06) species between 4050 and 4200 masl.
Beta diversity along the altitudinal gradient
Beta diversity along the elevational gradient showed a relatively high turnover in species composition in lower and higher intervals (Fig 5). The most marked shift in species occurred in the lowland Desert belt (1200–2700 masl) between the different habitat types. Species composition recorded in the absolute desert habitat (interval 1800–1950 masl) was distinct from all the other habitats. In this belt, columnar cactus, riparian, and riparian with arboreal habitat types formed three different clusters of species. In the midland Pre-Puna belt, species occurring between 2700 and 3450 masl formed a distinct cluster dominated by riparian and agricultural habitat types, with columnar cactus habitat at its higher and lower limits. A final cluster of species was recorded at the highest elevations of the gradient (>3450 masl), specifically in highland steppe habitats located in the Puna and High Andean belts.
The yellow clusters indicate desert belt intervals, the green cluster indicates Pre-Puna belt intervals, and the brown cluster indicates Puna and High Andean belts intervals. Agr: Agricultural habitat; Arb: Arboreal shrubland habitat; Col: Columnar cactus habitat; Des: Desert habitat; HiS: Highland steppe habitat; Rip: Riparian habitat.
The detectability of 17 out of 21 species analyzed in more detail was associated with survey-specific (weather or temporal) covariates (S2 Appendix). Relatively low temperature conditions (TEM) were positively associated with the detectability of nine species. By contrast, Asthenes dorbignyi, A. modesta and Ashenes pubidunda were positively associated with relatively high temperature conditions. The detectability of M. aymara, Phrygilus plebejus and X. concolor was positively associated with the wet season (SEA). Only Metropelia maculirostris, Conirostrum cinereum and Phrygilus atriceps detection rates was negatively associated with humidity (HUM).
Estimated bird density showed significant variations between elevational intervals (Kruskal-Wallis test; H = 328.2, d.f. = 18, p<0.05). Even so, species density did not show a strong association with elevation (p<0.05, r2 = 0.33, y = -2.1x3–1.5 x2 + 1.9x + 0.4; Fig 3B). No significant association between elevation and any bird species density was observed.
Density along the gradient was related to vegetation heterogeneity (HET; p<0.05, r2 = 0.63, y = 1.4 x2 + 4.2x + 0.4; Fig 4B). Overall, 11 (52.4%) species were associated with vegetation heterogeneity. Of these, 10 were positively associated with heterogeneity, with Spinus uropygialis (0.2 [SE ± 0.05]) and Pipraeidea bonariensis (0.2 [SE ± 0.04]) showing the highest slopes for the beta coefficient. Only Sicalis uropygialis (-0.36 [SE ± 0.05]) was negatively associated with vegetation heterogeneity.
Species density showed important differences between habitat types (Kruskal-Wallis test; H = 241.2, d.f. = 5, p<0.05; Fig 4B). Bird density was higher in agricultural and highland steppe habitats. In arboreal shrublands, desert and riparian habitats, density was lower compared to other habitat types (Wilcoxon post-hoc test p<0.05). At the species level, 15 (71.4%) were strongly associated with one or more habitat types (HAB). P. plebejus (1.69 [SE ± 0.4]) was positively associated with highland steppe habitats, while S. uropygialis (1.64 [0.3]) and Sicalis olivacens (0.72 [SE ± 0.2]) were associated with columnar cactus habitat.
In the lowland Desert belt (<2500 masl), maximum bird density was between 1200 and 1350 masl, with 0.33 (SE ± 0.01) individuals per hectare in the riparian habitat type. At higher elevation, the lowest density values were estimated between 1800 and 1950 masl, reaching 0.16 (SE ± 0.002) in the desert habitat type. In the midlands (2500–3300 masl), bird density increased from 0.32 (SE ± 0.004) individuals/ha between 2550 and 2700 masl to a peak of 0.82 (SE ± 0.05) individuals/ha between 3150 and 3300 masl. Above 3300 masl, bird density decreased again, reaching 0.41 (SE ± 0.02) individuals/ha at the highest elevations (Table 3).
This study reveals important variations in bird diversity along an elevational gradient in the Dry Tropical Andes of northern Chile, chiefly according to habitat type and vegetation heterogeneity. Existing local features at mid-elevations–such as traditional indigenous agriculture, which generates relatively high vegetation heterogeneity–may play a major role in structuring diversity and density. Human activity modifies biotic and abiotic factors, and anthropogenically constructed niches result in cascade effects that permeate the entire ecological community [6,25,74,75].
In contrast to Stevens’s rule , our analysis of elevational range size showed that birds with the broadest elevational ranges were more common at intermediate elevations. This finding is similar to that of a study in the Wet Tropical Andes and in the Himalayas, where a combination of biological and habitat-structural factors played a determining role in the pattern of bird elevation ranges [18,34]. In our study, 61% of the species detected showed ranges broader than 1000 m, with peaks in alpha diversity in both agricultural and riparian habitats, suggesting that these mid-elevation habitats act as a source of species that utilize a broad elevational range . One possible explanation for this pattern is the mass effect hypothesis , in which dispersal from a more suitable habitat–in this case, from agricultural and riparian areas at mid-elevations–may add species into a less suitable habitat . This is consistent with Quintero and Jetz , who suggest that high-elevation mountains harbor a great variety of habitats, and thus offer many opportunities for bird mixing and diversification.
On a regional scale, species turnover along the elevational gradient may be influenced by historical immigration processes in the Andes. Previous studies have suggested that bird diversity in mountain ecosystems of northern Chile was enriched by the flow of species from adjacent regions (e.g. Wet Tropical Andes and Southern Andes) [1,77–80]. These species from neighboring regions have most probably settled to specific habitat conditions along the elevational gradient, resulting in high species turnover between habitats [76,81,82].
Elevational patterns along the gradient
The relationship between elevation and both alpha diversity and density is similar to the “hump-shaped” pattern found in other studies on plants [54,83], herpetofauna [84,85], birds [18,55,86] and small mammals [87,88]. This type of elevational pattern is characteristic of dry-based mountains [14,15], where elevation often shows a non-linear association with diversity and density . Although our results showed shifts in alpha diversity and density along the elevational gradient, there was not a strong association between these parameters and elevation.
The “hump-shaped” pattern found in our study also supports the results of studies on bird diversity from other regions. For example, in the Himalayas, bird species richness and density at mid-elevations were positively associated with vegetation productivity, habitat diversity [18,55] and human settlements . In the Southern Alps, the peak of bird richness at mid-elevations was attributed to optimal climatic conditions, environmental heterogeneity and man-made habitats . By contrast, in the Andean forests of Bolivia, bird diversity decreased with elevation and the presence of human settlements . Kessler et al.  showed that bird diversity was favored by the presence of old-growth forest remnants at intermediate elevations. A similar finding was reported for the Andes of Colombia, where decreasing and unimodal patterns of bird diversity were influenced by productivity . For northern Chile, the unimodal pattern of terrestrial bird species was initially proposed by Vilina and Cofré , highlighting a diversity peak at mid-elevations. However, these authors did not explore the influence of environmental and/or anthropogenic factors on bird diversity. For their part, Gantz et al.  showed that peaks of bird diversity in the Atacama Desert depend on food availability and proximity to source habitats with crop vegetation. The higher number of species (n = 80) reported in the latter study compared to that of the present study (n = 49) may be explained by the former’s larger study area, as well as additional habitat types surveyed, such as littoral desert and highland wetlands.
Habitat type and the role of traditional indigenous agriculture
Habitat type and indigenous use of valley bottoms and slopes may be strong drivers of alpha diversity and density along elevational gradients in the Andes [24,31,91]. In our study, alpha diversity was positively associated with agricultural habitats, which also showed maximum values of vegetation heterogeneity. Meanwhile, bird density only showed a positive association with agricultural habitat. Historically, the Pre-Puna belt has been one of the zones preferred for Aymara agricultural activities [32,48,49]. The positive association between agricultural habitat and bird diversity and density highlights the potential–and perhaps historical–role of traditional indigenous agriculture in mountain biodiversity. Agricultural activities provide habitats with relatively high vegetation heterogeneity, creating food and shelter resources along the elevational gradient . The expansion and intensification of agriculture, along with the loss of traditional land-use in the Andes, may be the cause of rapid decline in local biodiversity [91,93–95].
In the Andes in general, the relationship of traditional mountain societies with nature has been based on coexistence rather than competition [48,96]. This relationship results in agricultural strategies based on low transformation of local geographies and resources, and thus the sustainable use of natural resources [24,96,97]. For example, practices such as the construction of a network of terraces that prevent erosion and maximize water availability, and the maintenance of unmanaged open areas for livestock grazing, allow for the growth of both native and agricultural species in an ecotone. These ecotones likely increase the diversity and abundance of food [98,99], as well as the likelihood for birds to nest close to reliable foraging habitats , and also address the multiple temporal requirements of species in terms of seasonal differences in vegetation phenology [101,102]. The notion that traditional Aymara agriculture has positive effects on bird diversity should be treated with caution. Our observational study was conducted along the length of a drainage basin with a particular set of socio-political and ecological characteristics, in which we found small-scale agricultural patches and communal management of streams. Future studies should implement experimental or pseudo-experimental approaches to define whether indigenous agricultural habitats have imposed an adaptive advantage for birds that utilize them.
Traditional agricultural habitats located in the midst of an arid region may support species that otherwise would not be present. For example, Norfolk et al,  reported that traditional agriculture supports a higher proportion of migratory and insectivorous species, and a greater number of birds associated with unmanaged habitats in the arid mountains of South Sinai. In line with this, we found Conirostrum tamarugense using Pre-Puna habitats during the non-breading season, possibly depending on arthropods found on cultivated and native vegetation . This threatened endemic insectivore breeds in lowlands, and its activity during the winter (non-breeding) season is poorly known [104–107]. In addition, hummingbirds such as C. coruscans and Patagona gigas peruviana may take advantage of the winter flowering season of native and non-native species in agricultural habitats [105,108,109]. Both species are common residents in the Wet Tropical Andes [106,108,110,111] and have been rarely recorded in Chile [112–116]. Our results support those of Montaño-Centellas and Garitano-Zavala , who suggest that ornamental and other cultivated plants may provide foraging habitat for nectarivorous birds in the Wet Tropical Andes. This may also occur in the Dry Tropical Andes.
Vegetation heterogeneity and seasonality
In our study, peaks in vegetation heterogeneity were associated with riparian and agricultural habitats at mid-elevations. Vegetation heterogeneity is often correlated with bird diversity [86,117–119], because heterogeneous vegetation offers more potential niches than homogeneous habitats [58,76,120]. Heterogeneity increases food and foraging opportunities [2,4], shelter and nesting substrates, and other conditions suitable for successful reproduction . Availability of highly heterogeneous vegetation in Pre-Puna intervals, resulting chiefly from the presence of agricultural habitats, may explain the fact that the highest level of species similarity was observed in this belt. This high degree of similarity may be the result of a greater number of species coexisting in communities with larger niche hyperspace, causing an increase in alpha diversity and density, and a decrease in turnover .
The combined effect of optimal local features and climatic conditions may cause productivity to peak at mid-elevations . In arid-based mountains (e.g. Dry Tropical Andes and South-Western US mountains), water availability is high at mid-elevations because rainfall and soil water retention are high, while evaporation is relatively low. In our study area, water availability decreases severely towards the lowlands–becoming concentrated in a narrow stream–where high temperature and near-absent rainfall produce extremely dry habitats. Water availability is also low in highland areas, with runoff increasing due to shallow soils and exposed rock towards mountaintops [14,43]. The presence of habitats with intrazonal vegetation in the Puna and High Andean belts, such as highland wetlands, would increase the diversity and density of birds in these elevational intervals [122–124]. However, our study basin does not comprise high-elevation wetlands. Basins with a presence of high-elevation wetlands may show a different pattern of bird diversity along the elevation gradient, as these habitats are known to be species-rich systems.
Seasonality was influential on alpha diversity, but not for species density. Only the detection rates of the granivorous M. aymara and P. plebejus, and the insectivore X. concolor were positively associated with the wet season. Studies in the Atacama Desert found that temporal variation in the diversity and density of granivorous birds may relate to variations in the supply of feeding resources determined by tropical rainfall and increases in primary productivity in northern Chile [41,125].
Implications for conservation
Bird diversity patterns along elevational gradients should not be attributed to a single universal explanation, but rather to a combination of “natural” and anthropogenic factors [54,95]. Association among species turnover and habitat diversity along the gradient, suggesting that conservation efforts should consider the whole gradient rather than just portions of it [1,18,24,89]. Traditional agricultural habitats at mid-elevations enhanced the vegetation heterogeneity that likely provided resources for resident and rare species throughout the year.
Historical and contemporary socio-economic changes in indigenous livelihoods can potentially drive changes in “anthropogenic habitats”, and thus in bird species assemblages in the Pre-Puna belt [24,126]. Traditional Aymara agriculture is an enduring cultural practice, but has gradually been modified due to the influence of State policies, industrial development, and growth of neighboring urban centers [95,97]. Many Aymara farmers who once subsisted on what they cultivated or traded with highland camelid pastoralists have become increasingly involved in a market economy that has greatly expanded their patterns of consumption. For example, many have shifted from small-scale traditional to intensive agriculture, increasing the use of agro-chemicals that can potentially affect the value of agricultural areas as critical habitats for birds [127–129].
Despite the relatively high number of species with restricted distributions in the Dry Tropical Andes, only C. tamarugense has been assigned to a conservation category [104,130]. Several of the species detected in small numbers by our study are largely unknown in the Dry Tropical Andes, making it difficult to estimate whether our records correspond to vagrants, migratory individuals or a breeding population [113,116]. Furthermore, Pre-Puna habitats of high ecological and cultural value are among the most poorly represented elevation belts in the National System of Protected Areas of Chile (SNASPE) . This situation is made worse by the fact that the mountains of northern Chile are increasingly subject to commercial interests in the form of intensive agriculture, road construction and mining operations [132–135]. Bird species occurring in the Dry Tropical Andes seem to be well adapted to their local environments; however, some of these species may represent the last stage in a taxon cycle , surviving as local remnant populations in their southernmost distribution range.
S1 Appendix. Birds observed along an elevational gradient in the Dry Tropical Andes of northern Chile, surveyed in wet (W) and dry (D) seasons between November 2016 and August 2017.
S2 Appendix. Bird species and covariates(a) used to estimate detectability (p) and density (D) in the elevational gradient, according to the selection of models based on the Akaike’s Information Criterion (AIC).
Positive (+) and negative (-) symbols indicate the direction of the relationship; values in parentheses ( ) indicate the standard error. See “S1 Appendix” for codes of bird species.
We thank to C. Carrasco, V. Herrera and M. Gaete from the Centro de Investigación en Medio Ambiente (CENIMA) of the Universidad Arturo Prat, for their logistical support of this research. The Chilean Forest Service (CONAF), J. Valenzuela, S. García, F. Montecinos and N. Kandalaft for the assistance during the field campaign. We thank D. Cayo, S. Larama and “Tata’Jachura” Chiapa Aymara Indigenous Community for allowing us to work in their lands. Finally, we thank M. Vallejos and A. Riffo for valuable support and advice.
- 1. Fjeldså J, Bowie RCK, Rahbek C. The role of mountain ranges in the diversification of birds. Annu Rev Ecol Evol Syst. 2012;43:249–65.
- 2. Körner C, Jetz W, Paulsen J, Payne D, Rudmann-Maurer K, M. Spehn E. A global inventory of mountains for bio-geographical applications. Alp Bot. 2016;127.
- 3. Kent R, Levanoni O, Banker E, Pe’er G, Kark S. Comparing the response of birds and butterflies to vegetation-based mountain ecotones using boundary detection approaches. PLoS One. 2013;8.
- 4. Messerli B, Ives JD. Mountains of the world: a global priority. London, UK: The Parthenon Publishing Group; 1997.
- 5. Sarmiento FO, Ibarra JT, Barreau A, Pizarro JC, Rozzi R, González JA, et al. Applied montology using critical biogeography in the Andes. Ann Am Assoc Geogr. 2017;4452:1–13.
- 6. Lomolino M V. Elevation gradients of species-density: historical and prospective views. Glob Ecol Biogeogr. 2001;10:3–13.
- 7. Stevens GC. The elevational gradient in altitudinal range: an extension of Rapoport’ s latitudinal rule to altitude. Am Nat. 1992;140:893–911. pmid:19426029
- 8. Rahbek C. The elevational gradient of species richness: a uniform pattern? Ecography. 1995;18:200–5.
- 9. Rahbek C. The relationship among area, elevation, and regional species richness in Neotropical birds. Am Nat. 1997;149:875–902. pmid:18811253
- 10. Kohler T, Maselli D. Mountains and climate change—From Understanding to Action. 3rd ed. Bern: Geographica Bernensia with the support of the Swiss Agency for Development and Cooperation (SDC), and an international team of contributors.; 2012.
- 11. Elewa AMT. Migration of organisms: climate geography ecology. Migr. Org. Clim. Geogr. Ecol. Berlin, Heidelberg: Springer; 2005.
- 12. Herzog SK, Soria a. R, Matthysen E. Seasonal variation in avian community composition in a high-Andean Polylepis (Rosaceae) forest fragment. Wilson Bull. 2003;115:438–47.
- 13. Katuwal HB, Basnet K, Khanal B, Devkota S, Rai SK, Gajurel JP, et al. Seasonal changes in bird species and feeding guilds along elevational gradients of the Central Himalayas, Nepal. PLoS One. 2016;11:1–18.
- 14. McCain CM. Global analysis of bird elevational diversity. Glob Ecol Biogeogr. 2009;18:346–60.
- 15. Kattan GH, Franco P. Bird diversity along elevational gradients in the Andes of Colombia: area and mass effects. Glob Ecol Biogeogr. 2004;13:451–8.
- 16. Rutten G, Ensslin A, Hemp A, Fischer M. Vertical and horizontal vegetation structure across natural and modified habitat types at Mount Kilimanjaro. PLoS One. 2015;10:1–15.
- 17. Cueto VR, Casenave JL De. Determinants of bird species richness: role of climate and vegetation structure at a regional scale. J Biogeogr. 1999;487–92.
- 18. Acharya BK, Sanders NJ, Vijayan L, Chettri B. Elevational gradients in bird diversity in the eastern Himalaya: an evaluation of distribution patterns and their underlying mechanisms. PLoS One. 2011;6:e29097. pmid:22174953
- 19. Jankowski JE, Ciecka AL, Meyer NY, Rabenold KN. Beta diversity along environmental gradients: implications of habitat specialization in tropical montane landscapes. J Anim Ecol. 2009;78:315–27. pmid:19040686
- 20. Quintero I, Jetz W. Global elevational diversity and diversification of birds. Nature. 2018;555:246–50. pmid:29466335
- 21. Vitousek PM, Mooney H, Lubchenco J, Melillo JM. Human domination of Earth’ s ecosystems. Science (80-). 1997;277:494–9.
- 22. Liu J. Integrating ecology with human demography, behavior, and socioeconomics: needs and approaches. Ecol Modell. 2001;140:1–8.
- 23. Mace GM. Whose conservation? Science (80-). 2014;345:1558–60.
- 24. Fjeldså J. The relationship between biodiversity and population centres: the high Andes region as an example. Biodivers Conserv. 2007;16:2739–51.
- 25. Kendal J, Tehrani JJ, Odling-Smee J. Human niche construction in interdisciplinary focus. Philos Trans R Soc B Biol Sci. 2011;366:785–92.
- 26. Albuquerque UP, Silva JS, Campos JLA, Sousa RS, Silva TC, Alves RRN. The current status of ethnobiological research in Latin America: gaps and perspectives. J Ethnobiol Ethnomed. 2013;9:72. pmid:24131758
- 27. Grötzbach S. Mountain peoples and cultures. In: Messerli B., Ives JD, editors. Mt world a Glob Prior. London: The Parthenon Publishing Group; 1997. p. 17–38.
- 28. Mittermeier RA, Turner WR, Larsen FW, Brooks TM, Gascon C. Global biodiversity conservation: the critical role of hotspots. In: Zachos EF, Habel CJ, editors. Biodivers hotspots Distrib Prot Conserv Prior areas. Berlin, Heidelberg: Springer Berlin Heidelberg; 2011. p. 3–22.
- 29. Mittermeier RA, van Dijk PP, Rhodin AGJ, Nash SD. Hotspots revisited: Earth’s biologically richest and most endangered ecoregions. Chelonian Conserv Biol. 2015;14:2–10.
- 30. Young B, Zador M, Comer PJ, Diego A, Rolando D, Moritan G. Ecosystem profile: Tropical Andes biodiversity hotspot. Crit. Ecosyst. Partnersh. Fund. 2015.
- 31. Erickson CL. Prehistoric landscape management in the Andean highlands: raised field agriculture and its environmental impact. Popul Environ. 1992;13:285–300.
- 32. Molyneux M, Pearce A. History and language in the Andes. New York: Palgrave Macmillan; 2011.
- 33. Terborgh J. Bird species diversity on an Andean elevational gradient. Ecology. 1977;58:1007–19.
- 34. Jankowski JE, Londoño GA, Robinson SK, Chappell MA. Exploring the role of physiology and biotic interactions in determining elevational ranges of tropical animals. Ecography. 2013;36:001–2.
- 35. Patterson BD, Stotz DF, Solari S, Fitzpatrick JW, Pacheco V. Contrasting patterns of elevational zonation for birds and mammals in the Andes of southeastern Peru. J Biogeogr. 1998;25:593–607.
- 36. Montaño-Centellas FA, Garitano-Zavala Á. Andean bird responses to human disturbances along an elevational gradient. Acta Oecologica. 2015;65–66:51–60.
- 37. Kessler M, Herzog SK, Fjeldsa J, Bach K. Species richness and endemism of plant and bird communties along two gradients of elevation, humdity and land use in the Bolivian Andes. Divers Distrib. 2001;7:61–77.
- 38. Rottmann J, Kuschel R. Observaciones ornitológicas en las provincias de Antofagasta y Tarapacá. Boletín Ornitológico. 1970;2:1–8.
- 39. Lavercombe B., Hill C. Aves observadas en la provincial Tarapacá, incluyendo el hallazgo de dos especies y una subespecie nuevas para Chile. Boletín Ornitológico. 1972;4:1–5.
- 40. Torres-Mura JC. Los canasteros (Furnariidae) de Tarapacá (I Región): confirmación de Asthenes pudibunda en Chile. Boletín Chil Ornitol. 1998;5:24–7.
- 41. Gantz A, Rau J, Couve E. Avian assemblages of Atacama Desert, far north of Chile. Gayana. 2009;73:172–9.
- 42. Marquet PA, Bozinovic F, Bradshaw G, Cornelius C, Gonzalez H, Gutierrez J, et al. Los ecosistemas del desierto de Atacama y área andina adyacente en el norte de Chile. Rev Chil Hist Nat. 1998;71:593–617.
- 43. Luebert F, Pliscoff P. Sinopsis bioclimática y vegetacional de Chile. Second ed. Santiago, Chile: Editorial Universitaria; 2017.
- 44. Trivelli MG, Valdivia V. Alcances sobre flora y vegetación de la Cordillera de los Andes. Región de Arica y Parinacota y Región de Tarapacá. Segunda ed. Santiago, Chile: Ministerio de Agricultura. Servicio Agrícola y Ganadero; 2009.
- 45. Placzek C, Quade J, Betancourt JL, Patchett PJ, Rech JA, Latorre C, et al. Climate in the Dry Central Andes over geologic, millennial, and interannual timescales. Ann Missouri Bot Gard. 2009;96:386–97.
- 46. DGA. Análisis de los recursos hídricos de la Quebrada de Aroma Región de Tarapacá. Santiago, Chile; 2013.
- 47. Martínez-Tilleria K, Núñez-Ávila M, León CA, Pliscoff P, Squeo FA, Armesto JJ. A framework for the classification Chilean terrestrial ecosystems as a tool for achieving global conservation targets. Biodivers Conserv. 2017;26:2857–2856.
- 48. Eisenberg A. Aymara Indian perspectives on development in the Andes. University of Alabama Press; 2013.
- 49. Bengoa J. La Memoria olvidada. Historia de los pueblos indígenas de Chile. Compilación del informe de la Comisión de Verdad Histórica y Nuevo Trato. Santiago, Chile: Comisión Bicentenario, Presidencia de la República; 2004.
- 50. Castro M, Bahamondes M, Albornoz P, Basaure MF, Cayo S, Larama S, et al. El Derecho consuetudinario en la gestión del riego en Chiapa. Las aguas del Tata Jachura. Bogotá: Fundación Konrad-Adenauer-Stiftung, Universidad de Chile, Facultad de Derecho; 2017.
- 51. Martinez G. Para una etnografía del riego en Chiapa. Rev Chungará. 1987;18:163–79.
- 52. Ibarra JT, Martin K. Biotic homogenization: loss of avian functional richness and habitat specialists in disturbed Andean temperate forests. Biol Conserv. 2015;192:418–27.
- 53. Ibarra JT, Martin K. Beyond species richness: an empirical test of top predators as surrogates for functional diversity and endemism. Ecosphere. 2015;6:1–15.
- 54. Dainese M, Poldini L. Plant and animal diversity in a region of the Southern Alps: the role of environmental and spatial processes. Landsc Ecol. 2012;27:417–31.
- 55. Pan X, Ding Z, Hu Y, Liang J, Wu Y, Si X, et al. Elevational pattern of bird species richness and its causes along a central Himalaya gradient, China. PeerJ. 2016;4:e2636. pmid:27833806
- 56. Terborgh J. Distribution on environmen gradients: theory and a preliminary interpretation of distributional patterns in the avifauna of the cordillera Vilcabamba, Perú. Ecology. 1971;52:23–40.
- 57. Rau J, Zuleta C, Gantz A, Saiz F, Cortés A, Yates L, et al. Biodiversidad de artrópodos y vertebrados terrestres del Norte Grande de Chile. Rev Chil Hist Nat. 1998;71:527–54.
- 58. August P V. The role of habitat complexity and heterogeneity in structuring tropical mammal communities. Ecology. 1983;64:1495–507.
- 59. Williams SE, Marsh H, Winter J, Ecology S, May N. Spatial scale, species diversity, and habitat structure: small mammals in Australian tropical rain forest. Ecology. 2002;83:1317–29.
- 60. Zuur A, Leno EN, Walker NJ, Saveliev AA, Smith GM. Mixed effects models and extensions in ecology with R. Stat. Biol. Heal. New York, USA: Springer; 2009.
- 61. Bates D, Maechler M, Bolker B, Walker S. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67:1:48.
- 62. Mazerolle MJ. AICcmodavg: model selection and multimodel inference based on (Q)AIC(c). R Packag version 21–1. 2017;
- 63. R Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2017.
- 64. Akaike H. Information theory and an extension of the maximum likelihood principle. In: Petran BN, Csàaki F, editors. Proceeding Second Int Symp Inf Theory. 2nd Ed. Budapest, Hungary: Akadèemiai Kiadi; 1973. p. 267–81.
- 65. Burnham KP, Anderson DR. Model selection and inference: a practical information-theoretic approach. 2nd ed. New York, USA: Springer-Verlag; 2002.
- 66. Legendre P, Legendre L. Numerical Ecology. Third ed. Amsterdam: Elservier Science; 2003.
- 67. Koleff P, Gaston KJ, Lennon JJ. Measuring beta diversity for presence-absence data. J Anim Ecol. 2003;72:367–82.
- 68. Royle JA, Dawson DK, Bates S. Modeling abundance effects in distance sampling. Ecology. 2004;85:1591–7.
- 69. Buckland ST, Anderson DR, Burnham KP, Laake JL, Borchers DL, Thomas L. Advanced distance sampling. First edit. New York, USA: Oxford University Press; 2004.
- 70. Marques TA, Thomas L, Fancy SG, Buckland ST. Improving estimates of bird density using multiple-covariate distance sampling. Auk. 2007;124:1229–43.
- 71. Fiske I, Chandler R. unmarked: an R package for fitting hierarchical models of wildlife occurrence and abundance. J Stat Softw. 2011;43.
- 72. Chandler R. Distance sampling analysis in unmarked. USGS Patuxent Wildlife Research Center; 2014. p. 1–8.
- 73. Anderson A, Marques TA, Shoo LP, Williams SE. Detectability in audio-visual surveys of tropical rainforest birds: the influence of species, weather and habitat characteristics. PLoS One. 2015;10:1–24.
- 74. Albuquerque UP, Gonçalves PHS, Ferreira Júnior WS, Chaves LS, Oliveira RC da S, Silva TLL da, et al. Humans as niche constructors: revisiting the concept of chronic anthropogenic disturbances in ecology. Perspect Ecol Conserv. 2018;16:1–11.
- 75. Rodríguez-Estrella R. Land use changes affect distributional patterns of desert birds in the Baja California peninsula, Mexico. Divers Distrib. 2007;13:877–89.
- 76. Shmida A, Wilson M V. Biological determinants of species diversity. J Biogeogr. 1985;12:1–20.
- 77. Victoriano PF, González AL, Schlatter R. Estado de conocimiento de las aves de aguas continentales de Chile. Gayana (Concepción). 2006;70:140–62.
- 78. Vuilleumier F. Biogeografia de aves en el Neotrópico. Rev Chil Hist Nat. 1993;66:11–51.
- 79. Vuilleumier F. Forest birds of Patagonia: ecological geography, speciation, endemism, and faunal history. Ornithol Monogr. 1985;36:255–304.
- 80. Cody ML. Chilean bird distribution. Ecology. 1970;51:455–64.
- 81. Gaston KJ, Davies RG, Orme CDL, Olson VA, Thomas GH, Ding T-S, et al. Spatial turnover in the global avifauna. Proc R Soc B Biol Sci. 2007;274:1567–74.
- 82. Jankowski JE, Merkord CL, Rios WF, Cabrera KG, Revilla NS, Silman MR. The relationship of tropical bird communities to tree species composition and vegetation structure along an Andean elevational gradient. J Biogeogr. 2012;40:950–62.
- 83. Jacquemyn H, Honnay O, Pailler T. Range size variation, nestedness and species turnover of orchid species along an altitudinal gradient on Réunion Island: implications for conservation. Biol Conserv. 2007;136:388–97.
- 84. Hofer U, Bersier L, Borcard D, Ecology S, Apr N, Hofer U, et al. Spatial organization of a herpetofauna on an elevational gradient revealed by null model tests stable. Ecology. 1999;80:976–88.
- 85. Fu C, Wang J, Pu Z, Zhang S, Chen H, Zhao B, et al. Elevational gradients of diversity for lizards and snakes in the Hengduan Mountains, China. Biodivers Conserv. 2007;16:707–26.
- 86. Kessler M, Herzog SK, Fjeldsa J, Bach K. Species richness and endemism of plant and bird communties along two gradients of elevation, humdity and land use in the Bolivian Andes. Divers Distrib. 2001;7:61–77.
- 87. Piksa K, Nowak J, Zmihorski M, Bogdanowicz W. Nonlinear distribution pattern of hibernating bats in caves along an elevational gradient in mountain (Carpathians, Southern Poland). PLoS One. 2013;8:1–10.
- 88. Nor SM. Elevational diversity patterns of small mammals on Mount Kinabalu, Sabah, Malaysia. Glob Ecol Biogeogr. 2001;10:41–62.
- 89. Paudel PK, Šipoš J. Conservation status affects elevational gradient in bird diversity in the Himalaya: a new perspective. Glob Ecol Conserv. Elsevier B.V.; 2014;2:338–48.
- 90. Vilina Y, Cofré H. Aves terrestres. Biodivers Chile, Patrim y Desafíos. 2008. p. 246–51.
- 91. Alwang J, Norton GW, Barrera V, Botello R. Conservation agriculture in the Andean highlands: promise and precautions. In: Mann S, editor. Futur Mt Agric. Berlin Heidelberg; 2013. p. 21–38.
- 92. Moguel P, Toledo VM. Biodiversity conservation in tradicional Coffee systems of Mexico. Conserv Biol. 1999;13:11–21.
- 93. Norris K. Agriculture and biodiversity conservation: opportunity knocks. Conserv Lett. 2008;1:2–11.
- 94. Chappell MJ, LaValle LA. Food security and biodiversity: can we have both? an agroecological analysis. Agric Human Values. 2011;28:3–26.
- 95. Hanspach J, Loos J, Dorresteijn I, Abson DJ, Fischer J. Characterizing social–ecological units to inform biodiversity conservation in cultural landscapes. Divers Distrib. 2016;22:853–64.
- 96. Ramakrishnan PS. Mountain biodiversity, land use dynamics, and traditional ecological knowledge. In: Huber UM, Bugmann HKM, Reasoner MA, editors. Glob Chang Mt Reg. Netherlands; 2005. p. 551–61.
- 97. Aldunate C, Villagrán C, Armesto JJ, Castro V. Ethnobotany of pre-altiplanic community in the Andes of northern Chile. Econ Bot. 1983;37:120–35.
- 98. Duelli P, Obrist MK. Regional biodiversity in an agricultural landscape: the contribution of seminatural habitat islands. Basic Appl Ecol. 2003;4:129–38.
- 99. Hagen M, Kraemer M. Agricultural surroundings support flower-visitor networks in an Afrotropical rain forest. Biol Conserv. Elsevier Ltd; 2010;143:1654–63.
- 100. Bruun M, Smith HG. Landscape composition affects habitat use and foraging flight distances in breeding European starlings. Biol Conserv. 2003;114:179–87.
- 101. Gilroy JJ, Anderson GQA, Grice P V., Vickery JA, Sutherland WJ. Mid-season shifts in the habitat associations of Yellow Wagtails motacilla flava breeding in arable farmland. Ibis (Lond 1859). 2010;152:90–104.
- 102. Wilson JD, Whittingham MJ, Bradbury RB. The management of crop structure: a general approach to reversing the impacts of agricultural intensification on birds? Ibis (Lond 1859). 2005;147:453–63.
- 103. Norfolk O, Power A, Eichhorn MP, Gilbert F. Migratory bird species benefit from traditional agricultural gardens in arid South Sinai. J Arid Environ. 2015;114:110–5.
- 104. Estades CF. Natural history and conservation status of the Tamarugo Conebill in northern Chile. Wilson Ornithol Soc. 1996;108:268–79.
- 105. Cofré H. Avifauna asociada a los bosques de queñoa (Polylepis spp.). Boletín Chil Ornitol. 2007;13:56–60.
- 106. Jaramillo Á. Aves de Chile: incluye la Península Antártica, las islas Malvinas y Georgia del Sur. Barcelona: Lynx Edicions; 2005.
- 107. Schulenberg TS. American Ornithological Society observations on two rare birds, Upucerthia albigula and Conirostrum tamarugense, from the Andes of southwestern Peru. Condor. 1987;89:654–8.
- 108. Del Hoyo J, Elliott A coed, coed Sargatal J, Altshuler D. Handbook of the birds of the world. Barcelona: Barcelona Lynx Eds.; 1992.
- 109. Villagrán C, Castro V. Ciencia indígena de los Andes del norte de Chile: programa interdisciplinario de estudios en biodiversidad (PIEB), Universidad de Chile. Santiago, Chile: Ed. Universitaria; 2004.
- 110. Salinas L, Arana C, Suni M. Nectar of Puya species like resource for high Andean hummingbirds of Ancash, Peru. Rev Peru Biol. 2007;14:129–34.
- 111. Hainsworth FR. Foraging efficiency and parental care in Colibri coruscans. Condor. 1977;79:69–75.
- 112. Sallaberry M, Aguirre J, Yáñez J. Adiciones a la lista de aves de Chile: descripción de especies nuevas para el país y otros datos ornitológicos. Not Mens Mus Hist Nat. 1992;321:3–10.
- 113. Marín M. Annotated checklist of the birds of Chile. Barcelona: Lynx Editions; 2004.
- 114. Marin M, Kiff LF, Pena L. Notes on Chilean birds, with descriptions of two new subspecies. Bull Br Ornithol Club. 1989;109:66–82.
- 115. Heynen I, Esman P, Kirwan GM. Giant Hummingbird (Patagona gigas). In: del Hoyo J., Elliott A., Sargatal J., Christie D.A. & de Juana E, editor. Handb Birds World Alive. Barcelona: Lynx Editions; 2018.
- 116. Züchner T, Boesman P, Kirwan GM. Sparkling violet-ear (Colibri coruscans). In: del Hoyo J., Elliott A., Sargatal J., Christie D.A. & de Juana E, editor. Handb Birds World Alive. Barcelona: Lynx Edicions; 2018.
- 117. Meynard CN, Quinn JF. Bird metacommunities in temperate South American forest: vegetation structure, area, and climate effects. Ecology. 2008;89:981–90. pmid:18481523
- 118. Koh C-N, Lee P-F, Lin R-S. Bird species richness patterns of northern Taiwan: primary productivity, human population density, and habitat heterogeneity. Divers Distrib. 2006;12:546–54.
- 119. Visco DM, Michel NL, Boyle WA, Sigel BJ, Woltmann S, Sherry TW. Patterns and causes of understory bird declines in human-disturbed tropical forest landscapes: a case study from Central America. Biol Conserv. 2015;191:117–29.
- 120. Herrnstadt Z, Howard PH, Oh C-O, Lindell CA. Consumer preferences for ‘natural’ agricultural practices: assessing methods to manage bird pests. Renew Agric Food Syst. 2016;31:516–23.
- 121. Wiens JA, Rotenberry JT. Habitat associations and community structure of birds in shrubsteppe environments. Ecol Monogr. 1981;51:21–42.
- 122. Fjeldså J. Origin, evolution, and status of the avifauna of Andean wetlands. Ornithol Monogr. 1985;85–112.
- 123. Caziani SM, Derlindati EJ, Tálamo A, Sureda AL, Trucco CE, Nicolossi G. Waterbird richness in altiplano wetlands of northwestern Argentina. Waterbirds. 2001;24:103–17.
- 124. Sielfeld , Amado N, Herreros J, Peredo R, Rebolledo A. La avifauna del Salar del Huasco: Primera Región, Chile. Boletín Chil Ornitol. 1996;3:17–24.
- 125. Muñoz AE, Bonacic C. Variación estacional de la flora y vegetación en la precordillera andina de la comuna de Putre. Periodo 2002–2003. Gayana Botánica. 2006;63:75–92.
- 126. Jodha NS. Economic globalisation and its repercussions for fragile mountains and communities in the Himalayas. In: Huber ViM, Bugmann HKM, Reasoner MA, ed. Glob Chang Mt Reg An Overv Curr Knowl. Netherlands: Springer; 2005. p. 583–91.
- 127. Green RE, Cornell SJ, Scharlemann JPW, Balmford A. Farming and the fate of wild nature. Science (80-). 2005;307:550–5.
- 128. Kajtoch Ł. The importance of traditional orchards for breeding birds: the preliminary study on Central European example. Acta Oecologica. 2017;78:53–60.
- 129. Gibbs KE, MacKey RL, Currie DJ. Human land use, agriculture, pesticides and losses of imperiled species. Divers Distrib. 2009;15:242–53.
- 130. International BirdLife. Conirostrum tamarugense. The IUCN red list of threatened species 2016. 2016.
- 131. Pliscoff P, Fuentes-Castillo T. Representativeness of terrestrial ecosystems in Chile’s protected area system. Environ Conserv. 2011;38:303–11.
- 132. Durán AP, Rauch J, Gaston KJ. Global spatial coincidence between protected areas and metal mining activities. Biol Conserv. 2013;160:272–8.
- 133. Maxwell SL, Fuller RA, Brooks TM, Watson JEM. The ravages of guns, nets and bulldozers. Nature. 2016;536:146–145.
- 134. Oyarzún J, Oyarzún R. Sustainable development threats, inter-sector conflicts and environmental policy requirements in the arid, mining rich, Northern Chile territory. Sustain Dev. 2011;19:263–74.
- 135. Fuentes-Allende N, Vielma A, Paulsen K, Arredondo C, Corti P, Estades CF, et al. Is human disturbance causing differential preference of agricultural landscapes by taruka and feral donkeys in high Andean deserts during the dry season? J Arid Environ. 2016;135:115–9.