Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Integrating niche and occupancy models to infer the distribution of an endemic fossorial snake (Atractus lasallei)

  • Camilo Alejandro Cruz-Arroyave,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing

    Affiliation Grupo Herpetológico de Antioquia, Instituto de Biología, Universidad de Antioquia, Medellín, Colombia

  • Felipe A. Toro-Cardona,

    Roles Conceptualization, Formal analysis, Investigation, Methodology, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing

    Affiliation Grupo de Ecología y Evolución de Vertebrados, Instituto de Biología, Universidad de Antioquia, Medellín, Colombia

  • Juan Luis Parra

    Roles Conceptualization, Formal analysis, Investigation, Methodology, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing

    juanl.parra@udea.edu.co

    Affiliation Grupo de Ecología y Evolución de Vertebrados, Instituto de Biología, Universidad de Antioquia, Medellín, Colombia

Abstract

Understanding species distribution and habitat preferences is crucial for effective conservation strategies. However, the lack of information about population responses to environmental change at different scales hinders effective conservation measures. In this study, we estimate the potential and realized distribution of Atractus lasallei, a semi-fossorial snake endemic to the northwestern region of Colombia. We modelled the potential distribution of A. lasallei based on ecological niche theory (using maxent), and habitat use was characterized while accounting for imperfect detection using a single-season occupancy model. Our results suggest that A. lasallei selects areas characterized by slopes below 10°, with high average annual precipitation (>2500mm/year) and herbaceous and shrubby vegetation. Its potential distribution encompasses the northern Central Cordillera and two smaller centers along the Western Cordillera, but its habitat is heavily fragmented within this potential distribution. When the two models are combined, the species’ realized distribution sums up to 935 km2, highlighting its vulnerability. We recommend approaches that focus on variability at different spatio-temporal scales to better comprehend the variables that affect species’ ranges and identify threats to vulnerable species. Prompt actions are needed to protect herbaceous and shrub vegetation in this region, highly demanded for agriculture and cattle grazing.

Introduction

Biodiversity loss is occurring at an unprecedented rate, with many species being lost before they are acknowledged [1]. This loss results from various threats that differentially impact organisms across distinct spatial and temporal scales. Climate change and habitat degradation are particularly influential for ectothermic organisms like reptiles and represent the main causes of biodiversity loss [26]. Therefore, an understanding of the determinants of species distribution is essential to identify threats and establish effective short- and long-term conservation and management plans [79]. The lack of information about threats to biodiversity hinders the optimization of conservation efforts and contributes to an absence of engagement from local communities in preserving their biotic richness [10]. This situation is particularly common for tropical snakes, because there is a considerable information gap about their distribution, ecology, and natural history [11, 12]. Threatened snakes are in a state of heightened vulnerability, not only due to indiscriminate killing resulting from widespread fear among people but also due to the lack of information about their natural history, which complicates the assessment of their conservation status [13]. This lack of knowledge is even more apparent in species with cryptic-fossorial habits, which typically spend a considerable portion of their lives beneath the surface [14].

With the advent of publicly available databases and remote sensing, several methodologies have been developed to approximate species distributions at broad scales, with ecological correlative niche modelling being one of the most widely used and effective methods [8, 15, 16]. Niche modelling can be used to predict potential distribution based on the relationship between geographic (where occurrences are registered) and environmental space (where the niche is defined) [17], typically using abiotic environmental predictors (climate) to estimate suitability [18]. However, a significant critique of these models is the lack of information about species’ absences. Presence-only models can take advantage of the copious number of detections stored in global repositories such as GBIF and iNaturalist but are conditioned to make coarse spatio-temporal predictions and are susceptible to sampling biases, that prevent them from robustly estimating prevalence or probabilities of occurrence [1922].

Methods to generate presence-nondetection data, which enable a direct estimation of the species occupancy (proportion of an area occupied by a species) require repeated visits to multiple localities to estimate absence. Occupancy models simultaneously estimate species occupancy and detectability (i.e., the probability of detecting a species if it present) based on systematic and repetitive sampling. Occupancy models offer a reliable alternative to estimate species distribution and habitat use at a fine scale [19, 23, 24]. Researchers must revisit a series of sites and acquire information during a specific time and spatial intervals. Sampling design must fulfil a series of assumptions (that can sometimes be relaxed), such as the closure assumption (sites are closed to changes in occupancy status within sampling period) and independent detection histories. Both approaches can be applied at various scales and have proven useful for studying the distribution of fossorial-cryptic snakes with conservation goals in mind [2527].

Atractus lasallei is a fossorial-cryptic Colombian endemic snake, characterized by its small size, that inhabits humid to very humid pre-montane forests in the northern Central and Western Andes [2830] and currently classified as "Least Concern" (LC) by the IUCN. Despite its common status, its populations are highly fragmented, and there is a presumed continuous decline in the numbers of adult individuals [29]. This decline is attributed to several threats that affect Colombia’s snake fauna, including indiscriminate killing, habitat alteration, local climate change, and reduction in prey availability [11, 13]. These threats are of particular concern for A. lasallei considering that the region it inhabits has historically suffered substantial deforestation, mainly for mining, livestock farming and agriculture [31], resulting in a landscape predominantly characterized by pastures and cultivated fields [32]. However, to date, most studies about A. lasallei have primarily focused on its taxonomy and systematics, and there are limited publications addressing ecological aspects such as habitat use [30, 33, 34]. Further, some species occurrences seem to be beyond its reported distribution range.

The fossorial and cryptic habits, coupled with the small distribution range of A. lasallei, make it an ideal model to explore the complementary implementation of ecological niche and occupancy modelling. We hypothesize that the geographic distribution of A. lasallei can be estimated through approaches implemented at different spatio-temporal scales. While niche models can make use of all records available but compromising temporal and spatial resolution, occupancy models require data collection under a strict sampling design that often limits the temporal and spatial extent. We take advantage of these trade-offs and predict that niche models will identify environmental requirements of the species at broad scales (e.g., Grinnelian niche), and occupancy models will identify variables more representative of essential interactions for the species (e.g., Eltonian niche). Given its fossorial-cryptic habits, we expect ground-level variables to be more informative for modelling its potential distribution, and a strong association with habitats characterized by low vegetation such as grasslands and shrubs. We estimated the geographic distribution area of A. lasallei through the combination of both modelling approaches, emphasizing the applicability of such results for conservation and management purposes.

Materials and methods

Ecological niche model and potential distribution

Presence data were acquired from three sources: 1) specimens from biological collections, obtained from the Global Biodiversity Information Facility (accessed 22 March 2022) [35] and most of them revised in situ in the following collections: MHUA-Museo de Herpetología Universidad de Antioquia (Curator: J.M. Daza-Rojas), CSJ-h-Museo de Ciencias Naturales de La Salle (D.Z. Urrego), CBUCES-D-Colecciones Biológicas Universidad CES (J.C. Duque); 2) iNaturalist records obtained directly from them (accessed March-May 2022; we did not use iNaturalist records from GBIF) by searching Atractus records in the northwest of Colombia that included pictures that allowed verification through morphology (coloration patterns, and scale counts when possible), and 3) individuals encountered during the field phase for occupancy models. Identification of individuals was based on the original species description and taxonomic revisions of the genus [28, 33]. Further, a geographical filter was applied to presence records that were within 1 km of each other to reduce spatial autocorrelation [36, 37].

We used the final database to delimit the species accessible area or M [38] based on the intersection between the minimum convex polygon generated with 50 km buffers around each presence record using QGIS v.3.10 [39], and the biogeographic regions of the northern Central Cordillera and the Western Cordillera of Colombia [40] (Fig 1).

thumbnail
Fig 1. Geographical location of study area.

(A) Study Area encompassing northern Western and Central Cordilleras. (B) Presence records of A. lasallei, where light blue dots depict specimens collected during the field phase of occupancy models, red dots represent iNaturalist records, and dark blue dots indicate specimens in biological collections. (C) Transects represented as black triangles. Green-Brown color-scale represents altitude. Reprinted from [56] and OpenStreetMap under a CC BY license, with permission from [56] and ©OpenStreetMap contributors, original copyright 2020 and 2024, respectively.

https://doi.org/10.1371/journal.pone.0308931.g001

The environmental variables for the niche model included topographic variables, atmospheric climate variables including temperature corrected to ground level [41] and soil variables [42] (S1 Table). Climate variables represent long-term averages (1980–2010 in the case of atmospheric variables, and 2000–2020 in the case of ground-level temperature; see S1 Table). These variables were selected based on previous research findings regarding the distribution of semi-fossorial reptiles [25, 4345]. All variables were used in the models at a spatial resolution of 1 km. Variables with finer resolutions were resampled using the bilinear method, with the R-package “raster” v.3.5 [46] in R v.4.2.1 [47]. Subsequently, a Spearman correlation test (S1 File) was conducted to select non-correlated variables (< 0.8), using R-package “corrplot” v.0.92 [48]. Finally, two sets of variables were created, one that included two ground-level temperature variables estimated at five centimeters above the ground (S1 Table) [41], and the second included the same two variables but measured at atmospheric level [49]. Models were calibrated with each data set independently, ensuring all variables used were not correlated.

The ecological niche model was generated using the maximum entropy algorithm [15] through the R-package “kuenm” [50]. This methodology allows the evaluation of different sets of environmental variables (set 1 and set 2, S1 Table) and various model parameterizations to ultimately identify the best model according to a set of criteria. We allowed the regularization parameter to vary from 0.1 (very strict in relation to observed values) to 5 (more flexible in relation to observed values), where 1 is the default value. We also evaluated across linear (l), quadratic (q), and linear-quadratic (lq) responses. The models were trained using a 20% random partition of the occurrence data for model evaluation. The evaluation criteria included omission rate (<5%), partial receiver operating characteristic (partial ROC), area under the curve (AUC ratio>1), and the Akaike Information Criterion corrected for sample size (AICc) [51]. In case several models achieved the evaluation criteria, we performed a consensus model using the median of the selected models. Finally, to obtain the geographic projection, a cutoff threshold was applied using the 10 percentile training presence criteria from the best model(s) to generate a presence/absence map.

Occupancy models

To identify fine-scale factors influencing the occupancy of A. lasallei within its known distribution area, single-season occupancy models were employed [24]. The sampling design followed the recommendations of a previous study for semi-fossorial species [52], wherein 30 linear transects of 100 m x 2 m were established within the sampling area, spaced at a minimum of 200 m apart to ensure independence of detection histories across sites (Fig 1). Each transect was equipped with nine artificial cover objects (three roof tiles, three boards, and three plastic sheets), which were installed a minimum of two months prior to sampling for the organisms to habituate to their presence (S1 Fig). The transects were surveyed between October 2021 and January 2022 to ensure consistent occupancy status during the sampling period (closed-site assumption) between 8 AM and 4 PM. Surveys involved searching beneath leaf litter and under cover objects (both artificial and natural). Each transect was surveyed a minimum of four times, with visits spaced at least two weeks apart to satisfy the assumption of temporal independence. Animals were photographed and examined in the field to ensure correct identification (Approval Act No. 138, February 9, 2021, granted by the Committee on Ethics for Animal Experimentation, Universidad de Antioquia).

Occupancy models were constructed using the R-package “unmarked” v.1.2.5 [53] implemented in the R software. All covariates were standardized (mean = 0, units in standard deviations) prior to modelling. To identify the best models, we first established the best detection model assuming constant occupancy, and then we used this detection model in all occupancy models [54]. To model detectability, we included as covariates, the number of cover objects, both natural and artificial (N_obj); vegetation height (Veg_H) [55]; soil moisture (Soil_moisture); and soil temperature (T_ground), both measured using a HOBO proV2 datalogger beneath a roof tile or under the object where an individual of the species was located at the time of each visit (vegetation height was not related to the number of cover objects, S5 Fig). As covariates for occupancy, we used vegetation height (Veg_H) [55]; terrain slope (Slope); topographic convergence (Con); compound topographic index (CTI) [56]; annual mean soil temperature (Tprom), maximum temperature of the warmest month (Tmax), and minimum temperature of the coldest month (Tmin) [41]; depth of leaf litter (Leaf_Dep) and depth of the 0 horizon (Hori0), both measured in the field using a soil auger; euclidean distance to the nearest house (D_house), nearest forest (D_forest), and nearest water body (D_water). These distances were estimated in QGIS [39], identifying the nearest houses and forests to the centroid of each transect using satellite imagery from GoogleEarth (https://www.google.com/intl/es/earth/). To calculate the distance to water bodies, it was necessary to construct a detailed hydrographic network for the area using a 12.5 m resolution DEM obtained from Alaska vertex (https://search.asf.alaska.edu/), utilizing the hydrology toolbox in ArcGIS Pro v.2.7 [57].

A total of 87 biologically plausible and simple models were evaluated, each including one or two variables (S2 Table), 20 of the models were for the detection component with constant occupancy, and the remaining models were for the occupancy component. Finally, to evaluate model fit to our data, we performed a parametric bootstrap test on the chosen model, using the parboot function of R package “unmarked” v.1.4.1 [53]. This test generates multiple sets of data iteratively from the best model and then compares these sets with the detection histories obtained in the field. A chi-squared test was employed to evaluate the null hypothesis that the observations are consistent with the proposed model.

Integration of models

To estimate the species’ realized distribution area [58], we used the binary (presence-absence) geographic projection from the consensus niche model to identify the areas where the macro conditions were suitable and applied the best occupancy model within those areas at a higher spatial resolution (0.00025° ≈ 27 meters). Finally, the resulting map was transformed into a binary outcome using a threshold of 0.78, based on the Q3 (third quartile) of the occupancy distribution values of that map; this threshold corresponds to 4 m of vegetation height according to the best occupancy model (Fig 2), which is biologically justified if we consider that all presence records obtained in the field phase were found in places with vegetation below 4 m.

thumbnail
Fig 2. Occupancy modeling results.

(A) Relationship between the number of cover objects and detectability. (B) Occupancy as a function of vegetation height measured in meters (m), where dashed red lines represent the 4 m vegetation height and its expected occupancy value. Both graphs constructed from the best model.

https://doi.org/10.1371/journal.pone.0308931.g002

Results

Ecological niche model and potential distribution

A total of 164 records were obtained, out of which 83 referred to specimens from biological collections. Two of these records were collected dead in the northern Western Cordillera, representing a new locality (collection permit under resolution 0524 by ANLA, https://ipt.biodiversidad.co/permisos/resource?r=udea_mhua_2022_atractuslasallei). Additionally, 25 records were harnessed from iNaturalist (S3 Table), and 56 individuals were encountered pre and during field sampling. Following the removal of spatially aggregated records, a dataset of 54 occurrence records for niche modelling was obtained.

We found four ecological niche models that met the established evaluation criteria. These models shared the same feature classes and set of variables but had slight variation in the magnitude of the regularization multiplier, thus a consensus model was created using the median of the selected models. These models included atmospheric level temperature variables (Table 1). The most informative variables for the best models included slope, average annual precipitation, evapotranspiration, and maximum temperature of the warmest month. The response curves for these variables indicated that the species is associated with flatter slopes (<10°), relatively high average annual precipitation (200–3500 mm), mean evapotranspiration values between 80 and 120 mm per month, and maximum temperatures between 17 and 22°C (S2 Fig).

thumbnail
Table 1. Selected models that met evaluation criteria: AUC ratio, probability value (Pval) of partial ROC (pROC), omission rate and AICc.

These models were used to create the consensus model. Model names follow this structure separated by underscores: Model (M), Regularization multiplier (0.5–5), Feature classes (F_l, F_q, and F_lq), and set of variables used (set_1 or set_2).

https://doi.org/10.1371/journal.pone.0308931.t001

The predicted potential distribution area for the species covers 6182 km² after applying the 0.096 threshold, obtained from averaging the 10 percentile training presence thresholds from the best models (Table 1), to generate a presence/absence map. The geographical projection reveals three primaries’ zones of distribution, including the northern Central Cordillera and two smaller centres along the Western Cordillera, supporting the few occurrence records of the species in this mountain range (Figs 1 and 3).

thumbnail
Fig 3. Potential and realized distribution models.

Comparison between (A) potential and (B) realized distribution area estimated from niche modelling and the integration of niche and occupancy models respectively. Green-Brown color-scale represents altitude. Reprinted from [56] under a CC BY license, with permission from [56], original copyright 2020.

https://doi.org/10.1371/journal.pone.0308931.g003

Occupancy models

Out of 145 surveys (30 sites x 5 repetitions, minus 5 surveys not conducted), A. lasallei individuals were recorded on 14 occasions (9.6%) at 7 sites (23% naïve occupancy) (S3 Fig).

The variable that best explained detection was the number of cover objects, which considers both natural and artificial objects (Table 2). The best occupancy model (deltaAICc < 2) only included vegetation height as a covariate (Table 3). Further, the parametric bootstrap test indicated that the observations fit well with this simple occupancy model (p-value = 0.119).

thumbnail
Table 2. Statistical comparison of detectability submodels using the Akaike information criterion corrected for sample size.

https://doi.org/10.1371/journal.pone.0308931.t002

thumbnail
Table 3. Statistical comparison of the best occupancy models.

The inclusion of vegetation height can be observed in the models with lower AICc.

https://doi.org/10.1371/journal.pone.0308931.t003

Integration of models

The estimated realized distribution area was 935.3 km2 encompassing the northern part of the Central Cordillera and a few areas in the Western Cordillera (Fig 3). This area represents 35.7% of the potential distribution area.

Discussion

We provide the first estimates of the geographic distribution of A. lasallei, whose ecology, including climatic and habitat requirements, were poorly known. The potential geographic distribution of this species was expanded to include the northern section of the Western Cordillera in Colombia, where new records were obtained during this project. Our results indicate that this snake is associated with flat terrain (<10°), characterized by cool weather (maximum temperatures below 27° C; S4 Fig) and high annual precipitation (between 2500–3500 mm). Occupancy models supported an association with habitats characterized by low vegetation cover and an increase in detection related to the number of artificial or natural cover objects. The integration of niche and occupancy models generated an estimate of 935.3 km2 as the actual distribution area for this species (Fig 3). This is well below the mean of the distribution of snake geographic ranges [59, 60] and low enough to consider the species as vulnerable according to IUCN guidelines [61].

The variables identified as important in the niche and occupancy models reveal interesting aspects of the species’ ecology, but also provide a means to highlight current threats. The best potential distribution model identified slope as the most informative variable. The species is associated with sites characterized by low slopes (< 10°) that in this region are highly transformed for cattle grazing and agriculture. Moreover, we found that atmospheric temperature variables were better predictors than ground-level corrected variables, which was unexpected for a semi-fossorial organism that experiences temperature at the level of its microhabitat. Ground-level variables were highly correlated with atmospheric maximum temperature (0.92 Pearson correlation), thus, there seemed to be no large differences between both [41]. Ground level variables should be more informative at finer spatial and temporal scales, which unfortunately are not currently available for the area of interest. We suggest exploring alternative sources of microclimatic data [62, 63].

Vegetation height best explained the species’ occupancy pattern, supporting the hypothesized strong association of A. lasallei with habitats characterized by low vegetation such as grasslands and shrubs (Fig 2). We hypothesized that the transformation of vegetation in most of the northern central Andes by gold mining and further introduction and propagation of grasses such as Pennisetum clandestinum for livestock farming, could represent a trade-off for A. lasallei because prey availability is higher in those places where the snakes are more likely to be killed by humans [64]. Changes in vegetation cover may benefit some reptiles due to better thermoregulation conditions [65], or an increase in prey availability because grasslands are often enriched with organic matter providing favourable conditions for earthworms [66, 67], the main prey of Atractus snakes [43, 68, 69]. Given that the distribution of A. lasallei is dominated by grasslands [32], we expect frequent encounters between these snakes and rural people, often resulting in the death of snakes. Indeed, conservation efforts should be concentrated on environmental education programs targeted towards local communities.

The previous results from the niche and occupancy models support two ideas well established in ecology in relation to the observed geographic distribution of species. One is that the distribution is the outcome of a hierarchy of environmental filters [7072] that may act at different spatio-temporal scales. A second idea is that variables that determine this distribution include requirements and impacts (static and dynamically linked variables) that may also act at different spatiotemporal scales. Selection at broader geographic scales is likely driven by physiological requirements of individuals to certain abiotic conditions, such as terrain relief and climate. In contrast, selection at finer scales might result from individual experiences and learning, leading to preferences for more dynamic niche variables such as prey and predator density. Even though we only used static variables in both models, we believe vegetation height might be thought as a proxy for the potential interactions at a site. It is also important to note that biotic variables such as vegetation height were only used in the occupancy models due to its relatively high temporal variability.

The inclusion of the northern section of the Western Cordillera in the geographic range of A. lasallei was confirmed based on the identification of various specimens (see lepidosis, S4 Table). This biogeographic pattern is consistent with that described in some disparate taxa such as plants and birds, suggesting a close relationship between the biotas of the northern Central and Western Cordilleras [40, 73, 74]. This pattern highlights the need for phylogenetic studies to clarify the relationship between the Atractus lineages in the Andean region and the continuous exploration of the complicated geography of the Andes.

One strategy to improve future explorations is to maximize our detection ability. Occupancy models revealed that the number of cover objects is directly related to detectability. This aligns well with findings from other snakes [52] and results for salamanders in the Great Smoky Mountains National Park, where a similar methodology identified detection varying with the sampling method [75]. However, it’s worth mentioning that out of the 14 specimens encountered during field sampling, only two were found under artificial cover objects—an adobe tile and a wooden board—placed in transects with low secondary vegetation and few natural cover objects. This could be explained by: (1) the relatively small size of the objects used, which can be associated to less thermoregulation opportunities [65] and (2) the time that the objects were left in place before data collection began, since snakes often take time to become accustomed to such objects [52, 76]. According to our findings and field observations, we suggest that future studies that focus on fossorial reptiles should consider the size, thickness, and material of the cover objects to improve the species´ thermoregulation opportunities and the detection probabilities.

Our results provide a reference to compare with the two geographic distribution methods used by the IUCN, the Extent of Occurrence (EOO) and the Area of Occupancy (AOO), that represent different aspects of geographic range size [61]. EOO measures the overall geographic spread of the localities at which a species is found, while AOO tends to be much smaller, usually a small fraction of the EOO, and is heavily dependent on the spatial resolution of both the distribution data and the method of measurement employed [59]. We used the same species presence records used for the niche models to estimate EOO and AOO through the R package “ConR” v.1.3.0 [77] (using default parameters) and compared both range sizes with our estimated realized distribution area. For IUCN criteria we obtained an area of 9383 km2 (EOO) and 200 km2 (AOO), while our model estimate of the realized species distribution area was between the two other estimates (935.3 km2). We consider this represents a better approach when trying to assign a conservation category.

Following our results, A. lasallei could be considered Vulnerable, not only because the distribution area based on IUCN threshold for Extent Of Occurrence and Area Of Occupancy (20000 km2 and 2000 km2 respectively) [61], but because their populations are highly restricted, fragmented, and with presumed decreasing tendencies [29]. However, these considerations and the ongoing growth of the human population and fast land transformation in this region could place A. lasallei at an even higher threat category.

Supporting information

S1 Table. Environmental variables for the niche models.

https://doi.org/10.1371/journal.pone.0308931.s001

(DOCX)

S2 Table. Models tested in the occupancy modeling process.

https://doi.org/10.1371/journal.pone.0308931.s002

(DOCX)

S3 Table. A. lasallei voucher and iNaturalist observations used for niche modelling.

https://doi.org/10.1371/journal.pone.0308931.s003

(DOCX)

S4 Table. Specimen scale count form Western Cordillera.

https://doi.org/10.1371/journal.pone.0308931.s004

(DOCX)

S1 File. Spearman correlation test for niche modeling variable selection.

https://doi.org/10.1371/journal.pone.0308931.s005

(DOCX)

S1 Fig. Artificial cover objects during the sample phase for occupancy models.

https://doi.org/10.1371/journal.pone.0308931.s006

(DOCX)

S2 Fig. Response curves of most important variables from selected niche model.

https://doi.org/10.1371/journal.pone.0308931.s007

(DOCX)

S3 Fig. Detection history from sampling phase.

https://doi.org/10.1371/journal.pone.0308931.s008

(DOCX)

S4 Fig. Elevation distribution of A. lasallei.

https://doi.org/10.1371/journal.pone.0308931.s009

(DOCX)

S5 Fig. Relationship between vegetation height and the number of cover objects.

https://doi.org/10.1371/journal.pone.0308931.s010

(DOCX)

Acknowledgments

This work wouldn’t have been possible without the selfless assistance of the local residents where the field phase was conducted, the tremendous patience and support of our family and friends, the collaboration of numerous researchers and biological collections, the relevant insights from professors, peers, and, of course, the realm of boundless learning that the Universidad de Antioquia provided, fostering the development of this project, and providing the capture permit (collection permit under resolution 0524 by ANLA, https://ipt.biodiversidad.co/permisos/resource?r=udea_mhua_2022_atractuslasallei). The authors are very grateful to Vivian P. Páez, Juan C. Arredondo and Jorge H. Valdez for her initial guidance, to Juan M. Daza for his comments, the editor Mark A. Davis and two anonymous referees, as well Brian J. Halstead for their critical reading, comments and suggestions in earlier versions of this work. a special mention to the GHA and EcoEvo colleagues for their selfless help.

References

  1. 1. Pimm SL, Jenkins CN, Abell R, Brooks TM, Gittleman JL, Joppa LN, et al. The biodiversity of species and their rates of extinction, distribution, and protection. Science. 2014;344: 1246752. doi: pmid:24876501
  2. 2. Whittaker RJ, Araújo MB, Jepson P, Ladle RJ, Watson JEM, Willis KJ. Conservation biogeography: Assessment and prospect. Divers Distrib. 2005;11: 3–23. https://doi.org/10.1111/j.1366-9516.2005.00143.x
  3. 3. Deutsch CA, Tewksbury JJ, Huey RB, Sheldon KS, Ghalambor CK, Haak DC, et al. Impacts of climate warming on terrestrial ectotherms across latitude. PNAS. 2008;105: 6668–6672. https://doi.org/10.1073/pnas.0709472105 pmid:18458348
  4. 4. Huey RB, Deutsch CA, Tewksbury JJ, Vitt LJ, Hertz PE, Pérez HJÁ, et al. Why tropical forest lizards are vulnerable to climate warming. Proc R Soc B: Biol Sci. 2009;276: 1939–1948. https://doi.org/10.1098/rspb.2008.1957
  5. 5. Theisinger O, Ratianarivo MC. Patterns of reptile diversity loss in response to degradation in the spiny forest of southern Madagascar. Herpetol Conserv Biol. 2015;10: 273–283.
  6. 6. Lara-Reséndiz RA, Galina-Tessaro P, Sinervo B, Miles DB, Valdez-Villavicencio JH, Valle-Jiménez FI, et al. How will climate change impact fossorial lizard species? Two examples in the Baja California Peninsula. J Therm Biol. 2021; 95. https://doi.org/10.1016/j.jtherbio.2020.102811 pmid:33454041
  7. 7. Loiselle BA, Howell CA, Graham CH, Goerck JM, Brooks T, Smith KG, et al. Avoiding Pitfalls of Using Species Distribution Models in Conservation Planning. Conserv Biol. 2003;17: 1591–1600. https://doi.org/10.1111/j.1523-1739.2003.00233.x
  8. 8. Mota-Vargas C, Rojas-Soto OR. The importance of defining the geographic distribution of species for conservation: The case of the Bearded Wood-Partridge. J Nat Conserv. 2012;20: 10–17. https://doi.org/10.1016/j.jnc.2011.07.002
  9. 9. Guisan A, Tingley R, Baumgartner JB, Naujokaitis-Lewis I, Sutcliffe PR, Tulloch AIT, et al. Predicting species distributions for conservation decisions. Ecol Lett. 2013;16: 1424–1435. https://doi.org/10.1111/ele.12189 pmid:24134332
  10. 10. McIntosh EJ, Pressey RL, Lloyd S, Smith RJ, Grenyer R. The Impact of Systematic Conservation Planning. Annu Rev Environ Resour. 2017;42: 677–697. https://doi.org/10.1146/annurev-environ-102016-060902
  11. 11. Lynch JD. El contexto de las serpientes de Colombia con un análisis de las amenazas en contra de su conservación. Rev Acad Colomb Cienc Exactas Fis Nat. 2012;36: 435–449.
  12. 12. Böhm M, Collen B, Baillie JEM, Bowles P, Chanson J, Cox N, et al. The conservation status of the world’s reptiles. Biol Conserv. 2013;157: 372–385. https://doi.org/10.1016/j.biocon.2012.07.015
  13. 13. Lynch JD, Angarita-Sierra T, Ruiz-Gómez FJ. Programa Nacional para la Conservación de las Serpientes Presentes en Colombia. Ministerio de Ambiente y Desarrollo Sostenible, Universidad Nacional de Colombia, Instituto Nacional de Salud. 2013.
  14. 14. How RA, Shine R. Ecological traits and conservation biology of five fossorial “sand-swimming” snake species (Simoselaps: Elapidae) in south-western Australia. J Zool. 1999;249: 269–282.
  15. 15. Phillips SJ, Anderson RP, Schapire RE. Maximum entropy modeling of species geographic distributions. Ecol. Modell. 2006;190: 231–259. https://doi.org/10.1016/j.ecolmodel.2005.03.026
  16. 16. Busby JR. BIOCLIM–a bioclimate analysis and prediction system. Nature conservation: cost effective biological surveys and data analysis (ed. by C.R. Margules and M.P. Austin), pp. 64–68. CSIRO, Melbourne; 1991.
  17. 17. Colwell RK, Rangel TF. Hutchinson’s duality: The once and future niche. PNAS. 2009;106: 19651–19658. https://doi.org/10.1073/pnas.0901650106 pmid:19805163
  18. 18. Hirzel AH, Le Lay G. Habitat suitability modelling and niche theory. J Appl Ecol. 2008;45: 1372–1381. https://doi.org/10.1111/j.1365-2664.2008.01524.x
  19. 19. Guillera-Arroita G. Modelling of species distributions, range dynamics and communities under imperfect detection: advances, challenges and opportunities. Ecography. 2017;40: 281–295.19. https://doi.org/10.1111/ecog.02445
  20. 20. Phillips SJ, Dudík M, Elith J, Graham CH, Lehmann A, Leathwick J, et al. Sample selection bias and presence-only distribution models: Implications for background and pseudo-absence data. Ecol Appl. 2009;19: 181–197. https://doi.org/10.1890/07-2153.1 pmid:19323182
  21. 21. Phillips SJ, Elith J. On estimating probability of presence from use—availability or presence—background data. Ecology. 2013;94: 1409–1419. https://doi.org/10.1890/12-1520.1 pmid:23923504
  22. 22. Yackulic CB, Chandler R, Zipkin EF, Royle JA, Nichols JD, Campbell-Grant EH, et al. Presence-only modelling using MAXENT: When can we trust the inferences? Methods Ecol Evol. 2013;4: 236–243. https://doi.org/10.1111/2041-210x.12004
  23. 23. Contreras Diaz RG. Sobre el uso del hábitat y su relación con la estructura interna del nicho ecológico sensu Grinnell: Un estudio de caso para especies de mamíferos con amplia distribución [Doctoral dissertation, Universidad Nacional Autónoma de México]. UNAM—Dirección General de Bibliotecas. 2022.
  24. 24. Mackenzie DI, Nichols JD, Lachman GB, Droege S, Royle JA, Langtimm CA. Estimating Site Occupancy Rates When Detection Probabilities Are Less Than One. Ecol. 2002;83: 2248–2255. https://doi.org/10.1890/0012-9658(2002)083[2248:ESORWD]2.0.CO;2
  25. 25. González-Fernández A, Manjarrez J, García-Vázquez U, D’Addario M, Sunny A. Present and future ecological niche modeling of garter snake species from the Trans-Mexican Volcanic Belt. PeerJ. 2018;6: 1–20. https://doi.org/10.7717/peerj.4618 pmid:29666767
  26. 26. Halstead BJ, Wylie GD, Casazza ML. Habitat suitability and conservation of the Giant Gartersnake (Thamnophis gigas) in the Sacramento Valley of California. Copeia. 2010;4: 591–599. https://doi.org/10.1643/CE-09-199
  27. 27. Hansen EC, Scherer RD, Fleishman E, Dickson BG, Krolick D. Relations between Environmental Attributes and Contemporary Occupancy of Threatened Giant Gartersnakes (Thamnophis gigas). J Herpetol. 2017;51: 274–283. https://doi.org/10.1670/16-143
  28. 28. Do Amaral A. Studies of neotropical Ophidia. XXVI. Ophidia of Colombia. Bull. Antivenin Inst. America. 1931;4: 87–88.
  29. 29. Arredondo JC. Atractus lasallei (errata version published in 2017). The IUCN Red List of Threatened Species 2015: E.T44581109A115390110.
  30. 30. Vásquez-Restrepo JD, Toro-Cardona FA, Alzate E, Rubio-Rocha L. Guía de las serpientes del Valle de Aburrá. Medellín. Editorial Universidad CES. 2018.
  31. 31. Sánchez-Cuervo AM, Aide TM. Identifying hotspots of deforestation and reforestation in Colombia (2001–2010): Implications for protected areas. Ecosphere. 2013;4: 1–21. https://doi.org/10.1890/ES13-00207.1
  32. 32. Orrego S. Economic Modeling of Tropical Deforestation in Antioquia (Colombia), 1980–2000: An Analysis at a Semi-Fine Scale with Spatially Explicit Data. [Doctoral dissertation, Oregon State University]. Oregon State University Graduate Theses and Dissertations. 2009. [Available from: https://ir.library.oregonstate.edu/concern/graduate_thesis_or_dissertations/5138jk440/]
  33. 33. Passos P, Arredondo JC, Fernandes R, Lynch JD. Three new Atractus (Serpentes: Dipsadidae) from the Andes of Colombia. Copeia. 2009;3: 425–436. https://doi.org/10.1643/CH-08-063
  34. 34. Passos P, Arredondo JC. Rediscovery and redescription of the Andean earth-snake Atractus wagleri (Reptilia: Serpentes: Colubridae). Zootaxa. 2009;68:59–68.
  35. 35. GBIF.org. GBIF Occurrence Download. 22 March 2022. [Available from: https://doi.org/10.15468/dl.jxppaq]
  36. 36. Chapman AD. Principles and Methods of Data Cleaning–Primary Species and Species- Occurrence Data, version 1.0. Copenhagen, Report for the Global Biodiversity Information Facility. 2005.
  37. 37. Cobos ME, Jiménez L, Nuñez-Penichet C, Romero-Alvarez D, Simoes M. Sample data and training modules for cleaning biodiversity information. Biodivers Inform. 2018;14: 49–50. http://orcid.org/0000-0001-7442-8593
  38. 38. Soberón J, Townsend Peterson A. Interpretation of models of fundamental ecological niches and species’ distributional areas. Biodivers Inform. 2005;2: 1–10. https://doi.org/10.17161/bi.v2i0.4
  39. 39. QGIS.org. QGIS 3.10 Geographic Information System. QGIS Association. 2022.
  40. 40. Hazzi NA, Moreno JS, Ortiz-Movliav C, Palacio RD. Biogeographic regions and events of isolation and diversification of the endemic biota of the tropical Andes. Proc Natl Acad Sci USA. 2018;115: 7985–7990. https://doi.org/10.1073/pnas.1803908115 pmid:30018064
  41. 41. Lembrechts JJ, van den Hoogen J, Aalto J, Ashcroft MB, De Frenne P, Kopecký M, et al. Global maps of soil temperature. Global Change Biol. 2021;28: 3110–3144. https://doi.org/10.1111/gcb.16060
  42. 42. Poggio L, de Sousa LM, Batjes NH, Heuvelink GBM, Kempen B, Ribeiro E, et al. SoilGrids 2.0: producing soil information for the globe with quantified spatial uncertainty. SOIL. 2021;7:217–240. https://doi.org/10.5194/soil-7-217-2021
  43. 43. Ferreira-Silva C, Ribeiro SC, De Alcantara EP, Ávila RW. Natural history of the rare and endangered snake Atractus ronnie (Serpentes: Colubridae) in northeastern Brazil. Phyllomedusa. 2019;18: 77–87. https://doi.org/10.11606/issn.2316-9079.v18i1p77-87
  44. 44. Passos P, Martins A, Pinto-Coelho D. Population Morphological Variation and Natural History of Atractus potschi (Serpentes: Dipsadidae) in Northeastern Brazil. South Am J Herpetol. 2016;11: 188–211. https://doi.org/10.2994/SAJH-D-16-00034.1
  45. 45. Sunny A, González-Fernández A, D’Addario M. Potential distribution of the endemic imbricate alligator lizard (Barisia imbricata imbricata) in highlands of central Mexico. Amphib-reptil. 2017;38: 225–231. https://doi.org/10.1163/15685381-00003092
  46. 46. Hijmans RJ. Raster-R package (3.5–21). 2022.
  47. 47. R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/; 2023.
  48. 48. Wei, T., Simko, V. corrplot-R package (0.92). https://github.com/taiyun/corrplot; 2021.
  49. 49. Karger, DN, Lange, S, Hari, C, Reyer, CPO, Zimmermann, NE. CHELSA-W5E5 v1.0: W5E5 v1.0 downscaled with CHELSA v2.0. ISIMIP Repository. https://doi.org/10.48364/ISIMIP.836809.2; 2022.
  50. 50. Cobos ME, Peterson AT, Barve N, Osorio-Olvera L. Kuenm: An R package for detailed development of ecological niche models using Maxent. PeerJ. 2019;7: 1–15. https://doi.org/10.7717/peerj.6281
  51. 51. Peterson AT, Papeş M, Soberón J. Rethinking receiver operating characteristic analysis applications in ecological niche modeling. Ecol Model. 2008; 213: 63–72. https://doi.org/10.1016/j.ecolmodel.2007.11.008
  52. 52. Ward RJ, Griffiths RA, Wilkinson JW, Cornish N. Optimizing monitoring efforts for secretive snakes: A comparison of occupancy and N-mixture models for assessment of population status. Sci Rep. 2017;7:1–12. https://doi.org/10.1038/s41598-017-18343-5
  53. 53. Fiske I, Chandler R. unmarked: An R Package for Fitting Hierarchical Models of Wildlife Occurrence and Abundance. J Stat Softw. 2011;43: 1–23.
  54. 54. Andrade-Ponce G, Cepeda-Duque JC, Mandujano S, Velásquez-C KL, Lizcano DJ, Gómez-Valencia B. Modelos de ocupación para datos de cámaras trampa. MaNo. 2021;7: 200. https://doi.org/10.18637/jss.v043.i10
  55. 55. Potapov P, Li X, Hernandez-Serna A, Tyukavina A, Hansen MC, Kommareddy A, et al. Mapping global forest canopy height through integration of GEDI and Landsat data. Remote Sens. Environ. 2020;253: 112165. https://doi.org/10.1016/j.rse.2020.112165
  56. 56. Amatulli G, Mclnerney D, Sethi T, Strobl P, Domisch S. Geomorpho 90m, empirical evaluation and accuracy assessment of global high- resolution geomorphometric layers. Sci Data. 2020;7: 1–18. https://doi.org/10.1038/s41597-020-0479-6
  57. 57. ESRI. ArcGIS pro (2.7). Hydrology Toolbox. Redlands, CA; 2022.
  58. 58. Jiménez-Valverde A, Lobo JM, Hortal J. Not as good as they seem: The importance of concepts in species distribution modelling. Divers. Distrib. 2008;14: 885–890. https://doi.org/10.1111/j.1472-4642.2008.00496.x
  59. 59. Gaston KJ, Fuller RA. The sizes of species’ geographic ranges. J. Appl. Ecol. 2009;46: 1–9. https://doi.org/10.1111/j.1365-2664.2008.01596.x
  60. 60. Böhm M, Kemp R, Williams R, Davidson AD, Garcia A, McMillan KM, et al. Rapoport’s rule and determinants of species range size in snakes. Diversity and Distributions. 2017;23: 1472–1481. https://doi.org/10.1111/ddi.12632
  61. 61. IUCN Standards and Petitions Committee. Guidelines for Using the IUCN Red List Categories and Criteria. Version 16. Prepared by the Standards and Petitions Committee. Downloadable from https://www.iucnredlist.org/documents/RedListGuidelines.pdf; 2024.
  62. 62. Maclean IMMD Mosedale JR, Bennie JJ. Microclima: An R package for modelling meso and microclimate. Methods Ecol. Evol. 2018;10: 280–290. https://doi.org/10.1111/2041-210X.13093
  63. 63. Toro‐Cardona FA, Parra JL, Rojas‐Soto OR. Predicting daily activity time through ecological niche modelling and microclimatic data. J. Anim. Ecol. 2023;92: 925–935. https://doi.org/10.1111/1365-2656.13895 pmid:36744653
  64. 64. Huang SP, Porter WP, Tu MC, Chiou CR. Forest cover reduces thermally suitable habitats and affects responses to a warmer climate predicted in a high-elevation lizard. Oecologia. 2014;175: 25–35. https://doi.org/10.1007/s00442-014-2882-1 pmid:24446103
  65. 65. Huey RB, Peterson CR, Arnold SJ, Porter WP. Hot rocks and not-so-hot rocks: retreat-site selection by garter snakes and its thermal consequences. Ecol. 1989;70: 931–944. https://doi.org/10.2307/1941360
  66. 66. Groth J, Irmler U. Spatial distribution of earthworms (Oligochaeta: Lumbricidae) and the relationship to environmental parameters in northern German wet grassland. Faun. Ökol Mitt. 2013;9: 301–310.
  67. 67. Poier KR, Richter J. Spatial distribution of earthworms and soil properties in an arable loess soil. Soil Biol Biochem. 1992;24: 1601–1608. https://doi.org/10.1016/0038-0717(92)90157-S
  68. 68. Balestrin RL, Di-Bernardo M, Moreno AG. Feeding ecology of the neotropical worm snake Atractus reticulatus in southern Brazil. Herpetol. J. 2007;17: 62–64.
  69. 69. Martins M, Oliveira ME. Natural history of snakes in forest of the Manaus region, Central Amazonia, Brazil. Herpetol. Nat. Hist. 1998;6: 78–150.
  70. 70. Keddy PA. Assembly and response rules: two goals for predictive community ecology. J Veg Sci. 1992;3: 157–164. https://doi.org/10.2307/3235676
  71. 71. Keddy PA, Laughlin DC. A framework for community ecology: species pools, filters and traits. Cambridge: Cambridge University Press. 2021.
  72. 72. Kraft NJ, Adler PB, Godoy O, James EC, Fuller S, Levine JM. Community assembly, coexistence and the environmental filtering metaphor. Funct Ecol. 2015;29: 592–599. https://doi.org/10.1111/1365-2435.12345
  73. 73. Gutiérrez-Zuluaga AM, González-Quevedo C, Oswald JA, Terrill RS, Pérez-Emán JL, Parra JL. Genetic data and niche differences suggest that disjunct populations of Diglossa brunneiventris are not sister lineages. J Ornithol. 2021;138: 1–14. https://doi.org/10.1093/ornithology/ukab015
  74. 74. Londoño C, Cleef A, Madriñán S. Angiosperm flora and biogeography of the páramo region of Colombia, Northern Andes. Flora: Morphol Distrib Funct Ecol. 2014;209: 81–87. https://doi.org/10.1016/j.flora.2013.11.006
  75. 75. Bailey L, Simons T, Pollock K. Estimating site occupancy and species detection probability parameters for terrestrial salamanders. Ecol Appl. 2004;14:692–702. https://doi.org/10.1890/03-5012
  76. 76. Hampton P. A comparison of the success of artificial cover types for capturing amphibians and reptiles. Amphib-reptil. 2007;28: 433–437. https://doi.org/10.1890/03-5012
  77. 77. Dauby G, Stévart T, Droissart V, Cosiaux A, Deblauwe V, Simo-Droissart M, et al. ConR: An R package to assist large-scale multispecies preliminary conservation assessments using distribution data. Ecol Evol. 2017;7: 11292–11303. https://doi.org/10.1002/ece3.3704 pmid:29299301