biogeografia-ago2025

David Walter Rössel-Ramírez ^a, Jorge Palacio-Núñez ^b, *, Santiago Espinosa ^a, Juan Felipe Martínez-Montoya ^b, Genaro Olmos-Oropeza ^b

^a Universidad Autónoma de San Luis Potosí, Facultad de Ciencias, Av. Parque Chapultepec 1570, Privadas del Pedregal, 78295 San Luis Potosí, San Luis Potosí, Mexico 

^b Colegio de Postgraduados, Campus San Luis Potosí, Calle de Iturbide 73, San Agustín, 78622 Salinas de Hidalgo, San Luis Potosí, Mexico 

*Corresponding author: jpalacio@colpos.mx (J. Palacio-Núñez)

Abstract

The wetland formed by the Media Luna Spring is used for tourism, generating landscape modifications to the habitat of fish populations. Given this, the present study focused on quantifying these changes and correlating them to the spatial distribution and habitat affinity of the endemic cichlid fish Herichthys labridens, over 2 decades. Its spatial distribution was modeled and its suitable core habitat was estimated, considering their adult and juvenile life stages during 3 summer events (1999, 2009, and 2019). Based on occurrence records and the variables of water depth and underwater coverage, the variation in the spatial distribution of this fish species was evaluated in each summer event using the DOMAIN Model. A suitable core habitat was also identified from habitat resistance analysis. The highest probability of distribution (DP > 0.70) of both stages always occurred in areas with a high density of underwater vegetation, with a maximum depth of 1.5 m. On a spatial-temporal scale, the core habitat with medium and high suitability (HR ≤ 0.01) for adults and juveniles was to the extent that maintained its underwater coverage characteristics. This information is important for delimiting a conservation area of H. labridens in the Media Luna Spring.

Keywords: DOMAIN model; Endemic freshwater fish; Habitat resistance; Predictor variables; Semi-arid zones; Spatial-temporal scale

Distribución espacial y hábitat núcleo idóneo de Herichthys labridens en el manantial Media Luna, San Luis Potosí, México

Resumen

El humedal formado por el manantial Media Luna tiene uso turístico, con modificaciones paisajísticas del hábitat de las poblaciones de peces. El presente estudio se enfocó en correlacionar y cuantificar estos cambios con la distribución espacial y afinidad al hábitat del pez endémico Herichthys labridens, en el transcurso de 2 décadas. Se modeló su distribución y se estimó su hábitat núcleo idóneo, considerando sus estadios de vida adulto y juvenil, en 3 eventos de verano (1999, 2009 y 2019). A partir de registros de presencia y las variables profundidad del agua y cobertura subacuática, la variación de la distribución espacial fue evaluada en cada verano mediante el modelo DOMAIN. También se identificó el hábitat núcleo idóneo a partir del análisis de resistencia al hábitat. La mayor probabilidad de distribución (DP > 0.70) de ambos estadios siempre se presentó en zonas con alta densidad de vegetación subacuática, con profundidad máxima de 1.5 m. A escala espacio-temporal, el hábitat núcleo con idoneidad media y alta (HR ≤ 0.01) para adultos y juveniles fue la extensión que conservó sus características de cobertura subacuática. Esta información es importante para delimitar un área de conservación de H. labridens en el manantial Media Luna.

Palabras clave: Modelo DOMAIN; Peces dulceacuícolas endémicos; Resistencia al hábitat; Variables predictoras; Zonas semiáridas; Escala espacio-temporal 

Introduction

In freshwater systems (e.g., rivers, lakes, springs) the distribution of each fish species is governed by environmental (e.g., temperature, dissolved oxygen, water chemistry) and habitat (e.g., water depth, underwater coverage) factors (Matthews et al., 2004), that act on a temporal and spatial scale. This has fueled research interest in spatial ecology (Elith & Leathwick, 2009), however, in Mexico, these kinds of studies applied to freshwater fish are scarce, especially for endemic and endangered fishes (Ceballos et al., 2018; de la Vega Salazar, 2003). A great richness of endemic fish inhabits the semi-arid springs (closed-type ecosystems) of the country (Contreras-Balderas, 1969; Contreras-MacBeath et al., 2014; Miller, 1984), having a variable and dynamic allopatric distribution on a temporal and spatial scale (Eby et al., 2003). These native fish communities are well-adapted to various environmental conditions and habitat compositions (Contreras-Balderas, 1969; Miller et al., 2005). Nonetheless, fish species are still susceptible to habitat modifications (e.g., elimination of underwater vegetation) and competition with exotic fishes (Lozano-Vilano et al., 2021; Ruiz-Campos et al., 2006), which constitutes a threat to their survival (de la Vega Salazar, 2003; Torres-Orozco & Pérez-Hernández, 2011). 

A clear example of a semiarid spring with habitat modification by anthropic pressure is the Media Luna Spring in San Luis Potosí, Mexico (Galván Meza et al., 2018; Palacio-Núñez et al., 2010). This spring stands out for its extension and cultural popularity (Damián-Santiago, 2015), as well as its particular landscape and crystalline-sulfate water of touristic interest (Galván Meza et al., 2018; Salazar et al., 2002). This site is also the habitat of 4 species of endemic fishes (Ceballos et al., 2018) and is the main reservoir of the endemic cichlid Herichthys labridens (Pellegrin, 1903) (Supplementary material: Figure Sup1). In this ecosystem, this species has been documented as a generalist species, with greater activity from ~ 0.1 m to over ~ 10.5 m depth, in sites ranked from highly vegetated to bare bottom (Miller et al., 2005; Palacio-Núñez et al., 2015; Soto-Galera et al., 2019). In addition, H. labridens current population conservation status is endangered (species status maintained by IUCN, 2024).

Although H. labridens is under a conservation category, it is still affected by habitat alterations caused by natural and/or man-made disturbances —e.g., drought, tourist pressure, changes in the composition of underwater vegetation— (Palacio-Núñez et al., 2010; Ruetz III et al., 2005). Therefore, attention must be paid to structural changes in the freshwater ecosystem and the ecological-spatial situation of this species, before reaching a population collapse and the extinction of the species, as has happened with several species in other sites (e.g., Hickley et al., 2004; Miranda et al., 2010). Especially, because the rates of biodiversity loss are higher in freshwater environments than in those of terrestrial and marine systems (Dudgeon et al., 2005; Ricciardi & Rasmussen, 1999). However, information about the degree of preference of H. labridens to its habitat and the variation of its spatial distribution in Media Luna is still scarce, particularly when life stages (i.e., juveniles and adults) are taken into account (Brandt, 1980). Moreover, due to possible habitat degradation, as has happened elsewhere, a study on a temporal scale allows us to determine the variability of habitat resistance for the species (Elith & Leathwick, 2009; Miranda et al., 2010; Ruetz III et al., 2005; Zeller et al., 2012). 

Taking this research approach, we hypothesized that in the Media Luna Spring, H. labridens has a heterogeneous distribution preferring geographic spaces with high underwater vegetation density that provide refuge habitat and food availability. In addition, we hypothesized that this species has a possible core habitat in conserved vegetated areas (i.e., low cost of resistance to the habitat). In the present study, we conducted the spatial distribution models for this species in adult and juvenile stages and estimated the suitable core habitat for both stages in 3 summer events separated by an interval of 10 years (years 1999, 2009, and 2019), in the Media Luna Spring. We used a combination of spatial distribution models and habitat resistance analysis to 1) determine the variation in H. labridens distribution, by life stage and to compare this distribution with the configuration of the underwater coverage, 2) convert the distribution probability into habitat resistance cost from the 3 summer events and estimate the suitable core habitat for H. labridens, and 3) also calculate the area of medium and high suitability of the core habitat. Such information is crucial to aid conservation measures for the key habitat of H. labridens. 

Materials and methods

The study was carried out at the Media Luna Spring, which is located in the semi-arid plain of Rioverde, San Luis Potosí, Mexico. We selected this site because it is one of the main reservoirs of H. labridens and is located within the known distribution range of the species (Miller et al., 2005; Palacio-Núñez et al., 2015). Together, the entire system covers ~ 7.49 ha of water surface (Fig. 1). Much of the water surface is covered by the macrophyte Nymphaea ampla (Wiersema et al., 2008), in association with other macrophytes of emergent, sub-emergent, and floating type (pers. obs., Palacio-Núñez, 2007). Given the ecological importance of the Media Luna Spring and the impact caused by tourism (Galván Meza et al., 2018; Macías-Andrade & Maldonado-González, 2011), on 7 June 2003, the site was cataloged as a protected natural area and categorized as a State Park (Periódico Oficial del Gobierno del Estado Libre y Soberano de San Luis Potosí, 2003; Segam, 2019).

To explore the variation in spatial distribution of H. labridens and to estimate its suitable core habitat, according to specific habitat conditions (e.g., water depth and underwater coverage), we studied a 20-year time interval, starting in 1999 (when the information records began) and ending in 2019. We sampled during 3 summer events separated by 10 years (1999, 2009, and 2019). Under these analysis conditions, assumptions about uniform spatial distribution and errors are avoided when evaluating temporal variation (Ruetz III et al., 2005). 

In the sampling design and systematization of the spatial-temporal analysis, we used the 14 sectors proposed by Palacio-Núñez et al. (2010), which were delimited according to the composition and structure of the underwater coverage, the depth of the water, and the degree of anthropogenic impact on the ecosystem (S1 to S14; Fig. 1; Supplementary material: Table Sup1). Despite the variations in structure and composition of the system over time, which were mainly due to tourism activities (e.g., diving, recreational swimming, splashing; Galván-Meza et al., 2018), we used the same extension and boundaries of the sectors during the study events. It is important to note that the spatial delimitation and validation of the sectors and the water surface were performed using the QGIS® software version 3.4.8 (Menke, 2019).

Specifically, during the summer periods, tourism increases at the study site (Galván-Meza et al., 2018), which influences the distributional range and habitat of H. labridens. Therefore, we obtained spatial records of this species, categorized by adult and juvenile life stages, for summer events of 1999, 2009, and 2019 (Table 1). The databases with spatial and habitat filtering were obtained from 2 sources: for 1999, from Palacio-Núñez et al. (2010), and for the subsequent events, from Rössel-Ramírez, Palacio-Núñez, Martínez-Montoya et al. (2024). These authors recorded each occurrence, by life stage, through direct observation method using snorkeling equipment and a GPS device. Also, to avoid registering the same organism twice in the fieldwork, they validated human observation with subaquatic cameras. 

Habitat conditions (e.g., vegetation, bottom types, elevation, and depth) influence the distribution of specialist or endemic species (Naoki et al., 2006). In the present study, we selected 2 predictor variables of the underwater habitat. Due to its relationship with the spatial distribution of various fish communities, the first predictor variable was water depth (McClain & Etter, 2005; Palacio-Núñez et al., 2010; Prchalová et al., 2009; Thrush et al., 2006). The second variable was the underwater coverage, which influences the food and habitat for fish fauna (Brosse & Lek, 2002; Brosse et al., 2007; Rozas & Odum, 1988). Both variables are related to abiotic and biotic factors important for fishes, such as temperature, dissolved oxygen (O₂), and biomass (Brandt, 1980; Brosse et al., 1999; Gido & Matthews, 2000; Matthews et al., 2004; Straškraba, 1974).

We obtained a digital bathymetric model for water depth (ranging from 0.0 to -36.2 m) and a raster with reclassification of underwater coverage for the 3 summer events (Rössel-Ramírez et al., 2023). The classification of underwater coverage included the “bare bottom” and “macrophyte vegetation” categories proposed by Palacio-Núñez et al. (2010) and Rössel-Ramírez et al. (2023) for the Media Luna Spring. The bare bottom was classified according to 1) depth, referring to bottom below 12 m, where underwater vegetation does not grow (this type of bottom is only present in S1); 2) tourism, which tends to eliminate vegetation from the sites; and 3) natural process, related to bottom clearings derived from sediment accumulation. 

The coverage provided by N. ampla was categorized as follows: 1) small carpet, plants smaller than 0.3 m and with high population density; 2) big carpet, plants with stems of variable length between 0.3 to 2.0 m and large ovoid leaves that overlap; and 3) mature shape, plants with floating and spaced leaves with white flowering. In terms of the presence of remnants of emergent vegetation of the big carpet and mature shape type (S10 to S14), the categories were: big carpet patch and mature shape patch. Thus, a total of 8 coverage categories were included. We loaded both water depth and underwater coverage variables into the geographic projection system WGS 84/Pseudo-Mercator (EPSG: 3857), by the summer event, keeping a Ground Sample Distance of 0.1 m. 

We used 2 methods in R^® and RStudio^® (RStudio Team, 2020) to evaluate both predictor variables in each summer event: the percentage of contribution (Dedman et al., 2017; Freund & Schapire, 1997) via a Gradient Boost Machine (GBM) model (Friedman, 2002), and the Pearson´s correlation to determine the multicollinearity or independence between variables (Guisan & Hofer, 2003). In the first method, this model allows us to estimate the relative contribution of predictor variables as a function of the occurrence of H. labridens (Dedman et al., 2017). For the parameterization, using the GBM package, we selected the Gaussian distribution with a cross-validation of 5,000 repetitions. Also, the model was trained with a random subgroup of training (80%) and testing (20%) from occurrence records (Friedman, 2002; Hijmans, 2012). Additionally, we generated 500 background points to validate the models.

In the 3 summer events, both water depth and underwater coverage had a percentage contribution ≥ 2%, so following the criteria of Awan et al. (2021) and Mohammadi et al. (2021), we selected both for our models (Supplementary material: Figure Sup2.). Subsequently, we evaluated the collinearity of the variables using the Pearson correlation coefficient, avoiding having a high redundancy between variables (Guisan & Hofer, 2003). We established a bivariate correlation below ± 0.70 for the 2 variables, based on a literature review (Awan et al., 2021). As a result, both predictor variables were independent in the 3 events and were included for modeling (Supplementary material: Table Sup2).

Distribution models are a suitable tool for understanding the geographic patterns of species over time (Castillo-Torres et al., 2017; Joy & Death, 2002; Maloney et al., 2013). Carpenter et al. (1993) and Naoki et al. (2006) propose DOMAIN as the appropriate model to project the distribution of endemic species. Therefore, to estimate the distribution of H. labridens, by life stage, we ran this model in Rstudio^® for the 3 summer events, loading the water depth and underwater coverage variables and the occurrence records. The packages used were: sp (Pebesma, 2021), dismo, raster, maptools (Hijmans & Elith, 2017), RColorBrewer (Neuwirth, 2014), pander (Farlane, 2013), Rcpp, lattice (Sarkar, 2008), caret (Kuhn et al., 2020), pROC (Robin et al., 2011), and sdm (Naimi & Araujo, 2016).

Table 1

Number of occurrence records of H. labridens in the Media Luna spring, San Luis Potosí, Mexico, organized by 3 summer events (years 1999, 2009, and 2019) and life stage (adult and juvenile) in each sector (S1 to S14). The total number of records is included.

Additionally, we generated 500 background points to validate the models; this number was a smaller amount than other studies (e.g., Barbet-Massin et al., 2012) due to the smaller extension of the Media Luna Spring (~7.49 ha). Then, both the occurrences and the background points were subdivided into a training group (80%) and a test group (20%). For modeling, we selected the training groups and the stack containing the variables of water depth and underwater coverage to estimate the DOMAIN probabilistic projections of adult and juvenile stages of H. labridens, in each summer event.

As a last step, we evaluated the prediction performance of each model using the area under the curve (AUC) of the receiver operating characteristic (ROC) curve (Fawcett, 2006). Likewise, we validated the models based on the ROC curve (Fielding & Bell, 1997). The values ​​of the ROC curve were evaluated, following the method of Pearce and Ferrier (2000); according to this method, AUC > 0.5 indicates a model that distinguishes between true presence and absence of a response variable (i.e., occurrence). For partial ROC validation, we followed the Manzanilla-Quiñones (2020) criterion, according to which an AUC ratio value > 1.0 indicates excellent model performance.

The raster layers with spatial distribution probability were loaded into the Qgis® version 3.28.4. We subsequently converted this probability to habitat resistance (Zeller et al., 2012), as a product of the Kernel algorithm, using Equation 1 (Wan et al., 2019): 

R = 1000^{(-1 x SDM)}

where R represents the habitat resistance cost assigned to each pixel value and SDM represents the spatial distribution derived from the previously projected models. Additionally, for all raster, the resistance values were rescaled to a range of 0 to 10 by linear interpolation. The minimum resistance (Rmin) was 0 when the distribution probability (DP) was 1, and the maximum resistance (Rmax) was 10 when the DP was 0, following the criterion of Mohammadi et al. (2021).

We calculated the suitability of the core habitat, for H. labridens juveniles and adults, from the accumulation of the habitat resistance cost for the 3 summer events, according to the method of Lawler et al. (2006). This analysis was performed to estimate the accumulation of habitat suitability based on the expected rate of distribution of the species under study (Kaszta et al., 2018).We thus explored the suitable core habitat of H. labridens by life stages.

Based on the habitat resistance (HR) and the new scale of the range of values (i.e., 1 to 10), for both life stages, we categorized the suitable core habitat by quantiles; so that the suitable habitat of the species is better adjusted (Anderson et al., 2003; Burneo et al., 2009). Therefore, we categorized the suitability of the habitat into 3 levels: high (adults, HR ≤ 0.01; juveniles, HR ≤ 0.03), medium (adults, 0.01 < HR ≤ 0.64; juveniles, 0.03 < HR ≤ 1.26); and low (adults, HR > 0.64; juveniles, HR > 1.26). The spatial results were used to identify and regionalize the core area, which was a combination of medium and high suitability for both stages (e.g., Brind’Amour et al., 2005; Poizat & Pont, 1996). Additionally, we calculated the area (ha) per sector of both suitability categories for each sector, by life stage, in the Media Luna Spring. 

Results

In terms of the performance of the models, based on the ROC/AUC curve, H. labridens adults had AUCs of 0.67, 0.60, and 0.66 in 1999, 2009, and 2019, respectively; and juveniles had AUCs of 0.73, 0.68, and 0.76 in 1999, 2009, and 2019, respectively (Supplementary material: Figure Sup3). In the validation, all models had a density > 2 of positive cases at a predicted value between 0.6 and 0.9 (Supplementary material: Figure Sup4) in the partial ROC plots. In addition, these models presented a Bandwidth < 0.05 and an AUC ratio > 1.0, per life stage and summer event.

For the adult and juvenile stages of H. labridens, we obtained similar spatial distribution models. In the summer of 1999, the highest probability of distribution (DP > 0.70) was projected in sectors S2-S10; only for adult, the highest probability was too presented in S12. In these sectors, all types of underwater vegetation coverage were present, together with bare bottom by tourism and natural bare bottom; the maximum depth was ~2.5 m. In the following 2 events, the DP > 0.70 was observed only from S2 to S9 which are the sites with the greatest vegetation coverage in big carpet and mature shape (Fig. 2). 

Based on the analysis of HR accumulation in adult and juvenile stages, we identified the highest concentration of high and medium suitability (i.e., medium and low categories of cost of resistance) in sectors S2 to S9 (Fig. 3). For the adult stage, high suitability (RH ≤ 0.01) was observed in sectors S2-S4 and S9. For the juvenile stage, high suitability (RH ≤ 0.03) was observed in sectors S2-S9, particularly close to the shores of the spring. Sectors S1 and S10-S14 presented low and medium suitability for both adult and juvenile stages. Sector S1 had the lowest suitability, which represented the highest HR cost for adults (HR > 0.64) and juveniles (HR > 1.26).

Finally, we calculated the area with high habitat suitability (sectors S2 to S9), with a total area of 0.752 ha for adults and 1.855 ha for juveniles. This represented ~15% and ~36% of the total system extent, respectively. Meanwhile, medium suitability covered an area of 3.987 ha (~ 77%) for adults and 2.834 ha (~ 55%) for juveniles of H. labridens (Table 2). Based on our calculations, the combination of high and medium suitability in sectors S2-S9 covered 4.739 ha for adults and 4.689 ha for juveniles. This represented ~64% of the total area of the spring, now identified as the suitable core habitat for H. labridens. The remaining ~ 36% of the total area of the spring corresponded to sectors S1 and S10-S14, from which ~ 20% corresponded to areas with low habitat suitability in S1 (adults, 1.180 ha; juveniles, 1.169 ha).

Discussion

We have found that, in the Media Luna Spring, in the 3 summer events assessed, the adult and juvenile stages of H. labridens had a higher probability of distribution (DP > 0.70) in sectors S2-S9, which are characterized by high and medium habitat suitability with depth between ~ 0.5 m and ~ 1.5 m. These observations confirm that this species, like the majority of the Perciformes, have dominant spatial patterns in shallow sites, mainly towards the water edge (McClain & Etter, 2005; Prchalová et al., 2009). 

In terms of underwater coverage, for a period of 20 years (1999-2019), both the juvenile and adultstages of H. labridens had a higher preference to sectors with high density of vegetation in the categories big carpet and mature shape. The affinity to vegetated sites is easily understood because it ensures both food and refuge (Rozas & Odum, 1988), especially for juveniles (Artigas-Azas, 2004). Furthermore, in these sectors, there are patches of natural bare bottom, which are important for the reproduction of H. labridens; as has been observed in other cichlids studied (Miller, 1984: Miller et al., 2005). Brandt (1980) found similar conditions on habitat selection and spatial segregation for Perciformes in Lake Michigan, USA. Likewise, Brosse and Lek (2002) reported the assembly of geographic patterns of Eurasian perch (Perca fluviatilis) by habitat affinity in the littoral zone of Lake Pareloup, France. 

In the evaluation of the distribution models by the ROC/AUC curve, Fielding and Bell (1997) and Fawcett (2006) indicate that an AUC > 0.90 represents excellent prediction performance for species distribution models. However, Deshpande (2020) indicates that an AUC between 0.60 and 0.80 constitutes a better prediction classifier, more in line with the ecological reality of the species. Based on this, our models had AUC values ranging from 0.60 to 0.76 ​​for both the adult and juvenile stages in the 3 summer events. We thus conclude that our models had high sensitivity and low specificity, features that correspond to a high discrimination power, in at least ~ 60% of positive cases of occurrence and prediction performance (false positives rate was < 40%). In addition, in the validation by partial ROC, the models had a bandwidth < 0.05 and an AUC ratio > 1.0. Therefore, according to Manzanilla-Quiñones (2020), our models had statistical and ecological significance, across the 3 events of analysis. Thus, we considered that there was a good correlation between habitat variables and the occurrence of H. labridens to allow accurate modeling of its spatial distribution.

Table 2

Area (ha) by suitability category (medium and high) of the core habitat for H. labridens adults and juveniles in the Media Luna Spring, San Luis Potosí, Mexico. The area was calculated for each one from 14 sectors. The values in bold denote the greatest area occupied based on the suitability probability. In most cases, the greatest and most representative areas were between sectors S2 and S9.

Our DOMAIN models provided a good overview of the variation in the spatial distribution of the endemic H. labridens, in both life stages. In addition to that, previous work (Hirzel et al., 2006; Zeller et al., 2012) indicates that estimating core habitats is more accurate when work is carried out over long temporal scales, because it was possible to discern between low, medium, and high resistance habitat, according to the relevant variables (Lawler et al., 2006). Following this, our maps and tables of the suitable core habitat for H. labridens were more conclusive in quantifying its cost of resistance according to the variables of water depth and underwater coverage. We confirmed that the sectors S2-S9 represent the suitable core habitat for H. labridens in the adult and juvenile stages for 20 years at least. These sectors presented low-cost resistance values (i.e., high suitability of core habitat) in adults (RH < 0.01) and juveniles (RH < 0.03), especially towards the shores of the spring, where the maximum depth was ~ 1.5 m and the underwater vegetation coverage was big carpet and mature shape.

It should be noted that the Media Luna Spring has been modified by tourism (Damián-Santiago, 2015; Galván-Meza et al., 2018), causing changes in the composition of underwater coverage for over 2 decades. This was reflected in the progressive decrease in vegetation coverage, although the depth was not noticeably affected, where more than 90% of the spring surface area maintained a depth of less than 3.5 m (pers. obs.). Thus, the distribution area and core habitat, as well as a higher number of occurrences, from S2 to S9 for both adult and juvenile stages, is likely related to habitat degradation in tourism sectors, as has happened in other sites (Hickley et al., 2004; Miranda et al., 2010). Particularly, 2009 was a critical year due to habitat pressure from tourism (Chávez, 2016), linked to the drastic decrease in occurrences for both life stages. Despite this habitat pressure, this speciessurvives and reproduces in this water system, but in a suboptimal population state. Therefore, it is important to keep H. labridens under the category of endangered species as a conservation measure (Miranda et al., 2009).

In conclusion, we found that the suitable core habitat for H. labridens in Media Luna Spring is the region between S2 and S9, which is characterized by a combination of shallow sites associated with dense coverage of emergent, sub-emergent, and floating vegetation patches. It is important to note that these coverage combinations provide resources for feeding and habitat refuge. This information, obtained on a temporal scale, is important to increase our knowledge about the variation in spatial distribution and core habitat of this endemic and threatened species. With this, it is now possible to compare and associate occurrence records with changes in vegetation coverage or anthropogenic pressure. Based on our findings, we urgently propose establishing this estimated core habitat (sectors S2-S9) as a priority conservation area for H. labridens. This means that these sectors, located in the eastern region of the spring, should be adequately managed without eliminating vegetation or modifying the natural system. Meanwhile, Sectors S1 and S10 to S14 would be categorized as suitable for tourism use, with management and monitoring by the local authorities and ejido leaders at the site.

In addition, we consider that our study provides relevant information on the abundance and habitat preferences of H. labridens to update the management plan and implement habitat care campaigns for the Media Luna Spring. At the same time, to include reports for other endemic species in this spring, as A. toweri (Rössel-Ramírez, Palacio-Núñez, Espinosa et al., 2024). Taken together, our spatial results reinforce the importance of identifying conservation areas or regions in this semi-arid spring.

Acknowledgments

We thank Jesús Enríquez for their participation in fieldwork during the sampling events. In addition, we thank to the Secretaría de Ecología y Gestión Ambiental (Segam; Oficio: ECO.05.1068), the local authorities of Rioverde municipality, and the Jabalí ejido council for allowing us to carry out this study in the Media Luna Spring.

References

Anderson, R. P., Lew, D., & Peterson, A. T. (2003). Evaluating predictive models of species’ distributions: criteria for selecting optimal models. Ecological Modelling, 162, 211–232. https://doi.org/10.1016/S0304-3800(02)00349-6

Artigas-Azas, J. M. 2004). “Herichthys labridens (Pellegrin, 1903)”. Cichlid Room Companion. Retrieved on January 1^rst, 2024 from: https://cichlidae.com/species.php?id=209&lang=es. (crc10342)

Awan, M. N., Saqib, Z., Buner, F., Lee, D. C., & Pervez, A. (2021). Using ensemble modeling to predict breeding habitat of the red-listed Western Tragopan (Tragopan melanocephalus) in the Western Himalayas of Pakistan. Global Ecology and Conservation, 31, e01864. https://doi.org/10.1016/j.gecco.2021.e01864 

Barbet-Massin, M., Jiguet, F., Albert, C. H., & Thuiller, W. (2012). Selecting pseudo-absences for species distribution models: how, where and how many? Methods in Ecology and Evolution, 3, 327–338. https://doi.org/10.1111/j.2041-210x.2011.00172.x 

Brandt, S. B. (1980). Spatial segregation of adult and young-of-the-year alewives across a thermocline in Lake Michigan. Transactions of the American Fisheries Society, 109, 469–478. https://doi.org/10.1577/1548-8659(1980)109<469:SSOAAY>2.0.CO;2 

Brind’Amour, A., Boisclair, D., Legendre, P., & Borcard, D. (2005). Multiscale spatial distribution of a littoral fish community in relation to environmental variables. Limnology and Oceanography, 50, 465–479. https://doi.org/10.4319/lo.2005.50.2.0465 

Brosse, S., Grossman, G. D., & Lek, S. (2007). Fish assemblage patterns in the littoral zone of a European reservoir. Freshwater Biology, 52, 448–458. https://doi.org/10.1111/j.1365-2427.2006.01704.x 

Brosse, S., & Lek, S. (2002). Relationships between environmental characteristics and the density of age-0 Eurasian perch Perca fluviatilis in the littoral zone of a lake: a nonlinear approach. Transactions of the American Fisheries Society, 131, 1033–1043. http://dx.doi.org/10.1577/1548-8659(2002)131<1033:RBECAT>2.0.CO;2 

Brosse, S., Lek, S., & Dauba, F. (1999). Predicting fish distribution in a mesotrophic lake by hydroacoustic survey and artificial neural networks. Limnology and Oceanography, 44, 1293–1303. https://doi.org/10.4319/lo.1999.44.5.1293 

Burneo, S., González-Maya, J. F., & Tirira, D. (2009). Distribution and habitat modelling for Colombian weasel Mustela felipei in the Northern Andes. Small Carnivore Conservation, 41, 41–45. 

Castillo-Torres, P. A., Martínez-Meyer, E., Córdova-Tapia, F., & Zambrano, L. (2017). Potential distribution of native freshwater fish in Tabasco, Mexico. Revista Mexicana de Biodiversidad, 88, 415–424. https://doi.org/10.1016/j.rmb.2017.03.001 

Carpenter, G., Gillison, A. N., & Winter, J. (1993). DOMAIN: a flexible modelling procedure for mapping potential distributions of plants and animals. Biodiversity & Conservation, 2, 667–680.

Ceballos, G., Pardo, E. D., Estévez, L. M., & Pérez, H. E. (2018). Los peces dulceacuícolas de México en peligro de extinción. Fondo de Cultura Económica. Cd. de México. 

Chávez, M. G. G. (2016). El impacto del turismo en la conservación de la biodiversidad en San Luis Potosí. Sociedad y Ambiente, 11, 148–159. 

Contreras-Balderas, S. (1969). Perspectivas de la ictiofauna en las zonas áridas del norte de México. In Simposium Internacional sobre Aumento de Producción de Alimentos en Zonas Áridas. Centro Internacional de Estudios sobre Tierras Áridas, Universidad de Texas Tech.

Contreras-MacBeath, T., Rodríguez, M. B., Sorani, V., Goldspink, C., & Reid, G. M. (2014). Richness and endemism of the freshwater fishes of Mexico. Journal of Threatened Taxa, 6, 5421–5433. https://doi.org/10.11609/JoTT.o3633.5421-33 

Damián-Santiago, F. I. (2015). Caracterización del perfil del ecoturista que visita el Parque Estatal Manantial de la Media Luna, en el Ejido El Jabalí, municipio de Rioverde, San Luis Potosí, México (Bachelor’s Thesis). Departamento Forestal, Universidad Autónoma Agraria Antonio Narro. Saltillo, Coahuila, México.

Dedman, S., Officer, R., Clarke, M., Reid, D. G., & Brophy, D. (2017). Gbm. auto: a software tool to simplify spatial modelling and Marine Protected Area planning. Plos One, 12, e0188955. https://doi.org/10.1371/journal.pone.0188955

De la Vega-Salazar, M. Y. (2003). Situación de los peces dulceacuícolas en México. Ciencias, 72, 20–30. 

Deshpande, R. (2020). ROC Curve and AUC in Machine learning and R pROC Package. Retrieved on September 08th, 2023 from: https://medium.com/swlh/roc-curve-and-auc-detailed-understanding-and-r-proc-package-86d1430a3191

Dudgeon, D., Arthington, A. H., Gessner, M. O., Kawabata, Z. I., Knowler, D. J., Lévêque, C. et al. (2005). Freshwater biodiversity: importance, threats, status and conservation challenges. Biological Reviews, 81, 163. https://doi.org/10.1017/s1464793105006950

Eby, L. A., Fagan, W. F., & Minckley, W. L. (2003). Variability and dynamics of a desert stream community. Ecological Applications, 13, 1566–1579. https://doi.org/10.1890/02-5211 

Elith, J., & Leathwick, J. R. (2009). Species distribution models: ecological explanation and prediction across space and time. Annual Review of Ecology, Evolution, and Systematics, 40, 677–697. https://doi.org/10.1146/annurev.ecolsys.110308.120159 

Farlane, J. M. (2013). Pandoc User’s Guide. Accessed on RStudio: Integrated Development for R. RStudio, PBC, Boston, MA URL: http://www.rstudio.com/

Fawcett, T. (2006). An introduction to ROC analysis. Pattern Recognition Letters, 27, 861–874. https://doi.org/10.1016/j.patrec.2005.10.010

Fielding, A. H., & Bell, J. F. (1997). A review of methods for the assessment of prediction errors in conservation presence/absence models. Environmental Conservation, 24, 38–49.

Freund, Y., & Schapire, R. E. (1997). A decision-theoretic generalization of on-line learning and an application to boosting. Journal of Computer and System Sciences, 55, 119–139. https://doi.org/10.1006/jcss.1997.1504 

Friedman, J. H. (2002). Stochastic gradient boosting. Computational Statistics and Data Analysis, 38, 367–378. https://doi.org/10.1016/S0167-9473(01)00065-2 

Galván-Meza, C. J., Flores-Castillo, E., Espericueta-Bocanegra, E. D., & Gutiérrez-Pérez, M. G. (2018). Análisis del impacto multidisciplinar del turismo dentro del ejido El Jabalí – Media Luna – San Luis Potosí. Medio Ambiente, Sustentabilidad y Vulnerabilidad Social, 5, 419–434.

Gido, K. B., & Matthews, W. J. (2000). Dynamics of the offshore fish assemblage in a southwestern reservoir (Lake Texoma, Oklahoma-Texas). Copeia, 2000, 917–930. http://dx.doi.org/10.1643/0045-8511(2000)000[0917:DOTOFA]2.0.CO;2 

Guisan, A., & Hofer, U. (2003). Predicting reptile distributions at the mesoscale: relation to climate and topography. Journal of Biogeography, 30, 1233–1243. https://doi.org/10.1046/j.1365-2699.2003.00914.x

Hickley, P., Muchiri, M., Boar, R., Britton, R., Adams, C., Gichuru, N. et al. (2004). Habitat degradation and subsequent fishery collapse in Lakes Naivasha and Baringo, Kenya. International Journal of Ecohydrology & Hydrobiology, 4, 503–517. http://karuspace.karu.ac.ke/handle/20.500.12092/1918 

Hijmans, R. J. (2012). Cross-validation of species distribution models: removing spatial sorting bias and calibration with a null model. Ecology, 93, 679–688. https://doi.org/10.1890/11-0826.1

Hijmans, R. J., & Elith, J. (2017). Species distribution modeling with R. R CRAN Project. https://www.researchgate.net/file.PostFileLoader.html?id=5730d34496b7e4bb46356f31&assetKey=AS%3A359897272209410%401462817604682 

Hirzel, A. H., Le Lay, G., Helfer, V., Randin, C., & Guisan, A. (2006). Evaluating the ability of habitat suitability models to predict species presences. Ecological Modelling, 199, 142–152. https://doi.org/10.1016/j.ecolmodel.2006.05.017 

IUCN (International Union for Conservation of Nature). (Version 2024.2). The IUCN red list of threatened species. Retrieved on January 25th, 2024 from: https://www.iucnredlist.org/

Joy, M. K., & Death, R. G. (2002). Predictive modelling of freshwater fish as a biomonitoring tool in New Zealand. Freshwater Biology, 47, 2261–2275. https://doi.org/10.1046/j.1365-2427.2002.00954.x 

Kaszta, Ż., Cushman, S. A., Sillero-Zubiri, C., Wolff, E., & Marino, J. (2018). Where buffalo and cattle meet: modelling interspecific contact risk using cumulative resistant kernels. Ecography, 41, 1616–1626. https://doi.org/10.1111/ecog.03039 

Kuhn, M., Wing, J., Weston, S., Williams, A., Keefer, C., Engelhardt, A. et al. (2020). Package ‘caret’: classification and regression training, R-package version 6.0-p. Misc Functions for Training and Plotting Classification and Regression Models.

Lawler, J. J., White, D., Neilson, R. P., & Blaustein, A. R. (2006). Predicting climate-induced range shifts: model differences and model reliability. Global Change Biology, 12, 1568–1584. https://doi.org/10.1111/j.1365-2486.2006.01191.x 

Lozano-Vilano, M. L., Contreras-Balderas, A. J., Ruiz-Campos, G., & García-Ramírez, M. E. (2021). Current conservation status of some freshwater species and their habitats in México. In D. L. Propst, J. E, Williams, K. R. Bestgen, & C. W. Hoagstrom (Eds.), Standing between life and extinction: ethics and ecology of conserving aquatic species in North American deserts (pp. 79–88). Chicago: The University of Chicago Press.

Macías-Andrade, M. E., & Maldonado-González, D. P. (2011). La práctica ecoturística en el Parque Estatal de la Media Luna en Rioverde SLP (Tesis). Facultad de Economía, Universidad Autónoma de San Luis Potosí. San Luis Potosí, México.

Maloney, K. O., Weller, D. E., Michaelson, D. E., & Ciccotto, P. J. (2013). Species distribution models of freshwater stream fishes in Maryland and their implications for management. Environmental Modeling & Assessment, 18, 1–12. https://doi.org/10.1007/s10666-012-9325-3 

Manzanilla-Quiñones, U. (2020). Validación de modelos de distribución con ayuda de la plataforma niche toolbox de Conabio. Retrieved on June 29th, 2023 from: https://www.researchgate.net/publication/341820933 

Matthews, W. J., Gido, K. B., & Gelwick, F. P. (2004). Fish assemblages of reservoirs, illustrated by Lake Texoma (Oklahoma-Texas, USA) as a representative system. Lake and Reservoir Management, 20, 219–239. https://doi.org/10.1080/07438140409354246. 

Menke, K. (2019). Discover QGIS 3. x. Locate Press.

McClain, C. R., & Etter, R. J. (2005). Mid-domain models as predictors of species diversity patterns: bathymetric diversity gradients in the deep sea. Oikos, 109, 555–566. https://doi.org/10.1111/j.0030-1299.2005.13529.x 

Miller, R. R. (1984). La Media Luna, San Luis Potosí, at edge of Chihuahua Desert, México. In Proceedings of the Desert Fishes Council Co. (Ed.). Volumes XVI-XVIII. Annual Symposium Bishop, CA. 67–72.

Miller, R. R., Minckley, W. L., Norris, S. M., & Gach, M. H. (2005). Freshwater fishes of Mexico (Núm. QL 629. M54 2005). Chicago: University of Chicago Press.

Miranda, L. E., Spickard, M., Dunn, T., Webb, K. M., Aycock, J. N., & Hunt, K. (2010). Fish habitat degradation in US reservoirs. Fisheries, 35, 175–184. https://doi.org/10.1577/1548-8446-35.4.175 

Miranda, R., Monks, S., Galicia, D., & Pulido-Flores, G. (2009). Threatened fishes of the world: Herichthys labridens (Pellegrin, 1903). Environmental Biology of Fishes, 86, 377. https://doi.org/10.1007/s10641-009-9528-x

Mohammadi, A., Almasieh, K., Nayeri, D., Ataei, F., Khani, A., López-Bao, J. V. et al. (2021). Identifying priority core habitats and corridors for effective conservation of brown bears in Iran. Scientific Reports, 11, 1044. https://doi.org/10.1038/s41598-020-79970-z 

Naimi, B., & Araujo, M. B. (2016). sdm: a reproducible and extensible R platform for species distribution modelling. Ecography, 39, 368–375. https://doi.org/10.1111/ecog.01881

Naoki, K., Gómez, M. I., López, R. P., Meneses, R. I., & Vargas, J. (2006). Comparación de modelos de distribución de especies para predecir la distribución potencial de vida silvestre en Bolivia. Ecología en Bolivia, 41, 65–78. 

Neuwirth, E. (2014). RColorBrewer: ColorBrewer palettes. R package version 1.1-2. The R Foundation.

Palacio-Núñez, J. (2007). Designación de zonas prioritarias de conservación en el Parque Estatal de la Media Luna (México): utilización de aves y peces como bioindicadores (Ph.D. Thesis). Instituto de Investigación de la Biodiversidad (CIBIO), Universidad de Alicante. Alicante, España.

Palacio-Núñez, J., Olmos-Oropeza, G., Verdú, J. R., Galante, E., Rosas-Rosas, O. C., Martínez-Montoya, J. F. et al. (2010). Traslape espacial de la comunidad de peces dulceacuícolas diurnos en el sistema de humedal Media Luna, Rioverde, SLP, México. Hidrobiológica, 20, 21–30. 

Palacio-Núñez, J., Martínez-Montoya, J. F., Olmos-Oropeza, G., Martínez-Calderas, J. M., Clemente-Sánchez, F., & Enríquez, J. (2015). Distribución poblacional y abundancia de los peces endémicos de la Llanura de Rioverde, SLP, México. Agroproductividad, 8, 17–24.

Pearce, J., & Ferrier, S. (2000). Evaluating the predictive performance of habitat models developed using logistic regression. Ecological Modelling, 133, 225–245. https://doi.org/10.1016/S0304-3800(00)00322-7 

Pebesma, E. (2021). Map overlay and spatial aggregation in sp, Report 253. Institute for Geoinformatics. University of Muenster, Weseler Strasse.

Pellegrin, J. (1903). Description de Cichlidés nouveax de la collection du Muséum. Bulletin du Musée National d’Histoire Naturelle, 9, 120–125.

Periódico Oficial del gobierno del estado libre y soberano de San Luis Potosí (07 de junio de 2003). Declaración de Área Natural Protegida bajo la modalidad de “Parque Estatal” denominado “Manantial de la Media Luna”. Retrieved on October 07^th, 2023 from: http://201.144.107.246/InfPubEstatal2/_SECRETAR%C3%8DA%20DE%20ECOLOG%C3%8DA%20Y%20GESTI%C3%93N%20AMBIENTAL/Art%C3%ADculo%2018.%20fracc.%20II/Normatividad/Decretos/Decreto%20Media%20Luna.pdf. Secretaría General de Gobierno, 2–12.

Poizat, G., & Pont, D. (1996). Multi-scale approach to species–habitat relationships: juvenile fish in a large river section. Freshwater Biology, 36, 611–622. https://doi.org/10.1046/j.1365-2427.1996.00129.x 

Prchalová, M., Kubečka, J., Čech, M., Frouzová, J., Draštík, V., Hohausová, E. et al. (2009). The effect of depth, distance from dam and habitat on spatial distribution of fish in an artificial reservoir. Ecology of Freshwater Fish, 18, 247–260. https://doi.org/10.1111/j.1600-0633.2008.00342.x 

Ricciardi, A., & Rasmussen, J. B. (1999). Extinction rates of North American freshwater fauna. Conservation Biology, 13, 1220–1222. https://doi.org/10.1046/j.1523-1739.1999.98380.x 

Robin, X., Turck, N., Hainard, A., Tiberti, N., Lisacek, F., Sánchez, J. et al. (2011). “pROC: an open-source package for R and S+ to analyze and compare ROC curves”. BMC Bioinformatics, 12, 1–77. https://doi.org/10.1186/1471-2105-12-77

Rössel-Ramírez, D. W., Palacio-Núñez, J., Espinosa, S., & Martínez-Montoya, J. F. (2023). Raster layers of underwater coverage and water depth of the Media Luna spring, Mexico, generated from data from three summer periods (in the years 1999, 2009, and 2019). Zenodo. https://doi.org/10.5281/zenodo.10253601 

Rössel-Ramírez, D. W., Palacio-Núñez, J., Martínez-Montoya, J. F., & Espinosa, S. (2024). Occurrences records of Herichthys labridens (Cichliformes: Cichlidae), with associated habitat information, in the Media Luna spring, San Luis Potosí, Mexico (Version 1) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.14231104 

Rössel-Ramírez, D. W., Palacio-Núñez, J., Espinosa, S., & Martínez-Montoya, J. F. (2024). Temporal variation in the relative abundance, suitable habitat selection, and distribution of Ataeniobius toweri (Meek, 1904) (Goodeidae), by life stages, in the Media Luna spring, Mexico. Environmental Biology of Fishes, 107, 173–188. https://doi.org/10.1007/s10641-024-01520-7 

Rozas, L. P., & Odum, W. E. (1988). Occupation of submerged aquatic vegetation by fishes: testing the roles of food and refuge. Oecologia, 77, 101–106. https://www.jstor.org/stable/4218746

RStudio Team (2020). RStudio: Integrated Development for R. RStudio, PBC, Boston, MA URL: http://www.rstudio.com/

Ruetz III, C. R., Trexler, J. C., Jordan, F., Loftus, W. F., & Perry, S. A. (2005). Population dynamics of wetland fishes: spatio-temporal patterns synchronized by hydrological disturbance? Journal of Animal Ecology, 74, 322–332. https://doi.org/10.1111/j.1365-2656.2005.00926.x 

Ruiz-Campos, G., Camarena-Rosales, F., Contreras-Balderas, S., Reyes-Valdez, C. A., De La Cruz-Agüero, J., & Torres-Balcázar, E. (2006). Distribution and abundance of the endangered killifish, Fundulus lima (Teleostei: Fundulidae), in oases of central Baja California Peninsula, Mexico. The Southwestern Naturalist, 51, 502–509.

Salazar, H. C., Sáenz, E. O., & Rivera, J. R. A. (2002). Agua de riego en la región de Rioverde, San Luis Potosí, México. Tecnología y Ciencias del Agua, 17, 37–56.

Sarkar, D. (2008). lattice: multivariate data visualization with R. Springer-Verlag, New York. http://lmdvr.r-forge.r-project.org/ 

Segam (Secretaría de Ecología y Gestión Ambiental). (2019). Declaratoria de Área Natural Protegida bajo la modalidad de Parque Estatal denominado “Manantial de la Media Luna”. Diario Oficial de San Luis Potosí. 7 de junio del 2003. Retrieved on May 12th, 2019 from: www.segam.gob.mx

Soto-Galera, E., Mejía-Guerrero, O., & Pérez-Miranda, F. (2019). Herichthys labridens. The IUCN Red List of Threatened Species 2019: e.T192897A2179729. Retrieved on 04 December 2023 from: http://dx.doi.org/10.2305/IUCN.UK.2019-2.RLTS.T192897A2179729

Straškraba, M. (1974). Seasonal cycle of the depth distribution of fish in the Klıcava Reservoir. Czech Journal of Animal Sciences, 19, 656–660. 

Thrush, S., Dayton, P., Cattaneo-Vietti, R., Chiantore, M., Cummings, V., Andrew, N. et al. (2006). Broad-scale factors influencing the biodiversity of coastal benthic communities of the Ross Sea. Deep Sea Research Part II: Topical Studies in Oceanography, 53, 959–971. https://doi.org/10.1016/j.dsr2.2006.02.006 

Torres-Orozco, R. E., & Pérez-Hernández, M. A. (2011). Los peces de México: una riqueza amenazada. Revista Digital Universitaria, 12, 1–15. 

Wan, H. Y., Cushman, S. A., & Ganey, J. L. (2019). Improving habitat and connectivity model predictions with multi-scale resource selection functions from two geographic areas. Landscape Ecology, 34, 503–519. https://doi.org/10.1007/s10980-019-00788-w 

Wiersema, J. H., Novelo, A. R., & Bonilla-Barbosa, J. R. (2008). Taxonomy and typification of Nymphaea ampla (Salisb.) DC. sensu lato (Nymphaeaceae). Taxon, 57, 967–974. https://doi.org/10.1002/tax.573025

Zeller, K. A., McGarigal, K., & Whiteley, A. R. (2012). Estimating landscape resistance to movement: a review. Landscape Ecology, 27, 777–797. https://doi.org/10.1007/s10980-012-9737-0

Héctor Alejandro Hernández-Castellanos ^a, *, Víctor Landa-Jaime ^a y Jesús Emilio Michel-Morfín ^b

^a Universidad de Guadalajara, Centro Universitario de la Costa Sur, Departamento de Estudios para el Desarrollo Sustentable de Zonas Costeras, Gómez Farías Núm. 82, 48980 San Patricio-Melaque, Jalisco, México

^b Universidad de Guadalajara, Centro Universitario de Ciencias Biológicas y Agropecuarias, Departamento de Ecología Aplicada, Camino Ramón Padilla Sánchez Núm. 2100, 45510 Zapopan, Jalisco, México 

*Autor para correspondencia: hecalehercas96@gmail.com (H.A. Hernández-Castellanos)

Resumen

El archipiélago de Revillagigedo se ubica en el Pacífico este tropical y consta de 4 islas volcánicas, isla Clarión es la más occidental y remota de este parque marino de México. Esta región está influenciada por aguas frías provenientes de la corriente de California y aguas cálidas de la corriente Norecuatorial, lo que resulta en una zona de convergencia, con una alta biodiversidad marina. Clarión es una isla pequeña y distante, dominada por acantilados, con playas arenosas-rocosas en su cara sur, lo que se traduce en áreas poco accesibles para llevar a cabo investigaciones. El objetivo de esta contribución es dar a conocer una lista de especies de babosas marinas presentes en la isla. El estudio se llevó a cabo entre enero y marzo de 2023, y febrero-marzo de 2024, a través del muestreo continuo en playas, pozas intermareales y zonas arrecifales someras de la cara sur de la isla. En total, se registraron 14 especies de babosas marinas, 13 registros nuevos para esta isla y 5 para el archipiélago de Revillagigedo. Este estudio contribuye significativamente al conocimiento de la biodiversidad de las babosas marinas heterobranquias insulares del Pacífico tropical.

Palabras clave: Moluscos; Gasterópodos; Pacífico este tropical; Zona intermareal; Isla oceánica; Biodiversidad marina

New records of sea slugs (Gastropoda: Heterobranchia) from Clarion Island, Revillagigedo Archipelago, Mexico

Abstract

The Revillagigedo Archipelago is located on the East Tropical Pacific; it consists of 4 volcanic islands. Clarion Island is the most western and remote island of this marine national park. The region is influenced by the cold currents from California and the warm waters of the North Equatorial current, creating a convergence zone, with high marine biodiversity. Clarion is a small and distant island, dominated by high cliffs with rocky sandy beaches on the south side, with hard-to-access areas for research. The objective of this contribution is to provide a species list of the sea slugs present on the island. The study was performed from January to March 2023, and February to March 2024, through a continual survey of the beaches, intertidal pools and shallow reef areas in the south side of the island. In total, we recorded 14 species of sea slugs, 13 were new records for this island and 5 new records for the Revillagigedo Archipelago. This study significantly contributes to the knowledge of the biodiversity of island heterobranch sea slugs of the Tropical Pacific.

Keywords: Mollusks; Gastropods; Eastern Tropical Pacific; Intertidal zone; Oceanic island; Marine biodiversity

Introducción

El archipiélago de Revillagigedo es un conjunto de 4 islas de origen volcánico ubicadas en la ecorregión del Pacífico Transicional Mexicano. Se localiza en la zona de encuentro de las provincias Panámica, Californiana y de Cortés, en un área de convergencia de aguas frías provenientes de la corriente de California y aguas cálidas de la corriente Norecuatorial, lo que propicia valores altos de diversidad, tanto en tierra como en el océano (Conanp, 2018). Comprende, de mayor a menor tamaño, a las islas Socorro, Clarión, San Benedicto y el islote Roca Partida.

Clarión es la isla más occidental del Archipiélago (18°21’ N, 114°45’ O) (fig. 1A). Se ubica a 390 km al oeste de isla Socorro, 707 km al suroeste de Baja California Sur (Cabo San Lucas) y 1,100 km al oeste de Colima (Manzanillo) (Conanp, 2018). Comprende un área de 1,925 ha, 8.5 km de largo y 3.6 km de ancho y su punto más alto es el monte Gallegos (335 m snm; fig. 1A). En la zona oceánica predomina la influencia de las corrientes de California y la corriente Norecuatorial, con temperaturas del agua entre 20 y 28 °C. Presenta un régimen de mareas de tipo semidiurno, con oleaje alto y periodicidad de surgencias estacionales (Conanp, 2018). La costa de la isla está conformada por acantilados de entre 3 y 250 m de alto en sus lados norte, este y oeste, mientras que en la porción sur se localizan 3 playas arenosas, separadas entre ellas por formaciones rocosas. Estas playas se extienden por 2.5 km aproximadamente, lo que constituye 10% del litoral de la isla (Conanp, 2018; Holroyd y Trefy, 2010).

Las babosas marinas son moluscos de la subclase Heterobranchia, clase Gastropoda, que suelen presentar las branquias y el intestino en la parte posterior del cuerpo, proceso que ocurre en fases tempranas del desarrollo, además de carecer de concha externa en su mayoría (salvo algunas especies que poseen una pequeña concha, interna o externa) (Behrens et al., 2022; Hermosillo-González, 2006). Habitan en todos los océanos, desde la zona intermareal hasta las ventilas hidrotermales. La mayoría de especies son bentónicas, aunque algunas se adaptaron a vivir en la zona pelágica y unas pocas pueden ser encontradas en agua salobre (Camacho-García et al., 2005). Su tamaño va desde los 3 a los 1,000 mm en algunas especies, como es el caso de Aplysia vaccaria (Winkler, 1955) (Behrens et al., 2022; Camacho-García et al., 2005; Hermosillo et al., 2006). A excepción de los órdenes Aplysiida, algunos Cephalaspidea y el superorden Sacoglossa (cuyas dietas se basan en algas), todas las especies se consideran carnívoras (Behrens et al., 2022). Todas las babosas marinas heterobranquias son hermafroditas y se reproducen sexualmente (Behrens et al., 2022; Camacho-García et al., 2005; Hermosillo et al., 2006). Para el Pacífico este, se tienen descritas alrededor de 360 especies con algunas más en proceso de descripción (Behrens et al., 2022).

En el Pacífico este tropical (PET) se han llevado a cabo importantes estudios sobre la riqueza y distribución de este grupo de moluscos. La guía de campo de babosas marinas del PET de Camacho-García et al. (2005) describe 163 especies bentónicas. Behrens y Hermosillo (2005) y Gosliner (2010) reportaron 122 especies para las costas de California. Por su parte, la guía de babosas marinas del Pacífico de Hermosillo et al. (2006) contempla 234 especies para el litoral mexicano y es el trabajo más completo para esta región. 

Las regiones insulares del PET han sido estudiadas por diferentes autores. Kaiser (1997) registra 54 especies para las islas Galápagos, Ecuador. Kaiser y Bryce (2001) registraron 27 especies en la isla de Malpelo, Colombia. Para el atolón francés de Clipperton, Kaiser (2007) registra la presencia de 35 especies. Finalmente, García-Méndez y Camacho-García (2016) incrementan en 13 el número de especies de babosas marinas en isla del Coco, Costa Rica, para compilar un total de 40 especies. En la península de Baja California destaca el trabajo de Sánchez-Ortíz (2000), quien registra 45 especies para las islas cercanas a Loreto y las localidades de La Paz, Cabo San Lucas, isla Cedros y bahía Tortugas; en el mismo año, Millen y Bertsch (2000) describen 3 especies nuevas para la misma región. Hermosillo y Valdés (2008) describieron 2 especies nuevas para el PET. Ortea y Llera (1981) describieron una especie endémica para isla Isabel. Para Bahía de Banderas, Hermosillo (2006) registró 140 especies y para la región compartida entre Colima, Michoacán y Guerrero, 76 especies fueron reportadas (Hermosillo y Behrens, 2005). Hermosillo-González (2006) reconoce que el Pacífico Transicional Mexicano comparte especies, tanto con la provincia Panámica, como con la provincia de California.

Por otra parte, los esfuerzos encaminados a incrementar el conocimiento de las babosas marinas en el archipiélago Revillagigedo han sido escasos. Hermosillo y Gosliner (2008) registraron 37 especies submareales para las 4 islas, sumando un total de 42 especies para este parque marino. En isla Socorro se obtuvo la mayor riqueza de especies, con 36 registros, mientras que fueron reportadas solamente 5 especies en isla Clarión. Behrens et al. (2009) describieron 1 especie nueva (Felimida socorroensis Behrens, Gosliner y Hermosillo, 2009) derivada del estudio de Hermosillo y Gosliner (2008). 

El objetivo de este estudio fue incrementar el conocimiento sobre las babosas marinas del Pacífico de México al presentar nuevos registros para isla Clarión, derivados de observaciones realizadas en la zona intermareal y submareal somera, así como actualizar el número de estas especies de moluscos presentes en el archipiélago de Revillagigedo.

Materiales y métodos

Se realizaron observaciones en diferentes zonas de la porción sur de isla Clarión: en la zona intermareal de las playas, la zona de pozas intermareales y el canal de desembarque (fig. 1B). La porción sur es la única área del litoral de la isla por la cual se puede acceder de manera segura a la zona intermareal. Dentro de esta porción, se definieron 5 sitios principales de muestreo: 1) playa este, con una superficie de 17 ha, se compone de fragmentos de coral y arena, es una planicie de aproximadamente 150 m de ancho en su parte más amplia, que queda expuesta durante la bajamar y en la cual hay algunas rocas y cabezas grandes de coral, así como pozas y canales intermareales; 2) bahía Azufre, se encuentra en el centro de la cara sur de la isla y se drena casi por completo durante la bajamar, exponiendo 19 ha de litoral conformado por sustrato arenoso en el centro, rocas en los extremos este y oeste, y canales y pozas intermareales en el centro-sur (fig. 2). En la entrada de la bahía se encuentra el área con mayor cobertura de arrecife de coral del archipiélago (Ketchum y Reyes-Bonilla, 2009), el cual queda parcialmente expuesto durante la bajamar; 3) playa oeste: se encuentra en el extremo oeste de la bahía Azufre. Es la más pequeña de las 3 playas (0.7 ha) y consiste en una caleta independiente, adyacente al canal de desembarque. Se conforma principalmente de sustrato rocoso con algunos parches arenosos y rocas grandes sueltas. Al igual que las otras playas, durante la bajamar queda expuesta un área considerable, donde hay canales y pozas intermareales; 4) pozas intermareales, consisten en una serie de 3 pozas relativamente grandes (~ 50-170 m² cada una) con comunicación intermitente con el océano. Se ubican debajo del “cerro del faro” y todas poseen profundidades entre 1 y 4 m, conteniendo sustrato rocoso y algunos parches arenosos; 5) canal de desembarque: consiste en una zanja tallada en el fondo rocoso-coralino, acondicionada por personal de la Marina mexicana para facilitar el desembarque en la isla. Tiene una longitud aproximada de 30 m y un ancho que va aumentando de 2 a 10 m conforme se aleja de la línea de costa. El sustrato es arenoso y rocoso, y se encuentra rodeado por arrecife coralino. La profundidad va de 20 cm a 10 m en la parte más alejada de tierra firme.

Los sitios de muestreo en intermareal somero fueron seleccionados por su accesibilidad durante los eventos de marea baja, al ser estos los momentos en los que se puede realizar una búsqueda activa de los especímenes en la zona. Por la misma razón, las búsquedas en el intermareal profundo se realizaron durante la pleamar, siendo éste el momento más accesible a estas zonas. Se llevaron a cabo 40 muestreos en la zona intermareal de las 3 playas arenosas-rocosas, el canal de desembarque y la zona de pozas intermareales adyacente al faro de la isla durante enero, febrero y marzo de 2023, y febrero y marzo de 2024. Para las playas este y oeste, bahía Azufre y la parte somera del canal de desembarque, se trazaron transectos de entre 80 y 350 m de longitud × 3 m de ancho (dependiendo de la morfología y extensión del sitio de muestreo), 2 h antes y 2 h después de la bajamar de las mareas vivas (3 días antes y 3 días después de las lunas nueva y llena), revisando exhaustivamente tanto en la superficie del sustrato como debajo de rocas y cabezas de coral sueltas, cada 2 a 5 m, aproximadamente. Para la zona de pozas intermareales y el canal de desembarque se realizó el mismo proceso 2 veces por semana, pero haciendo uso de equipo de buceo libre, entre 40 cm y 2 m de profundidad. Se seleccionaron horarios diurnos y vespertinos variables durante los eventos de marea alta, 2 h antes y 2 h después de la pleamar. Una vez localizado un organismo, se fotografió a detalle, mostrando tanto las características del ejemplar como de la fauna y flora acompañantes, así como el sustrato donde se encontró. Para ello, se utilizó una cámara Olympus Tough TG6 con el modo “Macro-Microscope”, con un aumento entre 2.9x y 11x. La profundidad donde se encontró cada organismo fue registrada con el profundímetro integrado en la cámara, el cual cuenta con una precisión de 10 cm. Se tomaron numerosas fotografías de cada ejemplar, las cuales fueron almacenadas y posteriormente revisadas a detalle. Todas las fotografías fueron tomadas por el primer autor de este estudio.

De acuerdo con los lineamientos consignados en el Programa de Manejo del Parque Nacional Archipiélago de Revillagigedo (Conanp, 2018), ningún ejemplar fue recolectado, manipulado o extraído de su hábitat para este estudio, por lo que la identificación de los ejemplares se realizó mediante las fotografías obtenidas, la observación posterior y la comparación de caracteres morfológicos (como la coloración principal del manto, rinóforos y branquias, así como la presencia o ausencia de estructuras [ceratas, apéndices] y marcas o patrones de coloración) con las guías de identificación disponibles para esta zona geográfica, así como listados históricos (Behrens et al., 2022; Camacho-García et al., 2005; Hermosillo y Gosliner, 2008; Hermosillo et al., 2006). La validez taxonómica que se presenta para cada una de las especies y géneros fue revisada de acuerdo con el World Register of Marine Species (WoRMS Editorial Board, 2024), así como con la autoridad taxonómica. El arreglo taxonómico desde orden y superorden hasta la especie fue el utilizado por Behrens et al. (2022) y por Bouchet et al. (2017).

Resultados

Se registraron 143 ejemplares de moluscos gasterópodos de la subclase Heterobranchia pertenecientes a 14 especies, en los 5 sitios del intermareal de isla Clarión. 

Phylum Mollusca Linnaeus, 1758

Clase Gastropoda Cuvier, G. 1795

Subclase Heterobranchia (Burmeister y Hermann, 1837)

Orden Cephalaspidea P. Fischer, 1883

Familia Aglajidae Pilsbry, 1895

Navanax aenigmaticus (Bergh, 1893). Fig. 3A.

Material examinado: 17 ejemplares.

Descripción. Concha interna. El cuerpo es café con manchas color crema y puede variar a rosa fuerte o café claro. Tiene una serie de puntos azul turquesa a lo largo del margen interior de los parapodios. La parte ventral del cuerpo es negra con puntos amarillos y blancos y alcanza los 75 mm de longitud (Behrens et al., 2022).

Ecología y hábitat. Todos los organismos fueron observados debajo de rocas, sobre sustrato arenoso y rocoso en las 3 playas y el canal de desembarque, a profundidades de entre 10 cm y 2 m de profundidad.

Distribución. Desde San Diego, California, Bahía Vizcaíno, Baja California, México, a Panamá; islas Galápagos, Ecuador, este y oeste del Atlántico tropical (Behrens et al., 2022).

Observaciones. Nuevo registro para isla Clarión. Visto en pareja en una ocasión. Seis variedades fueron identificadas gracias a las diferencias en la pigmentación (coloración y patrones), se clasificaron como marrón, café claro, café oscuro, negro, oliva y ocre (material suplementario: fig. S1). Carece de rádula. Se alimenta de otras babosas marinas heterobranquias, como notaspídeos y nudibranquios (Behrens et al., 2022).

Familia Haminoeidae Pilsbry, 1895

Haminoea virescens (G. B. Sowerby II, 1833). Fig. 3B.

Material examinado: 1 ejemplar.

Descripción. Concha expuesta, verde, con la espira hundida (material suplementario: fig. S2). Diferente de H. vesicula en la abertura mayor. Cuerpo verde con puntos amarillos en el escudo cefálico y manchas amarillo y blanco en los parapodios. Llega a medir 45 mm de longitud (Behrens et al., 2022).

Ecología y hábitat. Observado después del atardecer y sobre sustrato arenoso en la Bahía Azufre, aproximadamente a 15 cm de profundidad.

Distribución. Desde Alaska, EUA, hasta Panamá (Behrens et al., 2022).

Observaciones. Nuevo registro para isla Clarión y el Archipiélago. Habita en áreas rocosas intermareales en marea baja.

Superorden Sacoglossa Ihering, 1876

Familia Caliphyllidae Tiberi, 1881

Cyerce orteai Valdés y Camacho-García, 2000. Fig. 3C.

Material examinado: 68 ejemplares.

Descripción. El color varía desde crema pálido con algunas incrustaciones blancas, hasta verde oliva con puntos verde claro y café. La cabeza tiene puntos blancos opacos y un par de áreas color crema claro alrededor de los ojos. Los rinóforos y tentáculos orales apuntan hacia adelante. Las ceratas son grandes, redondas e infladas, cubiertas de tubérculos. Alcanza los 20 mm de longitud (Behrens et al., 2022).

Ecología y hábitat. Se observó en las 3 playas y el canal de desembarque, entre 5 y 50 cm de profundidad, sobre sustrato arenoso, rocoso y entre la vegetación y fragmentos de coral. Ocasionalmente en parejas y/o acompañada de masas ovígeras (material suplementario: fig. S3).

Distribución. Desde La Paz, Baja California, hasta Costa Rica, Hawaii y Okinawa (Behrens et al., 2022).

Observaciones. Nuevo registro para isla Clarión y el archipiélago. Hermosillo et al. (2006) mencionan que esta especie es de hábitos nocturnos y se alimenta del alga Halimeda. Fue la especie más abundante en este estudio. C. orteai ha sido registrada previamente en Baja California Sur, Costa Rica y en la isla de Oahu (Hawaii) (Behrens et al., 2022; Hermosillo et al., 2006; Valdés y Camacho-García, 2000). Sin embargo, no había sido observada en el archipiélago Revillagigedo, a pesar de encontrarse en el mismo corredor biogeográfico.

Polybranchia mexicana Medrano, Krug, Gosliner, Biju Kumar y Valdés, 2018. Fig. 3D.

Material examinado: 1 ejemplar.

Descripción. Cuerpo translúcido variable del amarillo al verde claro, con pequeños lunares blancos. Los rinóforos bifurcados están orientados al frente. Las ceratas tienen pequeñas manchas negras y las glándulas digestivas son visibles. Los rinóforos no poseen papilas. Llega a medir 12 mm (Behrens et al., 2022).

Ecología y hábitat. Observada a 1.5 m de profundidad en la zona de pozas intermareales.

Distribución. Desde el sur de Baja California Sur, México, hasta las islas Galápagos (Medrano et al., 2018).

Observaciones. Nuevo registro para isla Clarión y el archipiélago. Especie descrita recientemente por Medranoet al.(2018). Anteriormente pertenecía al complejo de especies dentro de P. viridis. Se trata de una especie nocturna con la capacidad de autotomizar sus ceratas cuando es molestada.

Orden Aplysiida [= Anaspidea] P. Fischer, 1883

Familia Aplysiidae Lamarck, 1809

Dolabella cf. auricularia (Lighfoot, 1786). Fig. 3E.

Material examinado: 1 ejemplar.

Descripción. La parte posterior del cuerpo forma un disco en diagonal, con papilas en toda la orilla y un sifón exhalante en el centro (material suplementario: fig. S4). Es de color muy variable, pero usualmente presenta tonos moteados de verde o café, lo que le da un muy buen camuflaje. Produce tinta cuando es molestada. Alcanza 500 mm de longitud (Behrens et al., 2022).

Ecología y hábitat. Solo fue observada en una ocasión, a 1.5 m de profundidad en una de las pozas intermareales, parcialmente enterrada en sustrato arenoso.

Distribución. Especie circumtropical (Behrens et al., 2022).

Observaciones. Nuevo registro para isla Clarión. Es citada como especie nocturna. Rara vez observada durante el día, aunque en este estudio se observó alrededor de las 10:00 h. Behrens et al. (2022) consideran que existe evidencia de que D. cf. auricularia es en realidad un complejo de varias especies no descritas en su totalidad; a la fecha de realización de esta contribución, el nombre Dolabella. cf. auricularia es válido para esta especie.

Stylocheilus ricketsii (MacFarland, 1966). Fig. 3F.

Material examinado: 3 ejemplares.

Descripción. Cuerpo alargado y pequeño en comparación con otras especies de la misma familia. Líneas oscuras irregulares longitudinales, con puntos azules dispersos. Posee apéndices ramificados por todo el cuerpo, así como una pequeña joroba y cola alargada. Alcanza 65 mm y libera tinta cuando es molestada (material suplementario: fig. S5) (Behrens et al., 2022).

Ecología y hábitat. Observado debajo de rocas en sustrato arenoso con presencia de algas del género Halimeda.

Distribución. Desde Baja California, México, hasta las islas Galápagos (Behrens et al., 2022).

Observaciones. Nuevo registro para isla Clarión. Especie críptica, difícil de ver entre las algas del fondo.

Orden Pleurobranchida [Pleurobranchoidea] Gray, 1827

Familia Pleurobranchidae Gray, 1827

Berthella martensi (Pilsbry, 1896). Fig. 4A.

Material examinado: 1 ejemplar.

Descripción. Variable en su coloración, desde blanco a anaranjado y marrón intenso. Branquia bajo el manto en el lado derecho del cuerpo. Esta especie tiene lunares característicos más oscuros que el cuerpo y frecuentemente una banda oscura en la orilla del manto y pie. Alcanza 70 mm de longitud total (Behrens et al., 2022).

Ecología y hábitat. Solo fue observada una vez, a 1.5 m de profundidad en el canal de desembarque, bajo una roca grande en sustrato rocoso.

Distribución. Desde Las Arenas, Baja California, hasta Panamá; también citada para el Indopacífico (Behrens et al., 2022).

Observaciones. Nuevo registro para isla Clarión. Al igual que otras especies de babosas, esta especie tiene la capacidad de autotomizar partes de su manto cuando es molestada.

Berthella grovesi Hermosillo y Valdés, 2008. Fig. 4B.

Material examinado: 1 ejemplar.

Descripción. El cuerpo es ovalado. El manto sobresale el ancho del pie, cubriendo completamente las branquias, el velo bucal y los rinóforos parcialmente. Estos últimos son enrollados, cortos, robustos y unidos basalmente. El pie es bilabiado. El color del fondo varía de un marrón rosado claro a uno más oscuro. Esta especie llega a medir 30 mm de longitud (Behrens et al., 2022).

Ecología y hábitat. Solo fue observada una vez bajo una roca pequeña sobre sustrato rocoso en la playa oeste, a 10 cm de profundidad.

Distribución. Registrada en isla Isabel y Bahía de Banderas, México y Panamá (Behrens et al., 2022).

Observaciones. Nuevo registro para isla Clarión y el archipiélago. Se encuentra bajo rocas en la zona intermareal y submareal somero. Esta es la observación más alejada del litoral continental publicada hasta ahora (707 km al suroeste de Baja California Sur y 1,100 km al oeste de Colima), reportada previamente en isla Isabel, Nayarit; en el Parque Nacional Coiba en Panamá y en el litoral continental mexicano (Bahía de Banderas en Jalisco-Nayarit, Manzanillo en Colima y Caleta de Campos en Michoacán) (Behrens et al., 2022; Hermosillo et al., 2006; Hermosillo y Valdés, 2008).

Berthella cf. agassizii (MacFarland, 1909). Fig. 4C.

Material examinado: 2 ejemplares.

Descripción. El cuerpo es translúcido y va del rosa a café claro, en ocasiones con puntos blancos dispersos. Los rinóforos están enrollados y la superficie dorsal se aprecia arrugada. Alcanza 12 mm (Behrens et al., 2022).

Ecología y hábitat. Solo observada en una ocasión, a 80 cm de profundidad en el canal de desembarque, debajo de una roca grande en sustrato rocoso.

Distribución. En el océano Pacífico, desde Punta Eugenia, Baja California, hasta Panamá. En el océano Atlántico, desde el mar Caribe hasta Santa Catarina, Brasil (Behrens et al., 2022).

Observaciones. Nuevo registro para isla Clarión. De acuerdo con Behrens et al. (2022), B. agassizii fue descrita originalmente para Brasil. Dado el distanciamiento geográfico, diferencias en tamaño, color y tracto reproductivo entre ambas especies (Pacífico y Atlántico), se consideran especies distintas (Alvim y Pimienta, 2015). Sin embargo, a la fecha de escritura de este documento, Berthella cf. agassizii es un nombre válido para la especie observada.

Orden Nudibranchia Cuvier, 1817

Familia Dendrodorididae O’Donoghue, 1924

Dendrodoris cf. fumata (Rüpell y Leuckart, 1831). Fig. 5A, B.

Material examinado: 21 ejemplares.

Descripción. Típico nudibranquio dórido, tiene el manto altamente ondulado. Se reconocen 3 variaciones de color: cuerpo negro aterciopelado con una línea roja en el margen, cuerpo blanquecino translúcido y cuerpo anaranjado o rojo. Todas las variaciones presentan puntas blancas en los rinóforos y en la pluma branquial, que es de gran tamaño. Alcanza una longitud total corporal de 30 mm (Behrens et al., 2022).

Ecología y hábitat. Se observó en las 3 playas en sustrato rocoso, arenoso y en zonas con coral, entre 10 y 50 cm de profundidad. Se identificaron las 3 coloraciones descritas para esta especie, además de la coloración propia de los juveniles: rojo (juvenil), negro, rojinegro y blanco, siendo este último el más abundante (material suplementario: fig. S6). En ocasiones fue visto en pareja.

Distribución. Desde bahía Vizcaíno, Baja California, hasta Costa Rica, Panamá y las islas Galápagos, Ecuador (Behrens et al., 2022).

Observaciones. Fue el nudibranquio más abundante, así como la única especie de este estudio reportada previamente para isla Clarión (Hermosillo y Gosliner, 2008). Nombrada así por el color de la variación oscura. Según Behrens et al. (2022), D. fumata es una especie ampliamente distribuida en el Indopacífico, genéticamente distinta a la presente en el Pacífico este tropical. Sin embargo, Dendrodoris cf. fumata es un nombre válido para esta especie a la fecha de escritura de este documento (Behrens et al., 2022; Valdés, Á., com. pers., 13 de julio de 2023). 

Dendrodoris nigromaculata (Cockerell, 1905). Fig. 5C.

Material examinado: 2 ejemplares. 

Descripción. Tiene el cuerpo alargado con márgenes ondulados. Su color va de blanco a crema con una serie de manchas de color chocolate o marrón de diferentes tamaños y cantidades. Las manchas se observan en 3 grupos, uno justo detrás de los rinóforos, otro en la parte central del cuerpo y una más anterior a la branquia, la cual es de color blanco o crema (al igual que los rinóforos). Alcanza una longitud corporal de 27 mm (Behrens et al., 2022).

Ecología y hábitat. Se registró en la playa oeste en sustrato rocoso, entre 10 y 20 cm de profundidad.

Distribución. Carmel, California; isla Guadalupe y las islas San Benito, Baja California (Behrens et al., 2022).

Observaciones. Nuevo registro para isla Clarión y el archipiélago. Los ejemplares fueron registrados en el intermareal y submareal somero. Este es el registro más meridional para esta especie, ya que anteriormente había sido registrada hasta las islas San Benito, Baja California (Behrens et al., 2022; Camacho-García et al., 2005; Goddard y Valdés, 2015; Hermosillo et al., 2006; Millen y Bertsch, 2005). Esta observación amplía su distribución y agrega a esta especie a la provincia del Pacífico Transicional Mexicano.

Familia Chromodorididae Bergh, 1891

Chromolaichma sedna (Ev. Marcus & Er. Marcus, 1967). Fig. 5D.

Material examinado: 17 ejemplares.

Descripción. El cuerpo es de color blanco, con una banda amarilla delgada y otra de color rojo brillante, un poco más ancha rodeando el manto, el pie y la cola. Los rinóforos y branquias son blancos con puntas rojas (material suplementario: fig. S7B). Alcanza 65 mm de longitud (Behrens et al., 2022).

Ecología y hábitat. Observada en las playas este y oeste, entre 5 y 15 cm de profundidad, siempre bajo rocas en sustrato rocoso y arenoso.

Distribución. Desde el golfo de California y el Pacífico mexicano, hasta Panamá, Colombia, Ecuador y el Caribe (Behrens et al., 2022).

Observaciones. Nuevo registro para isla Clarión. Especie gregaria que se alimenta de esponjas. Observada en cópula en una ocasión (fig. S7A). Su coloración brillante es probablemente aposemática (Behrens et al., 2022).

Familia Aeolidiidae Gray, 1827

Anteaeolidiella ireneae Carmona, Bhave, Salunkhe, Pola, Gosliner y Cervera, 2014. Fig. 5E.

Material examinado: 7 ejemplares. 

Descripción. El cuerpo es de color blanco translúcido, con pigmentación naranja dispersa sobre la espalda. Hay una línea blanca irregular desde la cabeza, que inicia entre los rinóforos y se extiende en forma de lágrima sobre el pericardio. Esta línea continua como varias manchas en forma de diamante de pigmento blanco o amarillo pálido a lo largo de la línea media. Los rinóforos y los tentáculos orales son de color naranja con puntas blancas. Las ceratas son largas y cilíndricas con puntas redondeadas, y se extienden desde la parte trasera de los rinóforos hasta la cola, dejando una zona desnuda sobre el dorso; son de color naranja brillante o gris oscuro, con una banda blanca y una punta blanca, y están dispuestas en hasta 36 filas, con hasta 12 ceratas en las filas anteriores y tan solo 4 en las últimas filas. Alcanza una longitud total de 12 mm (Behrens et al., 2022; Carmona et al., 2014).

Ecología y hábitat. Registrada en las playas este y oeste, bajo fragmentos de coral grandes y sobre sustrato rocoso, entre 5 y 20 cm de profundidad.

Distribución. En el Pacífico oriental. Ha sido registrada en México, Costa Rica y Panamá (Carmona et al., 2014).

Observaciones. Nuevo registro para isla Clarión. Esta especie fue descrita a partir de especímenes recolectados bajo coral suelto y arena en isla Clipperton. Recientemente, A. indica atravesó por una reestructuración taxonómica, pasando de ser una especie con una distribución circumtropical, a 8 nuevas especies. A. ireneae se distribuye en Clipperton, Revillagigedo, Panamá y Costa Rica (Behrens et al., 2022; Carmona et al., 2014).

Anteaeolidiella lurana (Ev. Marcus et Er. Marcus, 1967). Fig. 5F.

Material examinado: 2 ejemplares.

Descripción. Cuerpo delgado y alargado, con una cola relativamente corta. El cuerpo es de color blanco traslúcido, con una mancha en forma de herradura en la cabeza y que se extiende desde la base de los rinóforos hasta la base de los tentáculos orales. Posee 2 líneas onduladas anaranjadas discretas, que se extienden desde la parte posterior de la cabeza hasta encontrarse en la cola, formando diamantes blancos entre ellas. Los rinóforos y los tentáculos orales poseen pigmentación naranja en su base. Las ceratas son largas y más gruesas en la parte superior. Se extienden desde la parte posterior de los rinóforos hasta la cola. El epitelio de las ceratas está cubierto con pigmento naranja difuminado. Alcanza 12 mm de longitud (Carmona et al., 2014).

Ecología y hábitat. Observada en el canal de desembarque a 10 cm de profundidad.

Distribución. Posee una amplia distribución circumtropical: península de Yucatán y mar Caribe hasta Brasil; mar Mediterráneo y las islas Canarias; Australia (Carmona et al., 2014). En el Pacífico oriental ha sido registrada en el archipiélago de Revillagigedo (isla Socorro) y Bahía de Banderas, Nayarit, México (Carmona et al., 2014).

Observaciones. Nuevo registro para isla Clarión. Observada junto a masas ovígeras. Al igual que A. ireneae, esta especie fue recientemente separada de A. indica (Carmona et al., 2014).

Discusión

Todas las especies reportadas en este estudio constituyen nuevos registros para isla Clarión, a excepción de D. cf. fumata, que había sido previamente citada por Hermosillo y Gosliner (2008). Además, de las 14 especies observadas en este estudio, 5 representan registros nuevos para todo el archipiélago Revillagigedo (H. virescens, P. mexicana, C. orteai, B. grovesi y D. nigromaculata) (tabla 1). Este estudio incrementa de manera notable el conocimiento de las especies de babosas marinas de la zona intermareal en isla Clarión y el archipiélago de Revillagigedo. Constituye el primero de su tipo para isla Clarión y complementa el trabajo desarrollado por Hermosillo y Gosliner (2008), el cual se enfocó principalmente en especies de la zona submareal.

Con los registros nuevos presentados en esta investigación, se incrementa el número de babosas marinas de 42 a 47 especies para el Parque Nacional Archipiélago de Revillagigedo, y de 5 a 18 especies para isla Clarión. Esta riqueza de especies de babosas marinas en el archipiélago es equiparable a la riqueza registrada en otras islas y archipiélagos del PET, como isla Clipperton, Francia (Kaiser, 2007), isla del Coco, Costa Rica (García-Méndez y Camacho-García, 2015), isla de Malpelo, Colombia (Kaiser y Bryce, 2001) y las islas Galápagos, Ecuador (Kaiser, 1997) (tabla 2). No obstante, será necesario llevar a cabo estudios similares en las otras islas del archipiélago y a profundidades mayores a las de la zona submareal para conocer de mejor manera la biodiversidad de este grupo taxonómico en la región. De la misma manera, es importante complementar esta contribución con estudios moleculares para confirmar algunas de las incertidumbres taxonómicas actuales y poder realizar estudios de filogeografía y conectividad en el Pacífico este tropical. Además, es necesario incrementar el esfuerzo de estudio en las otras islas y zonas continentales del Pacífico de México.

Dedicar esfuerzos para incrementar el conocimiento de grupos de invertebrados poco estudiados en el archipiélago permitirá contar con una línea base más completa de la biodiversidad de este parque nacional y brindará elementos para determinar la efectividad de las medidas de manejo que buscan la protección de esta importante zona marina, considerada patrimonio de la humanidad (UNESCO, 2016). Finalmente, la ubicación del archipiélago, así como el hecho de ser la isla más alejada del territorio mexicano convierte a isla Clarión en un importante punto de investigación y un laboratorio natural para observar y comprender diferentes procesos ecológicos.

Agradecimientos

A la Secretaría de Marina y al Grupo de Ecología y Conservación de Islas por el apoyo logístico durante el viaje y la estancia en isla Clarión. A Ángel Valdés y a Alicia Hermosillo por su valioso apoyo en la identificación de algunos ejemplares y facilitar material bibliográfico. A los estudiantes de la licenciatura en Biología Marina de la Universidad de Guadalajara, Joshua Martín y Lizely Estrada por su colaboración en el trabajo de campo, y finalmente a la bióloga Elizabeth Piña Vera por su compañía y apoyo durante todo el proceso de elaboración de este trabajo.

Referencias

Alvim, J. y Pimienta, A. D. (2015). Taxonomic review of Berthella and Berthellina (Gastropoda: Pleurobranchoidea) from Brazil, with description of two new species. Zoologia, 32, 497–531. https://doi.org/10.1590/S1984-46702015000600010 

Behrens, D. W., Fletcher, K., Hermosillo, A. y Jensen, G. C. (2022). Nudibranch and sea slugs of the Eastern Pacific. Bremerton, Washington: MolaMarine.

Behrens, D. W., Gosliner, T. M. y Hermosillo, A. (2009). A new species of dorid nudibranch (Mollusca) from the Revillagigedo Islands of the Mexican Pacific. Proceedings of the California Academy of Sciences, 4, 423–429.

Bouchet, P., Rocroi, J. P., Hausdorf, B., Kaim, A., Kano, Y., Nützel, A. et al. (2017). Revised classification, nomenclator and typification of gastropod and monoplacophoran families. Malacologia, 61, 1–526. https://doi.org/10.4002/040.061.0201 

Camacho-García, Y. (2009). Benthic opisthobranchs. En I. S. Wehrtmann y J. Cortés (Eds.), Marine biodiversity of Costa Rica, Central America (pp. 371–386). Berlín: Springer Science and BussinesMedia. https://doi.org/10.1007/978-1-4020-8278-8_33 

Camacho-García, Y., Gosliner, T. M. y Valdés, A. (2005). Guía de campo de las babosas marinas del Pacífico este tropical. San Francisco, CA: California Academy of Sciences. 

Carmona, L., Bhave, V., Salunkhe, R., Pola, M., Gosliner, T. y Cervera, J. L. (2014). Systematic review of Anteaeolidiella (Mollusca, Nudibranchia, Aeolidiidae) based on morphological and molecular data, with a description of three new species. Zoological Journal of the Linnean Society, 171, 108–132. https://doi.org/10.1111/zoj.12129 

Conanp (Comisión Nacional de Áreas Naturales Protegidas). (2018). Programa de Manejo Parque Nacional Revillagigedo. Ciudad de México: Secretaría de Medio Ambiente y Recursos Naturales.

García-Méndez, K. y Camacho-García, Y. E. (2016). New records of heterobranch sea slugs (Mollusca: Gastropoda) from Isla del Coco National Park, Costa Rica. Revista de Biología Tropical, 64, 205–219. https://doi.org/10.15517/rbt.v64i1.23449 

Goddard, J. H. R. y Valdés, A. (2015). Reviving a cold case: two northeastern Pacific dendrodorid nudibranch reassessed (Gastropoda: Opisthobranchia). The Nautilus, 129, 31–42.

Gosliner, T. M. (2010). Two new species of nudibranch mollusks from the coast of California. Proceedings of the California Academy of Sciences, Ser., 4, 61, 623–631.

Hermosillo, A. y Behrens, D. W. (2005). The opisthobranch fauna (Gastropoda, Opistobranchia) of the Mexico states of Colima, Michoacán and Guerrero: filling in the faunal gap. Vita Malacológica, 3, 11–22.

Hermosillo, A., Behrens, D. W. y Ríos-Jara, E. (2006). Opistobranquios de México. Guía de babosas marinas del Pacífico, golfo de California y las islas oceánicas. Guadalajara, Jalisco: Comisión Nacional para el Conocimiento y Uso de la Biodiversidad/ Universidad de Guadalajara.

Hermosillo, A. y Gosliner, T. M. (2008). The opisthobranch fauna of the Archipiélago de Revillagigedo, Mexican Pacific. The Festivus, 40, 25–34.

Hermosillo, A. y Valdés, A. (2008). Two new species of opisthobranch mollusks from the Tropical Eastern Pacific. Proceedings of the California Academy of Sciences, Ser., 4, 59, 521–532.

Hermosillo-González, A. (2006). Ecología de los opistobranquios (Mollusca) de Bahía de Banderas, Jalisco-Nayarit, México (Tesis doctoral). Centro Universitario de Ciencias Biológicas y Agropecuarias. Posgrado en Ciencias Biológicas. Universidad de Guadalajara. Guadalajara, Jalisco.

Holroyd, G. L. y Trefry, H. E. (2010). The importance of Isla Clarión, Archipelago Revillagigedo, Mexico, for green
turtle (Chelonia mydas) nesting. Chelonian Conservation and Biology, 9, 305–309. https://doi.org/10.2744/CCB-0831.1 

Kaiser, K. L. (1997). The recent molluscan marine fauna of the Islas Galapagos. The Festivus, XXIX (Supl.), 1–67. https://doi.org/10.5962/bhl.title.129924 

Kaiser, K. L. (2007) The recent molluscan fauna of Île Clipperton (Tropical Eastern Pacific). The Festivus, XXXIX (Supl.), 1–162. https://doi.org/10.5962/bhl.title.129868 

Kaiser, K. L. y Bryce, C. W. (2001) The recent molluscan marine fauna of Isla Malpelo, Colombia. The Festivus, XXXIII (Occasional paper 1), 1–49.

Ketchum, J. T. y Reyes-Bonilla, H. (2009). Taxonomía y distribución de los corales hermatípicos (Scleractinia) del archipiélago de Revillagigedo, México. Revista de Biología Tropical, 49, 803–848.

Medrano, S., Krug, P. J., Gosliner, T. M., Biju-Kumar, A. y Valdés, A. (2019). Systematics of Polybranchia Pease, 1860 (Mollusca: Gastropoda: Sacoglossa) based on molecular and morphological data. Zoological Journal of the Linnean Society, 186, 76–115. https://doi.org/10.1093/zoolinnean/zly050 

Millen, S. V. y Bertsch, H. (2000). Three new species of dorid nudibranchs from Southern California, USA, and the Baja California Peninsula, Mexico. The Veliger, 43, 354–366.

Millen, S. V. y Bertsch, H. (2005). Two new species of porostome nudibranch (family Dendrodorididae) from the coast of California (USA) and Baja California (Mexico). Proceedings of the California Academy of Sciences, Ser. 4, 56, 189–199.

Ortea, J. y Llera, E. M. (1981). Un nuevo dórido (Mollusca: Nudibranchiata) de la Isla Isabel, Nayarit, México. Iberus, 1, 47–52.

Sánchez-Ortíz, C. A. (2000). Biodiversidad de moluscos opistobranquios (Mollusca: Opisthobranchiata), del Pacífico mexicano: isla Cedros, Vizcaíno e islas del golfo de California parte Sur. Universidad Autónoma de Baja California Sur. Informe final SNIB-Conabio proyecto Núm. L136. México D.F.

Valdés, A. y Camacho-García, Y. (2000). A new species of Cyerce Bergh, 1871 (Mollusca, Sacoglossa, Polybranchiidae)
from the pacific coast of Costa Rica. Bulletin of Marine Science, 66, 445–456.

UNESCO (2016). Lista del Patrimonio Mundial. United Nations Educational Scientific and Cultural Organization. Recuperado el 27 de enero, 2025 de: https://whc.unesco.org/es/list/1510 

WoRMS Editorial Board (2024). World Register of Marine Species. Disponible en http://www.marinespecies.org

Carlos Fernando Castillo-García ^{a, b}, Juan J. Morrone ^c, Isaías Hazarmabeth Salgado-Ugarte ^a, David Espinosa ^a, *

^a Universidad Nacional Autónoma de México, Facultad de Estudios Superiores Zaragoza, Batalla del 5 de mayo s/n, Ejército de Oriente Zona Peñón, Iztapalapa, 09230 Ciudad de México, Mexico

^b Universidad Nacional Autónoma de México, Posgrado en Ciencias Biológicas, Unidad de Posgrado, Edificio D, 1er. Piso, Circuito de Posgrados, Ciudad Universitaria, Coyoacán, 04510 Ciudad de México, Mexico

^c Universidad Nacional Autónoma de México, Facultad de Ciencias, Departamento de Biología Evolutiva, Museo de Zoología “Alfonso L. Herrera”, Circuito Exterior s/n, Ciudad Universitaria, Coyoacán, 04510 Ciudad de México, Mexico

*Corresponding author: despinos@unam.mx (D. Espinosa)

Received: 22 January 2024; accepted: 25 February 2025

Abstract

A software program for undertaking track analysis named Panbiotracks is introduced. It aims to solve various issues that currently are present in similar software packages. Panbiotracks is intended: 1) to be a fast, accurate, and reliable tool to generate individual tracks, generalized tracks and panbiogeographical nodes; 2) to solve the dependency on old and obsolete software that other packages have; and 3) to be a free and open source program that can be used in a variety of environments and that can be updated, modified, and improved continuously.

Keywords: Panbiogeography; Track analysis; Python; Software; Systematics

Panbiotracks: programa para análisis de trazos

Resumen

Se presenta un programa para llevar a cabo el análisis de trazos llamado Panbiotracks. Su objetivo es resolver varios problemas que están presentes en paquetes de software similares. Panbiotracks intenta: 1) ser una herramienta rápida, precisa y confiable para generar trazos individuales, trazos generalizados y nodos panbiogeográficos; 2) resolver la dependencia en software anticuado y obsoleto que tienen otros paquetes; y 3) ser un programa libre y de código abierto que pueda ser utilizado en una variedad de entornos y que puede ser actualizado, modificado y mejorado continuamente.

Palabras clave: Panbiogeografía; Análisis de trazos; Python; Software; Sistemática

Introduction

Biological evolution is closely linked to geological history. In the second half of the past century, León Croizat developed a new theoretical approach that he called “panbiogeography”. According to panbiogeography, the geographical distribution of taxa is mainly influenced by geological changes that lead to vicariance processes, where geographical barriers are the main cause of biota fragmentation, which in turn leads to speciation (Morrone, 2015).

Track analysis is panbiogeography’s primary methodological tool (Craw, 1988; Morrone, 2015). It is based on the spatial congruence between tracks, which are geometrical representations of taxon distribution. The 3 components of track analysis are individual tracks, generalized tracks, and panbiogeographical nodes (Morrone, 2015; Page, 1987). An individual track (Fig. 1a-d) is the basic unit of a panbiogeographical analysis, and is defined as an open, non-cyclic graph, also known as a minimum spanning tree (MST), whose vertices represent the geographical locations of the taxon, and the edges are the shortest paths that connect each of them. Mathematically speaking, it is a graph with n localities that are connected through n-1 edges, and whose total length is the shortest possible (Page, 1987). It can be thought of as a graphical representation of the main coordinates of a taxon in space. The basic method to draw an individual track consists of selecting a random locality and linking it to the nearest one using a straight line, then connecting both to another one, also the nearest, and repeating these steps until all the localities are connected, but without any loop between any of them. The most straightforward way to automate this is by building an adjacency matrix to find the edges, then saving the track to a file. However, some difficulty arises when measuring the distances between points, because this needs to be done considering the shape of the earth. There are multiple methods to do this, which are explained with more detail in the Algorithms subsection.

A generalized track (Fig. 1e) is a graph built from the statistically significant superposition between 2 or more individual tracks from different taxa, and it represents the history of ancestral biotas that were fragmented in the past (Craw, 1988; Morrone, 2015). However, there isn’t a formalized implementation of this concept, resulting in multiple methods that can be used to quantify it, like Parsimony Analysis of Endemicity (PAE) (Morrone, 2014; Rosen, 1988), Track Compatibility Analysis (Craw, 1988, 1989), or geometrical methods. A major issue is the difficulty of automating the process to find the generalized tracks, since the superposition between individual tracks cannot be calculated easily. Some programs, like MartiTracks (Echeverría-Londoño & Miranda-Esquivel, 2011), have approached the problem from a purely geometrical standpoint, using distances between segments to decide whether they are congruent with each other or not, but the results have not been satisfactory (Ferrari et al., 2013). The software Trazos2004 (Rojas-Parra, 2007) also uses a geometrical method, where the program finds the intersections between individual tracks and builds a generalized track from those intersections. This method does not follow the formal definition of a generalized track, since it considers only the superpositions between tracks and ignores those cases where 2 segments may be very close to each other, but without overlapping, which could also be considered as spatial congruence (Zunino & Zullini, 2003; Morrone, 2015). Even then, this method is fast and can be used to get an overall congruence approximation, since in many cases the generalized tracks obtained will reflect a general degree of similarity between individual tracks. To differentiate this approach from others, we opted to refer to the generalized tracks obtained by this method as “internal generalized tracks” (IGT). 

A panbiogeographical node (Fig. 1f), marked by the superposition between the terminal segments of 2 or more generalized tracks, indicates the location of a complex area of tectonic or biotic convergence (Craw, 1988; Morrone, 2015); however, this concept may have multiple interpretations. Heads (2004) mentions that a node might represent the location of endemism, high diversity, distribution boundaries, areas of disjunction and “anomalous” absences, but also states that there are more meanings to it than these. One problem arises from the existing ways to quantify a node. As mentioned, a node is located at the ends of a generalized track, or more specifically, where 2 terminal vertices (those that are connected only to another vertex) from different generalized tracks overlap. However, other authors had considered the intersections between individual tracks, or the intersections between non-terminal segments of 2 or more generalized tracks, as nodes (Grehan, 2001; Henderson, 1989). Morrone (2015) mentions that, to avoid confusion, the former might be called “individual nodes” and the latter “generalized nodes”. In this regard, panbiogeographical nodes can be considered as a subset of generalized nodes, since the latter accounts for all the intersections between generalized tracks, while the former takes into account only those that occur at the terminal segments of generalized tracks. This means that a tool that finds generalized nodes can be used to find panbiogeographical nodes, if only those that fulfill the proper definition will be counted. Out of the 3 concepts, nodes have received the least attention from a computational point of view, possibly because the process to limit the search of the intersections between tracks to only those that happen at the terminal segments has not been properly explored.

The usefulness of panbiogeography and track analysis has been questioned many times, mainly regarding its ability to generate meaningful results about the evolution of species through space and time (Seberg, 1986; Waters et al., 2013). Others have criticized how track analysis is done, like the criteria used for track orientation (Seberg, 1986), or the lack of a true quantitative process to obtain generalized tracks and nodes (Ferrari et al., 2013). However, many studies have used the panbiogeographical method to identify distributional patterns of many taxa, like mammals (Escalante et al., 2018; Florentin et al., 2016), mollusks (Aguilar-Estrada & Morrone, 2022; López et al., 2022), plants (García-Díaz et al., 2023; Puga-Jiménez et al., 2013), birds (Beauchamp, 1989), fungi (González-Ávila et al., 2017), arthropods (Maya-Martínez et al., 2011), and fossil taxa (Gallo et al., 2013; Hernández-Cisneros & Vélez-Juarbe, 2021). The ability to detect ancestral biotas through generalized tracks and nodes has been mentioned by other researchers and panbiogeography has also been considered useful from other points of view, such as searching for areas of biological complexity like areas of endemism and priority conservation areas (Craw et al., 1999; Escalante et al., 2018; Miguel-Talonia & Escalante, 2013; Morrone, 2015). Because of this, we consider that the development and improvement of new methods and tools for track analysis are important and should be continued. 

In recent times, several software packages have been published to automate track analysis using different algorithms and programming languages to calculate both the MSTs and the superposition between tracks and nodes. Some of them, like Trazos2004 (Rojas-Parra, 2007), Croizat (Cavalcanti, 2009), and fossil (Vavrek, 2011), use Prim’s algorithm (Sedgewick & Wayne, 2011). PASSaGE (Rosenberg & Anderson, 2011) uses its own algorithm, while SAM (Rangel et al., 2010) and MartiTracks (Echeverría-Londoño & Miranda-Esquivel, 2011) do not specify which algorithm they use. A list of the aforementioned software packages and the algorithms used by each is presented in Table 1. These programs, however, suffer from one or various of several issues: lack of documentation about what internal methods or algorithms they use, lack of accuracy in their results (Escalante et al., 2017, 2018), dependency on obsolete and unsupported software, difficulty in obtaining or reviewing their source code to improve or modify it, and lack of support for different operating systems other than Microsoft Windows and, sometimes, Apple macOS. 

Table 1

Comparison between different software packages for performing track analysis and their algorithms.

Program	Algorithms used	Measuring distance method	Platform
Croizat	Prim (1957), Bron and Kerbosch (1973), Wormwald (1984)	Not specified	Windows, macOS, Linux
Fossil	Not specified	Not specified	Independent
MartiTracks	Proprietary	Not specified	Windows, Linux
PASSaGE	Proprietary	Proprietary	Windows, macOS, Linux
SAM	Not specified	Not specified	Windows
Trazos2004	Prim (1957)	Bessel formulae	Windows

Panbiotracks

Panbiotracks is a Python program for track analysis that aims to overcome some of the issues mentioned above, namely the accuracy of results, dependency on obsolete software, analysis speed, and portability between operating systems. Panbiotracks addresses these issues in the following manner: 1) Accuracy and speed, by using an improved set of functions, Panbiotracks generates results in a shorter time and in more accurately than its predecessors. Moreover, since its code is smaller and is based on modern tool kits, it can be updated and improved faster than other solutions. 2) Dependency on obsolete software, since Panbiotracks uses modern releases of the Python programming language and its code is constantly reviewed to ensure that it can be used with newer versions, it is guaranteed that it will not be made obsolete in the short term. Additionally, its code can be ported to future versions of Python more easily. 3) Portability and compatibility, releases of Panbiotracks are published currently for the Windows and Linux operating systems, and a version for macOS is planned. Moreover, its source code is available publicly, which allows anyone to download it and build their own executable.

Technical information

Panbiotracks is programmed in the Python programming language version 3. It consists of one main command-line executable that can be configured at runtime to do different analyses by passing a flag that enables a particular function. Current functions are: -m I flag, generates individual tracks from a list of taxon names and localities, established by their geographical coordinates (latitude and longitude), and saves them to an ESRI shape file (SHP); -m P flag, generates an internal generalized track (IGT) from a set of individual tracks. It identifies the intersections between individual tracks and it marks the associated coordinates, then builds a new track based on those coordinates and saves it to an ESRI shape file (SHP); -m N flag, locates and identifies “generalized nodes”, which result from the intersections between 2 or more IGTs and saves them to an ESRI shape file (SHP).

Panbiotracks can be downloaded from its GitHub repository and its source code is freely available and licensed under the GNU General Public License version 3: https://github.com/cfnnmcg/panbiotracks

Algorithms

The first component of a track analysis is the individual track, which, according to Page (1987), is defined as a minimum spanning tree (MST) with n vertices and n-1 edges, that does not contain any loop, and whose total length (the sum of all of the lengths of its edges) is not greater than the length of any other graph built from the same vertices (Sedgewick & Wayne, 2011). To build an individual track, the starting point are presence records from the taxa to be evaluated, which will be treated as the vertices of the MST. To create the MST, Panbiotracks builds an adjacency matrix to save the list of the nodes and the connections between them, with as many rows and columns as there are localities in the input file. In an adjacency matrix, each row and column crossing represents a node, and the cells or intersections between them are the graph’s edges. In each cell there can be a numeric value indicating whether there is a connection between nodes or not. Since an MST is a weighted graph, that is, a graph whose edges have a length value, presences are coded with the length of each edge, while absences are coded with a zero (Table 2).

Table 2

Example of an adjacency matrix used by Panbiotracks to represent a MST.

	1	2	3	4	5
1	0	66.827884	450.588278	467.852057	67.569065
2	66.827884	0	486.027102	499.716705	1.152371
3	450.588278	486.027102	0	33.259944	487.176345
4	467.852057	499.716705	33.259944	0	500.858559
5	67.569065	1.152371	487.176345	500.858559	0

Since Earth is not a 2-dimensional plane but a spheroid, it is necessary to devise a method to calculate the correct minimal distances between localities. Several approximations have been used by many authors. Rojas-Parra (2007) used Bessel’s trigonometric formulae, which are somewhat similar to the haversine formula (Brummelen, 2013) in that they both calculate the distance between 2 points located over a sphere. In the software PASSaGE (Rosenberg & Anderson, 2011), the authors used a formula to measure the great circle distance over the earth’s surface. However, since the shape of our planet is not a perfect sphere, these methods can lead to inaccurate results. To address this issue, Panbiotracks makes use of Vincenty’s formulae (Vincenty, 1975), which take into account Earth’s spheroid shape, providing a more accurate distance measurement between points. Using this method, Panbiotracks calculates the distance in kilometers between each pair of localities and stores the result in the corresponding cell. To build properly the MST there are numerous algorithms; Prim’s (Prim, 1957) and Kruskal’s (Kruskal, 1956) algorithms are the most used (Graham & Hell, 1985). Panbiotracks uses Prim’s algorithm to build the MST from the adjacency matrix, where a random vertex is selected, then the program checks the distance between it and the others. The shortest distance is selected and the associated vertex added to a temporal matrix. The software then repeats the process until there are no remaining vertices. Finally, a function is used to draw the track from this matrix of vertices and distances and save it to a SHP file.

To build an internal generalized track (IGT), Panbiotracks starts from a set of individual tracks belonging to different taxa and, upon loading the files, the software removes any rows with duplicate data, based on the combination of the taxon name and its coordinates. This is done to ensure that no duplicate points are present. Then, it uses a function where each individual track is compared with all the others to assess if there is any overlapping point between them. An overlap is considered positive when any segment or segments from the first individual track superimpose with any other segment from the second individual track. When this happens, the program stores the point or points where the overlap take place, as well as their coordinates, in a list, then repeats the process for each of the other individual tracks until no more overlapping points are found. If a segment from one track coincides in its totality with another segment from another track, then their starting and ending vertices and the associated coordinates are added to the list. This list is then processed with the same algorithm as with the individual tracks, where the shortest distance between vertices is calculated, then a MST is built from that data and the IGT is saved to a SHP file.

To detect generalized nodes, Panbiotracks uses a modified version of the method used for the IGTs. First, it loads a set of 2 or more IGTs or generalized tracks into memory. A function to compare each track with all the others to find any intersecting point between them is used. This function will detect any overlapping points between segments and will store them and their coordinates in a list. Then another function is used to transform this set of points into a multi-point geometry. This geometry is then saved to a SHP file.

Input and output data

Panbiotracks is programmed with ease of use in mind, so its operation is straightforward. The initial input data needed to do a track analysis is a list of taxa and localities contained within a comma-separated (CSV) file that is loaded to Panbiotracks using the individual tracks flag. This file must have 3 columns: species, lat, and lon, with the taxon name, its latitude, and its longitude, respectively:

species,lat,lon

taxon_A,19.432637,-99.133205

taxon_A,19.1498,-99.0275

taxon_B,20.676215,-103.346979

taxon_B,21.00498,-102.9752

taxon_C,25.671798,-100.309384

taxon_C,24.61789,-98.762498

For IGT generation and node identification, the input data are SHP files containing tracks. The output data for all 3 functions are SHP files as well, containing MultiLineString data in the case of individual tracks and IGTs, and MultiPoint data in the case of nodes. Panbiotracks has been tested with CSV files containing thousands of records, with minimal memory impact and generating the corresponding individual tracks in less than 60 seconds. When computing IGTs or nodes, running time escalated according to the number of input files being evaluated, since the program needs to assess each possible combination of tracks.

Example analysis

To demonstrate the capabilities and use of Panbiotracks, an example analysis is presented. Detailed instructions and explanations about running and using the software can be found in the project’s documentation, available at the GitHub repository. For this example, occurrence data from the Global Biodiversity Information System (GBIF) were used, corresponding to several species of pine (genus Pinus) and oak (genus Quercus), mainly from the Mexican mountain ranges (GBIF.org, 2016a, 2016b), and totaling over 3 thousand records. These data were saved to a file named pinus_quercus_simplified.csv and formatted according to the example specified in the previous section. The following code block shows a sample of this file and its format:

species,lat,lon

Pinus_douglasiana,
20.358329999999999,-104.058329999999998

Pinus_douglasiana,
19.391670000000001,-103.325000000000003

Pinus_douglasiana,
24.074999999999999,-106.091669999999993

Pinus_herrerae,
18.800280000000001,-102.771109999999993

Pinus_herrerae,
16.191669999999998,-96.541669999999996

Pinus_herrerae,
16.263330000000000,-96.593329999999995

Quercus_crassifolia,
19.588110000000000,-99.404280000000000

Quercus_crassifolia,
19.823329999999999,-99.519999999999996

Quercus_crassifolia,
19.445830000000001,-101.346390000000000

Quercus_candicans,
19.228330000000000,-101.030000000000001

Quercus_candicans,
19.866109999999999,-97.700000000000003

Quercus_candicans,
19.844439999999999,-97.579719999999995

Quercus_laxa,
23.949439999999999,-105.064170000000004

Quercus_laxa,
22.533860000000001,-103.760930000000002

Quercus_laxa,
20.957500000000000,-99.336939999999998

…

Individual tracks

The CSV file is loaded into Panbiotracks using the following command, assuming that it is in the same directory as the program:

panbiotracks -m I -i ./pinus_quercus_simplified.csv -o ./pinus_quercus/

Where panbiotracks is the name of the executable program. Option -m I configures the individual tracks function. The -i flag defines the name of the input file and where it is located, and the -o flag defines the path where the individual tracks file or files will be saved. Panbiotracks generated a separate SHP file for each of the taxon names contained in the CSV file (Fig. 2). These files can be projected by any GIS software, like QGIS (Fig. 3).

Internal Generalized Tracks

To generate the IGTs, the individual tracks were first grouped based on their geographical proximity and record density (where the majority of records from a given species was located). For this example, we focused on 4 groups (Table 3), 2 from each genus, mainly located in the Sierra Madre Occidental (SMOcc), the Sierra Madre del Sur (SMS), and the Eje Volcánico Transversal (EVT). Pinus groups covered mainly the SMS and the SMOcc (Fig. 4), while Quercus groups spanned across the SMS, the EVT, and the SMOcc (Fig. 5). For each group, IGTs were built using a command similar to the following:

panbiotracks -m P -i ./pinus_quercus/Quercus_castanea.shp ./pinus_quercus/Quercus_crassifolia.shp ./pinus_quercus/Quercus_deserticola.shp ./pinus_quercus/Quercus_laxa.shp ./pinus_quercus/Quercus_subspathulata.shp -o ./pinus_quercus/igt/quercus_30_pac_smocc-evt.shp

Option -m P configures the program to build IGTs. The individual tracks files needed are defined after the -i flag and must be separated by spaces. After analyzing the individual tracks, Panbiotracks will save the IGT to the output file name defined after the -o flag. Please note that, unlike the command for individual tracks, the name defined here is not the name of the directory where you want to save the IGT, but the name of the file itself. This process is repeated for each IGT that needs to be done. As with the individual tracks, the SHP files of the IGTs can be projected in QGIS or other GIS software (Fig. 6). 

Table 3

Table of the species used in the example analysis and their grouping.

Genus	Group	Species
Pinus	SMS	P. devoniana, P. lawsonii, P. maximinoi, P. oocarpa, P. pringlei, P. rzedowskii
	SMOcc	P. durangensis, P. engelmannii, P. jaliscana, P. leiophylla, P. lumholtzii, P. luzmariae, P. maximartinezii, P. praetermissa
Quercus	SMOcc-EVT	Q. castanea, Q. crassifolia, Q. deserticola, Q. laxa, Q. subspathulata
	SMOcc-SMS-EVT	Q. candicans, Q. elliptica, Q. insignis, Q. scytophylla, Q. urbani

Generalized nodes

The 4 sets of individual tracks gave 4 IGTs, which were then loaded into Panbiotracks to find the generalized nodes between them. The syntax follows the same logic as the previous cases:

panbiotracks -m N -i ./pinus_quercus/igt/pinus_30_pac_smocc-sms.shp ./pinus_quercus/igt/quercus_30_pac_smocc-evt.shp ./pinus_quercus/igt/quercus_30_pac_smocc-sms-evt.shp -o ./pinus_quercus/gn/pinus30-smocc-sms_quercus30-smocc-evt-smocc-sms-evt

Where -m N configures the program to find generalized nodes. The rules to add the input and output files are the same as with the IGTs. In this case, each input path and file name (pinus_30_pac_smocc-sms.shp, quercus_30_pac_smocc-evt.shp, quercus_30_pac_smocc-sms-evt.shp) corresponds to an IGT and is separated from the others by a space, whereas the output file name is defined after the -o flag. As with the IGTs, the name corresponds to the actual file and not a directory. Panbiotracks will generate a single SHP file with all the generalized nodes in it (Fig. 7).

Conclusions and future developments

Though many software tools exist for assisting panbiogeography’s track analysis, most of them are difficult to use, difficult to distribute, rely on outdated software or are outdated themselves, lack precision in their algorithms or results, their results are difficult to reproduce, or their development has been halted for more than a decade in some cases. Panbiotracks aims to solve those problems by using efficient algorithms in constant optimization, by developing methods that are fast and precise, and by using a modern software stack that is extensible, easy to implement, and potent.

Panbiotracks’ algorithm for building individual tracks is faster and, thanks to its use of Vincenty’s formulae, more precise. In tests with CSV files with more than 3 thousand records, Panbiotracks generated the corresponding individual tracks in less than thirty seconds. When compared with other programs, like Trazos2004 (Rojas-Parra, 2007), the tracks generated with Panbiotracks were more precise, meaning that the distances between their vertices were more accurately calculated, and they did not have issues in their construction of loops (closed segments within a track), problems that have been observed with Trazos2004.

As mentioned, the algorithm for generating IGTs finds any intersection between pair of tracks to get a list of vertices from which the IGT will be built. This same algorithm is used to find the generalized nodes within IGTs. This method has various advantages, like its speed and straightforward approach. An IGT is a type of generalized track that only takes into account “true” connections, that is, intersections between individual tracks. From this point of view, it can be said that an IGT is a form of generalized track with the least ambiguity, since an overlap can be identified without any uncertainty. Considering this, a tool that can identify and generate these features is very valuable and useful.

There are, however, some pending issues regarding the IGTs and the method to identify nodes. If we consider the formal definition of a generalized track, it is necessary to add improvements to the method used and allow the program to consider other degrees of congruence between individual tracks besides the direct superposition. For example, with the current method, segments of individual tracks that are very close to each other but that do not intersect, will not be marked as part of the generalized track. This can lead to the dismissal of areas where 2 or more taxa share a similar evolutionary path, which can be represented by a generalized track, but since they do not overlap, the program does not consider their congruence. An example of this can be seen in Figure 4a, where the 2 westernmost segments are almost parallel, but the algorithm does not count them for the IGT (Fig. 6a). A potential solution is to add an option to define a buffer or area of influence around the individual tracks, and use that to compute the generalized track, but this method requires more testing to be implemented properly. There are other methods that are currently being researched for their inclusion in Panbiotracks. One is the Fréchet distance, which measures the degree of similarity between curves (Alt & Godau, 1995; Aronov et al., 2006). Another is the Hausdorff distance, which measures the distance of 2 subsets of a metric space (Bai et al., 2011; van Kreveld et al., 2022). Both concepts can potentially be used to detect and build generalized tracks with more accuracy than with present methods.

It also should be noted that Panbiotracks currently does not automatically separate the ITs that will be used for the IGTs. Because of this, it is strongly advised to first use methods like PAE-PCE (Luna-Vega et al., 1999, 2000) or NDM/NVDM (Goloboff, 2005) to segregate those ITs that are useful for building an IGT. These methods have been used to identify areas of endemism (García-Barros et al., 2002; Santiago-Alvarado et al., 2022), but they can also be used to define sets of species whose ITs are related enough to form an IGT.

The formal definition of a panbiogeographical node indicates that it is present only at the intersection of 2 or more endpoint vertices from 2 or more generalized tracks, that is, those vertices located at the periphery of the MST that only have one edge connecting them to another node (Henderson, 1989; Morrone, 2015). Since Panbiotracks takes into account all the intersections between generalized tracks, it does not quite follow this definition. However, as mentioned before, the generalized nodes can be seen as a set of which the panbiogeographical nodes are a part. As such, it is necessary to develop an improved algorithm that can filter and locate only the desired localities. This also applies for those cases where 2 terminal segments from different IGTs do not overlap. Under certain circumstances, a panbiogeographical node can be said to be present at a given location, but the IGTs might not overlap and, consequently, the software will not consider this area as a node. In Fig. 6, for example, the area towards the northwest of the map, where there are terminal ends of the IGTs, might be considered a node, but the program does not mark it. To solve this, a similar solution to that of the IGTs may be useful: define a buffer and evaluate the degree of overlap not between tracks, but between these buffers.

Automating track analysis and other methods in biogeography is essential and of utmost importance to the future of this field of knowledge. With the introduction of faster computer systems and programming languages specialized in data manipulation and analysis, like Python and Julia, it is necessary to develop new algorithms and improve existing ones. Moreover, another issue of great importance is to ensure that the software can be updated and improved and not let it become obsolete or dependent on outdated systems and software.

Acknowledgments

This paper serves as a fulfillment of CFC-G for obtaining a M.Sc. degree in the Posgrado en Ciencias Biológicas, UNAM. We thank to the Secretaría de Ciencia, Humanidades, Tecnología e Innovación (Secihti) for the support of this research through a graduate scholarship (No. 1178796) to CFC-G. This work was funded by the project IN215321 (DGAPA-PAPIIT).

References

Aguilar-Estrada, L. G., & Morrone, J. J. (2022). Distributional patterns of Vetigastropoda (Mollusca) all over the world: a track analysis. Zoological Journal of the Linnean Society, 196, 442–452. https://doi.org/10.1093/zoolinnean/zlac004

Alt, H., & Godau, M. (1995). Computing the Fréchet distance between two polygonal curves. International Journal of Computational Geometry & Applications, 5, 75–91. https://doi.org/10.1142/S0218195995000064

Aronov, B., Har-Peled, S., Knauer, C., Wang, Y., & Wenk, C. (2006). Fréchet distance for curves, revisited. In Y. Azar, & T. Erlebach (Eds.), Algorithms – ESA 2006 (pp. 52–63). Springer. https://doi.org/10.1007/11841036_8

Bai, Y. B., Yong, J. H., Liu, C. Y., Liu, X. M., & Meng, Y. (2011). Polyline approach for approximating Hausdorff distance between planar free-form curves. Computer-Aided Design, 43, 687–698. https://doi.org/10.1016/j.cad.2011.02.008

Beauchamp, A. J. (1989). Panbiogeography and rails of the genus Gallirallus. New Zealand Journal of Zoology, 16, 763–772. https://doi.org/10.1080/03014223.1989.10422933

Bron, C., & Kerbosch, J. (1973). Algorithm 457: Finding all cliques of an undirected graph. Communications of the ACM, 16, 575–577. https://doi.org/10.1145/362342.362367

Brummelen, G. V. (2013). Heavenly mathematics: the forgotten art of spherical trigonometry. Princeton, NJ: Princeton University Press.

Cavalcanti, M. J. (2009). Croizat: a software package for quantitative analysis in panbiogeography. Biogeografía, 4, 4–6.

Craw, R. (1988). Continuing the synthesis between panbiogeography, phylogenetic systematics and geology as illustrated by empirical studies on the biogeography of New Zealand and the Chatham Islands. Systematic Biology, 37, 291–310. https://doi.org/10.1093/sysbio/37.3.291

Craw, R. (1989). New Zealand biogeography: a panbiogeographic approach. New Zealand Journal of Zoology, 16, 527–547. https://doi.org/10.1080/03014223.1989.10422921

Craw, R., Grehan, J., & Heads, M. (1999). Panbiogeography: tracking the history of life. Oxford, UK: Oxford University Press.

Echeverría-Londoño, S., & Miranda-Esquivel, D. R. (2011). MartiTracks: a geometrical approach for identifying geographical patterns of distribution. Plos One, 6, e18460. https://doi.org/10.1371/journal.pone.0018460

Escalante, T., Noguera-Urbano, E. A., & Corona, W. (2018). Track analysis of the Nearctic region: identifying complex areas with mammals. Journal of Zoological Systematics and Evolutionary Research, 56, 466–477. https://doi.org/
10.1111/jzs.12211

Escalante, T., Noguera-Urbano, E. A., Pimentel, B., & Aguado-Bautista, O. (2017). Methodological issues in modern track analysis. Evolutionary Biology, 44, 284–293. https://doi.org/10.1007/s11692-016-9401-8

Ferrari, A., Barão, K. R., & Simões, F. L. (2013). Quantitative panbiogeography: Was the congruence problem solved? Systematics and Biodiversity, 11, 285–302. https://doi.org/10.1080/14772000.2013.834488

Florentin, J. E., Arana, M. D., & Salas, R. M. (2016). Panbiogeographic analysis of Galianthe subgenus Ebelia (Rubiaceae). Rodriguésia, 67, 437–444. https://doi.org/10.
1590/2175-7860201667214

Gallo, V., Avilla, L. S., Pereira, R. C. L., & Absolon, B. A. (2013). Distributional patterns of herbivore megamammals during the Late Pleistocene of South America. Anais Da Academia Brasileira de Ciências, 85, 533–546. https://doi.org/10.1590/S0001-37652013000200005

García-Barros, E., Gurrea, P., Luciáñez, M. J., Cano, J. M., Munguira, M. L., Moreno, J. C. et al. (2002). Parsimony analysis of endemicity and its application to animal and plant geographical distributions in the Ibero-Balearic region (western Mediterranean). Journal of Biogeography, 29, 109–124. https://doi.org/10.1046/j.1365-2699.2002.00653.x

García-Díaz, R. F., Valdez-Hernández, E. F., Martínez-Cárdenas, L., Díaz-Nájera, F., & Ayvar-Serna, S. (2023). Diversity and distribution of Andean tubers: an agrogeographic analysis. Investigaciones y Estudios – UNA, 14, 59–70. https://doi.org/10.57201/IEUNA2313312

GBIF.org. (2016a). GBIF occurrence download. The Global Biodiversity Information Facility. https://doi.org/10.15468/DL.DGUZHT

GBIF.org. (2016b). GBIF occurrence download. The Global Biodiversity Information Facility. https://doi.org/10.15468/DL.ORDDD1

Goloboff, P. (2005). NDM/VNDM (Versión 2.5). Programs for identification of areas of endemism. Programs and documentation available at: http://www.zmuc.dk/public/phylogeny/endemism

González-Ávila, A., Contreras-Medina, R., Espinosa, D., & Luna-Vega, I. (2017). Track analysis of the order Gomphales (Fungi: Basidiomycota) in Mexico. Phytotaxa, 316, 22. https://doi.org/10.11646/phytotaxa.316.1.2

Graham, R. L., & Hell, P. (1985). On the history of the minimum spanning tree problem. Annals of the History of Computing, 7, 43–57. https://doi.org/10.1109/MAHC.1985.10011

Grehan, J. (2001). Biogeography and evolution of the Galapagos: Integration of the biological and geological evidence. Biological Journal of the Linnean Society, 74, 267–287. https://doi.org/10.1111/j.1095-8312.2001.tb01392.x

Heads, M. (2004). What is a node? Journal of Biogeography, 31, 1883–1891. https://doi.org/10.1111/j.1365-2699.2004.01201.x

Henderson, I. M. (1989). Quantitative panbiogeography: an investigation into concepts and methods. New Zealand Journal of Zoology, 16, 495–510. https://doi.org/10.1080/03014223.1989.10422918

Hernández-Cisneros, A. E., & Vélez-Juarbe, J. (2021). Palaeobiogeography of the North Pacific toothed mysticetes (Cetacea, Aetiocetidae): a key to Oligocene cetacean distributional patterns. Palaeontology, 64, 51–61. https://doi.org/10.1111/pala.12507

Kruskal, J. B. (1956). On the shortest spanning subtree of a graph and the traveling salesman problem. Proceedings of the American Mathematical Society, 7, 48–50. https://community.ams.org/journals/proc/1956-007-01/S0002-9939-1956-0078686-7/S0002-9939-1956-0078686-7.pdf

López, B., Naranjo-García, E., & Mejía, O. (2022). Diversity patterns of Mexican land and freshwater snails: a spatiotemporal approach. Revista Mexicana de Biodiversidad, 93, e933966. https://doi.org/10.22201/ib.20078706e.2022.93.3966

Luna-Vega, I., Alcántara-Ayala, O., Espinosa-Organista, D., & Morrone, J. J. (1999). Historical relationships of the Mexican cloud forests: A preliminary vicariance model applying Parsimony Analysis of Endemicity to vascular plant taxa. Journal of Biogeography, 26, 1299–1305. https://doi.org/10.1046/j.1365-2699.1999.00361.x

Luna-Vega, I., Alcántara-Ayala, O., Morrone, J. J., & Espinosa, D. (2000). Track analysis and conservation priorities in the cloud forests of Hidalgo, Mexico. Diversity and Distributions, 6, 137–143. https://doi.org/10.1046/j.1472-4642.2000.00079.x

Maya-Martínez, A., Schmitter-Soto, J. J., & Pozo, C. (2011). Panbiogeography of the Yucatán Peninsula based on Charaxinae (Lepidoptera: Nymphalidae). Florida Entomologist, 94, 527–533. https://doi.org/10.1653/024.094.0317

Miguel-Talonia, C., & Escalante, T. (2013). Los nodos: el aporte de la panbiogeografía al entendimiento de la biodiversidad. Biogeografía, 6, 30–42.

Morrone, J. J. (2014). Parsimony analysis of endemicity (PAE) revisited. Journal of Biogeography, 41, 842–854. https://
doi.org/10.1111/jbi.12251

Morrone, J. J. (2015). Track analysis beyond panbiogeography. Journal of Biogeography, 42, 413–425. https://doi.org/10.
1111/jbi.12467

Page, R. D. M. (1987). Graphs and generalized tracks: quantifying Croizat’s panbiogeography. Systematic Zoology, 36, 1. https://doi.org/10.2307/2413304

Prim, R. C. (1957). Shortest connection networks and some generalizations. The Bell System Technical Journal, 36, 1389–1401. https://doi.org/10.1002/j.1538-7305.1957.tb01515.x

Puga-Jiménez, A. L., Andrés-Hernández, A. R., Carrillo-Ruiz, H., Espinosa, D., & Rivas-Arancibia, S. P. (2013). Patrones de distribución del género Zanthoxylum L. (Rutaceae) en México. Revista Mexicana de Biodiversidad, 84, 1179–1188. https://doi.org/10.7550/rmb.32047 

Rangel, T. F., Diniz-Filho, J. A. F., & Bini, L. M. (2010). SAM: A comprehensive application for Spatial Analysis in Macroecology. Ecography, 33, 46–50. https://doi.org/10.1111/j.1600-0587.2009.06299.x

Rojas-Parra, C. A. (2007). Una herramienta automatizada para realizar análisis panbiogeográficos. Biogeografía, 1, 31–33.

Rosen, B. R. (1988). From fossils to earth history: applied historical biogeography. In A. A. Myers, & P. S. Giller (Eds.), Analytical biogeography: an integrated approach to the study of animal and plant distributions (pp. 437–481). Netherlands: Springer. https://doi.org/10.1007/978-94-009-1199-4_17

Rosenberg, M. S., & Anderson, C. D. (2011). PASSaGE: Pattern Analysis, Spatial Statistics and Geographic Exegesis. Version 2. Methods in Ecology and Evolution, 2, 229–232. https://doi.org/10.1111/j.2041-210X.2010.00081.x

Santiago-Alvarado, M., Luna-Vega, I., Rivas, G., & Espinosa, D. (2022). Effect of cell size and thresholds in NDM/NVDM methods on recognizing areas of endemism. Zootaxa, 5134, 1–33. https://doi.org/10.11646/zootaxa.5134.1.1

Seberg, O. (1986). A critique of the theory and methods of panbiogeography. Systematic Zoology, 35, 369–380. https://doi.org/10.2307/2413388

Sedgewick, R., & Wayne, K. (2011). Algorithms (4th ed). Boston, Massachusetts: Addison-Wesley Professional.

van Kreveld, M., Miltzow, T., Ophelders, T., Sonke, W., & Vermeulen, J. L. (2022). Between shapes, using the Hausdorff distance. Computational Geometry, 100, 101817. https://doi.org/10.1016/j.comgeo.2021.101817

Vavrek, M. J. (2011). Fossil: palaeoecological and palaeogeographical analysis tools. Palaeontologia Electronica, 14, 16.

Vincenty, T. (1975). Direct and inverse solutions of geodesics on the ellipsoid with application of nested equations. Survey Review, 23, 88–93. https://doi.org/10.1179/sre.1975.23.176.88

Waters, J. M., Trewick, S. A., Paterson, A. M., Spencer, H. G., Kennedy, M., Craw, D. et al. (2013). Biogeography off the tracks. Systematic Biology, 62, 494–498. https://doi.org/10.1093/sysbio/syt013

Wormald, N. C. (1984). Generating random regular graphs. Journal of Algorithms, 5, 247–280. https://doi.org/10.1016/
0196-6774(84)90030-0

Zunino, M., & Zullini, A. (2003). Biogeografía: la dimensión espacial de la evolución (Vol. 259). Ciudad de México: Fondo de Cultura Económica.