Introduction

Biodiversity is an essential component for the balance and health of ecosystems. Mammals play central roles, acting as predators, herbivores, and seed dispersers, to mention a few. In Mexico - one of the most diverse countries worldwide (^{Ceballos and Oliva 2005}) -, the study of mammals is essential given the rapid change in land use and habitat fragmentation that, together with poaching, threaten their survival (^{Aguilar et al. 2000}). Significant changes in the composition of mammal communities have been documented around the world, due to the elimination or displacement of species (^{Laurance and Yensen 1991}; ^{Krikpatrick and Jarne 2000}; ^{Janecka et al. 2014}).

Consequently, inventories of mammals should be elaborated because they are essential to acquire comprehensive knowledge, essential for carrying out ecological, conservation, and management studies (^{Pacheco et al. 2004}). Its importance has grown considerably in the face of the increasing environmental deterioration driven by human population growth and the associated urban, industrial, agricultural, and livestock-raising activities, all of which generate adverse impacts on natural environments (^{Chávez and Ceballos 1998}). In this context, strategies for the management and conservation of natural resources, particularly flora and fauna, largely depend on the availability of information on biological diversity (^{Chávez and Ceballos 1998}; ^{Romero and Ceballos 2006}). Inventories provide a crucial starting point for understanding changes in the structure and composition of mammal communities in different areas, conserved and disturbed, as well as at different times, facilitating the implementation of conservation and management strategies (^{López-Ramírez et al. 2020}; ^{Mezhua-Velázquez et al. 2022}).

In Tamaulipas, studies on mammals are scarce, and most focus on protected natural areas (^{Vargas-Contreras and Hernández-Huerta 2001}; ^{Carvajal-Villareal et al. 2012}; ^{Carrera-Treviño et al. 2018}; ^{Branney et al. 2023}; ^{Ochoa-Espinoza et al. 2023}), leaving aside other forest areas that, despite being affected by anthropogenic activities, function as refuges and biological corridors for biodiversity (^{López-Ramírez et al. 2020}). Such is the case of the submontane scrubland, characteristic of northeastern Mexico, including Tamaulipas. This vegetation type is characterized by a mixture of thorny shrubs, small trees, and perennial herbs (^{Rzedowski 2005}). It covers 8.3 % of the surface area of Tamaulipas (^{INEGI 2017}) and undergoes constant changes due to the human activities already mentioned (^{Estrada-Castillón et al. 2012}), which affect wild mammal populations.

Therefore, studying this ecosystem is essential to describe its ecological aspects, determine the factors that influence their populations, and generate better conservation, management, and use strategies (^{Buenrostro-Silva et al. 2017}; ^{Salazar-Ortiz et al. 2020}). The objective of this study was to evaluate the diversity, structure, and composition of medium and large mammals in a submontane scrubland.

Materials and methods

Study Area. The study area is located southwest of the municipality of Casas, Tamaulipas, Mexico, between coordinates 23° 24’ and 23° 21’ N, -98° 44’ and -98° 38’ O, at an altitude of 240 to 420 meters above sea level. (^{INEGI 2010}). The area is located outside the limits of the western slope of the Sierra de Tamaulipas Biosphere Reserve (Figure 1). The regional climate is semi-warm and semi-dry, with minimum temperatures of 16 °C to 19 °C and maximum temperatures of 34 °C to 36 °C (^{INEGI 2021}), and a mean annual precipitation of 600 mm to 800 mm. The dominant vegetation type is submontane scrubland (^{INEGI 2021}).

Figure 1 Location of the study area, Casas, Tamaulipas, México. 

Sampling design. Ten simple stations (camera traps) were placed with a minimum separation of 3 km and a maximum of 3.5 km from each other. The stations were set on nature trails or close to trails and roads with traces of mammals, such as footprints and feces. Camera traps were installed at 30 cm to 50 cm above the ground and set to capture three photographs at five-second intervals, operating 24 hours a day. They were checked each month to change memory cards and batteries (^{Chávez et al. 2013}; ^{Maffei et al. 2002}; ^{Mattey et al. 2022}). The sampling campaign covered 12 months, from July 2020 to June 2021; months were grouped according to climatic seasons (dry and rainy). To determine the dry and rainy seasons, the mean historical precipitation was calculated for the period between 1982 and 2013 based on climatological statistical information available from the national meteorological service of the National Water Commission (^{CONAGUA 2010}). As a result, the dry season included January, February, March, April, November, and December, while the rainy season spanned from May to October.

Analysis of photographic and taxonomic identification and nomenclature. To determine the independence of the records, the following criteria were considered: a) consecutive photographs of the same species should be separated by 24 h and b) in photographs of gregarious species, each was considered as a separate record (^{Monroy-Vilchis et al. 2011}; ^{Chávez et al. 2013}; ^{Ávila-Nájera et al. 2015}; ^{Pozo-Montuy et al. 2019}). Individuals were identified based on ^{Ceballos and Oliva (2005}) using the nomenclature according to ^{Ramírez-Pulido et al. (2014}).

Data analysis. The potential number of species was calculated according to ^{Moreno (2001}) and ^{Magurran (2004}) by using the non-parametric estimators Chao 1, which uses abundance data, and Jacknife 1, based on species incidence. Estimators were calculated using 100 randomizations with no replacement in the program EstimateS 9.1.0 (^{Colwell 2013}).

Diversity numbers were calculated using the Hill series of first (q1) and second (q2) orders. These were obtained from the exponential of the Shannon-Wiener index: q1 = e ^h ' (where: q1 = first-order Hill number, and e ^h' = Shannon index) and Simpson’s reciprocal: q2 = 1/D (where: q2 = second-order Hill number and D = Simpson’s dominance index); these indicate the number of effective species (^{Moreno 2001}; ^{Magurran 2004}; ^{Magurran 2021}).

The species composition of mammal communities and their abundances between seasons were compared with a permutation-based analysis of variance (PERMANOVA); this results in the sum of squares within groups (SS) and the sum of squares within groups (Ss), using the Bray-Curtis index as a measure of distance, with 9999 random permutations (^{Anderson 2001}). A percentage similarity analysis (SIMPER) was also used to determine which species contributed most to the differences between the seasons (^{Clarke 1993}).

Results

Species richness. A total of 457 independent records were obtained with a sampling effort of 3,650 trap-days; the mammals recorded belong to three orders, seven families, and 12 species. The order Carnivora presented the highest richness with four families: Canidae, Felidae, Mephitidae, and Procyonidae; of these, Felidae was the best-represented family, with five species (Table 1). The highest species richness was recorded during the rainy season (12 species) vs. the dry season (nine species), while the potential richness for each season was in rainfall from 13.99 (Chao 1) to 16.17 (Jacknife 1) and for the dry season from 10.06 to 11.50. Thus, the observed richness relative to the potential richness ranged from 74.21 % to 97.95 % in the rainy season and from 78.26 % to 89.46 % in the dry season (Figure 2).

Figure 2 Species accumulation curves during seasons. Light blue = observed species; Green = Chao 1; Orange = Jack 1. 

Records and diversity. The mammal species with the highest number of records were Odocoileus virginianus (white-tailed deer) with 205 records (44.86 %), followed by Dicotyles tajacu (collared peccary) with 103 (22.5 %) and Leopardus pardalis (ocelot) with 58 (12.7 %) (Figure 3). On the other hand, the species with a single record were Herpailurus yagouaroundi, Panthera onca, and Procyon lotor (Table 1).

For the study area, we estimated q1 = 4.908 abundant species and q2 = 3.584 dominant species that contribute to diversity. The highest number of effective species was observed in the dry season, with 4.468 (Table 2).

The PERMANOVA determined significant differences between the seasons [SS = 1.331; Ss = 0.9798; F = 3.586; p < 0.019] (Table 3). The SIMPER analysis indicated that the species that determine the main differences between seasons are D. tajacu (34.87 %), O. virginianus (28.36 %), N. narica (11.26 %), and L. pardalis (10.84 %), accounting for 85.32 % of the differences between the seasons. Specifically, O. virginianus and D. tajacu showed the highest mean abundance per season, with 14.5 and 14.3 in the rainy seasons, while O. virginianus (19.7), L. pardalis (6.67), and N. narica (4.17) showed the highest mean abundance in the dry season; Table 4).

Discussion

The present study reports approximately 8 % of the wild mammal species known for the state of Tamaulipas (152 species) and 24.5 % of the medium and large species (49 species.; ^{Ceballos and Oliva 2005}; ^{Moreno 2024}). Compared with the study by ^{Branney et al. (2023}), which recorded 15 species of the Order Carnivora in the Sierra de Tamaulipas Biosphere Reserve (RBST), an adjacent area, the present study reports nine species of this order despite the impact of livestock ranching. This variation can be attributed to the number of sampling stations, as more camera traps were used in the RBST and a larger area was covered. The presence of Sylvilagus sp. Is worth noting, detected by direct observation, although it was not included in the study because it was not recorded in the sampling stations. The high species richness in the study area may be associated with the vegetation type and food availability (^{SEMARNAT 2018}; ^{Pozo-Montuy et al. 2019}). Diverse habitats, such as submontane scrubland, provide a wide range of resources and microhabitats that offer shelter and suitable feeding areas (^{Alanís-Rodríguez et al. 2015}).

This study reports five of the six Mexican felines. This finding could indicate that the ecosystem has fragments of vegetation with a good degree of conservation (^{Aranda et al. 2012}; ^{Hernández-Pérez et al. 2024}). According to ^{Ceballos and Olivia (2005}), Aranda et al. (2012), ^{Velazco-Macías and Peña-Mondragón (2015)}, the feline species identified in the area thrive preferentially in arid, xeric (submontane), and subtropical scrublands. These areas, covered by dense vegetation, provide an ideal habitat for these taxa (^{Buenrostro-Silva et al. 2015}), as they provide them with shelter and camouflage, contributing to their successful hunting.

The submontane scrubland is of great importance because it is home to medium and large mammal species with a high cultural and ecosystem value (^{Cortes-Marcial and Briones-Salas 2014}). Among these species, there are six with declining populations (H. yagouaroundi, L. pardalis, P. concolor, P. onca, C. leuconotus, N. narica), four with stable populations (D. marsupialis, L. rufus, D. tajacu, and O. virginianus), and two with growing populations (C. latrans and P. lotor;^{IUCN 2023}). In addition, 11 of these species are classified as Least Concern (LC) and only P. onca is listed as Near Threatened (NT) by the ^{IUCN (2023)}; three of these species are protected by NOM-059-SEMARNAT-2010 (^{SEMARNAT 2010}). Likewise, H. yagouaroundi, L. pardalis, and P. onca are included in Appendix I of CITES, while L. rufus and P. concolor are in Appendix II (^{CITES 2023}).

Figure 3 Mammals recorded in the southwest of the municipality of Casas, Tamaulipas, México. 

In both seasons, a similar diversity (orders 1 and 2) was recorded in the mammal communities, with uniform values in terms of the abundant and dominant species in the site. This is consistent with the study conducted by ^{Ríos-Solís et al. (2021)} in El Gavilán, Oaxaca, an area covered by tropical dry forest with dense vegetation in some period of the year, similar to some elements of the submontane scrubland. These ecosystems are suitable for the diversity of medium and large mammals.

On the other hand, the mountain cloud forest (BMM, in Spanish) of Tamaulipas shows a higher diversity during the dry season, similar to the submontane scrubland. In the submontane scrubland, diversity values also show a greater presence of abundant and dominant species during the dry season, with a minimal variation between seasons.

It is important to note that the BMM is located in the El Cielo Biosphere Reserve, which could be stabilizing the diversity levels, i.e., it fosters a stable structure and composition of mammals over time, since, being a protected natural area, activities such as hunting are prohibited. In contrast, the submontane scrubland, which lacks this type of protection, shows a constant diversity between the rainy and dry seasons. This suggests an ecosystem that maintains a greater resilience in the face of temporal variations (^{De Mazancourt et al. 2013}; ^{Loreau and De Mazancourt 2013}).

The distribution of mammal abundance varies between seasons. During the dry season, the species with the highest number of records were O. virginianus, L. pardalis, and N. narica; during the rainy season, the abundant species were O. virginianus, D. tajacu, and L. pardalis. To note, O. virginianus, D. tajacu, and L. pardalis maintain reproductive populations, since they were recorded with offspring.

Of the recorded species, O. virginianus was very abundant obtained in both seasons, likely due to its high plasticity to different environments. These generalist herbivores thrive in various types of vegetation and, according to several authors (^{Ceballos and Oliva 2005}; ^{Weber 2014}; ^{Gallina and López 2016}; ^{Jiménez-Sánchez et al. 2024}), are common in arid and scrub areas. In addition, their diet usually includes plants from the families Fabaceae and Asteraceae (^{Navarro-Cardona et al. 2018}), which are abundant in submontane scrubland areas (^{Rzedowski 2005}). However, a trend of declining abundance of O. virginianus was observed during the rainy season, while the abundance of D. tajacu increased. This suggests a more intense competition for food in the rainy season, forcing O. virginianus to travel greater distances in search of food (^{Sánchez-Pinzón and Arias 2022}).

Like O. virginianus, D. tajacu is adapted to a wide variety of ecosystems (^{Zaldivar et al. 2022}). In the present study, its abundance was higher during the rainy season. This finding is consistent with the observations reported by ^{Reyna-Hurtado et al. (2014}) and ^{Sánchez-Pinzón et al. (2020}), who highlighted that water availability is essential for the presence of this species and also influences the rolling behavior for grooming, to regulate temperature, or to eliminate ectoparasites (^{García-Marmolejo et al. 2015}; ^{Sánchez-Pinzón et al. 2020}).

The fact that L. pardalis was the feline with the highest number of records may indicate that it is the top predator in the study area, and its presence may lead to the "pardalis effect". In other words, the presence of the ocelot influences the dynamics of the populations of its prey and other predators, affecting the structure and composition of the ecological community (^{Silva-Magaña and Santos-Moreno 2020}). This may explain the low number of records of P. onca, P. concolor, Lynx rufus, and H. yagouaroundi, so they may be occasional visitors. Likewise, H. yagouaroundi is a cryptic and rare species, so it is difficult to detect it (^{Giordano 2015}), and its presence is influenced by the pardalis effect (^{De Olivera et al. 2010}; ^{Caso et al. 2015}).

Compared to other studies, ocelot records were more frequent in the present study. For example, in the El Cielo Biosphere Reserve in Tamaulipas, 40 records were documented over a 24-month period (^{Ochoa-Espinoza et al. 2023}); in the northeastern Sierra de Puebla, 33 records were captured over 21 months (^{Ordoñez-Pardo et al. 2023}); in Tequila, Veracruz, a single record was recorded in eight months (^{Salazar-Ortiz et al. 2020}); and in the Lagunas de Chacahua National Park, Oaxaca, four records were captured during 12 months (^{Buenrostro-Silva et al. 2015}). These differences can be attributed to variables such as the vegetation type, degree of human activities, or presence of big cats, in contrast with the area studied in the present work, which is covered by a dense submontane scrubland vegetation (^{Rzedowski 2005}), which favors the presence of L. pardalis (^{Ceballos and Oliva 2005}; ^{Aranda et al. 2012}; ^{Galindo-Aguilar et al. 2019}). In addition, the study was carried out during the COVID-19 pandemic during which human activities were limited, a circumstance that may have also favored the presence of this feline.

^{Ramírez-Bravo et al. (2010}) and ^{Galindo-Aguilar et al. (2019}) point out that the ocelot tolerates fragmented environments that are usually close to mountainous areas within protected areas. Such is the case of this study, which was carried out in an area adjacent to the Sierra de Tamaulipas Biosphere Reserve. In addition, habitat modification and fragmentation are detrimental to feline populations, with ocelots being most affected by the decline in vegetation cover (^{Hernández-Pérez et al. 2024}).

It should be mentioned that the study area is being affected by the introduction of free-range cattle. The abundance of Leopardus pardalis in this area indicates that there are still vegetation remnants or fragments that are suitable for the subsistence of species with a high ecosystem value. Therefore, the area should be considered for monitoring, and federal and state authorities should establish conservation strategies.

The record of a raccoon was interesting because this species thrives in a wide variety of environments associated with permanent water bodies (^{Guerrero et al. 2000}; ^{Timm et al. 2016}). One of the reasons of this distribution is that, as racoons lack salivary glands, they need to moisten the food to ingest it (^{Ceballos and Oliva 2005}). In the study area, water bodies are temporary from April to September, so this habitat is not suitable for the species.

Additional research and inventories on mammals should be conducted, especially in unprotected areas, to gain a more complete understanding of mammal diversity and their conservation status. This may contribute to identify priority areas and develop effective strategies for species conservation.