texto en 
Inglés (pdf)
Artículo en XML
Referencias del artículo
Traducción automática
Enviar artículo por email
Citado por SciELO 
Similares en
SciELO 
Permalink
Introduction
The white tailed deer Odocoileus virginianus is a highly valued species in subsistence hunting in the neotropical region. The use of this species has notably increased in Wildlife Management Units (UMA) in Mexico (Villarreal-Espino, 2006). The management of UMAs has originated a greater need for reliable biological information in order to sustainably manage the communities. Among several biological aspects, knowledge of the food habits is important for the management of deer both in the wild and in captivity (Fulbright & Ortega-Santos, 2007; Ramírez-Lozano, 2004; Villarreal, 1999).
Food habit studies of white tailed deer have been carried out in Mexico, in pine-oak temperate forests (Aguilera-Reyes et al., 2013; Gallina, Maury, & Serrano, 1981; González & Briones-Salas, 2012), in xeric shrublands of the northeast (Martínez, Molina, González, Marroquín, & Navar, 1997; Ramírez-Lozano, 2004) and in tropical regions (Arceo, Mandujano, Gallina, & Pérez, 2005; Granados-Rivera et al., 2014; López-Pérez, Serrano, Aguilar, & Herrera, 2012; Plata et al., 2009). The white tailed deer inhabits the Tehuacán-Cuicatlán Biosphere Reserve (RBTC) (Ortiz-García, Ramos-Robles, Pérez-Solano, & Mandujano, 2012) where there is a particular interest in sustainably making use of it in extensive UMAs. For this reason, the objective of this work was to determine and compare the botanical composition of the white tailed deer's diet in two communities of the RBTC during the dry and rainy seasons.
Materials and methods
Study area
The study area was comprised of the communities of San Pedro Chicozapotes in the municipality of Cuicatlán, and San Gabriel Casa Blanca in the municipality of San Antonio Nanahuatipam in the state of Oaxaca, Mexico, which are part of the Tehuacán-Cuicatlán Biosphere Reserve (RBTC). This region is one of the zones with the most biological and cultural diversity (Dávila et al., 2002) in the state. The RBTC is part of the Sierra Madre del Sur and occupies the northwestern region of the sub-county of Oaxaca's plateau. The reserve is located in the extreme southeast of the state of Puebla and northeast of Oaxaca between 7° 39' - 18° 53' N latitude and 96° 55' - 97° 44' W longitude. The RBTC has a surface area of 490,187 ha and its altitude varies from 600 to 2,950 m. The medium annual temperature in the valley of Tehuacán, Puebla varies between 18 and 22 °C, and increases to 24.5 °C in Cuicatlán, Oaxaca (Instituto Nacional de Estadística y Geografía [INEGI], 2002).
According to the Comisión Nacional de Áreas Naturales Protegidas (CONANP, 2010), the main types of vegetation and soil use of the region that comprises the protected area and the primary land region of the valley of Tehuacán-Cuicatlán are: low deciduous forest or dry tropical forest (29 % of the reserve territory); land dedicated to agriculture, livestock and forest exploitation (22 %); oak and pine forest (21 %); desert rosette scrubland with thorny bushes and a significant presence of cacti (10 %); crasicaule scrubland dominated by large cacti (8 %); and other types of vegetation (10 %). The type of vegetation that prevails in the community of San Pedro Chicozapotes is the tropical dry forest (TDF) with a higher and more dense tree layer in the areas with higher slopes, while in San Gabriel Casa Blanca it is the tropical dry forest and the crasicaule shrubland (TDF/CS) with a predominance of stalks of the Neobuxbaumia genus (Barrera-Salazar, Mandujano, Villarreal, & Jiménez-García, 2015).
Experimental development
Fifty three fecal groups of the white tailed deer were collected during the dry season (November to May) and 37 groups during the rainy season (June to October) from 2011 to 2013. Fecal groups were collected in transects of 500 x 2 m: 25 in TDF and 32 in TDF/CS. The fecal samples were dried in a forced air oven at 60 °C during 24 h. A fraction of the processed fecal sample was placed on the glass slide and was left to sit and solidify during two weeks. On each slide (particle population), 20 micro fields (micro parcels) of 7 mm in diameter each were drawn systematically (on the slide) and then they were read with a photomicroscope at 100x (Nikon®, LABOPHOT-2A model, Japan). Each circular area at 100x constituted a field, which was called a micro parcel and represented the sampling unit. One slide was prepared and identified per group of feces collected in the study area and each slide was considered an experimental unit with 20 repetitions. 1,060 fields were observed for the dry season (400 fields for BTC/MC and 660 fields for BTC/CS) and 740 for the rainy season (320 fields for BTC and 420 for BTC/CS).
Simultaneous to the search for fecal samples, a scientific collection was done (leaves, stems, flowers and fruit) of the plant species per season (rainy and dry) in order to integrate the reference material. For this, the sweeping method was used, which comprised serial and exhaustive collections in the study area. The material was stored in botanical presses, writing down the plant information and the location of the collection on a card. Subsequently, the material was transferred to the laboratory to do a taxonomic determination of the samples. The epidermal patterns of the plants were obtained with the method described by González and Améndola (2010). The epidermis was photographed with a Nikon® photomicroscope, with a 35 mm photographic camera incorporated. The photographs were taken at 10x and 40x, in order to facilitate the comparison of the reference patterns with the slides that corresponded to the collected fecal groups. The plants that comprised the diet of the deer per season and per community were identified when observing the epidermis of the fecal samples and when comparing it to the photographic patterns of the reference plants. The comparison was based on the histological characteristics such as size, shape and the provision of cells on the epidermal tissue.
During the microhistological analysis, the fecal material and the life form were identified (epiphyte, succulent, tree, shrub, broad-leaved herbs, thin blade grass and cyperaceae). The fragments found such as leaves, stems, flowers and fruit were separated in order to determine if the deer consumed more tender or more fibrous plants. The younger branches and tree leaves, broad-leaved herbs in vegetative state, fruit and shrubs of deciduous leaves were considered tender, while the mature or dry leaves, thin blade grass and leaves and branches and shrubs with evergreens were considered woody parts.
The contribution of each plant species to the diet was estimated based on the accumulated frequency, which in this case was the total of the fields analyzed in which a specific species or component was registered (Sparks & Malechek, 1968). In this study, a component can be understood as the life form and consumed parts of the plants. Based on the accumulated frequency (AF) and the total number of analyzed fields (N), the frequency (F) was calculated as follows:
Once the frequency was calculated, the density (D) was calculated as:
D = -In(1 - F)
Finally, the relative density (%) was calculated in order to express the botanical composition of the diet through the following equation:
where:
RDa = Relative density of the species
Da = Density of the species
Di = Density of each individual species (González & Améndola, 2010).
This calculation was obtained by seasons (rainy and dry) and study location (TDF/CS and TDF).
Statistical Analysis
Considering that the methods based on the lineal models are not appropriate for count data, in this case the number of species, generalized lineal models were used (known as GLMs) with a Poisson error type and a log type link function. The log type link function ensures that all the values adjusted are positive; while the Poisson error considers that the count data are integral values and have a mean variance (Crawley, 2013). In order to prove this, the residual values of the deviation were compared in the analyses against the residuals of the degrees of freedom. If the former is different, it indicates an overdispersion, and in this case the quasipoisson type error function can be used. These models are also known as log-linear models. The differences between seasons and locations for the variables on the number of species, number of species per life form and number of species per consumed parts of the plant were tested through GLMs. The types of vegetation and seasons were considered fixed factors in these models. For all the analyses, the statistical program R version 3.1.3 (R Development Core Team, 2015) was used.
Results and discussion
The annual diet of the white tailed deer was comprised of 83 species classified in 36 botanical families (Appendix 1). The Malvaceae, Commelinaceae and Myrtaceae families were the ones that contributed the most to the deer's diet in the rainy season; while the Fabaceae, Poaceae, Rubiaceae, Begoniaceae and Malpighiaceae contributed most in the dry season. Of the total species consumed, only 13 contributed with more than 50 % of the seasonal diet in each location. These species were: Bursera schlechtendalii Engl., Caesalpiniaceae sp., Lysiloma divaricatum (Jacq.) J. F. Macbr., Opuntia lasiacantha Pfeiff., Agave macroacantha Zucc., Ceiba parvifolia Rose, Euphorbia sp., Talinum paniculatum (Jacq.) Gaertn., Bursera fagaroides Ballock, Ziziphus sp., Solanum lanceolatum Cav., Senna wislizeni (Rose) Irwin & Barneby and Karwinskia humboldtiana (Schult.) Zucc. (Table 1). During the rainy and dry seasons, the number of species consumed in the TDF/CS was 37 and 51, while in the TDF it was 32 and 37, respectively. These differences were not statistically significant (deviance = 0.29, df = 1, P = 0.59), which indicates that the deer relatively consume a similar number of species in both locations and seasons.
Table 1. Main botanical species consumed by the white tailed deer with a contribution higher than 50 % to the seasonal diet in two types of vegetation in the state of Oaxaca. There is a complete detailed list in Appendix 1.
The tree and shrub species represented more than 63 % of the deer's diet in both types of vegetation and seasons; the herbal, succulent, grass and cyperaceae followed in order of importance (Figure 1). Only in the TDF did the deer consume epiphytes. The generalized linear model indicates that there are statistically significant differences (P < 0.05) in the consumption of the different life forms depending on the type of habitat and the season (Table 2). Particularly, in the consumption of herbaceous plants (P = 0.0001) and the collection of succulent, grass, cyperaceae and epiphytes (P = 0.008) differed between the habitat and season types; while the white tailed deer consume a constant percentage of tree and shrub species during the entire year in both locations, the consumption of other life forms was more variable.
Figure 1. Life forms of the plants consumed by the white tailed deer during the rainy (■) and dry (□)seasons in two types of vegetation in the state of Oaxaca: a) Tropical dry forest/crasicaule shrubland (San Gabriel Casa Blanca), b) Tropical dry forest (San Pedro Chicozapotes). Trees (Arb), shrubs (arbt), herbaceous plants (herb), succulent (suc), zateas and cyperaceae (zat/cip) and epiphytes (epi).
Table 2. Generalized Linear Model (GLM) Poisson type to analyze the consumption of plant species by the white tailed deer dependent on the type of habitat, season and life form. The "interception" represent the fixed value of the linear model for the trees of the tropical dry forest (TDF) during the rainy season; the estimated coefficients indicate the difference in respect to the interception.
TDF/CS: Tropical Dry Forest /Crasicaule shrubland
AIC: Akaike Information Criterion
In the "others" life form, the succulent, grass, cyperaceae and epiphytes were grouped.
On the other hand, the consumption of leaves and tender branches was higher than the consumption of flowers and fruit (Figure 2). In the generalized linear model for this case, it was necessary to use a quasipoisson type error considering that the residuals of the deviation were smaller than the degrees of freedom (Table 3). According to this analysis, in the TDF/CS there was a marginally significant tendency (P = 0.051) of a higher consumption of branches and leaves in comparison to the TDF. Particularly, the consumption of branches and leaves (P = 0.0001) increased in the dry season (P = 0.038).
Figure 2. Consumption of the leaves/branches (■) and flowers/fruit (□) by the white tailed deer during the rainy and dry seasons in San Gabriel Casa Blanca (tropical dry forest/crasicaule shrubland) and San Pedro Chicozapotes (tropical dry forest) in the state of Oaxaca.
Table 3. Generalized Linear Model (GLM) Poisson type to analyze the consumption of plant parts (leaves/branches, flowers/fruit) by the white tailed deer dependent on the type of habitat and season. The "interception" represents the fixed value of the linear model for the consumption of flowers and fruit in the tropical dry forest (TDF) during the rainy season; the estimated coefficients indicate the difference in regards to the interception.
TDF/CS: Tropical Dry Forest /Crasicaule shrubland
AIC: Akaike Information Criterion
The feeding behavior of the white tailed deer in the communities of San Pedro Chicozapotes and San Gabriel Casa Blanca is relatively similar to what is indicated in the bibliography for other TDF. For example, Arceo et al. (2005) indicate that there are 82 plants selected by the white tailed deer in the TDF of Chamela, Jalisco, and only 12 contributed to more than 50 % of the annual diet. On the other hand, González and Briones-Salas (2012) found that the annual diet of this hoofed animal in the region of the Sierra Madre of Oaxaca was comprised of 42 plant species included in 23 botanical families. For the Mixteca region of Puebla, Villarreal-Espino (2006) reports 133 species grouped in 50 botanical families, but once again only a few contributed to the largest percentage. Granados-Rivera et al. (2014) report 40 species of 15 families in a study done in Campeche. In general, the white tailed deer selected a high number of species from different families, but it particularly consumes a higher percentage of fewer species.
The tender leaves and branches of tree and shrub species were the most consumed by the white tailed deer both in season and study locations. It has been reported that the white tailed deer prefer the shrub and tree species during the entire year and the herbaceous plants in rainy seasons when there is greater availability. Gallina et al. (1981) reported this consumption pattern in the temperate forests of La Michilía, Durango. This pattern has also been observed in the TDF, even with an increase in the consumption of flowers and fruit during the dry season (Arceo et al., 2005) when the green plants and water availability are scarce (Mandujano & Gallina, 1995).
The diversity of species consumed by the white tailed deer in the locations studied in the RBTC is possibly due to the differences in the botanical structure and composition of San Gabriel Casa Blanca dominated by columnar cacti in less pronounced slopes, in comparison to San Pedro Chicozapotes where the TDF is more diverse and dense in more rugged slopes. However, the number of plant species consumed could notably increase if the vast expanse of the RBTC and the different types of environment which promote a greater plant diversity are considered (Dávila et al., 2002); for example, in the northeast part of the RBTC in the region of Zapotitlán in Puebla, a type of even drier vegetation dominates. As a consequence, when increasing the number of locations and environment sampled, a significant increase of the plant species consumed by the herbivore is expected, as has been reported in other regions of the country (Gallina et al., 1981; Ramírez-Lozano, 2004; Villarreal-Espino, 2006).
Conclusions
The results of this study helped expand the knowledge on the foraging habits of the white tailed deer in dry tropical regions where the habitat seasons are contrasting. The results clearly indicate that this herbivore selects a wide diversity of species for its consumption, but it also shows that the white tailed deer consumes a higher percentage of a reduced number of species, as happens in other regions. The results could have an implication on the management of the white tailed deer in the study region; for example, this could serve to provide the deer kept in backyards or in intensive UMAs with seasonal food. However, it is suggested to carry out studies to assess the nutritional values of the plants and thus know in which way they contribute to the growth, preservation and reproduction of the hoofed animal. This could allow for the elaboration of silages with native plants for the feeding of the deer through family micro-enterprises.
