Rapid ecological assessment of mammals from a locality of middle basin at Palmar River, Zulia state, Venezuela

. Here we present the results from a rapid ecological assessment of mammals within Campo Boscán, located in the middle basin of the Palmar River in Zulia state, Venezuela. Records for species inventories were obtained from mist-nets and traps surveys and direct observation across three-vegetation types; we also include data from surveys to local communities. We identified 47 mammal species–representing 24 families and nine orders–, including the first record of the Hairy Big-eared Bat ( Micronycteris hirsuta ) for the Maracaibo Lake Depression region. Orders Chiroptera, Rodentia and Carnivora represented the most diverse and abundant taxa. Insectivores and carnivores were the most diverse trophic groups. Approximately 13% of the species are included within some threat category. Our results can be used for developing future monitoring programs in order to increase the knowledge on terrestrial vertebrates and biodiversity in general within the region.


Introduction
The first step for studying biodiversity is the assessment of species richness -and its space and time dimensions-given that this information represents the most important knowledge resource for developing systematic conservation planning; including its management and monitoring (Sánchez et al. 2004, Martins et al. 2016. In this sense, information regarding presence/absence, abundance and diversity of species, especially mammals, has been fundamental for further understanding other important ecological parameters such as population dynamics, communities' structure and ecological processes (e.g., dispersion, pollination), as well as to identify and assess biodiversity loss (Voss & Emmons 1996, Wright et al. 2000, Prieto-Torres et al. 2011, Martins et al. 2016.
In Venezuela, knowledge about the taxonomy and biogeography of mammals has considerably increased recently, including a complete and updated list (Linares 1998, Sánchez & Lew 2012. According to a recent update on mammals of the worldfrom Wilson and Reeder (2005)-the total number of mammal species reported for Venezuela is 391; this includes 14 orders, 47 families and 184 genera (Sánchez & Lew 2012, Quiroga-Carmona & Woodman 2015. However, despite these contributions, diverse studies have shown that species diversity values are evidently an underestimation (e.g., Ceballos et al. 2002, Sánchez andLew 2012), suggesting the existence of significant knowledge gaps and absence of studies in numerous regions and habitats, especially in the northwest region of the country (e.g., Prieto-Torres et al. 2011, 2015a. In this sense, the Maracaibo Lake Depression (MLD; Figure 1) -located in the Zulia state (northwest Venezuelan)-is considered a unique and important bioregion in the country, due to its climatic, geologic and ecological (including species diversity) characteristics (e.g., PDVSA 1993, MARN 2000, Rivas 2006, Rodriguez et al. 2010, Prieto-Torres et al. 2011. The most up-to-date information about this region indicated an approximate of 160 mammal species (representing 35 families and 11 orders), which represent ~83 and ~41% of all mammal species reported for Zulia state and Venezuela, respectively (Rivas 2006, Prieto-Torres et al. 2015b. Despite this bioregion offers great potential -from a taxonomic perspective-it has received Data collection. In order to obtain a preliminary diversity estimation of mammals within the area, we performed a rapid ecological assessment (from 20 to 30 November 2004, with 8 days duration). Sampling methods included traps and mist-nets, as well as records obtained from direct observations and surveys to local communities. Collecting procedures were authorized by the Venezuelan Ministry of Environment under a scientific collection license and followed the ethical guidelines for animal research established by Latin American Mammalogy Network (RELAM, in Spanish).
First, for capturing small and medium non-volant mammals, we implemented three lineal transects (two in the lowland dry forests and one in the grassland/scrubs; Figure 1), where three types of catch traps (i.e., Tomahawk, Sherman and Victor [i.e., blow]) were placed properly baited (Tirira 1998, Prieto-Torres et al. 2011. Each transect had 36 Tomahawks, 36 Sherman and 36 blows traps, 10-15 m apart, alternately arranged. The traps were baited with a mixture made of peanut butter, sardines, and cornmeal (Tirira 1998, Rossi et al. 2016. These were checked daily in the morning, and all captured individuals were collected. For this method, the sampling effort was of 342 to 537 trap-nights in the lowland dry forests and the grassland/scrubs, respectively. Complementarily, a total of 11 mist nets (of 9-12 m in length) were used in order to capture bat species; these were placed only in the lowland dry forests and the riparian evergreen forests ( Figure 1) with a sampling effort of 64 hours/mist nest. Third, medium-and large-sized species having diurnal and/or nocturnal activity were inventoried from censuses carried out at a constant speed (i.e., ~1-1.5 km/h) along the transects at each sampling site. For each record, the following data was registered: species observed, number of individuals within the group (in the case of social species), detection method (visual or auditory), place and time of observation. In addition, occasional observations and indirect evidence (footprints, scats, nests, and carcasses) were also considered (e.g., Tirira 1998, Prieto-Torres et al. 2011, Rossi et al. 2016. Further, interviews with local people were conducted based on illustrations found in Linares (1998).
Each collected specimen was either prepared as skin, skull, and skeleton or fixed in 10% formalin and later preserved in 70% alcohol. These specimens were deposited in the mammal collection of the Museo de Historia Natural La Salle (MHNLS), Caracas, Venezuela (Table 1). All specimens were identified according to Sánchez and Lew (2012), which represent the most updated list of mammals for Venezuela since Wilson and Reeder (2005 descriptions (e.g., Sánchez-H et al. 2005), new records of species for the country, as well as the recent reviews of species and description of new genera (e.g., Emmons 2005, Larsen et al. 2007, Quiroga-Carmona & Woodman 2015. Data analyses. We estimated species rarefaction curves for species collected by both methods together (i.e., traps and mist nests). To test sampling sufficiency during the study, we estimated the species richness using the first order Jackknife and its respective confidence intervals, using program EstimateS 9.1.0 (Colwell 2013), considering as sampling unit each collecting day. These curves represent the cumulative number of species against the increase of collecting effort, obtained after 50 randomizations. The Jackniffe was used because it is easy to interpret, as it expresses the total estimated species richness through a relatively simple function of the number of rare species (Heltshe & Forrester 1983).

Figure 1.
Geographical location of Campo Boscán locality, middle basin of the Palmar River, Zulia state-Venezuela. Numbers in map correspond to sampling localities: the lowland dry forests (1 and 3), grassland/scrubs (2), and the riparian evergreen forests (4). Blue line corresponds to the Palmar River. Photos (by Giuseppe Colonnello) correspond to sampling sites within Campo Boscán: a) lowland dry forests; b) riparian evergreen forests; and c) grassland/scrubs. We did not include the flooded and non-flooded grasslands due that raining season produced the inaccessibility to that vegetation type during the study.
In order to briefly characterize the mammal species, we provided information about the distribution, trophic levels and conservation categories for each species. First, all species were categorized based on their national distributional range (see Linares 1998) in: a) Widely Distributed (WD; those that occupy large and continuous extensions in the country); b) Broad Distributed (BD; those that occupy large but discontinuous extensions); c) Broad-Restricted (BR; those species that -despite occupy large areas-are restricted to a particular habitat or ecosystem); and d) Restricted (R; those distributed throughout two or three continuous bioregions). Then, we calculated the percentage of species corresponding to each trophic level (e.g., omnivore, frugivorous, granivorous, others). Finally, we indicated the conservation status to species at national (Rodríguez et al. 2015) and

Results
Species richness. For this study, we recorded 47 species belonging to nine orders and 24 families (Table 1). From these species, 32 were recorded during field work (28 collected and 4 observed), while 15 corresponded to species referred by local people. The greatest richness was observed for the Chiroptera order with 20 species (42.55%), followed by Rodentia and Carnivora with nine (19.14%) and eight species (17.02%), respectively. However, the rarefaction curves ( Figure 2) showed that the survey did not reach the asymptote, which indicates that more species are expected in the sampling site. The first order Jackknife estimator of species richness predicted a richness of 59.95 species, which was beyond the 95% confidence intervals of the observed cumulative curve ( Figure 2).

Discussion
For MLD's mammals, studies on inventories, population dynamics and community structure are practically non-existent in the middle basin at Palmar River. In this context, the present study contributes to increase the current knowledge of biodiversity and threatened species in region. Mammals families and species reported in Campo Boscán represent 52.2 % and 12.7 % of the total reported for Venezuela, respectively (Sánchez & Lew 2012); as well as the 63.2 % and 33.6 %, respectively, for the Zulia state (Rivas 2006, Prieto-Torres et al. 2015b. Despite that these results can be described as satisfactory, it is important to note that only approximately ~78 % (considering the 15 species reported by local people) of richness predicted by the Jackknife richness estimator were recorded. This last fact was due to the short sample period (a characteristic of REA's studies; Lasso et al. 2009), the climatic conditions that produced floods in the riparian evergreen forests (producing a lower sampling effort during the study), and capturing methods used (e.g., Voss et al. 2001, Martins et al. 2016. Unlike other vertebrates -that can be effectively poisoned, captured and/or hunted-most of mammals can only be inventoried by trapping, which could require in most cases long-term efforts to obtain a complete data (e.g., Voss et al. 2001, Prieto-Torres et al. 2011, García et al. 2015, Martins et al. 2016. This is the case of some forest canopy species (e.g., arboreal marsupials and rodents), which require of other additional techniques such as acoustic methods and camera traps (Ochoa et al. 2005, García et al. 2015, Martins et al. 2016); here not used due the logistical difficulties and resource unavailability.
Despite the differences on sampling efforts for all vegetation types, the percentage of species reported associated with the riparian evergreen (n = 20; 42 %) and lowland dry (n = 13; 28 %) forests could be explained, on one hand, by the bats sampling (see below; Table 1) during the study. However, alternatively, results can be explained by the greater spatial heterogeneity of riparian habitats, which provide greater diversity of refuges and food (e.g., Machado et al. 2016). It is also possible to infer that the species richness observed in Campo Boscán is considerably high due the overlap of different environments (i.e., ecotones) in this region; this is evidenced by the presence of four species (Sylvilagus floridanus, Calomys hummelincki, Sigmodon alstoni, and Urocyon cinereoargenteus) defined as characteristic elements of dry forests and savannas. This mixed composition increases the biological importance of region (e.g., Costa et al. 2000, Joshi 2016, Machado et al. 2016, which emphasizes the need for its conservation in order to preserve the biodiversity of these fragments. Most species of mammals in our study have wide distributions along Venezuela and some medium-size and large species were strongly associated with diverse ecotones -mainly dependent on water-for example L. longicaudis, Procyon cancrivorus, and Hydrochaeris isthmius. Felid species (Carnivora), such as Puma concolor and P. yagouaroundi, are mainly associated with riparian evergreen forest areas, but can expand their home range given their ability to occupy open areas with intervened covers such as pastures and crops (MARN-PROFAUNA 1997, Prieto-Torres et al. 2011, Machado et al. 2016. Furthermore, despite Primates are among the most endangered groups, it is important to note that there is few information on the ecology and conservation status of species such as Alouatta seniculus and Cebus albifrons in MLD' areas (Portillo & Velásquez 2006). These species have broad phenotypic flexibility and occur in small forest fragments, which demonstrates a high capacity to adapt to environmental changes (Boubli et al. 2008, de la Torre et al. 2015. Therefore, although these species usually inhabit larger areas, its presence in areas outside close Protected Areas (i.e., Sierra Perijá and Juan Manuel National Parks) could be important for its preservation (Portillo & Velásquez 2006).
Bat species composition and richness varied along sampling sites at Campo Boscán (Table 1). The high abundance of most common species -such as Noctilio albiventris, Desmodus rotundus and Glossophaga longirostris-could indicate the availability of resources for generalists such as roost sites and food sources (e.g., livestock for D. rotundus and artificial lakes stocked with fish for N. albiventris). This last idea is supported by the presence of varied species of insectivores, frugivores, nectivores, hematophages, and piscivores (Novaes & Nobre 2009, Shapiro & Bordignon 2014. In addition, the presence of individuals from the subfamily Phyllostominae -which are considered bio-indicators (Fenton et al.1992 insect populations´ control (including agricultural pests), seed dispersal, and plant pollination (e.g., Kelm et al. 2008, Kunz et al. 2011, Shapiro & Bordignon 2014. However, further sampling is necessary at this site to assess the role of these species in the secondary succession and the regeneration of degraded forests in the area (Kelm et al. 2008, Kunz et al. 2011). Further, based on the high relative frequency observed for the common vampire bat (D. rotundus), a transmitter of bovine rabies (Baer 1982;Greenhall et al. 1983), future studies should be implemented regarding sanitary control to avoid economic loss.
In terms of conservation, unfortunately there is no other information about the extant wildlife found in the area; however, it is evident the negative effects of habitat loss and hunting over local wildlife populations. Forests have been reduced considerably, only leaving isolated relicts in almost the entire area (Duarte 1991, Romero & Monasterios 1996, Rodríguez et al. 2010, Colonnello & Lasso-Alcala 2011. Many mammal species are dependent on native vegetation cover, thus, fragmented and reduced habitats -such as Campo Boscán-can lose the original mammal community if species are not able to find the resources for long-term survival (Chiarello 1999, Machado et al. 2016. This is probably the case for species such as T. terrestris -considered locally extinct (IUCN 2015, Rodríguez et al. 2015 and the Brown Spider Monkey (Ateles hibrydus), a common species in other MLD's localities (Portillo & Velásquez 2006), which was not observed or referenced during the study, as well as neither the armadillo (Cabassous centralis), deer (Mazama amaricana, M. bricenii, and Odocoileus cariacou) and peccaries (Pecari tajacu and Tayassu pecari) species -all species reported as heavily hunted in the region (e.g., Prieto-Torres et al. 2011). In this sense, the information obtained suggests the need for implementation of improved conservation strategies focusing on the importance of minimizing impacts in the area.
Definitely, a single-species approach to conservation, management, and monitoring is insufficient to combat the threat to the overall biological diversity of an area (Joshi 2016). Thus, rapid ecological assessment and multi-species-based monitoring approaches are believed to be more reliable, timely, and informative in describing and measuring changes in populations, communities and biological diversity in general (Joshi 2016, Martins et al. 2016. In this sense, our results provide a valuable database in order to recognize and promote this area as an important wildlife resource; which currently have been affected by important problems related to the loss, fragmentation and degradation of habitats (Romero & Monasterios 1996, Rodríguez et al. 2008, Colonnello & Lasso-Alcalá 2011, Colonnello et al. 2014). This information can be used for the development of monitoring and conservation programs, promoting the recovery, reforestation and connectivity of forest patches -important for endangered species with large home-range such as L. pardalis, L. tigrinus and L. longicaudis. Additional studies on the abundance/distribution and status of wildlife species, also involving local communities and all stakeholders, from MLD is of supreme importance for securing long-term persistence of mammalian fauna in the area. Hydrochoeris isthmius Goldman, 1912--Handley (1966; Mones (1991) Description: Size small (Smaller than H. hydrochaeris), head and body length < 1 m. Pelage color ranges from dark reddish to dull clay color, usually darker above than below. The ears and feet are brownish. Eye rings, sides of muzzle, and spots at base of ears paler. Skull with pterygoid processes shorter, thicker, and more rounded, less produced posteriorly. Premaxillae with posteriorly extended portions reaching to near posterior plane of incisive foramina.

Family Cuniculidae Miller et Gidley, 1918
Cuniculus paca (Linnaeus, 1766) --Cabrera (1961Hall (1981); Pérez (1992) Description: Chestnut red to dark brown upperparts with three or four lines of large white spots, often coalescing into stripes, from sides of neck to rump; lower cheeks, throat, chest, and venter white. Pelage is short and sparse; spots aligned in rows largely on sides; plantar surfaces smooth. Zygoma more greatly expanded laterally, compressing the infraorbital foramen as viewed from the front; nasals short.

Family Erethizontidae Bonaparte, 1845
Coendou prehensilis (Linnaeus, 1758) -- Emmons & Feer (1997)  Description: Is a mid-sized rabbit weighing approximately 1 -2 kg. Five toes are present on the forefeet, and four on the hind feet. The dorsal portion of the body varies in color from grays to browns with the tips of each hair being white or silver giving the pelage a frosty look. A distinct rust-colored patch can be seen on the nape of the neck and fronts of the forearms. The ventral portions are cleaner than back. The short, fluffy tail is grayish above and white below.

Order Carnivora Bowdich, 1821
Family Canidae Fischer, 1817  (2000) Description: Large size (> 25 kg). It has a uniform dorsal pelage, with colorations ranging from light brown to dark brown, while ventral pelage is fur cream color. It has white color under the nose and throat, black on the side and upper side of the muzzle and at the posterior base of the ears. The tail is large and cylindrical, with a dark brown or black tip and approximately 30% of the total length of the animal. The skull is rounded and short.
Puma yagouaroundi (E. Geoffroy Saint-Hilaire, 1803) --Lacépède (1809); Desmarest (1816); Weigel (1961); Hemmer (1978); Kratochvíl (1982); Oliveira (1998) Description: It has short, dense and rigid pelage. Its body has homogenous coloration, which may be brown, grayish brown, reddish brown, yellow or black. Its tail length is greater than 60 % of body length. The legs and hands are short and digitrified. It has narrow skull and short facial area (about 20 % of skull length). They have a differentiated and deep pit between the internal back and the anterior part of the interfrontal sagittal sutures.

Family Mephitidae Bonaparte, 1845
Conepatus semistriatus (Boddaert, 1784) --Cabrera (1957) Description: They have long body with short legs, with dark brown to black color. Their pelage is dense and rough. Their feet and hands have long, strong and black claws. They have a white band from the head to the shoulders, and two parallel bands on the lower back, separated by a black band. The tail is usually erect, short (about half head-to-body length), spongy, and wholly white with black base. Their head are conical, long and bare muzzle. Their short, black ears with white hairs at base.

Family Mustelidae Fischer, 1817
Lontra longicaudis (Olfers, 1818) --Pohle (1920); Cabrera (1957) Description: It has medium size with height about 23 cm. Dorsal pelage short and dark brown to gray; short and thick pelage on the neck. Its face has a white line followed by and mask (white and black) that pierces the eyes. Its tail is cylindrical and shorter than body length, with dark rings. Slightly pointed ears, and dark colored legs.