Biological survey of a cloud forest in southwestern Mexico : plants , amphibians , reptiles , birds , and mammals

Cloud forest ecosystems contain unique flora and fauna characterized by high levels of richness and endemism. However, this ecosystem is one of the most threatened because of land-use changes stemming from anthropogenic activities. Therefore, biological inventories are necessary to adequately assess the effects of land-use changes on species now and in the future. In this study, we conducted an inventory of plants and terrestrial vertebrates (amphibians, reptiles, birds, and mammals) in three fragments of cloud forest in southwestern Mexico. Field work was carried out for 15 days per biological group during distinct time periods (2005-2008). Conventional methods of species capture and observation were employed to record species. Recorded species were then categorized based on their endemism and risk category. A total of 67 species of plants, 17 species of amphibians, 25 species of reptiles, 93 species of birds, and 46 species of mammals were recorded. The species accumulation curves for most taxa, except for birds and mammals, showed an asymptotic trend. A total of 56 species endemic to Mexico and four quasi-endemic species were recorded. Plants, amphibians, and reptiles presented the greatest number of species exclusive to Mexico (13 species). Six species of herpetofauna endemic to Guerrero were recorded. According to Mexican laws, 24 of the encountered species are under special protection, while 16 are categorized as threatened and seven as endangered. Reptiles and birds presented the greatest number of at-risk species (14 species). Bird and mammal richness in this study is high in comparison to that recorded in the cloud forests of the entire Mexican state of Guerrero (157 and 75 species, respectively). This data highlights the importance of cloud forests in the study area for local and regional biodiversity. Effective conservation strategies should be prioritized in cloud forests, as this ecosystem is poorly represented in natural protected areas.


Introduction
Mexico is a country that contains a large portion of the world's biodiversity and is well represented in worldwide diversity lists, usually occupying one of the first places.Overall, Mexico contains approximately 10% of global biological richness (Mittermeier & Goettsch de Mittermeier 1992, Rammamorthy et al. 1998, Martínez-Meyer et al. 2014).Among higher organisms, Mexico occupies fourth place in plant richness (Villaseñor 2016), fifth in amphibian richness (Parra-Olea et al. 2014), second in reptile richness (Flores-Villela & García-Vázquez 2014), eleventh in bird richness (Navarro-Sigüenza et al. 2014), and third in mammal richness (Ramírez-Pulido et al. 2014).These high levels of biodiversity are partly due to the convergence of the Nearctic and Neotropical biogeographical regions within Mexico's territory, giving rise to species with both temperate and tropical affinities (Plascencia et al. 2011).Despite the vast biological resources of Mexico, the flora and fauna of this country have not been fully documented (Martínez-Meyer et al. 2014), and several regions with potentially high levels of diversity and endemism remain relatively unexplored.
The Sierra Madre del Sur (SMS) in western Mexico is an important region from a biological perspective.This region is characterized by a complex orography and geological history that has promoted the development of a wide range of environments and ecosystems (Ferrusquía 1998, Luna-Vega et al. 2016), including cloud forests (CF).In the SMS, CF covers approximately 1,765 km 2 and corresponds to 20% of the territory (INEGI 2010).As previously mentioned, this ecosystem is highly diverse and contains many exclusive flora and fauna species (Challenger 1998, Gual-Díaz & Rendón-Correa 2014).Notably, various fragments of this ecosystem are distributed in the form of an archipelago across the mountainous environments of the SMS, and each fragment has a particular biotic composition (Gual-Díaz & Rendón-Correa 2014).However, limited biological information is available on the CF of this region.
The lack of biological studies on the CF of western Mexico is related to the physical inaccessibility of some sites (e.g., because of difficult terrain or lack of roads) and social problems.As few biotic inventories have been conducted in this region to date, our study serves as an initial approximation for increasing knowledge on the biological diversity of CF in the SMS of the Mexican state of Guerrero.We performed biological inventories of plants and terrestrial vertebrates and provided information on the endemism and risk status of the recorded species.Also, our study represents an important contribution to the national biotic inventory and serves as a baseline of biodiversity with which future scenarios and studies can be compared.Ultimately, this inventory serves as a tool for conservation efforts in the study region, especially considering that CF are highly threatened by human activities.

Study area
The study area is located to the west of the biotic province of SMS in the state of Guerrero in southwestern Mexico between 101°6'30.84" to 100°57'57.86" W and 17°46'16.87" to 17°43'24.45"N (Figure 1).The studied CF fragments comprise a total area of 11 km 2 and present an altitude between 1700 and 2200 masl (Figure 1).The climate is humid and sub-humid temperate, with a mean annual temperature between 16 and 20 °C and a mean annual precipitation between 1200 and 2000 mm (García 2004).The study area is characterized by rugged orography with slopes of 4° to 60°.Several types of land management are practiced in the CF of the study area, including agriculture (e.g., mainly corn, bean, and squash crops) and forestry, wherein species of economic importance (e.g., Quercus glaucoides and Q. candicans) are harvested.mostraron un comportamiento asintótico para la mayoría de los taxa, excepto aves y mamíferos.Se obtuvo un total de 56 especies endémicas a México y cuatro cuasiendémicas.Los grupos de plantas, anfibios y reptiles presentaron el mayor número de especies exclusivas al país (13 especies), y se obtuvo un total de seis especies de herpetofauna endémicas a Guerrero.De acuerdo con las leyes mexicanas, se registró un total 24 especies en protección especial, 16 amenazadas y siete en peligro de extinción, de los cuales los grupos de los reptiles y aves presentaron el mayor número de especies en categoría de riesgo (14 especies).La riqueza de especies de aves y mamíferos en este estudio representa un número importante comparado con el total de especies registradas en los bosques mesófilos del estado de Guerrero (157 y 75 especies, respectivamente).Estos datos resaltan la importancia de los bosques mesófilos de montaña del área de estudio para la biodiversidad local y regional, por lo que deben priorizarse estrategias de conservación efectivas para este ecosistema poco representado en áreas naturales protegidas.Palabras clave: bosque mesófilo, conservación, diversidad, vertebrados terrestres, flora, riqueza de especies.

Sampling effort, inventory completeness, and identification of species at risk
Different conventional methods were employed for the inventory of each biological group.A total of 15 days of field work, carried out from 2005-2008, were dedicated to each group.Species accumulation curves were calculated to evaluate the completeness of the inventories of each taxonomic group.Potential species richness was estimated using the incidence-based coverage estimator (ICE) method, wherein the unit effort corresponded with the number of sampling days.This estimator is based on estimated sample coverage (i.e., the proportion of richness represented in a set of replicated incidence samples) and is insensitive to sample size (Gotelli & Colwell 2011).These analyses were performed using the "specaccum" function in the "vegan" package (Oksanen et al. 2017) in R 3.3.3(R Development Core Team 2017).Following the surveys, a comparative analysis was carried out to evaluate differences in the richness of each biological group in comparison to the richness levels recorded in CF at both the state and national level.For these comparisons, data on plants, amphibians, reptiles, and birds at the state and national level were taken from Villaseñor (2010) and Gual-Díaz & Rendón-Correa (2014).Mammal data at the national level were taken from González-Ruiz et al. ( 2014).In addition, the studies of León-Paniagua & Romo-Vázquez (1993), Jiménez-Almaraz et al. (1993), Ávila-Nájera (2006), Ruiz-Gutiérrez (2012), andAlmazán-Catalán et al. (2013) were used to compare mammals at the state level because no single source has reported the total number of mammal species in the CF of the state of Guerrero.
For each group, the risk status of species was assigned according to the following categories listed in the Mexican Official Standard NOM-059-SEMARNAT-2010 (SEMARNAT 2010): (1) special protection, which includes those species or populations that could be threatened by factors that adversely affect their viability and determine the need to facilitate their recovery (this category may include the lower risk categories of the IUCN classification), (2) threatened, which includes those species or populations that could be in danger of disappearance in the short to medium term if the factors that adversely affect their viability, such as habitat modification or disturbance, directly reduce the size of their populations (this category overlaps with the "vulnerable" category in the IUCN classification), and (3) endangered, which includes those species whose range or population size has dramatically decreased within Mexico as a result of drastic habitat modification or destruction, overharvesting, disease, or predation, among others, thereby threatening the biological viability of these species throughout their natural habitat (this category overlaps with the categories "critically endangered" and "endangered" in the IUCN classification).

Plants
To inventory plant species, we delimited 30 quadrants of 30 x 30 m (0.027 km 2 ) at different sites and counted and identified plant species with flowers and/or fruits.When we were unable to identify plants in situ, we collected botanical samples (three specimens per species) for subsequent identification.Collected specimens were pressed and labeled for herborization according to the protocol of Wendt (1986) and identified through comparison with existing specimens in the Herbarium of the Instituto de Investigación Científica Área Ciencias Naturales (IICACN) of the Universidad Autónoma de Guerrero (UAGro) and in the Vascular Plant Laboratory of the Facultad de Ciencias of the Universidad Nacional Autónoma de México (UNAM).Nomenclature was verified in the database of the Missouri Botanical Garden (W3Tropicos 2010).Species endemism was determined based on Villaseñor et al. (2016).The systematic arrangement of species follows Wearn et al. (2013) and the guidelines of the Angiosperm Phylogeny Group (APG IV 2016).

Amphibians and reptiles
To inventory amphibian and reptile species, we conducted walks along linear transects during three daily observation episodes (diurnal, crepuscular, and nocturnal) to cover the peak hours of activity.To cover a high proportion of amphibian and reptile microhabitats, we placed transects along diverse environment, including rivers and streams in CF as well as those associated with coffee plantations.Sampling consisted of direct observation along roads, wetlands, streams, and trunks as well as under rocks and in caves.Specimens were collected employing traditional capture techniques (Casas-Andreu et al. 1996).Individuals were directly collected by hand, and for particularly poisonous species, a herpetological hook was used.For each specimen captured, we recorded meristic data, coloration, and microhabitat.Amphibian and reptile specimens were identified using field guides by Casas & McCoy (1987) and Flores-Villela et al. (1995).

Birds
To inventory bird species, we carried out bird observations during the hours of highest bird activity in the morning (07:00 to 10:30 h) and afternoon (16:00 to 18:30 h).In addition, 10 mist nets 12 m long and 2.5 m wide were placed to complement the species inventory of the study area.A total of 1800 h/net were obtained.Nets were placed in different sites than those used for the bird observations.Captured specimens were identified and subsequently released at the same capture site.Binoculars (8 x 40 and 10 x 50) and field guides (Howell & Webb 1995, National Geographic Society 1999, Sibley 2000) were used to identify species.Each observed bird species was categorized according to its seasonality (Howell & Webb 1995) and endemism (González-García & Gómez de Silva 2003).The scientific nomenclature and systematic arrangement of species followed the guidelines of the American Ornithologists' Union (AOU 1998) and its most recent update (Chesser et al. 2017).

Mammals
To inventory bats, five mist nets 12 m long and 2.5 m wide were used.Nets were placed across rivers and streams and in open areas between wooded areas.All nets were opened daily from 19:00 to 06:00 h; a total of 550 h/net were obtained.To collect rodents, 50 Sherman traps baited with oatmeal and vanilla were placed along two transects.To collect shrews, 1-L pitfall traps were buried at soil level.The traps for rodents and shrews were placed in the afternoon and checked the following morning.Also, the occurrence of medium-and large-sized mammals was confirmed through searching for excreta and footprints, mainly along riversides and walking trails.All excreta and footprints were determined using the guide by Aranda (2000).In addition, data on sightings of medium-sized species in the field were recorded.The taxonomic guides by Hall (1981), Álvarez et al. (1994) 4b) and reptiles (Figure 4c) showed an asymptotic trend, yet the curve for amphibians began to stabilize at the middle of the sampling period.The ICE method estimated 17 amphibian species (CI = 11 to 23 species; α = 0.05) and 26 reptile species for the study area (CI = 8 to 44; α = 0.05).

Birds
A total of 93 bird species belonging to 32 families were recorded (Figure 2 The number of species continuously increased with sampling, and the accumulation curve was not asymptotic (Figure 4d).The ICE method estimated 101 bird species for the study area (CI = 83 to 119; α = 0.05).

Comparisons with CF at the state and national level
The species richness obtained for each taxonomic group in the CF of our study area is not low in comparison to that of CF ecosystems at the state level (Figure 5).For example, 52% of amphibian species, 63% of reptile species, 59% of bird species, and 61% of mammal species recorded in CF of the state of Guerrero were also recorded in the study area.Also, in comparison to the species richness of each group in CF at the national level, the species recorded in our study area represent 1.1% of plant species, 9% of amphibian species, 10% of reptile species, 17% of bird species, and 18% of mammal species (Figure 5).et al. (1997) were used to identify mammal species.After identification, mammals were released at the same capture site.Endemism was determined following Gutiérrez-Blando et al. (2016).Scientific nomenclature and systematic arrangement followed the guidelines by Ramírez-Pulido et al. (2014).

Discussion
With the exception of plants, the species richness obtained in our study area for the four evaluated animal groups represented more than half of the total species richness of CF in the state of Guerrero.In addition, this species richness represented at least one-quarter of the total species richness of the entire SMS (Espinosa et al. 2016, Flores-Villela & Ochoa-Ochoa 2016, Gutiérrez-Blando et al. 2016, Navarro-Sigüenza et al. 2016).In the case of the plants, the inventory is apparently small, as the sampling was conducted across only 0.027 km 2 (0.25% of the entire study area).In comparison with other studies in CF across the SMS, the species richness of our study area is lower than that obtained by Fonseca et al. (2001), Catalán-Heverástico et al. (2003), and Lozada-Perez et al. (2003); however, these studies were carried out in larger areas (35.5 km 2 , 0.045 km 2 , and 7 km 2 , respectively).This suggests the biological importance of the CF in our study area and provides a baseline of species richness in this region.Existing species must be continuously monitored, yet this baseline can be used to develop effective conservation strategies.In addition, these data are of particular importance considering that studies on flora and fauna in the SMS biotic province have been scarce (Luna-Vega et al. 2016), although several descriptive studies of the biological resources of this region have been carried out in the last decade (e.g., Almazán-Núñez et al. 2007, Almazán-Catalán et al. 2009, Almazán-Núñez et al. 2009, Flores-Villela & Ocho-Ochoa 2016, Gutiérrez-Blando et al. 2016, Navarro-Sigüenza et al. 2016).
However, biological studies have not been performed in several areas, particularly CF, that likely have high levels of biodiversity.Cloud forests are one of the ecosystems with the highest levels of biodiversity in Mesoamerica (Gual-Díaz & Rendón-Correa 2014).In Mexico, the best-studied CF in biological terms are located in the central-eastern portion of the country, particularly in the states of Veracruz and Hidalgo (e.g., Martínez-Morales 2007, García-Franco et al. 2008, Álvarez-Zúñiga et al. 2012, Aguilar-López et al. 2013, Rueda-Hernández et al. 2015).Meanwhile, other areas with an extensive presence of this ecosystem, such the southwestern portion of the country (e.g., the state of Guerrero), remain relatively unexplored.In comparative terms, the total species richness of the five taxa in our study area is higher than that reported for Cerro Piedra Larga, Oaxaca, Mexico (Peterson et al. 2004), even though the present study only focused on CF.However, such comparisons should be performed with caution because these geographic areas as well as the corresponding sampling efforts and environmental factors are not similar (Watson & Peterson 1999).The importance of several species recorded in our study area can also be highlighted from a conservation biology perspective.For instance, several species are endemic to western Mexico (e.g., Eleutherodactylus nitidus, Rhadinaea hesperia, Micrurus distans, Piranga erythrocephala, and Amazona finschi).Others are restricted to the state of Guerrero (e.g.Mixcoatlus barbouri, Sceloporus adleri, and Thorius grandis) or to small portions of the SMS (e.g., Charadrahyla tecuani, Eupherusa poliocerca, Cyanolyca mirabilis, and Cryptotis goldmani).In fact, our study area is part of the biotic province of the SMS, which is considered a priority terrestrial region because of its high biological diversity and the presence of a significant number of endemic species (Arriaga et al. 2000).
The distribution of species richness in biological groups is not homogeneous across CF at the state and national level, as some areas contain higher concentrations of certain groups of species (Gual-Díaz & Rendón-Correa 2014).The differential presence of distinct biological groups is largely the result of in situ speciation processes.The physical characteristics of CF ecosystems, such as climate and historical factors (i.e., rugged orography that acts as a barrier) have favored the presence of many endemic species in CF (Watson & Peterson 1999) and a significant number of species that are restricted to this habitat (e.g., Cyanolyca mirabilis, Megadontomys thomasi, and Eleutherodactylus augusti).Several of these species form genetically differentiated populations that are widely distributed, for example, Charadrahyla tecuani, Thorius grandis, Lampornis amethystinus, Aulacorynchus prasinus, and Chlorospingus flavopectus (Hanken et al. 1999, Navarro-Sigüenza & Peterson 2004, Campbell et al. 2009).Overall, 75 mammal species have been recorded in other CF of Guerrero by León-Paniagua & Romo-Vázquez (1993), Jiménez-Almaraz et al. (1993), Ávila-Nájera (2006), Ruiz-Gutiérrez (2012), andAlmazán-Catalán et al. (2013).In our study area, a relatively large portion of these species were found (46 species).The absence of species characteristic of CF in certain regions may be due to different factors, for example, the variation in dominant floristic species from one region with CF to another.Also, latitude has been suggested to be one of the most important factors that determines differences in species richness and endemism throughout the CF of Mesoamerica (Watson & Peterson 1999).Other possible factor that can explain the absence of some species in the CF of our study area is the uneven sampling effort, as several species unregistered in this study are indeed common and relatively easy to detect in other CF ecosystems (e.g., Colaptes auricularis, Lepidocolaptes affinis, Glossophaga soricina, and Cryptotis mexicanus; Hernández-Baños et al. 1995, Peterson et al. 2004, Martínez-Morales 2007, González-Ruiz et al. 2014).Also, some latitudinal migratory species, particularly birds (e.g., Oreothlypis ruficapilla, Polioptila caerulea, and Selasphorus rufus), are likely poorly represented in our study, as the bird sampling period comprised only the months of January, April, and May; this was reflected in the accumulation curve for bird species.However, the lists of each biological group generated in our study may be considered representative based on the expected richness calculated by the ICE estimator.
A high concentration of endemism is present in different biological groups of western Mexico (García-Trejo & Navarro 2004, Ochoa-Ochoa & Flores-Villela 2006, Gutiérrez-Blando et al. 2016, Villaseñor 2016).The flora and fauna endemic to our study area (56 species) was notable compared to the total number of endemic species reported for the entire state of Guerrero (380 species; Navarro 1998, Pérez-Ramos et al. 2000, Villaseñor 2016).This richness of endemic species was found in an area that barely covers 0.02% and 0.019% of the area of Guerrero and the SMS, respectively.In comparison to the CF of Guerrero (Hernández-Baños et al. 1995, Gual-Díaz & Mayer-Goyenechea 2016, Mayer-Goyenechea & Gual-Díaz 2016), our study area contained a large proportion of endemic amphibian species (48%), reptiles (35%), and birds (73%).However, despite the high concentration of endemic plants in CF (Villaseñor 2010), no endemic plants were found in our study.
With respect to the at-risk status of the recorded species, a total of 47 species of flora and fauna were identified to be at risk.In particular, Carpinus caroliniana, Eupherusa poliocerca, Herpailurus yagouaroundi, and Trimorphodon quadruplex are threatened, and Aulacorhynchus prasinus, Craugastor uno, Cryptotis goldmani, Cyathea bicrenata, and Lithobates sierramadrensis are under special protection status.Species such as Cyanolyca mirabilis and Tamandua mexicana are endangered.These results indicate the importance of preserving the CF ecosystems, as it presents high levels of endemism, contains numerous at-risk species, and currently represents one of the most vulnerable and unprotected ecosystems in Mexico.
Despite the high species richness and the high number of at-risk species recorded in the CF of our study area according to NOM-059-SEMARNAT-2010, few studies have focused on the importance of biological richness in CF at the state level.A greater knowledge of wildlife and floristic resources at the local level can complement inventories at the state and national level and can be useful for promoting effective strategies for the conservation and use of biodiversity.For these reasons, it is imperative to continue these types of studies, especially in unexplored environments, in order to obtain greater information on the distribution and diversity of different taxa and to improve conservation efforts for these species.

Figure 1 .
Figure 1.Location of the study area in the SMS of the state of Guerrero in southwestern Mexico.

Figure 2 .
Figure 2. Species richness, endemic species, and species at risk per biological group in cloud forest fragments of southwestern Mexico.

Figure 5 .
Figure 5. Number of species per biological group in the cloud forests of Guerrero and Mexico.