Introduction

In Mexico, wetlands are examples of the great biodiversity of the country, not only due to their different types, from coastal lagoons and mangrove swamps to inland lakes and marshes, but also because they support a great diversity of species; likewise, they provide the local human societies with valuable natural resources (^{Wilson and Ryan 1997}). Nevertheless, they are also one of the most threatened environments due to anthropogenic pressures (^{Sebastián-González et al. 2013}).

Wetlands are essential for many species such as waterfowl, which depend ecologically on them and are one of the most remarkable elements of global biodiversity (^{Wetlands International 2012}). For a considerable proportion of North America’s waterfowl and other migratory bird populations, Mexican wetlands are essential stopover and wintering sites and their protection is critical to ensure the long-term survival of these birds (^{Wilson and Ryan 1997}, ^{DGVS 2008}). In response to the wetlands and waterfowl protection needs, Mexico has become a member of conservation agreements, participating internationally in the Ramsar Convention on Wetlands (^{Ramsar 2014}), regionally in the North American Wetlands Conservation Act (^{Wilson and Ryan 1997}) and the American Waterfowl Management Plan (^{Williams et al. 1999}, ^{Pérez-Arteaga et al. 2005}), and at the country level, the National Strategy for the Conservation, Management and Rational Use of Waterfowl and their Habitats in Mexico has been implemented (^{DGVS 2008}).

The study and characterization of wetlands and waterfowl populations are relevant for planning, adapting, and implementing conservation actions against the threats that they face (^{DGVS 2008}). Leading efforts have been achieved for assessing the status of wetlands and waterfowl at international and national levels (^{Pérez-Arteaga and Gaston 2004}, ^{Pérez-Arteaga et al. 2005}, ^{Wetlands International 2012}), but studies about regions or specific locations are still needed, especially in the Neotropical region (^{Rojas-Soto and Olivares de Ita 2005}). Hence, it is necessary to monitor the waterfowl populations to gather information that can be used for the establishment of priority sites and protection strategies at the local and regional levels, especially in relation to the modification of natural habitats and the creation of new ones by human activities, like artificial wetlands which could become alternatives for biodiversity conservation (^{Carmona et al. 2011}, ^{Sebastián-González et al. 2013}). Here, we make the first description of the waterfowl community that inhabits an artificial and protected wetland in the central highlands of Mexico, during a period of one year, and in that way, we provide a reference for further conservation decisions and ecological research.

Methods

Study area

The study area is encompassed in the river basin of Lerma-Chapala-Santiago (^{Obregón 2011}), in the middle of the trans-volcanic belt, where intense volcanic activity created several lakes and marshes (^{Mitsch and Hernández 2013}). It is near the Ciénegas de Lerma, a natural protected area which is the biggest wetland in the central zone of Mexico and it is a crucial area for the migration of North American waterfowl (^{Colón-Quezada 2009}, ^{Ramsar 2014}). This area is under considerable pressure for agricultural production and due to the loss of the original vegetation, the ground has a high evaporation rate and many natural wetlands have been lost. For that reasons, several artificial water reservoirs were constructed to irrigate the fields. The climate of the region is temperate with summer rains and frosts, the average annual temperature is 12.6°C and the average precipitation is 878.4 mm (^{López et al. 2007}).

The artificial wetland that we studied is called “Bordo las Maravillas”, is part of the University Campus El Cerrillo, Piedras Blancas of the Autonomous University of Mexico State, and it is located in the municipality of Toluca in Mexico State, at 2600 m above sea level (19°24’26.71”N; 99°41’30.71’’ W) (Figure 1). It was constructed for agricultural purposes, although it is a natural area under the scheme of Unit for the Management and Conservation of Wildlife (UMA for its abbreviation in Spanish) without natural resources extraction. Its surface is 16.59 ha and it has a floodgate and a channel that delivers domestic sewage. The depth is 70 cm on average, but it varies across the year due to the input of rainwater, the evaporation and the extraction of water for agricultural irrigation; the deepest areas are toward the floodgate and the shallowest ones are near the channel (^{Obregón 2011}). The perimeter is 1,577 m, and is secured with cyclonic mesh and surrounded almost entirely by willows (Salix babilonica); the aquatic vegetation is composed of species like Scirpus sp., Juncus microcephalus and Typha latifolia (^{Colón 2004}, ^{López et al. 2007}), which create several islands at the interior of the wetland.

Figure 1 Map of the artificial wetland “Bordo las Maravillas”, Mexico State. The numbers indicate the observation points, the dotted line is the route followed in the transect and the shaded areas are the water surface observed from the observation point and the transect of the corresponding color. 

Results

We recorded the presence of 23 species, which belong to five orders, six families, and fifteen genera. The most represented family was Anatidae, from which we observed ten species, and the second one was Ardeidae with six species. From the families Threskiornithidae and Scolopacidae we observed one species each, thus they were the least represented. By contrast, considering the observed species in relation to the total of species for the country per family, the highest proportion was in the Ardeidae family in which we recorded six of the sixteen species (37.5%). Most of the species that we observed (14/23) are considered migratory in the region of our study (^{Berlanga et al. 2008}). The complete list of species, their proportions per family and migratory status can be consulted in Table 1.

The species accumulation curve that we recorded had typical growth and on the December 13^th observation, reached its maximum value of 23 species. Accordingly, the curve calculated with Clench’s model was very similar; its asymptote had a value of 23.98, predicting that the ecosystem may harbor 24 species and shows that we found 95% of them and could only register one additional species (Figure 2). Likewise, the determination coefficient (R2) is close to one (0.99) showing that the model had a good adjustment to the data, and the curve slope was less than 0.1 (0.04) which let us conclude that the species list is reliable (^{Jiménez-Valverde and Hortal 2003}). Therefore, we had an accurate description of the species richness.

Figure 2 Observed and estimated species accumulation curves. The shadow area indicates the upper and lower 95% confidence intervals. 

The number of species varied during the observation period but the increase due to the addition of migratory species in the wintering months (October-March) was evident. At the beginning, we registered an average of seven species in August; then, from October to April, we registered over ten species. The highest record was during January (n=18), and afterwards, the species richness decreased to eight in August the next year; nevertheless, we recorded four species on November 29th which was the lowest value (Figure 3-A). The most frequent species were the Mexican Duck (Anas platyrhynchos diazi) and the American Coot (Fulica americana), which we observed in all but one of our visits and both had a relative abundance of 100%; by contrast, the species with the lowest relative abundance (6%) was the Sora (Porzana carolina) (Table 1).

Figure 3 Species richness (A) and abundance of individuals (B) observed in the artificial wetland “Bordo las Maravillas”, Mexico State. The green shading indicates the reproductive period and the blue one, the wintering period. 

The abundance of individuals in the site varied throughout the year, which was expected for the arrival of migratory species. Figure 3-B shows the increase of individuals during the winter period and the subsequent decrease in the summer months. We recorded the maximum abundance during January 24^th (n = 8,808) and the minimum, during August 2^nd of the first year (n = 79) (Table 2).

Excluding the group of unidentified ducks and teals, the Cattle Egret (Bubulcus ibis) was the most abundant species reaching an accumulative abundance of 1,943 individuals; followed by the Green-winged Teal (Anas crecca) with 1,141. The least represented species were the Sora (Porzana carolina), from which we just saw one individual on one occasion, the Least Grebe (Tachybaptus dominicus) (one individual in two visits), and the Great Blue Heron (Ardea herodias) (one individual in three visits). The accumulated abundance that we documented at the end of the study, considering only the maximum record per month, was 25,220 individuals. In Table 2 the abundance per visit, species and total is displayed.

Regarding the resident species, we confirmed that some were reproducing in this wetland. We observed Black-crowned Night-Heron (Nycticorax nycticorax) younglings and American Coot (Fulica americana) chicks. Additionally, we found a Mexican duck (Anas platyrhynchos diazi) nest with two eggs and later on the season we observed the younglings of this species.

Two species that we recorded have a category of protection in the Mexican law for the protection of endangered species. The Least Grebe (Tachybaptus dominicus) is included in the “special protection” category and the Mexican Duck (Anas platyrhynchos diazi), in the “threatened” category (^{SEMARNAT 2010}).

Discussion

The species richness, distribution and abundance are necessary parameters for making conservation decisions and for the sustainable use of waterfowl (^{DGVS 2008}, ^{Ruelas et al. 2009}).

Performing a systematic description by covering most of the water surface of the wetland of our study allowed us to obtain a comprehensive description of the waterfowl community. We had an almost complete record of the species richness, given that we registered 23 of the 24 predicted species; likewise, we could estimate the number of birds present and in this way, we observed the changes in the compositions of species and in the abundance of individuals in a year basis.

The particular low species count that we had on November 29^th, was associated with an unknown disturbance that made the birds leave at the beginning of our visit before we were able to identify them, although we could count the individuals as they flew over the wetland and therefore that fluctuation is not strictly reflected in the corresponding abundance. Likewise, we also faced a disturbance that scared the birds on December 13^th; on that occasion, it happened in the middle of our visit, so we identified many individuals, but we could not count all of them because they flew away from the wetland. To avoid this, digital recording methods like high definition photographs and videos could be used for achieving a better identification; although we did not have access to them during our study. Contrastingly, the outstanding abundance that we observed on January 24^th highly exceeded the previous and subsequent records. In this case, there is the possibility that other disturbances had scared the birds in the surrounding wetlands and they took refuge in our study wetland. Despite these peculiar situations, both the species richness and the abundance of individuals in the wetland showed a considerable increase in the wintering months, showing that this wetland can be a refuge for waterfowl during migration, similar to other important resting areas in the country (^{Ramírez-Bastida et al. 2008}).

An interesting observation was the presence of the Fulvous Whistling-Duck (Dendrocygna bicolor) and the Least Grebe (Tachybaptus dominicus), both of which are considered to be infrequent in the central highlands (^{Colón 2004}, ^{Ortega-Álvarez 2013}). In a previous study in this wetland, ^{Colón (2004)} reported the presence of the Fulvous Whistling-Duck (Dendrocyna bicolor) and Black-bellied Whistling-Duck (Dendrocygna autumnalis) which are unusual in the region or not present at all (^{Colón 2004}, ^{Ramírez-Bastida et al. 2008}). Although we confirmed the presence of the Fulvous Whistling-Duck during the summer of both years, we did not observe the Black-bellied Whistling-Duck. Similarly, the sight of the Least Grebe (Tachybaptus dominicus) is considered to be uncommon in the region (^{Ortega-Álvarez 2013}). In a similar way, the unique observation that we had of the Sora (Porzana carolina) was notable because certain species of waterfowl, like some from the Rallidae family, are elusive, cryptic and difficult to watch; therefore the studies of their abundance can be complemented with other techniques like recorded calling playback (^{Brambilla and Jenkins 2009}, ^{Conway and Nadeau 2010}). These observations suggest that this wetland offers specific characteristics to rare species, which increase its ecological value.

We confirmed that the Mexican Duck is reproducing here through our observation of a nest with eggs and juveniles. The Mallard (Anas platyrhynchos) has a broad distribution and big populations in America, but the Mexican Duck (Anas platyrhynchos diazi) is the only member of the genus Anas that has adapted as a resident in Mexican highlands, where 98% of its population resides (^{Pérez-Arteaga and Gaston 2004}) and it showed a significant decline in its populations in the Lerma region from 1991 to 2000 (^{Pérez-Arteaga et al. 2002}). The lack of suitable vegetation for nesting and refuge has being proposed as one of the limiting factors for its distribution (^{Pérez-Arteaga et al. 2002}); therefore, the preservation of wetlands that offer refuge, food sources, and nesting resources, is relevant for the conservation of this threatened duck.

In comparison to other similar descriptions of waterbird communities in the central zone of Mexico, the wetland that we studied has a high abundance of birds and species richness considering its surface (16.59 ha). For instance, ^{Fonseca et al. (2012)} reported for a one year period, an accumulated abundance of 48,794 individuals and 36 species of waterfowl (ten belonging to the Anatidae family) in the Acuitlapilco lagoon in Tlaxcala State, which surface varies from 30 to 70 ha through the year. In Río Grande de Morelia basin in Michoacán State (700 ha surface), during one year, ^{Barragán et al. (2002)} recorded 23 species of aquatic birds and the Anatidae and Ardeidae families were the best represented with six and five species respectively. From the 69 species (including others than waterfowl) observed by ^{Pineda-López and Arellano-Sanaphre (2010)} during two years in 12 sites with a combined surface of over 1500 ha in the Queretaro State, 18 belong to the Anatidae family. Like the mentioned authors, we also found that the most abundant family was Anatidae and that its input of migratory species has a strong influence on the composition and size of the waterfowl communities; hence, this family has a constant presence in the region despite the wetlands’ differing conditions. In our study and in the cited ones, the proportion of migratory and transient species was elevated; considering the regional residence status described in ^{Berlanga et al. (2008)}, 14 (61%) of the species that we observed, 27 (75%) of the species reported by ^{Fonseca et al. (2012)}, 16 (70%) from ^{Barragán et al. (2002)}, and 51 (74%) listed by ^{Pineda-López and Arellano-Sanaphre (2010)} belong to these categories. This confirms that the wetlands of the country are essential as resting and wintering areas during migration as mentioned previously (^{Ramírez-Bastida et al. 2008}).

Wetlands are highly dynamic ecosystems where resources vary seasonally (^{Barragán et al. 2002}) and these variations have a direct influence on waterfowl distribution because it depends on the characteristics, extensiveness, and quality of the wetlands (^{DGVS 2008}). Waterfowl choose sites with good food availability and shelter, especially for nesting and wintering (^{DGVS 2008}) and they use wetlands with high eutrophication levels (^{Barragán et al. 2002}); therefore, it is likely that migratory ducks adjust their time of stopover according to the foraging quality of the habitats (^{O’Neal et al. 2012}). For instance, ^{Obregón (2011)} found that this particular reservoir has a high level of eutrophication, tending to be hyper-eutrophicated. Also, the diet of some waterfowl species is mainly based on plants associated with wetlands and it can be complemented with crops, especially in regions where the wetlands have suffered from continuous degradation, the nearby fields can provide an alternative food resource for waterfowl (^{DGVS 2008}, ^{Colón-Quezada 2009}); for example, ^{Colón-Quezada (2009)} found that corn was one of the main diet components for the Mexican duck (Anas platyrhynchos diazi) in Ciénegas del Lerma. These could be some of the reasons why we observed migratory ducks in all but one of our visits and we recorded high concentrations of them during the wintering period; however, a comprehensive analysis of their diet in situ is needed.

Our study location is close to Ciénegas del Lerma Flora and Fauna Protected Area (approximately 25 linear km) and it is also surrounded by several artificial wetlands and dams; this creates a variety of close alternative habitats for the waterfowl with different quality and changing available resources; therefore, the species richness that we observed may be related to the regional diversity created by those wetlands. The features of this particular wetland, such as its protection, food availability, roosting sites, and refuge in the surrounding and aquatic vegetation, favor the presence and abundance of the species that we observed. Contrastingly, we did a reconnaissance visit to the six nearby reservoirs, less than 1.6 Km away, and we saw that they have different surfaces, depths, water sources, do not have any kind of protection, and the vegetation at their interior and edges was scarce, allowing people and cattle access to them creating disturbances constantly. The exception was the wetland situated Northwest from our study site which had similar vegetation islands; however, the number of aquatic birds that we observed in all of them was limited.

It should be also considered that the inter and intraspecific interactions, natural variations such as the vegetation succession, human usage of wetlands, and catastrophic events like fires, diseases or natural disasters, have a strong influence on the community composition (^{Sebastián-González et al. 2013}). During the period of our visits, a fire affected the South edge of the wetland on January 2011 and the vegetation recovered entirely in the following months; likewise, during December 2010 and January 2011 the water level dropped gradually around 0.5 m because the water was used for irrigation. Nevertheless, we did not observe changes in the species richness or abundance throughout these events; therefore, also the magnitude of these alterations is relevant. Thus, in order to have an integrated understanding of the waterfowl community in Bordo las Maravillas wetland and in the region, including population tendencies, changes in the species distribution and abundance, the long term analysis of the habitat selection is needed.

Artificial wetlands, like those constructed in the central highland region to satisfy the human demand for water (^{Barragán et al. 2002}), have a recognized importance for waterfowl because they provide functions and resources, like food and refuge, and they can partially mitigate the loss or reduction of natural wetlands (^{Carmona et al. 2011}). Large numbers of migratory and resident birds use artificial wetlands, at least temporally, which makes them beneficial to the bird communities, although they should not be considered as substitutes because natural wetlands are capable of supporting a greater diversity of species (^{Carmona et al. 2011}, ^{Sebastián-González et al. 2013}). By considering not only the wetlands’ origin but also their intrinsic properties, artificial wetlands become relevant and their protection should continue to be encouraged (^{Carmona et al. 2011}). For instance, from the 142 Ramsar Sites (Wetlands of International Importance) that cover 8.6 million ha in Mexico, 37 are human-made wetlands (^{Ramsar 2014}). Regarding the waterfowl, conservation programs should consider it as a priority group (^{Pineda-López and Arellano-Sanaphre 2010}) and in that sense, the improvement of wetland conditions, especially concerning water quality, the increase of vegetation, nesting resources, and protection from disturbances such as cattle, benefits the diversity of waterfowl and other associated species (^{Pérez-Arteaga et al. 2002}, ^{Pineda-López and Arellano-Sanaphre 2010}).

Although the conservation efforts traditionally take place in zones with high ecologic value, this does not necessarily mean that only large and diverse habitats should be prioritized; the importance of small habitats (10 - 40 ha) as refuge for certain species or as connecting nodes has being highlighted previously (^{Pineda-López and Arellano-Sanaphre 2010}, ^{Sebastián-González et al. 2013}). This is especially relevant in Mexico, as most of its 14,000 wetlands are small in size (<10 ha) (^{Ramírez-Bastida et al. 2008}). Thus, the extensive system of small wetlands from the central highlands can represent a broad habitat for waterfowl as a whole (^{Ramírez-Bastida et al. 2008}) and the wetland of our study show just a part of it. The species richness and abundance that we recorded show that even small artificial wetlands could be relevant for the conservation of waterfowl, as this wetland is a refuge for migratory species, can host rare species, and is a suitable place for the reproduction of threatened species.

The study of biodiversity on a local scale is an essential requirement for conservation interests (^{Martínez-Morales et al. 2013}). It can influence management decisions, inform particular scientific inquiries, detect specific situations of interest, and together with other studies, can provide a valuable aid for the regional management of natural resources. Therefore, it is recommended that other studies be performed in the Lerma-Chapala-Santiago basin including several natural and artificial wetlands comparing their characteristics, particularly the vegetal resources for roosting, covering and nesting, the foraging features, water quality, the degree of perturbation, and the distribution, richness, and abundance of waterfowl.