INTRODUCTION

The main objective of studying parasites of wild mammals is to de termine their effect on the health and ecology of populations (^{Samuel et al., 2001}). In manatees (Sirenia: Trichaechidae), however, studies of parasites are relatively scarce (^{Marsh et al., 2012}). The Florida manatee (Trichechus manatus latirostris, Harlan 1824), in subtropical latitudes, has been the most studied (^{Bando et al., 2014}), because many dead manatees are recovered and available for examination. Some studies have been conducted on Antillean manatees (Trichechus m. manatus, Linnaeus 1758) in Cuba (^{Coy-Otero, 1989}), Puerto Rico (^{Mignucci-Giannoni et al., 1999a},^b), and the Dominican Republic (^{Colón-Llavina et al., 2009}). Only six species of helminth parasites have been reported in T. m. manatus: five digenetic and one Nematoda (^{Marsh et al., 2012}). Searching for eggs in feces is a viable and non-invasive technique to describe helminth diversity in manatees (^{Bando et al., 2014}).

The species Trichechus m.manatus is listed as endangered, both internationally (^{Self-Sullivan & Mignucci-Giannoni, 2008}) and in Mexico (^{SEMARNAT, 2010}).The largest manatee population in Mexico inhabits the southern region of the Gulf of Mexico, within the states of Tabasco, northern Chiapas, southern Campeche (^{Colmenero & Hoz, 1986}), and the Caribbean coast of Quintana Roo (^{Morales-Vela & Olivera-Gómez, 1997}; ^{Morales-Vela et al., 2000}). Recently, ^{Nourisson et al. (2011)} demonstrated genetic evidence of discreteness between these two populations in Mexico. In this study, we examined the presence and prevalence of helminth eggs in the feces of T. m. manatus in Mexico, from populations of the Gulf of Mexico (inland population, inhabiting freshwater systems) and the Caribbean (inhabiting mostly brackish and marine water).

MATERIALS AND METHODS

From 2005 to 2008, we collected fecal samples from 31 T. m. manatus in different localities in the Southern Gulf of Mexico (GM) and Caribbean coasts (CAR) in Mexico (Fig. 1). We analyzed 22 samples from 10 loca lities in the state of Tabasco in GM and 9 samples from two localities of the state of Quintana Roo in CAR (Table 1, Fig. 1). Within GM, samples came from two river basins, Grijalva (GV) and Usumacinta (USU) and within CAR, samples came from two coastal bays, Bahía de Chetumal (BCH) and Bahía de Ascensión (BA). All CAR samples and nine from GM were taken from manatees captured for health-assessment projects, immediately after the individual excreted during handling. For the other samples, one was taken from a rescued animal, four came from necrop sies, four were from animals in captivity, and four samples were from fresh feces found floating on waterways. Fecal samples collected in dead stranded animals were taken from the latest newly formed stools with a greenish consistency, directly from the large intestine. Collected feces were placed in sterile containers with a capacity of 100 ml, with a tight lid and fixed and stored in 70% ethylic alcohol until analyses.

Figure 1 Study area at Southeastern Mexico. Points are sampling localities within the Gulf of Mexico (GM) and Caribbean coasts (CAR). Subregions of GM are the Usumacinta River Basin (USU) and Grijalva River Basin (GV). Subregions of CAR are Bahía de Chetumal (BCH) and Bahía de Ascensión (BA). 

A qualitative flotation technique (saline saturated solution) was used to isolate eggs from stool samples. This technique involves a so lution with a higher density than the eggs in the sample. We used 2 to 5 g of homogenized stool sample and mixed it with 20 ml of saline saturated solution in a stirring device and filtered same through a mesh to a test tube until forming a convex meniscus. We then gently placed a coverslip on top of the test tube. After 20 min we carefully lifted off the coverslip from the tube and placed it on a microscope slide to observe it. We also used the sedimentation (formalin-ether) technique where the eggs are not distorted but rather deposited by gravity at the bottom of the container. We used 2 to 5 g of stool sample in 10 ml of 10% forma lin, filtered through a mesh. We added 6 ml of the suspension in 3 ml of ether in a centrifuge tube and mixed at 3000 rpm for 3 m. Then we de canted the final solution and an aliquot was taken with a Pasteur pipette and placed on a microscope slide to observe it (^{Lamothe-Argumedo, 1997}; ^{Cruz-Reyes & Camargo-Camargo, 2001}; ^{Foreyt, 2001}).

Eggs were identified under a microscope, using specialized litera ture and graphic catalogs (^{Travassos et al., 1969}; ^{Dailey et al., 1988}; ^{Coy-Otero, 1989}; ^{Foreyt, 2001}; ^{Bando et al., 2014}) and measured to µm with a ruled ocular. Microphotographs of representatives were taken with a digital camera. Empiric prevalence was calculated following the concepts of ^{Bush et al. (1997)}, i.e., the number of hosts infected with one or more species of parasites divided by the number of hosts exa mined. Qualitative similitude within the sampling region was estimated by cluster analysis, using the UPGMA algorithm and Jaccard´s index (^{Moreno, 2001}), with MVSP software 3.22 (^{Kovach Computing Services, Kovach, 2010}).

RESULTS

Five species of helminths were identified, three digenetic and two ne matode. Digenea: Chiorchis fabaceus (Diesing, 1938) Fischoider, 1901 (n=12, Fig. 2A), a large ovoid, almost elliptical, egg, 169.0 x 101.3 (148-180 x 90-120), a thin membrane covering the egg, opercula ted side, barrel shaped with symmetrical walls. Chiorchis groschafti Otero, 1989 (n=18, Fig. 2B) a smaller egg, 97.5 x 69.5 (80-108 x 60-71), same characteristics of C. fabaceus. Pulmonicola cochleotrema (Travassos&Vogelsang, 1931) Batron& Blair, 2005, (n = 23, Fig. 2C) very small eggs, ovoid, thin covered, 18.6 x 10.2 (17-22 x 10-12), with a filament extended on each pole, the operculum is located on the base of the filament of the wider pole.

Figures 2a-e Helminth eggs found in feces of Trichechus manatus manatus: a) Chiorchis fabaceus (169.0x101.3 μm). b) Chiorchis groschafti (97.5 x 69.5 μm). c) Pulmonicola cochleotrema (18.6x10.2 μm). d) Heterocheilus tunicatus (160.4x149.1 μm). e) Ascarididae gen. sp. (50x30 μm).  

Nematodes, Heterocheilus tunicatus Diesing, 1839 (n = 24, Fig. 2D) a reddish egg, slightly ovoid with a thick, rough wall, 160.4 x 149.1 (150-170 x 140-160), content does not fill the full eggshell. Ascarididae gen. sp. Baird, 1853 (n = 1, Fig. 2E) an ovoid egg, compressed on both poles, 50 x 30, thick and smooth reddish wall.

From the 31 individuals samples, 19 (61.2%) were infected by at least one helminth. Ten animals from GM (45.4%) and the whole CAR sample (100%) were infected. Four helminth species were found in GM samples and three from CAR. Among the species found, in the CAR sample, C. groschafti had the highest prevalence (66.6%), followed by H. tunicatus (44.4%) and P. cochleotrema (33.3%) (Fig. 3). In GM sam ple, P. cochleotrema had the highest prevalence (22.7%) followed by C. fabaceus (18.1%) and H. tunicatus (13.6%) (Fig. 3). Parasites in the individual samples varied from 0 to two species (mean = 0.83 species/sample). Seven samples had two species (22.5%), twelve samples had one species (38.7%), and twelve samples were free (38.7%) of para sites.

Figure 3 Prevalences of five species of helminth eggs found in feces of Trichechus manatus manatus in México on subregions of the Gulf of Mexico (Usumacinta River basin (USU), Grijalva River basin (GV), Caribbean Coasts (Bahia de Chetumal (BCH) and Bahia de Ascension (BA ). 

In terms of similarity, the generated cluster separated GM and CAR regions at an index ca. 0.30 (Fig. 4). Subregions in CAR (BA and BCH) were more similar than those in GM (GV and USU) (Fig. 4). Three species were shared by both populations, two were found only in GM samples, and one species only in CAR samples (Fig. 3). The species with higher prevalence differed among populations.

Figure 4 Similarity of subregions of the Gulf of Mexico (GM) and Caribean coasts (CAR) of Mexico, based on the egg prevalences of five species of parasite helminth in feces of Trichechus manatus manatus. The method used was UPGMA.  

DISCUSSION

Helminth parasite diversity in T. m. manatus is consistently low. There are reports of only six species in the literature (^{Marsh et al., 2012}). This could be a result of small samples, limited research, and a focus on feces or content of the digestive track, but also due to the habits of the species, whose distribution tends toward freshwater environments, while its herbivorous diet is mainly fibrous plants. As a comparison, 22 species of helminths are reported for the dugong Dugong dugon (Müller, 1776) (Sirenia: Dugongidae) (^{Beck & Forrester, 1988}), which is a marine species. Four of the species found in this study, C. fabaceus, C. groschafti, P. cochleotrema, and H. tunicatus were reported as adult parasites of dead manatees in Florida (^{Beck & Forrester, 1988}), Cuba (^{Coy-Otero, 1989}), Puerto Rico (^{Mignucci-Giannoni et al., 1999a}), the Dominican Republic (^{Mignucci-Giannoni et al., 1999b}), Brazil (^{Carvalho et al., 2009}), and Mexico (^{González-Solís & Vidal-Martínez, 2004}).

^{Bando et al. (2014)}described eggs from six helminth species in T. m. latirostris: H. tunicatus; C. fabaceus; C. groschafti; P. cochleotrema; Moniligerum blairi^{Dailey, Vogelbein & Forrester, 1988}, and Nudacotyle undicola ^{Dailey, Vogelbein & Forrester, 1988}. The eggs we analyzed coincided with descriptions and measurements taken by ^{Bando et al. (2014)} for C. fabaceus 169x110 (160.4-178.2 x 98-118.8), C. groschafti 98x70 (93.5-104.5 x 49.5-74.8), P. cochleotrema 22.4x10.5 (20.9-24.2 x 8.8-11), and H. tunicatus 154x155 (124.7-190.1 x 127.7-193.1).

Members of the Paramphistomidae family are often found in T. m. manatus, C.fabaceus, and C. groschafti in Florida, Cuba, Puerto Rico, the Dominican Republic, and the east coast of the Yucatan Peninsula. C. fabaceus had been reported in the three species of the Trichechi dae family: T. manatus, T. inunguis (Natterer, 1883), and T. senegalensis (Link, 1795) (^{Marsh et al., 2012}). Individuals of Chiorchis sp. inhabit the digestive tract, mainly in the colon and cecum. Their life cycle has not yet been studied, but it is thought that manatees ingest them by consuming small snails with cercarias (digenea larvae) that are on the aquatic plants they feed on (^{Raga et al., 2009}).

Digenea P. cochleotrema from the Opisthotrematidae family is des cribed from nostrils, trachea, and bronchi of manatees (^{Marsh et al., 2012}). This species is registered from T. m. latirostris in Florida (^{Beck & Forrester, 1988}) and from T. m. manatus in México, Puerto Rico, the Dominican Republic, and Brazil (^{Mignucci-Giannoni et al., 1990b}; ^{Gon zález-Solís & Vidal-Martínez, 2004}; ^{Colon-Llavina et al., 2009}; ^{Carvalho et al., 2009})

The ascarid nematode H. tunicatus belongs to the Heterocheilidae subfamily, which is common in aquatic reptiles (especially in crocodi les) and in sirenian mammals (^{Anderson, 2000}). These adult helminths have been recorded from Florida (^{Beck & Forrester 1988}), Puerto Rico (^{Mignucci-Gianonni et al., 1999a}), the Dominican Republic (^{Mignucci-Gianonni et al., 1999b}), and the eastern Yucatan Peninsula (^{González-Solís & Vidal-Martínez, 2004}), in great numbers throughout the digesti ve tract, mainly in the stomach. ^{Sanvicente (2005)} reported eggs of H. tunicatus in feces found along the Caribbean coasts of Mexico (Quin tana Roo) and from the Gulf of Mexico (Tabasco and Campeche). ^{Jueco (1977)} previously suggested that manatees feed on the intermediate hosts when consuming vegetation.

Eggs of the parasites that we attribute solely to the Ascarididae gen. sp. family have an unknown origin. In the Gulf of Mexico localities, domestic animals like cows, horses, dogs, and cats are often found along the waterways and lakes; many lakes and freshwater systems are lentic (low energy) and receive wastes from urban and rural com munities, which could favor some species of parasites.

Necropsies of manatees from the Caribbean coast of Mexico, whe re manatees inhabit coastal environments, often found greater num bers of nematodes and digenea in the digestive tract. However, in the Gulf of Mexico, where manatees are found almost exclusively within freshwater, the parasite load is low and just one adult species had been registered, the C. fabaceus (Hernández-Olascoaga & Olivera-Gómez, pers. com.) digenea.

Other reported species of digenea, Nudacotyleundicola and Mo niliger umblairi or coccidea, Eimeria manatus^{Upton, Odell, Bossart & Walsh, 1989} and E. nodulosa^{Upton, Odell, Bossart & Walsh, 1989} (^{Beck & Forrester, 1988}; ^{Upton et al., 1989}) were not observed as eggs or as other forms in our study. The sample size could influence the diversi ty found, since more samples were analyzed for the GM population. However, this study supports evidence that manatee helminth fauna is not diverse (^{Raga et al., 2009}) compared to other aquatic mammals like cetacean or pinnipeds (^{Foreyt, 2001}; ^{Raga et al., 2009}).

The differences of helminth fauna between the two regions are in keeping with the genetic information of the limited movements of indi viduals between the regions (^{Nourisson et al., 2011}) and indicate that different ecological pressures act upon the populations. These diffe rences also support the idea that both populations should be managed separately. Despite their low diversity, future studies on local helminth differences that used larger sample numbers could generate informa tion on their habitat.