Introduction

Sigmodontine rodents (Cricetidae) are a fundamental component of the fauna of South America, representing about 20 % of the species of mammals inhabiting the subcontinent. Seven families of rodents are represented within the Misiones Province, situated in the far northeast of Argentina, limiting with Paraguay and Brazil (^{Patton et al. 2015}; ^{Lanzone et al. 2018}). An analysis of the species richness considering only the rodents of the family Cricetidae (Sigmodontinae), indicates that spatially the greatest values are concentrated towards the central-southern part of the province, in a region where the Bosque Atlántico Interior overlap with the ecoregion of the Fields and Malezales. The high richness in the area may be related to the occurrence of a greater environmental heterogeneity with the presence of forest-dwelling rodents and others adapted to pasture environments. In addition, out of the four endemic species of rodents reported for the province of Misiones, one (Akodon philipmyersi Pardiñas, D’ Elía, Cirignoli and Suárez) is reported for the ecoregion of the Fields and Malezales (^{Galliari and Goin 1993}; ^{Pardiñas et al. 2005}; ^{Patton et al. 2015}; ^{Lanzone et al. 2018}).

Usually sigmodontine rodents are associated with a variety of ectoparasites, such as mites, ticks and fleas. A parasite component community represents all of the parasites associated with some subset of a host species, such as a population (^{Bush et al. 1997}). Host-parasite associations are the result of evolutionary and ecological processes, since individuals in a rodent community or population vary in ways that may affect their interactions with their parasites. For example, host specimens may vary in their sex, age, physiology, morphology, ethology, ecology, etc., and all these features may influence their ectoparasite populations (^{Marshall 1981}; ^{Kim 1985}; ^{Morand et al. 2006}; ^{Krasnov 2008}). In addition, some ectoparasites are epidemiologically important because they are involved in the transmission of pathogens that cause diseases in humans and domestic and wild animals, and whose reservoirs are rodent hosts.

Studies on component communities of ectoparasites of sigmodotines are not abundant worldwide, and Latin America is not an exception. Most of the studies were conducted in Brazil and Argentina, and in the last country, researches were concentrated in Buenos Aires Province (^{Linardi et al. 1991}; ^{Barros-Battesti et al. 1998}; ^{Lareschi et al. 2007}; ^{Lareschi and Krasnov 2010}; ^{Sponchiado et al. 2015}; ^{Liljesthröm and Lareschi 2018}). Despite the biological relevance of northeastern Argentina, only a few ectoparasite species have been reported associated with sigmodontines, and there are not ecological analyses characterizing ectoparasite component communities (^{Lareschi 2010}, ²⁰¹¹; ^{Lareschi et al. 2016}; ^{Pardiñas et al. 2016}). Thus, the aim of our study is focused on the ecological study of the communities of ectoparasites of rodents of the subfamily Sigmodontinae in southern Misiones Province in northeastern Argentina.

Materials and Methods

Study area. The study was conducted in the south of the Misiones Province. We selected two localities situated in the southern limit of the Fields and Malezales ecoregion (^{Burkart et al. 1999}), very close to the the Bosque Atlántico Interior, and about less than 20 km far from Posadas City (Figure 1). Samples were taken in Estancia Santa Inés (-27° 31´ S; -55° 52´ W) during April-2007 and May-2009 and 2018, and in the Estación Experimental del Instituto Nacional de Tecnología Agropecuaria (EEA INTA) Villa Miguel Lanús (-27° 25’ S; -55° 53’ W) in April, 2007. The traps were placed in dirty (not carpid) fields of yerba mate, in fields invaded by chilcas (Baccharis sp.), in Andropogon sp. grasslands, in grasslands of cuttings in low areas and in forests implanted with native species located in the surroundings of the hull of the stay.

Figure 1 Sites of collection of ectoparasites. Locality 1: Estación Experimental del Instituto Nacional de Tecnología Agropecuaria (EEA INTA) Villa Miguel Lanús (27°25’S; 55°53’W). Locality 2: Estancia Santa Inés (-27° 31´ S; -55° 52´ W). Grey indicates the ecoregion Fields y Malezales. 

Rodents. Ninety-seven rodents (Cricetidae: Sigmodontinae) captured alive by using Sherman traps baited with oats were examined for ectoparasites. Ulyses F. J. Pardiñas (IDEAus Centro Nacional Patagónico, Puerto Madryn, Chubut, Argentina) and Carlos Galliari (Centro de Estudios Parasitológicos y de Vectores, La Plata, Argentina) taxonomically identified the rodents as: Cricetidae, Sigmodontinae, Akodontini: Necromys lasiurus (Lund) (n = 61), A. philipmyersi (n = 13), and Oxymycterus rufus Fischer (n = 1); Phyllotini: Calomys sp. (n = 10); Oryzomyini: Oligoryzomys flavescens (Waterhouse) (n = 7) and O. nigripes Desmarest (n = 5). Representative individuals of each species of rodents were deposited at the Colección de Mamíferos del Centro Nacional Patagónico (CNP; Puerto Madryn, Chubut Province, Argentina), some of them still have the field number: CNP742, CNP3041, CNP6007, CNP6020, CNP5705, and CNP4950.

Ectoparasites. The fur of the hosts was searched for ectoparasites with the use of combs, tooth brushes and forceps. The ectoparasites were preserved in 96 % ethanol. For taxonomic identification, mites were cleared in lactophenol and individually mounted in Hoyer´s medium; fleas were cleared by using KOH and mounted in Canadian balsam; ticks were identified directly under stereoscopic microscope. Ectoparasites were identified in accordance with the keys, drawings and descriptions given by ^{Furman (1972)}, ^{Smit (1987)} and ^{Martins et al. (2014)}. Representative specimens of ectoparasites of each species were deposited at the Colección del Departamento de Entomología, Museo de La Plata (MLP; La Plata, Argentina). The catalogue number consists of the number of the host followed by a script and the number corresponding to the individual ectoparasite; some specimens still have field numbers: MLP-CNP742-3, MLP-CNP3041-1, MLP-CNP6007-1, MLP-CNP6020-1, MLP-CNP5705-1, MLP-LTU697-1/2, MLP-LTU709-1/2, and MLP-LTU717-1.

Data analyses. Since only 12 individuals of N. lasiurus were captured at EEA INTA, and both sites of collection are close and similar in their vegetation, data from both localities was analyzed together. Indices and parameters were calculated as follow: Ectoparasite specific richness (S = number of species), Shannon specific diversity index [H = -∑ (pi ln pi)], equitability index (J = H/ln S), mean abundance (MA = total number of individuals of a particular parasite species in a sample of a particular host species/total number of hosts of that species, including both infected and non-infected hosts) and prevalence [P = (number of hosts infected with one or more individuals of a particular parasite species/the number of hosts examined for that parasite species) x 100] (^{Begon et al. 1988}; ^{Bush et al. 1997}). We tested significance (P) of differences between mean abundances and prevalences using Fisher’s exact test and Student´s t-test, respectively. The variance to mean ratio (V/M) was also calculated to examine the distribution of every ectoparasite within each host species. Analyses were calculated using parasitology software Quantitative Parasitology 3.0 (^{Rózsa et al. 2000}). Species accumulation curves were made with PAST Program (^{Hammer and Harper 2001}), which implements the analytical solution known as “Mao’s tau”, with 95 percent confidence intervals. We use a matrix of presence-absence data. Species richness of ectoparasites is estimated as a function of number of samples or hosts (number of rodent individuals of the same species).

Results

A total of 834 ectoparasites were collected from the 97 rodents examined. Out of them, 782 specimens were mites (Acari, Mesostigmata), 50 were fleas (Hexapoda, Siphonaptera), and two were ticks (Acari, Ixodida). Total species richness was S = 10; and diversity was H = 0.68. In comparison, mites were highly significant more prevalent and abundant (S = 7; P = 78.35 %; MA = 8.06) than fleas (S = 2; P = 21.65 %; MA = 0.53) and ticks (S = 1; P = 2.06 %; MA = 0.02, P < 0.005).

The collected ectoparasites were identified as: Acari, Mesostigmata, Laelapidae, Laelapinae: Androlaelaps fahrenholzi (Berlese; n = 94), A. rotundus (Fonseca; n = 366), A. ulysespardinasi Lareschi (n = 204), Gigantolaelaps wolffsohni Oudemansi (n = 21), Laelaps manguinhosi Fonseca (n = 6), L. paulistanensis Fonseca (n = 56), Mysolaelaps parvispinosus Fonseca (n = 35); Acari, Ixodida, Ixodidae: Amblyomma ovale Koch (n = 2, nymphs); Hexapoda, Siphonaptera, Rhopalopsyllidae, Rohopalopsyllinae: Polygenis (Polygenis) tripus (Jordan; n = 48), Polygenis sp. (n = 2).

Ectoparasite component communities are compared in Tables 1 and 2. Oxymycterus rufus and Calomys sp. were not included, since O. rufus was not parasitized, and only one specimen of Calomys sp. was associated with a single flea (P. (P.) tripus). Akodon philipmyersi shows the lowest values of S, H and J (Table 1), with a remarkable dominance of A. ulysespardinasi (92 %; Table 2).

Species accumulation curves considering species richness of ectoparasites as a function of number of individual hosts of each species are shown in Figures 2 a-d. Although the number of host specimens of some species is low, the curve seems to stabilize for most of the species. The number of specimens of every ectoparasite species, as well as the values of mean abundance, prevalence and their distribution are shown in Table 2.

Figure 2 (a-d). Species accumulation curves considering species richness of ectoparasites as a function of number of individual hosts of each species. 

Ectoparasite component communities associated with A. philipmyersi showed the highest MA (17.08), which was significantly different from the remaining ones (P < 0.05). On the contrary, component communities of O. flavescens showed the lowest MA (4.86), which differed from those of A. philipmyersi and O. nigripes (P < 0.005), but not from N. lasiurus (P > 0.05). Differences among total prevalences did not show significance (P > 0.05). Akodon philipmyersi shows the higher value of dominance (92 %, A. ulysespardinasi), and O. nigripes the lowest one (35 %, M. parvispinosus). Considering those ectoparasite species associated with two or more host species (Table 2), only G. wolffsohni was significantly more abundant in O. nigripes (3.00) than in O. flavescens (0.86; P < 0.05).

When comparing ectoparasite species within every component community (Table 2), the differences observed between prevalences and mean abundances of A. fahrenholzi and A. rotundus associated with N. lasiurus; and A. fahrenholzi and A. ulysespardinasi parasitizing A. philipmyersi were highly significant (P < 0.005). Considering high taxa of ectoparasites, differences in prevalences and mean abundances between mites (P = 85.2; MA = 7.3) and fleas associated with N. lasiurus, were highly significant (P < 0.005).

When comparing ectoparasite component communities between host tribes, Akodontini vs Oryzomini, differences between prevalences (P = 90.50 vs P = 91.70; P > 1) and mean abundances (MA = 10.66 vs MA = 10.82; P = 0.92), were not significant. Comparing host species included in every tribe, O. nigripes MA (13.4) show differences highly significant from that of O. flavescens (MA = 4.86; P < 0.05), while differences in their prevalences were not significant (P = 100 % vs P = 85 %; P = 1). Results are similar when comparing N. lasiurus vs A. philipmyersi: differences in their mean abundance (MA = 8.11 vs MA = 17.11, P < 0.005) were highly significant, while those in prevalences were not (P = 88.5 % vs P = 100; P > 1). Considering all ectoparasites in every component community, the distribution was aggregate (V/M > 1). The same results were observed when considering every ectoparasite species, with some exceptions in ticks and P. tripus collected from O. nigripes. In these cases randomless distribution (V/M = 1) were shown.

Discussion

In the present study, 10 ectoparasite species were identified parasitizing four sigmondontine species in southern Misiones Province in Argentina. With the exception of Polygenis sp., probably a new species deserving further study, the remaining species were previously reported for Argentina. Out of them, the following five are mentioned for the first time for the northeastern: A. rotundus, G. wolffsohni, L. manguinhosi, L. paulistanensis, and P. (P.) tripus (^{Lareschi and Mauri 1998}; ^{Lareschi et al. 2016}). Previously, the following mites and fleas have been mentioned parasitizing sigmodontines in Misiones Province: Adoratopsylla (Adoratopsylla) antiquorum antiquorum (Rothschild) (Ctenophthalmidae) and Polygenis (Polygenis) rimatus (Jordan; Rhopalopsyllidae, Rhopalopsyllinae), both from Abrawayaomys chebezi Pardiñas, Teta and D’Elía, A. fahrenholzi from A. philipmyersi and A. chebezi; Androlaelaps misionalis Lareschi from Akodon montensis Thomas, and M. parvispinosus from Oligoryzomys sp. (^{Lareschi 2010}, ²⁰¹¹; ^{Lareschi and Mauri 1998}; ^{Pardiñas et al. 2016}). Thus, the results obtained increase to ten the biodiversity of ectoparasites associated with sigmodontines in Misiones Province.

Although the number of captured specimens of some species is low, the accumulated curves suggest that the specific ectoparasite richness reported is representative of most of every component community. The purpose of the study is more descriptive than analytical and is the first study on ectoparasites of rodents from northeastern Argentina that considers usual ecological parameters in parasitology. We consider that in this sense, our results are novel.

Considering every ectoparasite high taxa, only mites were associated with the four host species. In addition, mites were dominant, and the most prevalent and specious taxa. These results are in agreement with studies from central Argentina and Brazil (e. g.,^{Linardi et al. 1991}; ^{Barros-Battesti et al. 1998}; ^{Lareschi et al. 2007}; ^{Lareschi and Krasnov 2010}; ^{Sponchiado et al. 2015}; ^{Liljesthröm and Lareschi 2018}). On the contrary, dominance of mites was not so remarkable observed in Argentinean Chaco (^{Nava and Lareschi 2012}), and fleas are dominant in Patagonia (^{Sanchez and Lareschi 2018}).

A tendency toward host aggregation was observed for most of the ectoparasites. Out of 10 ectoparasite species identified in the present study, five were collected from a unique host species, and so, species richness varied between three and four in every component community. Core species, characterized by high prevalence and abundance (^{Bush et al. 1997}), was detected in every community.

Every mite-host associated reported herein was previously reported for other areas. Androlaelaps ulysespardinasi and A. rotundus, belong to the Androlaelaps rotundus species group, which includes host specific species associated with different akodontine rodents (^{Lareschi and Galliari 2014}; ^{Lareschi 2018}). Laelaps manguinhosi, L. paulistanensis, Mysolaelaps spp., and Gigantolaelaps spp. were reported mainly parasitizing oryzomines (^{Furman 1972}; ^{Lareschi and Mauri 1998}). On the contrary, usually fleas and nymphs of ticks are not host specific (^{Linardi and Guimarães 2000}; ^{Lareschi et al. 2016}; ^{Nava et al. 2017}). Among the genera of fleas included into the family Rhopalopsyllidae, Polygenis Jordan is the largest and most widely distributed, parasitizing mainly a broad range of sigmodontine rodents (^{Linardi and Guimarães 2000}; ^{Lareschi et al. 2016}). Herein, P. (P.) tripus and Polygenis sp. were identified exclusively associated with N. lasiurus and A. philipmyersi, respectively. Further studies are necessary to understand these host-parasite associations.

In the same way, nymphs of ticks were only associated to N. lasiurus, although this relationship seems to be hazardous (V/M = 0.98). Amblyomma ovale has a wide distribution from southern United States to northern Argentina. Our findings are consistent with literature, since larvae and nymphs of A. ovale have the ability to parasitize a wide variety of hosts, including small rodents and birds, in order to increase the probability to reach the adult stage, which parasitize Carnivora (^{Nava et al. 2017}).

In the last years for the northeast of Argentina, new species of mites and the unknown males of these species were described (^{Lareschi 2010}, ²⁰¹¹, ²⁰¹⁸). These studies suggest that the diversity of the area is underestimated and that more studies are needed to know its biodiversity.

In addition, A. ovale is a common parasite of dogs in rural and forested areas, and the records of A. ovale adults biting humans in South America are numerous. Besides, this tick species has capacity to transmit Hepatozoon canis, the causative agent of a serious dog disease, and it is also a potential vector of the human pathogen Rickettsia sp. strain Atlantic rainforest (^{Nava et al. 2017}). Other ectoparasites might play an important role in epizootic diseases and their perpetuation among those rodents (^{Morand et al. 2006}). Such is the case of Polygenis spp., (^{Linardi and Guimarães 2000}), and Androlaelaps spp. (^{Lareschi and Mauri 1998}; ^{González et al. 2005}; ^{Chaisiri et al. 2015}).

Since some of the ectoparasites identified may play a role in the transmition of pathogens, the results obtained contribute to a better understanding of the ectoparasite-host relationship, which may have epidemiological implications.