Diet and food ontogeny of the lizard Tupinambis matipu

The tegus are generalists lizards that use large amounts of prey in its diet, providing environmental services as a biological controller and seed disperser, which reveals how important diet studies are to understand ecological relationships related to a particular species. So the objective of this study was to analyze diet and food ontogeny of T. matipu, investigating changes in the pattern and composition of food items in different age classes and Research, Society and Development, v. 9, n. 11, e52391110073, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i11.10073 3 how the species shares its intraspecific niche. The captured specimens had the contents of their digestive tracts were analysed qualitatively and quantitatively. Our results indicate that T. matipu is a generalist lizard, consuming many food items, which fruits are the most important item in its diet. However, the species uses food resources in different importance proportions, according to its age class. Fruit consumption tends to increase and arthropods consumption decline as the age class increase. Thus, T. matipu performs an intraspecific sharing of feeding niche between the age classes and constitutes potential seed dispersers in its populations distributed along the Upper Course of the Paraguai River.

wetlands in the Brazilian Pantanal (Silva et al., 2018). Ecological data on the species are rare, and their diet has not yet been investigated.
Food is a fundamental component in structuring the interaction dynamics of a given lizards population with its habitat (Duffield & Bull, 1998), making it a fundamental aspect of these organisms (Colli et al., 1992). Therefore, knowing the role of this species and how it uses the environment to obtain its food resources is essential to understand its relationship with the ecosystem. However, in the diet, the items used by a given species may have qualitative and quantitative variations (Van Sluys, 1993;Duffield & Bull, 1998), according to different influences. These influences can be caused by biotic factors such as physiological or resource availability (Vitt & Caldwell, 2009), and abiotic factors, such as temperature (Rocha et al., 2009). Furthermore, tegus present an ontogenetic change in their dentition (Dessem, 1985;Presch, 1974), this fact has been associated with differences in the composition of the diet, between different age classes (Presch, 1974).
In any case, the environment structure in which these animals live is fundamental to the nutritional life of the species, since the environment is the food source. Therefore, changes in the biotic structure of different environments can promote changes in food availability for different organisms in the community. Thus, the environmental changes caused by the implementation of agricultural activities, for instance, are described as the main threats to Brazilian reptiles (ICMbio, 2018), not only by the direct changes in its preferred habitats but indirectly by disrupting important food sources.
However, reptiles such as tegus can contribute to agricultural activities, since they

Methodology
The study area is located in the Upper Course of the Paraguay River, from the municipality of Barra do Bugres (15° 05' 41.66" S; 57° 14' 30.08" W) to Taiamã Ecological Station (16° 51' 54.20" S; 57° 33' 11" W) (Figure 1), in a gradient of approximately 200 km extension. The region's climate is Aw, with a rainy season between October and April and a dry season between May and September (Amaral & Fonzar, 1982). The vegetation present in this location differs between Cerrado, Amazon Forest, Pantanal and ecotonal zones (Mourão et al., 2002). Six modules were established along the sampling area, in which each module had four sampling points, two on the right bank and two on the left bank of the Paraguay River, with the exception of Taiamã, which for logistical reasons and the absence of riparian forest on the left bank of the river, the four areas were implemented on the right bank (Figure 1), with a minimum distance of 2 km from each other. The traps were arranged in a "Y" shape, using four 60 L buckets, 15 meters from the center and interconnected by a 70 cm height guide Development, v. 9, n. 11, e52391110073, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i11.10073 8 fence. The traps were open for ten days at each sampling point and were inspected daily in the morning. In the same areas, at points close to the pitfalls, there was also mammals sampling, using live traps (Sherman and Tomahawk), where the T. matipu captured in these traps were included as an occasional capture in our sample. In the laboratory, all collected individuals were measured using tape to obtain head length (HL), rostrum-cloacal length (RCL) and tail length (TL). To identify the age group of T. matipu individuals, we observed the reproductive development of the specimens, together with morphological characters reported by Silva et al. (2018). Where adult males have welldeveloped testicles, female follicles are well developed, in juveniles, these reproductive structures are in development. In Silva et al. (2018) the authors relate that the species has the dorsum predominantly dark brown, with small elongated spots with black edges, forming irregular paravertebral and dorsolateral stripes from the nape to the base tail. This pattern can be overlapped on cream and black transverse stripes, which is visible in juveniles and subadults and disappears or becomes less evident in adults. Considering these criteria, we observed that juvenile had a maximum SVL of 250 mm and weight up to 320 g, sub-adults with SVL between 251 and 290 mm and weight between 321 to 490 g and adults with SVL above 290 mm and weight above 490 g.
The animals were dissected through a ventral incision to remove the digestive tract.
Subsequently, the content analysis was made with the help of a stereoscopic microscope. The food items found were quantified and organized into categories, identified at the lowest taxonomic level possible and stored separately in ergometric flasks with a solution containing 70 % alcohol. The arthropods were identified by the morphological characteristics of the exoskeleton and specific literature. To identify the fruits consumed by the species, including seed fragments, we used literature and consulted the collection of fruits from the Herbarium of Pantanal "Vali Joana Pott" (HPAN) of the State University of Mato Grosso, in addition to Research, Society and Development, v. 9, n. 11, e52391110073, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i11.10073 9 specialists in the area (Maria Antonia Carniello). The volume of food items was obtained using the liquid displacement in a graduated cylinder, adapted from Magnusson et al. (2003).
To determine the importance of each prey category, three different quantities were used: (i) numerical percentage (N %); (ii) percentage of frequency (F %) and (iii) percentage of volume (V %). To show the most important items in the diet, the Importance Value Index (IVI) of the food item was calculated based on the model proposed by Meira et al. (2007), IVI = (N % + F % + V %)/3. The Niche Overlap Index (Pianka, 1973) was calculated using the Spaa package (Zhang & Zhang, 2016) in R (R Core Team, 2016). To demonstrate the species feeding strategy in its different age groups (juvenile, sub-adult and adult) we used Costello's (1990) graphic method.

Results
With a sampling effort of 2.880 buckets/nights, 46 specimens of Tupinambis matipu were collected, of which 26 were females and 20 males (Figure 2A and 2B). The specimens were identified according to age, being 17 juveniles, 10 sub-adults and 19 adults. The specimens were collected in five of the six modules sampled: Taiamã Ecological Station (n = 15), Morrinhos Farm (n = 15), Recanto do Dourado (n = 7), Sepotuba River mouth (n = 3) and Barra do Bugres (n = 6). The exception was the Porto Estrela module, however, the species was observed at this site. All analysed stomachs showed at least one item inside, totalling 435 items registered, distributed in 36 different food categories (Table 1). In the composition of the species diet, considering the 46 specimens, the arthropod category was the most frequent, occurring in Research, Society and Development, v. 9, n. 11, e52391110073, 2020 (CC BY 4 (Table 1).
In Table 1, we present the food categories consumed by the species T. matipu, as well as the total of items consumed by the different age groups throughout the study area. Research, Society and Development, v. 9, n. 11, e52391110073, 2020 (CC BY 4.  As for the feeding strategy, the species showed dominant fruit consumption, rare consumption of vertebrates and specialist consumption of arthropods, as showed by the analysis of the Costello diagram (Figure 4).
Juveniles showed specialist and dominant consumption of arthropods, while fruits and vertebrates were rare items. The sub-adults showed a diet specialized in arthropods, while fruits and vertebrates were rare, with a slight predominance of vertebrates. Among adult individuals, fruit consumption was dominant, while vertebrates were rare and the specialist consumption of arthropods (Figure 4). For this development stage, the vertebrate category constituted the second-largest volume (39.68 %), and the second-largest IVI was fruits, with (45.51). For adults (n = 19), the largest volume (80.51 %) and IVI (98.60) were fruits, followed by arthropods with a volume of 14.26 % and IVI = 47.98 (Table 1).
The Pianka Niche Overlap Index showed that juvenile and sub-adults had greater niche overlap (0.79), followed by the overlap between juvenile and adults (0.74). Sub-adults and adults showed the lowest index value (0.40).

Discussion
Our results show that in the riparian forest of upper Paraguay River region, T. matipu has a generalist feeding habit, with a diversified diet, including items such as vertebrates, invertebrates, and fruits (Table 1). This diversification may be related to the Teiidae's active foraging manner (Pianka, 1966;Schoener 1971;Huey & Pianka, 1981;Pough, 1999), which the predator consumes a large number of food items compensating the energy expenditure in the search of food (Anderson & Karasove, 1981;Nagy et al., 1984). In this way, the species reset the expenses used in the food search by diversifying its diet with the resources available in the environment.
Although amphibians are commonly reported in the tegus' diet (Mercolli & Yanosky, 1994;Kiefer & Sazima, 2002;Silva, 2013), in our sample we did not find it. This fact may be related to the unpalatability that some anurans have or the total digestion of bones of these prey (Yáñez et al., 1980). However, Huey & Pianka (1981), reported that there may be plasticity in the foraging mode of a given species of lizard, influenced by factors such as predation risk and/or resources availability. Greff & Whiting (2000) describe that, if food is abundant, lizards can be more selective and focus on more rewarding prey. Thus, T. matipu can have a certain selectivity, discarding items such as amphibians, consuming items of easier capture and greater compensation, such as fruits, once amphibians were abundant in the sampled areas (Silva-Alves, 2019; personal observation).
As regards fish consumption, Mercolli & Yanosky (1994) have already observed its matipu diet can offer advantage related to the different nutritional needs during the stages of its development, as some lizards such as Tropidurus, in the early stages of their development, invest more during the growth (Meira et al., 2007), already after the sexual maturity a change occurs, and its energy is turned to reproduction activities (Fitch, 1981).
Thus, for T. matipu, in the early stages (juvenile and sub-adult) arthropods provide a greater amount of protein, while fruits provide more energy for reproductive activities in adulthood. The sub-adults presented fruits and arthropods consumption in similar proportion.
However, this age class presented the highest index of importance for vertebrate consumption.
It may be related to the fact that this age group is an intermediate phase between juvenile and adult, so their protein needs may increase, focusing on the increase in vertebrate intake, as an important extra source of proteins.
Although there are changes in the proportion of food intake in different age groups, our results showed a marked sharing of food resources. Considering that at the species level the resources sharing occurs according to the type of food, habitat and time (Pianka, 1969), differentiating their needs in minimal quantities (Pianka, 1974), our intraspecific approach shows that the segregation of trophic niche also occurs on this scale.

Conclusion
We observed that T. matipu shares food resources between the age classes adopting physiological/temporal criteria, with strategies where the different age classes use the resources in different proportions throughout their development. This mechanism can be an important strategy for the recruitment of new individuals in these populations, since it relieves competition for resources, and can even avoid cannibalism, considering that reptiles were the most important item within vertebrates for the species.
The T. matipu diet reveals that the species uses a wide variety of items, taking advantage of available resources in the environment, and may have selectivity according to the development phase, opting for items that have a higher cost-benefit. Fruits were the most important items, showing that the species can be an important seed disperser. T. matipu also consumes many arthropods that can be considered agricultural pests, reinforcing its role as a natural controller of these organisms in the environment. In short, this species plays an important role in the transport of native seeds, contributing to the regeneration of the disturbed area and can assist production systems by consuming potentially harmful arthropods. Despite sharing food resources between the different stages of development, T.