Does nutrition influence sexual dimorphism in Triatoma infestans (Hemiptera: Reduviidae) of natural habitats?

: Triatoma infestans is the main vector of the parasite that causes the Chagas disease in South America. It is known that T. infestans has diﬀerent reproductive and development patterns depending on whether they feed on birds or mammals. Using the head of adult insects as an estimator of the specimen development, we attempt to determine if there are any diﬀerences in the sexual size dimorphism associated with the availability of the food resource in T. infestans of natural habitats in the Llanos Riojanos region (Argentina). e nutritional status resulted higher in chicken coops and, in both habitats, it was higher for females in relation to males. e centroid size was larger in females than in males from chicken coops, but not in the specimens from goat corrals. Centroid sizes revealed smaller medians in goat insects in comparison to those coming from chicken coops. Sexual size dimorphism occurs associated with diﬀerences in the nutritional status only for triatomines from chicken coops in natural habitats. e heads shape was not inﬂuenced by the nutritional status. e sexual morphophysiological diﬀerences found in T. infestans help us understand aspects of the behavior of the species in diverse environments and its implications in the vectorial transmission of Trypanosoma cruzi


Introduction
One of the most significant differences in the morphology of some animals is the body size of males and females.is difference is known as sexual size dimorphism and it is oen affected by ecological, physiological, developmental, and environmental factors, among others (Fairbairn et al., 2007).e large amount of scientific information and literature related to the body size of animals reflects the relevance of this feature in biology (LaBarbera,1986(LaBarbera, , 1989;;Calder, 1996;Smith & Lyons, 2013).It is known that body size is strongly influenced by both biotic and abiotic environmental factors (Gaston, 1991;Chown & Gaston, 2010, 2013;Price et al., 2011), although it is not always easy to understand the adaptive significance of dimorphism.Triatoma infestans is the main vector of the parasite that causes the Chagas disease in South America and its vectorial capacity is closely associated to the feeding behavior of this insect (Wisnivesky-Colli et al., 1982).Triatoma infestans is an almost exclusively domestic species, although it is also commonly found in peridomestic habitats such as chicken coops, goat or pig corrals, or storage rooms, which could act as a source of T. infestans that can disperse towards human dwellings aer control interventions such as residual spraying (Gürtler et al., 1993;Cecere et al., 1997).Laboratory experiments revealed that T. infestans have different reproductive and development patterns depending on whether they feed on birds or mammals (Guarneri et al., 2000;Nattero et al., 2011).In natural conditions, this species maintains significant populations in chicken coops and goat corrals, particularly in the Arid and Semi-arid Chaco region, as it is associated with mammals of the Andean region (Noireau et al., 2005).In the Llanos Riojanos region (Argentina), successive collections showed that chicken coop insects reached higher nutritional status than those collected in goat corrals (Abrahan et al., 2011;Hernández, 2012), and, in turn, females were the largest blood consumers, promoting egg laying (Catalá, 1989(Catalá, , 1994;;Catalá et al., 1992).e probability of achieving good nutritional status depends on the interaction with their host, since, being large insects, they require considerable blood intake.Nymphs gain between eight and nine times their own weight, while adults do so by two to four times (Schofield, 1994).e irritability of the host during feeding is a crucial factor, since it hampers the procurement of resources.Successive interruptions in feeding condition the nutritional status by a limitation in the accessibility to food (Rabinovich, 1985).e nutritional status is a factor that can affect reproduction, moulting, dispersion, vectorial capacity and may show differences between sexes.is sexual difference in nutritional status may occur due to the feeding priority of one sex, which guarantees population growth, thus making the most of the available resources.In some cases, the resource is shared equally among members of the population, and in other cases it is distributed unequally, giving rise to hierarchical or non-hierarchical competition mechanisms for the exploitation of a resource as in the case of food (Townsend et al., 2008).e head of the triatomines is an organ that does not undergo any modifications once the insect reaches its adult stage and its phenotype reflects the ontogenetic development in a particular environment.In addition, since adult body size reflects the growth of the insect in its nymphal stages, it is possible to use adult head sizes as estimators of the insect's development.e head of triatomines has proven to be a useful organ for solving identification problems of reinfestants (Dujardin et al., 1997;Borges et al., 2005;Hernández et al., 2013) for the differentiation of species (Gurgel-Gonçalves et al., 2011), and where variables associated with flight potential can be recognized (Hernández et al., 2015).e existence of a sexual dimorphism in Triatomines is a known fact, with females being, on average, larger than males (Lent & Wygodzinsky, 1979), although what happens in natural habitats and how it relates to their nutritional status has not yet been studied.We attempt to determine if there are any differences in the sexual size dimorphism associated to the availability of the food resource for T. infestans in natural habitats (chicken coops and goat corrals).e aim is to verify whether differences in the nutritional status correlate with a sexual head size dimorphism and analyze its implications regarding the utilization of food.

Study area
e work was conducted in the Independencia Department, located West of Los Llanos (La Rioja, Argentina).is department is located at the South end of the Gran Chaco region (Arid Chaco).In the Llanos Riojanos region, the rearing of goats and chickens is common for the subsistence of small producers.is activity is associated with the construction of precarious corrals and chicken coops which become the main refuge for peridomestic populations of T. infestans (Cecere et al., 1997;Gürtler et al., 2004;Porcasi et al., 2006).e sampling was conducted in three localities of the said department: La Torre, La Aguadita, and Patquía Viejo.In this study, we chose chicken coops and goat corrals, which are the main peridomestic structures in the study area.

Insects
For morphometric analysis, 68 females (34 from goat corrals and 34 from chicken coops) and 82 males (32 from goat corrals and 50 from chicken coops) from all three localities were used.ese habitats were actively searched by hand during 30 minutes, using tetramethrin 0.2% to dislodge the insects (Spacial 0.2, Ministerio de Salud de la Nación/ National Health Ministry).e samplings were conducted in January, April, and October 2007 and 2009, in each of the peridomestic structures.e weight and length of the insects were recorded in the laboratory immediately aer the collection.e insects' heads were detached and kept in 70% ethanol until the time of the morphometric analysis.
Geometric morphometrics of the heads e heads were separated at the collar and mounted on a pin attached to a metal support.Photographs were taken with a Kodak C613 (6.2 MP) digital camera, and using a stereo microscope (10X).Ten type II landmarks were selected on the ventral surface of the adult T. infestans heads (Fig. 1).e average of the landmarks on both sides of the head was used (five landmarks), which allowed to reduce intraindividual variation and minimize digitization errors.For head size comparisons between groups, the isometric size estimator or centroid size (CS), derived from the coordinates, was used.is is defined as the square root of the sum of the squared distances between the centre of the landmark configuration and each individual landmark (Bookstein, 1991).e statistical significance of the CS differences was assessed using a nonparametric test based on permutations (5000 runs).
e shape variables were obtained through a generalized Procrustes analysis and the subsequent projection of the residuals into the Euclidean space (Rohlf, 1999).
Principal and partial warps were used as conformation variables.Principal warps describe the global variation such as elongation and compression, while partial warps correspond to the local variation (Zelditch et al., 2004).ese two components describe the differences in the conformation, as well as the deviations from the average landmark configuration.For the landmark digitization and morphometric data analysis, the CLIC99 module developed by J.P. Dujardin (2013) was used.e statistical significance of the head size and conformation analysis was assessed using a non-parametric test based on permutations (5000 runs) included in the CLIC module.e equality of variances analysis was conducted using an F-Test with InfoStat soware (Di Rienzo et al., 2016).

Nutritional status
e nutritional status of the adult T. infestans was estimated through the proportion between body weight (P) and total length (L), (P/L index), as suggested by Schofield (1980).Each insect was weighed on an analytical balance accurate to 0.01mg, and measured from the clypeus to the end of the abdomen using a caliper.e nutritional status was considered low for those insects with values of P/L<8 mg/mm, as suggested by Lehane and Schofield (1982).e comparison of the P/L index in males and females from both ecotopes was accomplished using STATISTICA (StatSo Inc, 2005).

RESULTS
Nutritional status by sex and habitat e nutritional status proved to be significantly different between habitats, comparing each sex separately, with higher means in T. infestans from chicken coops than those from goat corrals (p<0.05).For each habitat, females showed higher nutritional status than males (p<0.05) (Fig. 2).

Comparison of head centroid size by sex and habitat.
e centroid size of the heads resulted significantly higher (p#0.05) in females than in males for chicken coop T. infestans, but not in insects from the goat corrals.
e centroid sizes of the heads displayed smaller medians in goat corral insects, compared to those collected in chicken coops.As for variances, chicken coop males and females show similar values (p =0.72; F= 1.11), while in goat corrals, variability in females resulted higher than that observed in males (p =0.03; F =1.85) (Table I, Fig. 3).

Comparison of head shape between sexes and habitats
e analyzed groups showed sexual differentiation in their head shape (males were significantly different from females in insects from both chicken coops and goat corrals, p#0.05, using the Bonferroni correction), but there were no differences between the two habitats for each sex.
e Mahalanobis distances were larger for habitat differences than for sexual differentiation (Table II).e allometric effect was 0% and 9% for the first two canonical factors, evidencing that the conformation of sexual dimorphism is not a consequence of sexual size dimorphism.e first factor explains 71% of the total variation, while the second factor explains the remaining 29%.
e metric disparity did not reveal, statistically, significant differences between males and females of both habitats (Table III).ese results show that both sexes show no significant variability in the head shape.Head shape is therefore not influenced by the nutritional status.

Table II
Mahalanobis distances between groups extracted from discriminant multivariate analysis.Mahalanobis.e Mahalanobis distances were larger for habitat differences than for sexual differentiation Fem: Females, Mal: Males, CC: Chicken coops, GC: Goat corrals.
Table III Metric disparity of male and female T. infestans collected in chicken coops and goat corrals.

DISCUSSION
Body size is probably the most important quantitative variable that can be recorded in animals since it has a strong influence on ecological and adaptive aspects (Peters, 1983;Schmidt-Nielsen, 1984).Sexual dimorphism constitutes a manifestation of sexual selection and is one of the most conspicuous differences in the morphology of males and females, as described by Darwin (1871).Phenotypic responses depend on both the environmental features, particularly those that act during the development of the individual, and on the genetic properties of species (Daly, 1985;Williams, 2001).Our results show that sexual size dimorphism occurs associated with differences in the nutritional status only for triatomines from chicken coops in natural habitats.Results also show that T. infestans do not behave equally in both habitats, since sexual size dimorphism could only be observed in chicken coop insects.It is possible that in chicken coops, given a lower irritability of the host, females get the maximum amount of blood their body size allows.Although no specific studies were conducted on the digestive contents to assure the unique source of blood of the individuals of each habitat, the number of hosts in each of the analyzed habitats and the distance between them allow to consider that the primary food source came from the animals present in the habitat where the insects were collected.e lack of differentiation in the head shape of the two habitats could be interpreted as an exchange of triatomines between both habitats, although four years of studies in the Department showed a low rate of inter-habitat exchange (Hernández, 2012).For this reason, the idea of the exchange of individuals between habitats cannot be ratified based on the existing data of populations in the same area.Chicken coop insects achieve higher nutritional status than those that feed on goat blood, and this fact influences their size, as reflected by the centroid sizes of their heads.Studies conducted in eight localities of the Llanos Riojanos region revealed the same results in terms of nutritional status (Hernández, 2012).ese results are similar to those found for other species under laboratory conditions (Crocco & Catalá, 1997;Guarneri et al., 2000;Sant'Anna et al., 2010;Nattero et al., 2011).
e observed intrapopulation morphometric variations can be due to phenotypic variations in response to environmental features, particularly those that act during the development of the individual, as well as the genetic properties of the species (Daly, 1985;Williams, 2001).ere are several hypotheses that could explain the morphometric differences between females and males in T. infestans.According to the fecundity hypothesis, the greater size of the female could be aimed at building up energy reserves prone to achieve high ovipositions (Higgins, 2000).Sexual selection factors could also be occurring since, usually, in insects, larger females present higher fecundity, as well as dominance in confrontations over resources, and male preference towards this type of females (Clutton-Brock, 2017).In laboratory-bred triatomines, a general downsizing of the insects has been observed over the generations (Szumlewicz, 1976;Zeledón, 1983).In T. infestans, the lack of sexual dimorphism is usually associated to a domiciliation process.It has been suggested that, as a result of incomplete feeding in domiciles, the periods between generations are longer, and this results in a smaller body size (Zeledón et al., 1970;Zeledón & Ravinovich, 1981).What could be happening in goat corrals is an incomplete feeding in both sexes, probably due to a lower availability of hosts.In our study area, the flock of goats is moved to grazing lands for several days and the moments with greatest host availability, and therefore available blood, occur during goatling season (twice a year).Under these conditions, neither sex competitively displaces the other.A non-hierarchical competition would take place, where each individual (regardless of its sex) perceives the limitation in the food supply in the same way.In addition, in the case of triatomines that feed on goats, the thickness of the goatskin may difficult the access to blood capillaries.Based on observations of chicken coop triatomines, there would be sexual differences in the assignment of the available blood.In chicken coops, where there seems to be no restriction for T. infestans to access blood, each sex obtains as much as it can ingest according to its body size.In goat corrals, where there would be a restriction to access the food resource, both sexes obtain a similar amount of blood, without evident competitive displacement between them.Conversely, it is possible that the difference is not given solely by the amount of blood available in both habitats, but it may also be influenced by the quality of the blood or the differential utilization of the blood of both hosts by T. infestans.In addition to the aforementioned, the possible irritability of the host cannot be ruled out, since it would make T. infestans achieve higher nutritional status when feeding on birds, which are less irritable than mammals.
e variability of the centroid size of the head in both habitats allows the discussion on intraspecific competition.It has been observed that the females found in goat corrals show more variation in the size of their heads in comparison to those of chicken coops.is could indicate competition within the female group when the resource is scarce.e size variability in females of goat corral would be explained by a differential access to food in less resourceful habitats.As a consequence of this competition among females, larger and smaller head centroid sizes are observed in this habitat.
It has been evidenced that in these two habitats there are different head conformations in both sexes, although this difference would not be a passive consequence of the difference in size.
Further studies might prove useful to understand what happens with these features in other geographic areas or even examine sexual size dimorphism of T. infestans considering other uncommon ecotopes for the vector.Morphophysiological sexual differences in T. infestans may help understand behavioral aspects of the species in different environments such as dispersive capacity, female fertility influenced by size, population density, blood intake volume and its implication in the vectorial transmission of Trypanosoma cruzi.Zeledón, R., Guardia, V.M., Zuñiga, A., & Swartzwelder, J.C. (1970) Biology and ethology of Triatoma Dimidiata (Latreille, 1811) II.Life span of adults and fecundity and fertility of females.Journal of Medical Entomology, 7, 462-469.

Fig. 1
Fig. 1 Landmarks measured as coordinates of heads of Triatoma infestans.e number indicates the order of landmark capture

Fig. 2
Fig. 2 Nutritional status by sex and habitat.e markers and the lines represent the mean and standard error.CC: Chicken Coops, GC: Goat Corral.W/L: Weight/Length in mg/mm.Different letters indicate, statistically, significant differences between groups.