The trophic ecology of fishes in the genus Rhamdia remains poorly reported. Here, we aim to describe the diet composition and infer the trophic levels, niche breadth, and niche overlap between Rhamdia guatemalensis and Rhamdia laticauda in northeast Oaxaca, Mexico. Fishes were sampled using an electrofishing device between 2016 and 2017. We calculated the Quotient index and the diet composition was analyzed using the percentage of the Index of Relative Importance (%IRI) to analyze possible ontogenetic and sexual differences on feeding ecology. We then calculated the trophic niche breadth using the Levins´ standardized index (B_A). Trophic levels (TL_k) of each species were calculated using the trophic index and trophic niche overlap using Pianka´s index (O_jk). The results indicate that both species are carnivorous and tend to prey on aquatic insects, with R. guatemalensis feeding on nine different orders of animals and R. laticauda displaying a broader diet spectrum, encompassing 19 orders. Our data showed ontogeny and sex differences in the trophic ecology of R. guatemalensis. They were secondary consumers (TL_k > 3) and showed a specialized diet (B_A < 0.4). We found evidence of trophic niche partitioning between adults (O_jk = 0.40) and females (O_jk = 0.42). Our results highlight the fundamental role of this species in the area as a carnivorous/insectivorous, predatory fishes, secondary consumers, and biological control for Diptera, Ephemeroptera, and Megaloptera. Both species minimize the trophic overlap through food resource partitioning mediated by fish body size. Additionally, adults and females consume different prey, which allows the coexistence of closely related species. Our findings regarding trophic ecology could be crucial for devising conservation and management plans for these particular Rhamdia species.

Coexistence, diet, feeding ecology, freshwater fishes, natural history, niche breadth

Ecological differences are essential to coexistence, and closely related, sympatric (co-occurring) species need to differ in at least one of the three dimensions of the ecological niche (i.e., temporal, spatial, and trophic) to avoid competition (Pianka 1973; Schoener 1974). The trophic dimension is the most important factor in the confirmation of the ecological niche (Krebs 1999; Bolnick et al. 2003), and in fishes, it can be analyzed through the analysis of stomach contents data (Hyslop 1980; Amundsen et al. 1996; Cortés 1997). Data on the trophic ecology of fishes is essential to understanding their role in the habitat, evolutionary biology, natural history, as well as for developing management and conservation strategies (Gerking 1994; Amundsen et al. 1996; Cortés 1997). Additionally, studies on trophic niche differentiation are crucial for exploring interspecific competition and stable coexistence among morphologically similar sympatric species (Brazil-Sousa et al. 2009; Silva et al. 2012; Bonato and Fialho 2014). Sympatric species can adjust their resource use to minimize niche overlap (Pianka 1974), a phenomenon known as resource partitioning (Schoener 1974). The partitioning of food resources can be influenced by various factors such as habitat, sex, and ontogeny (Ross 1986). It is imperative to consider one of these factors when studying the mechanism of coexistence among closely related species (Bonato and Fialho 2014).

The genus Rhamdia Bleeker, 1858 (Siluriformes, Heptapteridae), is found in almost every freshwater environment in the Neotropical region from southern Mexico to central Argentina (Hernández et al. 2015; Bockmann and Slobodian 2017; Thomas and Sabaj 2020). However, aspects of its natural history are poorly known (Bockmann and Slobodian 2017; Thomas and Sabaj 2020), and few quantitative data have been reported on the trophic ecology of the genus Rhamdia, with most studies focusing on Rhamdia quelen (Quoy and Gaimard 1824) (Brazil-Sousa et al. 2009; Kütter et al. 2009; Pagotto et al. 2011; Olaya-Nieto et al. 2012; Silva et al. 2012; Bonato and Fialho 2014; Villares-Junior and Goitein 2015; López-Rodríguez et al. 2019). Rhamdia catfishes are nocturnal and show crevice-seeking habits, with the vast majority of species sharing similarities in body shape, color pattern, and habitat use (Miller 1984; Miller et al. 2005; Ribolli et al. 2017).

So far, seven species of Rhamdia are known from Mexico (Perdices et al. 2002; Miller et al. 2005; Hernández et al. 2015). The Pale catfish Rhamdia guatemalensis (Günther 1864) and the Filespin chulín Rhamdia laticauda (Kner 1857), which occur together in the southern Tehuacán-Cuicatlán Biosphere Reserve (TCBR), Mexico (Martínez-Ramírez et al. 2013; Julián-Caballero et al. 2017), were selected for a study of trophic ecology. Rhamdia guatemalensis is distributed on both the Pacific and Atlantic slopes, extending from southern Mexico to the trans-Andean region of Colombia (Hernández et al. 2015), while R. laticauda inhabits the waters of the Atlantic slope from Río Jamapa, Veracruz, Mexico, to central Panama (Perdices et al. 2002). Due to this distribution, sympatry exists between these two catfishes at many Central American sites (Miller 1984; Perdices et al. 2002; Hernández et al. 2015), offering an excellent opportunity to study trophic resource partitioning under such ecological conditions. Both species are classified as Least Concern by the IUCN (Arroyave 2019; Schmitter-Soto 2019). However, some populations may be threatened by water pollution, water extraction for human activities, habitat degradation, and invasive species (Martínez-Ramírez et al. 2013; Julián-Caballero et al. 2017). In Mexico, R. guatemalensis is listed as “subject to special protection” (Pr) under the Official Mexican Legislation NOM-ECOL-059-SEMARNAT-2010 (SEMARNAT 2019), while R. laticauda is not included. Both species play a crucial role in the food chain, acting as both prey (Solís and Guerrero 2016; Lavariega et al. 2020; Carranza and Anderson 2023) and predators (Schmitter-Soto 1998; Pichler and Schiemer 2008), and are important food resources for local human communities (Martínez-Ramírez et al. 2013; Julián-Caballero et al. 2017).

Previous research in Southern Mexico has reported that R. guatemalensis exhibits a diverse and opportunistic diet that encompasses aquatic invertebrates, detritus, fishes, terrestrial arthropods, and terrestrial plant remains (Chávez-Lomelí et al. 1989; Schmitter-Soto 1998; Soto-Galera 2006; Pichler and Schiemer 2008; Anzueto-Calvo et al. 2013), whereas R. laticauda primarily subsists on aquatic insects and decapods in the Mexican state of Chiapas (Soto-Galera 2006; Anzueto-Calvo et al. 2013). These studies have shown that the diet of these species varies based on their size classes (Soto-Galera 2006; Anzueto-Calvo et al. 2013). However, there are no available studies describing sexual differences in diet for either of these species. Furthermore, details regarding niche breadth, trophic levels, and resource partitioning between these species remain unexplored.

To fill these knowledge gaps, we analyzed quantitatively the overall diet composition of R. guatemalensis and R. laticauda, and compared the dietary composition between size classes and sexes. We determine the niche breadth and trophic levels to quantify the degree of dietary specialization and to determine their positions within the food web, respectively. Finally, we assess the overlap in food items between the species to examine possible feeding mechanisms that allow them to coexist. We propose that differences in the consumption of prey items could be attributed to food resource partitioning between these coexisting species.

We found significant difference in SL (U = 15748; P < 0.05), W (U = 15614; P < 0.05), and IL (U = 5047; P < 0.05) between the species. The mean IQ obtained (using all specimens) for R. guatemalensis was 0.69 and 0.59 for R. laticauda, with IQ < 1 for both sexual and ontogeny groups (Table 2), indicating that both species are classified as carnivores. We analyzed the stomach contents of 72 specimens of R. guatemalensis (27 females, 32 males, and 13 undifferentiated) and 112 of R. laticauda (52 females, 56 males, and 4 undifferentiated). We identified nine prey orders for R. guatemalensis and 19 for R. laticauda (Table 2). The most important orders in the diet of the overall composition of both species were Diptera, Ephemeroptera, Megaloptera, and Trichoptera (Fig. 2). Diptera was the most important prey order for R. guatemalensis (42.2%IRI), followed by Megaloptera (29.80%IRI), and Ephemeroptera (22.3%IRI). For R. laticauda, the most important category in the diet was Ephemeroptera (50.1%IRI), followed by Diptera (37.7%IRI). The TL_k obtained (using all specimens) for R. guatemalensis was TL_k = 3.25 and TL_k = 3.19 for R. laticauda, indicating their positions as secondary consumers within the food web.

Figure 2. 

Species-prey network consumed by two Rhamdia species. The vertices sizes are proportional to the number of interactions according to %IRI data. P1 = Diptera, P2 = Coleoptera, P3 = Ephemeroptera, P4 = Trichoptera, P5= Hymenoptera, P6 = Megaloptera, P10 = Lepidoptera, P12 = Bassommatophora, and P18 = Cyprinodontiformes. For the rest of the prey orders see Table 2.

The diet composition of both species varies as they grow (Table 2). Rhamdia guatemalensis showed an increased proportion of Cyprinodontiformes in adult specimens, while R. laticauda showed an increase in Decapoda, Characiformes, and Cyprinodontiformes. Significant differences in diet were found between size classes for R. guatemalensis (Chi = 88.7; df = 4; P < 0.05) and R. laticauda (Chi = 17; df = 5; P < 0.05).

Dietary difference in %IRI between sexes were marked in R. guatemalensis, but not in R. laticauda. Female specimens of R. guatemalensis consumed a greater proportion of Megaloptera and Cyprinodontiformes compared to males (Table 2). In contrast, R. laticauda females only showed a higher proportion of Diptera and Ephemeroptera compared to males (Table 2). Significant difference in the diet were found between sexes for R. guatemalensis (Chi = 61; df = 4; P < 0.05), but no significant differences were found for R. laticauda (Chi = 8.8; df = 5; P > 0.05). The overall trophic niche breath was B_A = 0.22 for R. guatemalensis and B_A = 0.12 for R. laticauda, with B_A < 0.4 for both sexual and ontogeny groups (Fig. 3). There was a significant trophic overlap between juveniles (P < 0.05), with evidence of food partitioning between adult and female specimens (P > 0.5) (Fig. 4).

Figure 3. 

Niche breadth of two Rhamdia species calculated by size classes and sexes.

Figure 4. 

Niche overlap based on Pianka index (O_jk) between size classes and sexes of Rhamdia guatemalensis and Rhamdia laticauda. Significant codes: *** indicates P < 0.001 and ** indicates P < 0.01.

We present data on morphometric and trophic niche separation between R. guatemalensis and R. laticauda, which co-occur in the TCBR, supported by evidence of body measurements (SL, W, and IL) and analysis of stomach contents data. Rhamdia guatemalensis and R. laticauda differ significantly in standard length and body mass (Table 1), which could avoid direct overlap in food and microhabitat use (Gerking 1994; Wootton 1999). A previous study indicated that these fishes show a negative allometric growth pattern and uniform body shape (Julián-Caballero et al. 2017); however, our data suggest that at the species level, R. guatemalensis is larger in terms of standard length and weight compared to R. laticauda. This information could be useful for accurate identification in the field, since it is often collected in the same sites in TCBR (Martínez-Ramírez et al. 2013; Julián-Caballero et al. 2017) and at many Central American sites (Hernández et al. 2015).

The mean IQ values (IQ < 1) recorded for both species in all samples (Table 2) suggest carnivorous habits. Fishes with carnivorous habits typically have shorter intestines compared to herbivorous or omnivorous fishes, presumably due to lower food ingestion amounts and higher quality, resulting in relatively rapid digestion (Pouilly et al. 2003; Becker et al. 2010; Olaya-Nieto et al. 2012). Our results confirm previous findings on the diet composition of these species in other Mexican states (Chávez-Lomelí et al. 1989; Schmitter-Soto 1998; Soto-Galera 2006; Anzueto-Calvo et al. 2013) and Costa Rica (Bussing 1998; Pichler and Schiemer 2008). Additionally, our results align with the carnivorous habit reported for R. quelen (IQ range 0.65–0.76) in Brazil (Barbieri et al. 1994; Becker et al. 2010) and Colombia (Olaya-Nieto et al. 2012). Most of Heptapteridae species showed a carnivorous feeding habit, with aquatic invertebrates and fishes being the fundamental dietary items of some species (Pichler and Schiemer 2008; Mazzoni et al. 2010; Pagotto et al. 2011; Bonato and Fialho 2014; López-Rodríguez et al. 2019). Heptapteridae species can vary greatly in their fish consumption; for instance, R. quelen, has been reported to have values greater than 81%IIR (Olaya-Nieto et al. 2012), while for Imparfinis usmai, the value is less than one (Villa-Navarro et al. 2014).

The diet composition of R. guatemalensis and R. laticauda in this area can be interpreted as aquatic invertivores, according to Goldstein and Simon (1998), and could behave as predatory fishes, as suggested by Fagade (1983). Our data showed that R. laticauda preys on a wide variety of fishes, anurans, ostracods, crabs, gastropods, arachnids, insects, and bivalves, while R. guatemalensis primarily preys on gastropods, insects, and fishes. We identified three insect orders crucial to the diets of both species based on %IRI from different samples (Table 2), Diptera, Ephemeroptera, and Megaloptera.

Larvae of lower Diptera vary in length from a few millimeters to many centimeters, depending on the species (Fusari et al. 2018), and are found in gravel, trunks, branches, and stems, living in colonies with slow movements, making them easy prey for predators (Merritt et al. 2008). Ephemeroptera are significant components during their nymph stages and are constantly present throughout the year (Barber-James et al. 2008; Merritt et al. 2008), serving as essential prey for Heptapteridae species (Rondineli et al. 2011; Villa-Navarro et al. 2014; Do Carmo and Souza 2019). Megaloptera larvae are known to be hunted by Heptapteridae species (Brazil-Sousa et al. 2009; Mazzoni et al. 2010; Bonato and Fialho 2014). Larval corydalids may be large (20–80 mm in body length) and are conspicuous benthic components in lotic waters, where they can achieve high abundances in specific ecosystems and regions, making them important prey for predatory fishes (Ardila-Camacho and Contreras-Ramos 2018; López-López et al. 2019; Rivera-Gasperín et al. 2019).

The obtained trophic level values (TL_k > 3) for R. guatemalensis and R. laticauda suggest that they are secondary consumers. These results are close to the values reported for the order Siluriformes (TL_k = 3.42) (Romanuk et al. 2011) and smaller than those reported for R. quelen (TL_k = 3.9) (Froese and Pauly 2023), the largest species of Rhamdia (> 300 mm SL) (Bockmann and Slobodian 2017). This could be explained by the positive relationship between trophic level and body size in carnivorous fishes (Romanuk et al. 2011; Keppeler et al. 2020). This data highlights the ecological role of R. guatemalensis and R. laticauda in the TCBR, recorded here for the first time using stomach content data, representing an advance in the knowledge of the natural history of genus Rhamdia.

Our analysis indicates that the diet composition of both species varies depending on size, with an increase in the proportions of fishes for R. guatemalensis and fishes and decapods for R. laticauda as they grow (Table 2). Our results align with those reported in other studies for R. guatemalensis and R. laticauda in different states of Mexico (Soto-Galera 2006; Anzueto-Calvo et al. 2013), as well as for other members of Heptapteridae (Moraes et al. 2013; Bonato and Fialho 2014). As fish grow, their diversity and trophic range expand due to morphological limitations in capturing certain foods. Juveniles are unable to consume fishes and decapods because of their limited capacity and size to capture and process these types of prey. The ability of juveniles to consume fishes and decapods may be restricted by the size of their mouth opening (Gerking 1994; Wootton 1999). This diet variation by size of the fish is also related to the growth and maturity stages of these species, because the energy and nutritional requirements may differ throughout their life history as predators (Gerking 1994; Wootton 1999).

We found a dietary difference between sexes in R. guatemalensis, but not in R. laticauda. In the case of R. guatemalensis, this difference could be related to a broader range of body measurements (Table 1) or different IQ values found in females, and niche breadth values between the sexes (Table 2). In contrast, data obtained for R. laticauda suggest that there is less variation in the range of body measurements or alimentary index (Tables 1, 2). Diet difference between sexes in genus Rhamdia has not been published. However, variation in diet between sexes in fishes may occur due to differences in behavior and energy allocation (Gerking 1994; Wootton 1999), or it might be influenced by size of individuals or sexual size dimorphism (Gerking 1994). However, this information requires further investigation because no sexual size dimorphism has been published for Heptapteridae species (Bockmann and Slobodian 2017).

Values of Levins´ standardized index (B_A < 0.4) suggest that both species showed a tendency towards dietary specialization (Grossman 1986; Bonato and Fialho 2014), probably associated with the morphological adaptations and a benthophagous feeding habit. Similar dietary specialization has been reported in R. quelen (Pouilly et al. 2006; Brazil-Sousa et al. 2009; Kütter et al. 2009; Olaya-Nieto et al. 2012; Silva et al. 2012; Bonato and Fialho 2014; López-Rodríguez et al. 2019). However, individuals of the genus Rhamdia could be adapted to exploit prey that becomes abundant seasonally, locally, or due to stochastically limitations to a narrow range of resources in their environment (Villares-Junior and Goitein 2015). To consume prey from the benthos, R. guatemalensis and R. laticauda possess three pairs of sensory and tactile barbels, which are used for detection (through taste buds) and selection of prey before ingestion (Miller 1984; Kobelkowsky and Castillo-Rivera 1996; Wilkens 2001; Hernández et al. 2009). Understanding the dietary specialization or habitat of organisms is crucial for comprehending population declines in specific species and implementing effective conservation measures (Costa et al. 2015). Specialist species often have very specific diets and habitat requirements, making them more vulnerable to environmental fragmentation and food depletion. Consequently, they are more prone to extinction compared to generalist species (Costa et al. 2015). Preserving the rivers and streams in this area is essential because certain orders of insects (Ephemeroptera, Plecoptera, and Trichoptera) that are part of the diet of Rhamdia species are sensitive to contamination (Barber-James et al. 2008; Merritt et al. 2008).

The Pianka index indicated a low to high dietary overlap, primarily driven by the consumption of Diptera and Ephemeroptera (Fig. 2), two abundant orders of aquatic macroinvertebrates in the study area (López-López et al. 2019), suggesting opportunistic exploitation. According to Taylor et al. (2011), if the composition and overlapping of the diets is based on very abundant trophic resources, then it is most likely that opportunistic exploitation of these resources is occurring. This is consistent with the previous knowledge of Rhamdia species, which exhibit an opportunistic diet pattern and high food plasticity that seems to be the strategy employed (Miller 1984; Wilkens 2001; Wilkens and Strecker 2017). Thus, their diet varies depending on food availability and geographical location (Miller 1984). Furthermore, our results do not necessarily indicate competition for food between the catfish species, as we recorded evidence of trophic niche partitioning among adults and between females (Fig. 4). Previous studies have documented food resource partitioning among Heptapteridae species (Brazil-Sousa et al. 2009; Silva et al. 2012; Bonato and Fialho 2014). Microhabitat segregation was also observed in the area: R. guatemalensis occurs in pools or rivers with slow currents and plant material such as algae, trunks, leaves and branches, while R. laticauda is found in rocky riffles and stream currents (Julián-Caballero et al. 2023). Similar microhabitat segregation has been reported for these species in Colombia (Hernández et al. 2015). However, it is imperative to conduct studies on spatial niche dimension (see Ferreira et al. 2023) to verify if partitioning occurs in this dimension (Schoener 1974).

These two species show overlapping distributions in the Neotropical region (Miller 1984; Perdices et al. 2002; Hernández et al. 2015) and can probably coexist in rivers as long as there are optimal conditions for both. Future studies should include an analysis of prey availability in the habitat where Rhamdia species occur. This will help assess their electivity (Vanderploeg and Scavia 1979) and provide a greater understanding of their feeding strategies.

Rhamdia guatemalensis and R. laticauda differ in standard length and weight, which could avoid direct overlap in food usage. Our result highlights the fundamental role of these species in the area as carnivores and biological controls for Diptera, Ephemeroptera, and Megaloptera. Our quantitative analysis of stomach contents allows us to classify the two species as aquatic invertivores and potential predators. The diet of both species varies in composition according to size and shows sexual differentiation for R. guatemalensis. Both species are secondary consumers and tend to specialize in consuming insects. We also provide evidence that adults and females consume different prey orders, enabling closely related species to coexist through trophic partitioning. This information is essential for understanding the tropic ecology of two Rhamdia species that coexist in the TCBR and could be crucial for developing conservation and management plans for these nocturnal predator fishes in a protected natural area of Mexico.

The authors are grateful to the authorities of the Tehuacán-Cuicatlán Biosphere Reserve for providing a research permit to carry out this investigation. The Comisión Nacional para el Conocimiento y Uso de la Biodiversidad and the Instituto Politécnico Nacional (grant SIP-2015-RE/054) supported this study throughout the project LI007. In addition, the Instituto Politécnico Nacional (México) provided funding for carry out this work (grant SIP 20164780). The first author is very grateful to the Master Scholarship granted by CONAHCYT (CVU 717180). We thank Miguel Calixto-Rojas, Jose Barrios-Gutiérrez, Rubí Marcos-Gómez, Carlos Pinacho-Pinacho, Gabriel Cruz-Ruiz, David Cruz, Lourdes Hernández, and Angel García for helping in field work. We are grateful to the anonymous reviewers for their valuable comments, which have greatly improved this manuscript.