Elsevier

Zoology

Volume 120, February 2017, Pages 92-100
Zoology

Review
Diet of Amazon river turtles (Podocnemididae): a review of the effects of body size, phylogeny, season and habitat

https://doi.org/10.1016/j.zool.2016.07.003Get rights and content

Highlights

  • Diet among the Amazon podocnemidids varies depending on species, sex, season and location.

  • There is no relationship between maximum carapace length and vegetable matter consumed.

  • Habitat (water type) indirectly influences the average volume of total vegetable matter consumed.

  • Species with no specialised stomach adaptations consume vegetable matter less hard to digest.

  • Plant secondary metabolites most effectively prevent consumption by turtles that lack adaptations to herbivory.

Abstract

Amazon rivers can be divided into three groups (black, white and clear waters) according to the origin of their sediment, dissolved nutrient content, and vegetation. White water rivers have high sediment loads and primary productivity, with abundant aquatic and terrestrial plant life. In contrast, black water rivers are acid and nutrient-poor, with infertile floodplains that support plant species exceptionally rich in secondary chemical defences against herbivory. In this study, we reviewed available information on the diet of Amazon sideneck river turtles (Family Podocnemididae). Our aim was to test the relationship between water type and diet of podocnemidids. We also took into account the effects of season, size, age, sex and phylogeny. Based on our review, turtles of this family are primarily herbivorous but opportunistic, consuming from 46 to 99% (percent volume) of vegetable matter depending on species, sex, season and location. There was no significant correlation between the maximum carapace size of a species and vegetable matter consumed. When the available information on diet, size and habitat was arranged on the podocnemidid phylogeny, no obvious evolutionary trend was evident. The physicochemical properties of the inhabited water type indirectly influence the average volume of total vegetable matter consumed. Species with no specialised stomach adaptations for herbivory consumed smaller amounts of hard to digest vegetable matter (i.e. leaves, shoots and stems). We propose that turtles with specialized digestive tracts may have an advantage in black water rivers where plant chemical defences are more common. Despite limitations of the published data our review highlights the overall pattern of diet in the Podocnemididae and flags areas where more studies are needed.

Introduction

The diet of freshwater turtles reflects factors such as sex, age, size, habitat selection, prey choice and interspecific interactions (Moll, 1976, Tucker et al., 1995, Armstrong and Booth, 2005). However, many turtles feed opportunistically, and the composition of the diet may reflect resource availability (Mahmoud and Klicka, 1979, Moll and Moll, 2004, Souza, 2004). Phylogeny is also important in shaping dietary preferences, which can be influenced by present day interactions and historical influences (Vitt and Carvalho, 1995, Vitt and Zani, 1996, Lindeman, 2000). Closely related species are more likely to be ecologically similar (Ricklefs and Miller, 1999).

Small body size has been viewed as a major constraint to herbivory in reptiles. According to many hypotheses, larger species have a greater absolute capacity for fat storage (enhanced buffer against seasonal variability in food suply), are less able to meet their metabolic requirements on a carnivorous diet, have a favourable ratio of mass-specific energy requirements to gut capacity, and can meet the high mass-specific energy and nutrient requirements on a nutrient-poor plant diet (Clark and Gibbons, 1969, Bell, 1971, Geist, 1974, Wilson and Lee, 1974, Pough, 1983, Penry and Jumars, 1987, Zimmerman and Tracy, 1989, Parmenter and Avery, 1990). These hypotheses have also been extended to explain ontogenetic dietary shifts in several species of freshwater turtles (Moll, 1976, Ernst and Barbour, 1989).

Ecological interactions in tropical rivers are strongly affected by changes in habitat and resource availability associated with seasonal variation in hydrological conditions, particularly annual cycles of flooding and drying (Lowe-McConnell, 1979, Jepsen et al., 1997). Likewise, floodplain forest productivity is influenced by nutrients originating from erosion, flooding, and sedimentation. The chemical composition of leaves, bark, and wood of floodplain trees is influenced by the availability of chemical elements in the water (Irion, 1982, Irion, 1984, Victoria et al., 1989, Junk, 1993). In the Amazon basin there is a synchrony between plant phenology and the inundation pulse. During the high water season there is intense production of seeds and fruits (Goulding et al., 1988, Kubitzki and Ziburski, 1994).

Amazon rivers can be divided into three groups according to the origin of their sediment, dissolved nutrient content and vegetation (Prance, 1979, Klinge et al., 1983, Junk, 1985). White water or ‘muddy’ rivers (i.e., Madeira, Purus, Juruá, Japurá, Magdalena and Amazon; Fig. 1) have a neutral pH and high turbidity, sediment loads and levels of primary production. White water floodplains (Várzea) are rich in dissolved nutrients (especially Ca and Mg) and fertile sediments originating from the Andes. Aquatic and terrestrial herbaceous plants are abundant in white water rivers, as are annual and perennial grasses and aquatic macrophytes which may form floating meadows (Gibbs, 1967, Sioli, 1967, Junk and Howard-Williams, 1984, Martinelli et al., 1989, Piedade et al., 1992).

Tannin-stained black water rivers (i.e., the Negro River and its tributaries) originate from the Guyana and Central Brazilian Shields. The waters are acidic and poor in electrolytes and nutrients, with extremely low sediment loads and low primary production. Their infertile floodplains (Igapó) support forests on extremely impoverished substrates. Herbaceous plants occur in low abundance and aquatic plants are often absent in black water areas (Junk and Howard-Williams, 1984, Junk, 1993).

The third group, clear water rivers, such as the Tocantins, Tapajós and Xingu, have an intermediate nutrient status and are less common than white and black water rivers (Junk, 1993, Goulding et al., 2003). They vary seasonally from crystalline to murky and generally have low sediment loads. The main focus of this review is on turtle diets in black and white water rivers, where the majority of work has been undertaken.

According to Janzen (1974), habitats such as the black water Igapó with a climate favourable to animals, but an environment with extremely low primary productivity, should favour plants rich in secondary chemical defences, or plants that are gregarious and have synchronised (mast) fruiting. Due to the constraint of low nutrient availability, these plants delay leaf replacement and produce a high concentration of toxic alkaloids, phenols and tannins to reduce the incidence of herbivory (Coley and Barone, 1996). This high degree of protection might affect the amount of vegetable matter available which aquatic vertebrates, such as podocnemidid turtles, are able to digest.

The pleurodiran (sideneck) family Podocnemididae is currently represented by three genera and seven species in South America and one in Madagascar (Williams, 1954, Neill, 1965, Rhodin et al., 1978). Six species of Podocnemis and one species of Peltocephalus are found in the Amazon region. Although the two larger species, Podocnemis expansa and Podocnemis unifilis, are sympatric throughout most of the Amazon and Orinoco drainages (Iverson, 1992, Bock et al., 2001), all species have distinct distributions and habitat preferences (Pritchard and Trebbau, 1984, Rueda-Almonacid et al., 2007, Vogt, 2008). Podocnemidids are typically opportunistic, generalistic and omnivorous, tending to herbivory. The proportion of plant and animal matter consumed varies considerably between species, populations and periods (Ojasti, 1971, Mittermeier and Wilson, 1974, Ramo, 1982, Pritchard and Trebbau, 1984, Almeida et al., 1986, Fachín-Terán et al., 1995, Balensiefer and Vogt, 2006, De La Ossa et al., 2011).

The feeding habits of these turtles have been studied for the last 40 years, yet to our knowledge a review of this information has not been attempted. As is the case with the Amazon fish fauna (Araujo-Lima et al., 1995), many of the studies are unpublised masters and Ph.D. theses or reports in Portuguese and Spanish, and the information has not been summarised for an English language audience. By reviewing the available information, we aim to test the relationship between Amazon environments and the diet of podocnemidids, particularly in relation to habitat (water type). Additionally, size, sex, season, location and phylogenetic relationships may play an important role and should be taken into account (Vargas-Ramírez et al., 2008). Our aim was to review the data available on diet of South American Podocnemididae species in relation to habitat, season, size, age, sex and phylogeny.

Section snippets

Overview

The literature was reviewed to gather information on the diet of Podocnemis erythrocephala, P. expansa, P. lewyana, P. sextuberculata, P. unifilis, P. vogli and Peltocephalus dumerilianus. Diet studies of captive animals or information based solely on interviews with local communities are cited in Section 3 but were not considered in the analyses. The information was categorised by species studied, location of the study and water type (black or white water; Fig. 1). Where available, vegetable

Species summaries

Peltocephalus dumerilianus is the most omnivorous species (Pérez-Emán and Paolillo, 1997, De La Ossa et al., 2011), although it consumes a high proportion of fruits and seeds, which can represent over 85% of its diet (Ojasti, 1971). According to Vogt (2001), P. dumerilianus has a preference for apple snails, when available. This species inhabits preferentially black water rivers, creeks and lakes (Pritchard and Trebbau, 1984, Rueda-Almonacid et al., 2007).

Podocnemis erythrocephala is mainly a

Influence of phylogeny

Information on diet (percentage of animal or vegetable matter), size (LCLmax; Ernst and Barbour, 1989) and habitat (mostly black, white or both) was combined with the family phylogeny assembled by Vargas-Ramirez et al. (2008) in Fig. 2. Neither diet, habitat nor size reflected taxonomic affinities and no clear patterns emerged from the inspection of Fig. 2. This variance in ecologically influenced characteristics is probably an indicator that the evolutionary history might not play an important

Amazon floodplains

Most freshwater turtles are omnivorous and opportunistic (Legler, 1993, Moll and Moll, 2004). The spatial and temporal oscillations of tropical riverine aquatic environments promote the predominance of opportunistic species (Abelha et al., 2008). Omnivory is also common in floodplains with seasonal floral and faunal variations (Cooper and Vitt, 2002, Metzger and Herrel, 2005). The capacity of opportunistic omnivorous species to exploit a variety of food items acts as a buffer against adverse

Seasonal differences

The most frequently studied season was low water (50% of studies). This was largely due to logistical constraints; during low water turtles are easier to find and capture in the reduced habitat available. However, low water studies probably do not account for most energy and nutrient intake. Most feeding takes place during high water, when the turtles move to the flooded forest. Many Amazon aquatic vertebrates depend on the flooded forests to obtain food and accumulate energy (Junk et al., 1997

Regional differences

Few studies (n = 5) compared the diet of a single species between different areas or habitats. As expected (Moll and Moll, 2004), most of these studies (80%) found differences between areas (Table 1). This included the proportion and diversity of fruits, seeds and aquatic plants (P. dumerilianus; Pérez-Emán and Paolillo, 1997), fish and crustaceans in artificial ponds versus streams (P. vogli; Ramo, 1982), and seeds and fruits in flooded forests versus shoots, stems and leaves in oxbow lakes and

Differences between sexes

Ten studies compared diet between sexes. Among those, 60% found dietary differences related to sex (Table 1). Females of P. vogli consumed more fish and mollusks than males (Ramo, 1982), and female P. dumerilianus consumed less Mauritia flexuosa seeds (Pérez-Emán and Paolillo, 1997). There were no differences between the diet of females and males of P. erythrocephala (Jaú River; Santos-Júnior, 2009). Males of P. dumerilianus in the Negro River during high water had lower trophic diversity than

Ontogenetic and size variation

Dietary ontogenetic changes are common in turtles. In many omnivorous species, juveniles are preferentially carnivorous and shift towards herbivory as they grow (Clark and Gibbons, 1969, Moll and Legler, 1971, Moll, 1976, Georges, 1982, Lima et al., 1997, Souza, 2004). These changes can be linked to physiological needs (Plummer and Farrar, 1981, Hart, 1983, Bury, 1986, Moll, 1990), such as a requirement for a diet rich in calcium, nitrogen and energy to stimulate growth (Moll, 1976, Shealy, 1976

Interspecific trends in diet

There was no significant correlation between LCLmax and volume of vegetable matter consumed (ρ(6) = 0.03, p = 0.96; Fig. 3). The two mainly black water species, P. dulmerilianus and P. erythrocephala are more omnivorous in comparison to the other species which inhabit white water or both white and black water habitats. The hydrochemical conditions and fertility of a river ultimately dictate the amount of plant matter available for consumption. White water rivers are more fertile than clear and

Limitations and recommendations for future studies

Of the 16 studies on the diet of South American Podocnemididae, four studies had small sample sizes, which are unlikely to adequately represent the species diet (n < 20; Table 1). Information on P. expansa is particularly problematic due to their small sample size. It is important to note that methods to assess diet differed substantially, which can potentially impact the results. The most common methods used were stomach flushing (e.g., Balensiefer and Vogt, 2006), stomach and digestive tract

Acknowledgements

This study was financed by the National Council for the Scientific and Technological Development (CNPq). C.C.E. was financed by the Science without Borders Program of CNPq (process number 233418/2014-8). We thank Fernando A. Perini, Virginia C. D. Bernardes and Marcia L. Queiroz for help with data collection and analysis.

References (136)

  • R. Franz et al.

    Herbivorous reptiles and body mass: effects on food intake, digesta retention, digestibility and gut capacity, and a comparison with mammals

    Comp. Biochem. Phys. A

    (2011)
  • M.C.F. Abelha et al.

    Plasticidade trófica em peixes de água doce

    Acta Sci. Biol. Sci.

    (2008)
  • Acosta, A., Fachín-Terán, A., Vilchez, I., Taleixo, G., 1995. Alimentacion de crias de taricaya Podocnemis unifilis...
  • H. Alarcón Prado

    Contribución al conocimiento de la morfología, ecología, comportamiento and distribución geográfica de Podocnemis vogli Testudinata (Pelomedusidae)

    Acad. Colomb. Cienc. Exact. Fís. Nat.

    (1969)
  • C. Alho et al.

    Early growth of pen reared Amazon turtles Podocnemis expansa (Testudinata: Pelomedusidae)

    Rev. Bras. Biol.

    (1982)
  • C.J.R. Alho et al.

    Ecologia da tartaruga da Amazônia e avaliação de seu manejo na Reserva Biológica do Rio Trombetas

    Bras. Florestal

    (1979)
  • S.S. Almeida et al.

    Vegetais utilizados como alimento por Podocnemis (Chelonia) na região do baixo rio Xingu (Brasil–Pará)

    Bol. Mus. Para. Emílio Goeldi

    (1986)
  • Araujo-Lima, C.A.R.M., Agostinho, A.A., Fabré, N.N., 1995. Trophic aspects of fish communities in Brazilian rivers and...
  • G. Armstrong et al.

    Dietary ecology of the Australian freshwater turtle (Elseya sp.: Chelonia: Chelidae) in the Burnett River, Queensland

    Wildlife Res.

    (2005)
  • K.E. Arthur et al.

    Ontogenetic changes in diet and habitat use in green sea turtle (Chelonia mydas) life history

    Mar. Ecol. Prog. Ser.

    (2008)
  • J.M. Ayres

    As Matas de Várzea do Mamirauá

    (1993)
  • D.C. Balensiefer et al.

    Diet of Podocnemis unifilis (Testudines, Podocnemididae) during the dry season in the Mamirauá Sustainable Development Reserve, Amazonas, Brazil

    Chelonian Conserv. Biol.

    (2006)
  • D.A. Belkin et al.

    An unusual chelonian feeding niche

    Ecology

    (1968)
  • R.H. Bell

    A grazing ecosystem in the Serengeti

    Sci. Am.

    (1971)
  • J. Bertl et al.

    An osteological comparison of Graptemys caglei Haynes and McKown and Graptemys versa Stejneger (Testudines: Emydidae)

    Herpetologica

    (1983)
  • K.A. Bjorndal

    Diet mixing: nonadditive interactions of diet items in an omnivorous freshwater turtle

    Ecology

    (1991)
  • K.A. Bjorndal et al.

    Digestive efficiencies in herbivorous and omnivorous freshwater turtles on plant diets: do herbivores have a nutritional advantage?

    Physiol. Zool.

    (1993)
  • K.A. Bjorndal et al.

    Digestive fermentation in herbivores: effect of food particle size

    Physiol. Zool.

    (1990)
  • B.C. Bock et al.

    Genetic population structure of two threatened South American river turtle species, Podocnemis expansa and Podocnemis unifilis

    Chelonian Conserv. Biol.

    (2001)
  • S.S. Bouchard et al.

    Ontogenetic diet shifts and digestive constraints in the omnivorous freshwater turtle Trachemys scripta

    Physiol. Biochem. Zool.

    (2006)
  • R.B. Bury

    Feeding ecology of the turtle Clemmys marmorata

    J. Herpetol.

    (1986)
  • A.M. Cano

    Algunos de los Aspectos de la Ecología Reproductiva de la Tortuga en el Río Chicagua

    Depresión Momposina (Bolivar, Colombia). Honours Thesis

    (2007)
  • D.B. Clark et al.

    Dietary shift in the turtle Pseudemys scripta (Schoepff) from youth to maturity

    Copeia

    (1969)
  • P.D. Coley et al.

    Herbivory and plant defenses in tropical forests

    Annu. Rev. Ecol. Syst.

    (1996)
  • J. Congdon et al.

    Ontogenetic changes in habitat use by juvenile turtles (Chelydra serpentina and Chrysemys picta)

    Can. Field Nat.

    (1993)
  • W.E. Cooper et al.

    Distribution, extent, and evolution of plant consumption by lizards

    J. Zool.

    (2002)
  • G.M. Costa

    Padrões Alimentares Durante um Periodo de Seca e Investigação de Endozoocoria por Podocnemis expansa na Reserva Biológica do Rio Trombetas, PA, Brasil. Master Thesis

    (2012)
  • B. de-Mérona et al.

    Food resource partitioning in a fish community of the central Amazon floodplain

    Neotrop. Ichthyol.

    (2004)
  • J. De La Ossa et al.

    Feeding of Peltocephalus dulmerilianus (Testudines: Podocnemididae) in a natural environment

    Actual. Biol.

    (2011)
  • N.R. dos Santos et al.

    Effect of seasonality and trophic group on energy acquisition in Amazonian fish

    Ecol. Freshw. Fish

    (2008)
  • F.S. Duarte

    Diagnóstico da Criação de Quelônios e Incubação Artificial de Ovos de Tartaruga Podocnemis expansa no Amazonas. Honours Thesis

    (1998)
  • C.H. Ernst et al.

    Turtles of the World

    (1989)
  • A. Fachín-Terán

    Ecologia de Podocnemis sextuberculata (Testudines, Pelomedusidae), na Reserva de Desenvolvimento Sustentável Mamirauá, Amazonas, Brasil. Ph.D. Thesis

    (1999)
  • A. Fachín-Terán et al.

    Food habits of an assemblage of five species of turtles in the rio Guapore, Rondonia, Brazil

    J. Herpetol.

    (1995)
  • B.O. Ferronato et al.

    Feeding habits of Amazonian freshwater turtles (Podocnemididae and Chelidae) from Peru

    Chelonian Conserv. Biol.

    (2013)
  • I.C. Figueroa et al.

    Componentes alimenticios de Podocnemis unifilis y P. expansa (Testudines: Podocnemididae) en el resguardo Curare–Los Ingleses, Amazonas, Colombia

    Rev. Colomb. Cienc. Anim.

    (2012)
  • G. Garcia et al.

    Erymnochelys madagascariensis – diet

    Herpetol. Rev.

    (2007)
  • V. Geist

    On the relationship of social evolution and ecology in ungulates

    Am. Zool.

    (1974)
  • L.L. George et al.

    Histologia Comparada

    (1998)
  • A. Georges

    Diet of the Australian freshwater turtle Emydura krefftii (Chelonia: Chelidae), in an unproductive lentic environment

    Copeia

    (1982)
  • J. Gibbons et al.

    Unusual population size structure in freshwater turtles on barrier islands

    Geo. J. Sci.

    (1979)
  • R.J. Gibbs

    The geochemistry of the Amazon river system: I. The factors that control the salinity and the composition and concentration of the suspended solids

    Geol. Soc. Am. Bull.

    (1967)
  • A. González-Zárate

    Caracterización del Hábitat, Uso de Recursos y Estado de Conservación de la Tortuga de Río Podocnemis lewyana, Águas Abajo del Embalse de Hidroprado, Tolima, Colombia. Master Thesis

    (2010)
  • G. Gottsberger

    Seed dispersal by fish in the inundated regions of Humaita, Amazonia

    Biotropica

    (1978)
  • M. Goulding

    The Fishes and the Forest: Explorations in Amazonian Natural History

    (1980)
  • M. Goulding et al.

    Rio Negro, Rich Life in Poor Water

    (1988)
  • M. Goulding et al.

    The Smithsonian Atlas of the Amazon

    (2003)
  • A. Hailey

    Digestive efficiency and gut morphology of omnivorous and herbivorous African tortoises

    Can. J. Zool.

    (1997)
  • D.R. Hart

    Dietary and habitat shift with size of red-eared turtles (Pseudemys scripta) in a southern Louisiana population

    Herpetologica

    (1983)
  • F.A. Huntingford et al.

    Energy reserves and reproductive investment in male three-spined sticklebacks, Gasterosteus aculeatus

    Ecol. Freshw. Fish

    (2001)
  • Cited by (15)

    • Pelodiscus sinensis embryos infected with Aeromonas hydrophila show higher survival rates than infected Mauremys reevesiis embryos

      2020, Aquaculture
      Citation Excerpt :

      The evolution progress among Chelonia species is still a charming subject (Refsnider and Janzen, 2016). Great efforts have been made by exploring the difference between Emydidae and Trionychidae, focussing on the areas of ecology, metabolism and developmental biology, e.g. thermal stress reaction and sex determination (Eisemberg et al., 2017; Nagashima et al., 2012; Valenzuela et al., 2013). However, differences in immunity processes between the two groups were neglected, both for adults and embryos.

    • First report on the mutagenicity and cytotoxicity of Zno nanoparticles in reptiles

      2019, Chemosphere
      Citation Excerpt :

      Therefore, the simulated environmental exposure scenario was based on a fictional aquatic P. expansa habitat presenting sediments with ZnO NPs at concentration of 100 μg/kg (Gottschalk et al., 2013); assumingly, these sediments, and other food items, were ingested by animals living in this ecosystem. The amount of sediment ingested by individuals belonging to genus Podocnemis sp. can vary from 0.02% to 31.1% of the animals’ stomach content volume, as reported in the literature (Ojasti, 1971; Fachín-Terán et al., 1995; Figueroa et al., 2012; Ferronato et al., 2013; Eisemberg et al., 2017). Although there is no consensus about the amount of sediment found in the stomach of these animals, and by taking into consideration that many factors can influence such volume, assumingly, 10% of the stomach content of the evaluated animals could be composed of sediments.

    • EVALUATION OF COMMERCIAL TORTOISE AND TURTLE FEEDS

      2022, Slovenian Veterinary Research
    View all citing articles on Scopus
    View full text