Naturwissenschaften (2013) 100:299–310 DOI 10.1007/s00114-013-1027-4
ORIGINAL PAPER
Trophic interactions of the endangered Southern river otter (Lontra provocax) in a Chilean Ramsar wetland inferred from prey sampling, fecal analysis, and stable isotopes Marcela Franco & Giovany Guevara & Loreto Correa & Mauricio Soto-Gamboa
Received: 31 May 2012 / Revised: 16 February 2013 / Accepted: 20 February 2013 / Published online: 7 March 2013 # Springer-Verlag Berlin Heidelberg 2013
Abstract Non-invasive methodological approaches are highly recommended and commonly used to study the feeding ecology of elusive and threatened mammals. In this study, we use multiple lines of evidence to assess the feeding strategies of the endangered Southern river otter, by determining seasonal prey availability (electrofishing), analysis of undigested prey remains (spraints), and the use of stable isotopes (δ15N and δ13C) in otter spraints (n=262) and prey in a wetland ecosystem of southern Chile (39°49′S, 73°15′W). Fecal and isotopic analyses suggest that the otter diet is restricted to a few prey items, particularly the lessmobile, bottom-living, and larger prey such as crayfish (Samastacus spinifrons, 86.11 %) and crabs (Aegla spp., 32.45 %), supplemented opportunistically by cyprinids (Cyprinus carpio, 9.55 %) and catfish (Diplomystes camposensis, 5.66 %). The results suggest that the river otter is highly specialized in bottom foraging. Isotopic signatures of food sources and feces revealed a mid-upper trophic position for the Southern river otter, with either higher or lower δ15N values than their potential prey items. δ13C values for river otters were less enriched than their potential food resources. We suggest that due to their narrow trophic niche and possible dependence on only a few food items, this species may be highly vulnerable to the reduction Communicated by: Sven Thatje M. Franco (*) : L. Correa : M. Soto-Gamboa Instituto de Ciencias Ambientales y Evolutivas, Facultad de Ciencias, Universidad Austral de Chile, Campus Isla Teja, Valdivia, Chile e-mail:
[email protected] G. Guevara Departamento de Desarrollo Rural y Recursos Naturales, Facultad de Ciencias Agropecuariass, Universidad de Caldas, Manizales, Colombia
in its prey populations. Finally, maintaining the ecological interactions between Southern river otters and their prey is considered a central priority for the survival of this endangered carnivore mammal. Keywords Southern river otter . Stable isotopes . Trophic ecology . Wetland . Chilean freshwater
Introduction The Southern river otter, or huillín (Lontra provocax Thomas, 1908), is one of four South American otter species and is endemic to the Southern cone (Sielfeld 1983). This species is associated with freshwater ecosystems in the northern portions of its range (39°S), but further south, it inhabits the rocky seashores of the Patagonian archipelagos (Gomez et al. 2010). Populations started to originally decline due to high hunting pressure and, despite recent hunting bans (Aued et al. 2003), this semi-aquatic species remains classified as “endangered” on the IUCN Red List of Threatened Species (Sepúlveda et al. 2008). This is because the riparian forest is seriously threatened by a number of factors, particularly by habitat destruction and increasing river channelization, which has prevented recovery and has caused further declines in populations of this species (Cassini et al. 2010; Medina-Vogel et al. 2003). As with other freshwater otters, this species is difficult to observe in nature and is the reason why previous studies have relied upon direct as well as indirect methods to estimate distribution, population size, and conservation status (Sepúlveda et al. 2007, 2009; Cassini et al. 2010). L. provocax is an opportunistic predator that spends a great deal of its time budget searching and capturing prey (Reyes-Küppers 2007). Its diet has been studied mainly
300
through the analysis of spraints (Chehébar 1985; Fasola et al. 2009; Medina 1997). The typical diet of L. provocax primarily consists of crustaceans, followed by fish and other occasional prey such as insects, amphibians, reptiles, birds, and mammals (Medina 1997, 1998; Medina-Vogel and Gonzalez-Lagos 2008). The use of non-invasive methods (fecal analysis and food availability) to study carnivore feeding strategies and diet is common practice in ecological studies, and is key to understanding the species role in the ecosystem, potential competition with other species, and impact on prey populations, particularly when direct observations of feeding behavior are often impossible under field conditions (Klare et al. 2011). Therefore, these methods may have a far-reaching impact on the implementation of management plans, especially the collection of ecological information on the conservation status of endangered species (Kohn and Wayne 1997; Gese 2001; Ciucci et al. 1996; Davison et al. 2002; Gallant et al. 2007; Crimmins et al. 2009). Despite the inherent limitations of this approach (e.g., bias in samples size, total fecal production, and laboratory methods; Reynolds and Aebischer 1991; Carss and Parkinson 1996; Rühe et al. 2008), these analyses are relatively inexpensive, easy to apply, can include large sample sizes and, most importantly, are non-invasive and do not require animals to be sacrificed or disturbed (Wise 1980). Otters generally have high metabolic rates due to foraging in the frequently cold, aquatic environment, which entails significant costs associated with swimming and thermoregulation (Kruuk 2006). Therefore, they must consume a large amount of prey to obtain the necessary energy to support their elevated energetic demands (Kruuk 2006; Dekar et al. 2010). Estimating the energy content of prey is an approach which allows a more accurate assessment of the energy returns per unit of foraging time (Oftedal et al. 2007), and can explain foraging strategies and nutritional constraints that maintain the energy requirements of these animals (Dekar et al. 2010). However, the parameters influencing food intake (energy and nutrients) also depend on the quality, size, and species of available prey (Trumble et al. 2003). The use of stable isotopes is a powerful tool to elucidate the trophic relationship between prey and predators in an ecosystem (Michener and Lajtha 2007; Martínez del Rio et al. 2009; Newsome et al. 2009). It is increasingly thought that the most reliable and non-invasive approach currently available for feeding ecology studies is the use of stable isotope ratios, comparing the δ13C and δ15N values in tissues from predators and prey (Ben-David et al. 1997). Hence, stable isotope techniques offer the opportunity to overcome several logistical and sample size constraints, particularly in species of high conservation interest (McFadden et al. 2006).
Naturwissenschaften (2013) 100:299–310
Whereas most diet studies of Latin American otters have been based on spraint analysis (except for Lontra longicaudis; Pardini 1998), to our knowledge, stable isotopes have never been used in Chilean otters. In addition, despite some authors having highlighted the importance of estimating prey availability to gain insights into trophic preferences and selectivity in otter species (e.g., Pardini 1998; Anoop and Hussain 2005), we found only one such study in Chile (Medina-Vogel et al. 2004). Knowledge of L. provocax feeding ecology is critical within a conservation context, particularly in light of the current reduction of their range in Chile (see Sepúlveda et al. 2007). Although there has been an increasing number of studies in the feeding habits of Southern river otters throughout Chile and Argentina (e.g., Medina-Vogel and Gonzalez-Lagos 2008; Medina 1997, 1998), it is still not clear what kind of prey items, size, and prey quality are preferred by this species. Also, in order to validate future studies, it is critical to establish their potential role in the ecosystem, potential competition with other species, as well as the real relationship between prey availability, prey presence, and spraints (Franco and Soto-Gamboa 2011). To accomplish this, the aims of this study were (1) to determine the trophic interactions of L. provocax, (2) to assess the effect of prey availability on food consumption, and (2) to establish the position of this otter within the food web and its implications for conservation of freshwater communities.
Methods Study area Our study area covered 20 km2, and included approximately 30 km of river courses situated in the Carlos Anwandter Nature Sanctuary (hereafter CANS),within the borders of the Cruces River, a Ramsar wetland site, and a national water bird biodiversity reserve under Chilean law since 1981 (Artacho et al. 2007; Marín et al. 2009). This shallow water ecosystem (5 m were considered as separate units. Fecal analysis Spraint samples were collected using plastic bags, and were then transported to the laboratory, washed through a sieve (250 μm), and dried at 75 °C for 48 h (Medina 1998). Undigested prey remains were identified using a dissecting stereomicroscope (×10). They were then sorted into five general categories: crustaceans, fish, amphibians, birds, and mammals, and were subjected to further identification
using reference collections from the Institute of Zoology (Universidad Austral de Chile) and procedures suggested by Conroy et al. (2005). Dietary composition was registered for each spraint and reported as frequency of occurrence and relative frequency of individual food categories (Erlinge 1968; Conroy et al. 2005). We calculated the Shannon– Wiener diversity index (H) for diet, and compared the seasonal diet using Chi-square tests χ2 (Zar 1999). The assumptions for statistical tests were assessed using the Kolmogorov–Smirnov test for normality and the Levene test for homoscedasticity (Quinn and Keough 2002). Availability of potential otter prey in the CANS The availability of crustaceans and small fish (8 cm body length) was sampled across a 10-m trawl line with ten hooks per line. Five lines (with three replicates) were located in each habitat, with a total of 15 fishing lines per habitat. This method allows the capture of the full range of sizes. We compared the seasonal prey availability between habitats using Chi-square tests χ2 (Zar 1999), and used a multiple correspondence analysis (Quinn and Keough 2002) to evaluate the associations between prey, season, and sampled habitats. The assumptions of statistical tests were assessed. Prey selection Given that previous results on the feeding features of L. provocax in other aquatic systems of Chile and Argentina (Medina 1997; Cassini et al. 2009) revealed a high occurrence of crayfish (Samastacus spinifrons) in spraints, we built a calibration curve (body length vs. telson and uropod length) based on the captured crayfish from electrofishing (Correia 2001). Body length of captured individuals exhibited a significant positive relationship with telson length (r2 =0.98; P