Amazon protected areas and its ability to protect stream-dwelling fish fauna
Introduction
Protected areas (PAs) are designed to safeguard and conserve all environmental components that maintain biodiversity, including habitats, species, populations, and ecosystem services (Margules and Pressey, 2000). In order to ensure that conservation areas can feasibly reach their goals, Margules and Pressey (2000) proposed implementing systematic conservation planning (SCP) efforts. SCP was created to optimize the effectiveness of protected areas by maintaining natural processes and viable populations under scenarios of limited resources. To do so, the principle of biodiversity complementarity, rather than species richness, has been used when deciding on target areas for conservation (Justus and Sarkar, 2002). The complementarity principle in a network of PAs requires choosing areas that complement each other in relation to biodiversity composition. Given that one of the purposes of SCP is to optimize the limited funding allocated to PAs, this principle has increasingly been incorporated in conservation planning worldwide (Justus and Sarkar, 2002, Rodrigues et al., 2003).
Environmental niche models (ENM) are powerful tools for SCP, intending to address the usual lack of accurate information concerning species distributions (Araújo et al., 2005, Diniz-Filho et al., 2009, Strecker et al., 2011) by determining suitable areas for the occurrence of species (Jiménez-Valverde and Lobo, 2007). The ENM approach is based on a species' Grinnellian niche (Soberón, 2007) and provides data that can be useful to delineate areas harboring potential habitats for the species, considering assumptions related to dispersal. Thus, this method is an important tool for conservation of high levels of diversity in data-poor areas, such as the tropics (Esselman and Allan, 2011, Nóbrega and De Marco Jr, 2011, Vilar et al., 2015, Fagundes et al., 2016). Maps of species distribution and suitable habitat areas generated by ENM could also be used to allow gap analyses to identify unprotected species in a PA network (Diniz-Filho et al., 2009, Nóbrega and De Marco Jr, 2011, Fagundes et al., 2016). Moreover, this method can also be employed when prioritizing areas for conservation (Moilanen, 2005, Moilanen, 2012).
In the past decade the conservation of freshwater ecosystems starts to get more attention (Abell, 2002, Linke et al., 2007, Linke et al., 2011, Turak and Linke, 2011). The discussion of SCP in freshwater ecosystems is highlighting new methods that take into account some specific characteristics from these ecosystems (Berger et al., 2010, Linke et al., 2011, Hermoso et al., 2011, Hermoso et al., 2016, Turak and Linke, 2011). Freshwater systems are composed of hierarchical networks, which render them highly complex and vulnerable both to local direct and large-scale indirect impacts, which include disturbances directly in the water courses or in the surrounding terrestrial ecosystem (Castello et al., 2013). Thieme et al. (2016) also showed that the PAs design and correct designation for freshwater ecosystems needs are important for conservation of freshwater biodiversity, once PAs usually have freshwater ecosystems (rivers) as their boundaries. Despite all advances in the knowledge of the structure, dynamics, and biological importance of aquatic systems, the selection of priority areas for conservation is still established mainly based on data concerning terrestrial organisms and ecosystems. This holds true for Brazil, where some authors argue that large protected “terrestrial” areas, or so-called megareserves, may encompass the diversity of aquatic habitats, and thus, be considered surrogate areas for freshwater ecosystems conservation (Peres and Terborgh, 1995, Peres, 2005). According to Peres (2005), the areas needed to maintain viable populations of large terrestrial predators should be large enough to protect many other wide-ranging terrestrial and aquatic species. This could hold true and would be classified according to Amis et al. (2009) as “incidental integration”; however, the opposite could also be possible since protecting large river catchments could lead to the conservation of terrestrial environments and organisms.
Brazil established the standards for the creation, implementation, and management of its PAs only in 2000, the Sistema Nacional de Unidades de Conservação da Natureza - SNUC (National System of Conservation Units http://www.planalto.gov.br/ccivil_03/leis/l9985.htm). In 2003, the Brazilian government created the Programa Areas Protegidas da Amazônia - ARPA (Amazon Protected Areas Program), and since then approximately 46 conservation units have been created in the region, summing 24 million hectares (http://programaarpa.gov.br/en/arpa-in-numbers//). These PAs were created between 2003 and 2010 in areas at high risk of human impacts, mostly intending to decrease deforestation rates (Veríssimo et al., 2011). The conservation units in the Amazon include the largest PAs in Brazil, and currently approximately 43% of the Brazilian Amazon is formally protected by conservation units of indirect or direct use (Veríssimo et al., 2011). Since almost half of Brazilian Amazon territory is under protected areas, it could be expected that freshwater biodiversity would also be adequately represented. However, Fagundes et al. (2016) found that Amazonian PAs do not protect freshwater turtles satisfactorily, which they attributed to the fact that conservation planning is focused mostly in terrestrial vertebrates and plants as targets. Thus, the ability of those units to protect fish and other aquatic elements of biodiversity is also questionable.
It is important to highlight that, even with all the advances in preserving the Brazilian Amazon forest as well as in reducing carbon emission rates, little attention has been given to its freshwater ecosystems, and the regional planning and management of water resources remains unsatisfactory (Castello et al., 2013, Castello and Macedo, 2016). Currently in Brazil, freshwater fishes are among the most threatened vertebrates, due to progressive habitat loss and the scarcity of knowledge concerning the ecological needs of most species (Nogueira et al., 2010). The Amazon Basin comprises approximately 7 million km2, harbors the highest species richness of freshwater fish in the world, and 70% of this basin is in Brazil (Goulding et al., 2003, Abell et al., 2008). The majority knowledge concerning the Amazon fish fauna is concentrated on medium- to large-sized species that dwell in large river-floodplain systems and are exploited by commercial fisheries. However, below the vast canopy of upland forests there is a dense and complex hydrological network of small-order streams, regionally called igarapés, with a very rich and poorly known fish fauna. Bearing in mind the efforts taken to protect Amazon ecosystems, our main objective here was to verify the ability of the current Brazilian Amazon PAs network to conserve stream-dwelling fishes. Considering the strong ecological dependence between upland forest streams and the surrounding landscape, we selected only stream-dwelling species to test the hypothetical surrogacy stated by Peres (2005). In order to accomplish our goal, and facing the lack of knowledge on the ecological characteristics of most of these species, we performed a Gap Analysis and a spatial prioritization procedure using Environmental Niche Models.
Section snippets
Data
First- to third-order streams depend almost exclusively on allochthonous sources of energy and nutrients, with strong dependence on the surrounding terrestrial ecosystem (Faush et al., 2002, Soininen et al., 2015). These small headwater streams are not affected by the seasonal flood pulse or lateral expansion of the channel of large rivers, and are mainly influenced by local precipitation. In order to only select true stream-dwelling species and to work with quantitatively comparable data, we
Models
We built ENMs for 138 selected stream fish species from the Igarapés Project database. It is important to highlight that we are aware about uncertainties in the data regarding presence-only distribution models; however, as noted above, this is a useful tool in data-poor areas. The AUC values ranged from 0.589 (Jupiaba zonata) to 0.978 (Iguanodectes rachovii), which are two small characin species. Widespread species with a high number of occurrence points usually had lower AUC values. The TSS
Discussion
Our findings provide the first evaluation of Amazon protected areas network effectiveness for conservation of stream-dwelling fish species. Although our analyses did not identify any total gap species, the current design of Amazon PAs network fails to fully safeguard its stream fish biodiversity. We consider that the gap analyses results represent an incidental integration, once considering that the most inclusive scenario, when all species were considered protected, covers 43% of the Brazilian
Conclusion
The Amazon River basin is the biggest and biodiversity richest basin in the world, with the largest portion of its area in Brazil, which makes the country a key for conservation of the biodiversity and functioning of the Amazon River basin. Our prioritization analysis supports the need for a re-evaluation of strategies for implementing Amazon reserves regarding stream fish conservation, which should integrate the knowledge and efforts of scientists, public and private agencies, and local
Acknowledgments
We thank Rafael Loyola and the three anonymous reviewers for comments to improve the paper and Bruno Ribeiro for help using the Zonation software. RGF was funded by a PhD fellowship from the “Coordenação de Aperfeiçoamento de Pessoal de Nível Superior” (CAPES). JZ receives a productivity grant from CNPq (grant number 313183/2014-7). This article represents contribution #53 of the Igarapés Project, which benefited from several research grants from CNPq and Fundação de Amparo à Pesquisa do Estado
References (70)
- et al.
Unlocking the potential of protected areas for freshwaters
Biol. Conserv.
(2007) - et al.
Integrating freshwater and terrestrial priorities in conservation planning
Biol. Conserv.
(2009) - et al.
Replacement cost: a practical measure of site value for cost-effective reserve planning
Biol. Conserv.
(2006) - et al.
Threshold criteria for conversion of probability of species presence to either–or presence–absence
Acta Oecol.
(2007) Methods for reserve selection: interior point search
Biol. Conserv.
(2005)Landscape zonation, benefit functions and target-based planning: unifying reserve selection strategies
Biol. Conserv.
(2007)- et al.
Protected area downgrading, downsizing, and degazettement (PADDD) in the Amazon
Biol. Conserv.
(2016) - et al.
Maximum entropy modeling of species geographic distributions
Ecol. Model.
(2006) - et al.
Freshwater conservation planning in data-poor areas: an example from a remote Amazonian basin (Madre de Dios River, Peru and Bolivia)
Biol. Conserv.
(2007) - et al.
Setting priorities for the conservation of marine vertebrates in Brazilian waters
Ocean Coast. Manag.
(2015)