Fish fauna of small streams of the Catua-Ipixuna Extractive Reserve, State of Amazonas, Brazil

This study was conducted in an Amazonas state conservation unit, the Catuá-Ipixuna Extractive Reserve (Catuá-Ipixuna RESEX). The main purpose was to provide an ichthyological survey of its small streams, all them tributaries of the Solimões River. Nine small streams (up to 4 m width and 1 m depth) were sampled in September 2006. A total of 1,525 specimens were captured, belonging to 78 species, 24 families and eight orders. Eight species had higher abundances and represented altogether 61.4 % of all collected specimens (Hemigrammus belotii, Microphilypnus amazonicus, Physopyxis ananas, Apistogramma agassizii, Elachocharax pulcher, Apistogramma cf. cruzi, Gladioglanis conquistador and Copella nigrofasciata). Based on the high number of singletons and doubletons present in our samples, as well as the estimated number of species for those streams (106 spp.), we believe that the total fish species richness present in the Catuá-Ipixuna Extractive Reserve may be considerably higher than indicated by our samples. This seems especially true when considering the dimensions of the Catuá-Ipixuna RESEX and the dense hydrographic network present in the area. Introduction Although not so conspicuous as the large rivers that dominate the landscapes in the Amazon region, the vast numbers of small forest streams constitute one of the main components of that fluvial system (Fitkau 1964; Welcomme 1985). These small water courses are hydrologically and ecologically connected to larger streams and rivers, and are a source of many essential resources (e.g. nutrients, particulate and dissolved organic matter, invertebrates) for the maintenance of the physical, chemical, ecological and biological integrity of the whole system (Vannote et al. 1980; Nadeau and Rains 2007; Wipfli et al. 2007). Low order streams may also be used as temporary refuges, feeding and spawning areas to migratory fish species (Meyer et al. 2007). Despite their reduced area and apparent low structural complexity, small forest streams harbor a rich fish fauna (Walker and Henderson 1996; Sabino and Zuanon 1998; Lowe-McConnell 1999; Buhrnheim and Cox-Fernandes 2003; Mendonça et al. 2005; Anjos and Zuanon 2007). However, published studies concerning the ichthyofauna of Amazon streams are still scarce and mainly concentrated to the vicinity of Manaus in central Brazilian Amazon. Check List 5(2): 154–172, 2009.


Introduction
Although not so conspicuous as the large rivers that dominate the landscapes in the Amazon region, the vast numbers of small forest streams constitute one of the main components of that fluvial system (Fitkau 1964;Welcomme 1985).These small water courses are hydrologically and ecologically connected to larger streams and rivers, and are a source of many essential resources (e.g.nutrients, particulate and dissolved organic matter, invertebrates) for the maintenance of the physical, chemical, ecological and biological integrity of the whole system (Vannote et al. 1980;Nadeau and Rains 2007;Wipfli et al. 2007).Low order streams may also be used as temporary refuges, feeding and spawning areas to migratory fish species (Meyer et al. 2007).
Despite their reduced area and apparent low structural complexity, small forest streams harbor a rich fish fauna (Walker and Henderson 1996;Sabino and Zuanon 1998;Lowe-McConnell 1999;Buhrnheim and Cox-Fernandes 2003;Mendonça et al. 2005;Anjos and Zuanon 2007).However, published studies concerning the ichthyofauna of Amazon streams are still scarce and mainly concentrated to the vicinity of Manaus in central Brazilian Amazon.
The main purpose of this study is to provide an ichthyological survey of small, 2 nd and 3 rd order streams (according to Horton's scale, modified by Strahler; see Petts 1994) at the Catuá-Ipixuna Extractive Reserve (Catuá-Ipixuna RESEX), a conservation unit in the Solimões River basin west of Manaus.There are no studies about stream fish fauna in that area, and the information presented herein will help to increase the knowledge about fish species distribution and diversity in the Brazilian Amazon.
Each sample station consisted of a 50 m reach of stream.The sampling effort was standardized and based on the use of hand nets (0.4 m 2 area, 2 mm mesh) by two collectors during a period of 120 minutes in each stream (Mendonça et al. 2005).
Fishes were preserved in 10 % buffered formalin in the field and later transferred to 70 % ethanol.Specimens were sorted and identified at the Fish Collection of the Instituto Nacional de Pesquisas da Amazônia (INPA) in Manaus.Species were identified with the use of specific dichotomic keys and ichthyofaunal catalogues (e.g.Géry 1977;Weitzman and Géry 1980;Lucena 1987;Ploeg 1991;Buckup 1993;Mago-Leccia 1994 1).Scale bars in the photos of voucher specimens represents 1 cm.Fish species composition was compared for the whole set of streams and for the Catuá and Ipixuna basins using Jaccard's coefficient (for presence-absence data).The coefficient values varies from 0 (completely different) to 1 (highly similar; Krebs 1999).A Student's t test was employed to compare the mean similarity values among streams of Catuá and Ipixuna basins (Zar 1996).
An overall estimate of the fish species richness in the Catuá-Ipixuna RESEX was calculated by means of the Jackknife 1 method (Krebs 1999).
Species richness and similarity values are presented as mean ± standard deviation (sd).Statistical analyses were conducted using Past software (Hammer et al. 2001).Fish were collected with IBAMA authorization number 11696-2.

Results and Discussion
A total of 1,525 specimens were captured, belonging to 78 species, 24 families and eight orders (Appendix 1 -  Fernández-Yépez, 1968 andRivulus cf. compressus Henn, 1916) were found in five streams (Table 1).Besides, 52 species (67% of the total species richness) occurred in just one or two streams.
Species richness in the samples ranged 10 -26 (20 ± 7 sd).The overall fish richness estimated for the Catuá-Ipixiuna RESEX was 106 species.Overall, fish assemblage similarity values were low and ranged 0.03 -0.35 (0.18 ± 0.08 sd).There was no significant difference in mean similarity values among the streams of the Catuá (0.17 ± 0.01 sd, n=6) and Ipixuna (0.20 ± 0.01 sd; n=10) basins (t= -5.63; p=0.582).Comparing the streams of the two basins, 22 species occurred exclusively in the Catuá and 21 in the Ipixuna.Thirty five species were common to both basins.Among the fishes that occurred in just one of the sampled basins were species of the genera Gymnotus, Aequidens, Apistogramma and Nannostomus, with different species in each basin (Table 1).
Studies about the composition of fish communities of small streams in Central Amazon sometimes diverge in relation to the dominance of the taxonomic groups that compose the ichthyofauna.Most studies generally point out to the dominance of Characiformes and Siluriformes (Araújo-Lima et al. 1999;Mendonça et al. 2005), although sometimes Perciformes can be the second in number of representatives, with approximately two times the amount of species of Siluriformes and Gymnotiformes together (e.g.Silva 1995;Buhrnheim and Cox-Fernandes 2001;Espírito-Santo et al. 2009).In the small streams of the Catuá-Ipixuna Extractive Reserve, Characiformes constituted more than half of the specimens collected, mainly small midwater characins such as the ubiquitous Hemigrammus belotii.The abundance of Perciformes in our samples resulted from the large number of specimens of the minute gobiid Microphilypnus amazonicus, as well as the dwarf cichlid Apistogramma agassizii, an also very common species in the Amazon lowlands (Kullander 1986).
Fish species richness in Amazon streams is known from several studies developed in terra firme areas in Central Amazon, with numbers varying between 17 and 61 species albeit being obtained with the use of diverse sampling efforts and methodologies (Sabino and Zuanon 1998;Araújo-Lima et al. 1999;Buhrnheim and Cox-Fernandes 2001;Mendonça et al. 2005;Anjos and Zuanon 2007).As far as we know, only Henderson and Walker (1990) studied the fish fauna of a stream subjected to periodic flooding, although focusing on species associated to litter banks.The species richness found in the Catuá-Ipixuna streams (78 species) may be considered high due to the small number of streams sampled (nine) in a single occasion, and to the fact that only hand nets were employed as sampling equipment.However, this high number of species may have been influenced by the close proximity and sometimes direct connection of the sampled streams to larger, main water courses of those basins.For instance, specimens of the gobiid Microphilypnus amazonicus were commonly found close to the main streams of the basins.
Among the species collected there were some fishes that constitute typical inhabitants of well preserved forested streams such as Helogenes marmoratus Gunther, 1863 (Cetopsidae) and Monocirrhus polyacanthus Heckel, 1840 (Polycentridae).Some species were found in most of the streams sampled, such as Apistogramma spp.(Cichlidae) and Elachocharax pulcher (Crenuchidae), always associated to litter banks.
Based on the high number of singletons and doubletons present in our samples, as well as the estimated number of species for those streams (106 spp.), we believe that the total fish species richness present in the Catuá-Ipixuna Extractive Reserve may be considerably higher than indicated by our samples.This seems especially true when considering the dimensions of the Catuá-Ipixuna RESEX and the dense hydrographic network present in the area.Further efforts should be directed to the study of the ichthyofauna of the small streams of floodplain areas in the Amazon, which had been largely negleted so far.Figura 79.Total abundance (number of specimens) of fish species collected in the streams of Catuá-Ipixuna Extractive Reserve.

Figure 1 .
Figure 1.Geographic location of the study area at Catuá-Ipixuna Extractive Reserve.Red line depicts the Reserve limits.(Source: SRTM image, NASA).