Abstract:

To better understand the fish fauna of Amazonian streams, we assessed small streams in a landscape of forest fragments and savannah in the region of the lower Tapajós River, Brazil. We sampled the fish fauna using a well-tested active capture method during two dry seasons (October 2006 and 2018) and one rainy season (March to May 2018). Species richness was calculated using an abundance matrix and first-order jackknife estimator. Using qualitative and quantitative data, we present a cluster analysis in which each stream corresponds to a sampling unit. We collected 6,094 individuals of 43 species distributed in six orders. The sampling effort represents 73% of the estimated richness (58.69 ± 7.65). The most abundant species were Copella nattereri, Iguanodectes variatus and Laimosemion dibaphus that together represent almost half of the total sample (48.7%). The most frequent species were Aequidens epae, Helogenes marmoratus and Laimosemion dibaphus, which were collected in 11 of the 13 sampled streams. This is the first fish fauna list for small-order streams of savannah and forest fragments landscape in Amazonian Brazil. The richness of fish and the presence of many rare species underscore the contribution of small streams to the regional fish fauna composition, even in dynamic and spatially restricted landscapes.

Keywords:
Eastern Amazonia; Headwater streams; Neotropical fish; Riverscapes; Stream fish

Resumo:

Com o objetivo de aprimorar o conhecimento científico sobre a ocorrência da ictiofauna de riachos na Amazônia, acessamos pequenos riachos em uma paisagem composta por fragmentos florestais e savana, na região do baixo rio Tapajós. Amostramos a fauna de peixes com um método bem testado de captura ativa em duas estações secas, outubro de 2006 e 2018 e em uma estação chuvosa de março a maio de 2018. A riqueza de espécies foi calculada a partir da matriz de abundância com o estimador jackknife de primeira ordem. Usando dados qualitativos e quantitativos, apresentamos uma análise de agrupamento, onde cada riacho corresponde a uma unidade amostral. Coletamos 6094 indivíduos, de 43 espécies distribuídas em seis ordens. O esforço de amostragem representou 73% da riqueza estimada (58,69 ± 7,65). As espécies mais abundantes foram Copella nattereri, Iguanodectes variatus e Laimosemion dibaphus, representando em conjunto quase a metade do total amostrado (48,7%). As espécies mais frequentes foram Aequidens epae, Helogenes marmoratus e Laimosemion dibaphus, coletadas em 11 dos 13 riachos amostrados. Esta é a primeira lista de peixes de pequenos riachos em savana e fragmentos florestais da Amazônia brasileira. A riqueza de peixes e a presença de muitas espécies raras ressaltam a contribuição de riachos de pequena ordem para a composição regional da fauna de peixes, mesmo em paisagens dinâmicas e espacialmente restritas.

Palavras-chave:
Amazônia oriental; Paisagens de rios; Peixes neotropicais; Peixes de riachos; Riachos de cabeceira

Introduction

The Neotropical region has the most diverse freshwater fish fauna on the planet, with 5,160 described species for South America and over 9,100 estimated species (Reis et al. 2016REIS, R.E., ALBERT, J.S., DI DARIO, F., MINCARONE, M.M., PETRY, P. & ROCHA, L.A. 2016. Fish biodiversity and conservation in South America. J. Fish Biol. 89(1):12-47.). The Amazon basin contains more than 2,700 freshwater fish species, of which around 1,700 are endemic (Reis et al. 2016, Dagosta and de Pinna 2019DAGOSTA, F.C. & DE PINNA, M. 2019. The fishes of the Amazon: distribution and biogeographical patterns, with a comprehensive list of species. Bulletin of the American Museum of Natural History 2019(431): 1-163.; Oberdorff et al. 2019OBERDORFF, T., DIAS, M. S., JÉZÉQUEL, C., ALBERT, J. S., ARANTES, C. C., BIGORNE, R., CARVAJAL-VALLEROS, F. M., DE WEVER, A., FREDERICO, R. G., HIDALGO, M. , HUGUENY, B., LEPRIEUR, F., MALDONADO, M., MALDONADO-OCAMPO, J. , MARTENS, K., ORTEGA, H., SARMIENTO, J., TEDESCO, P. A., TORRENTE-VILARA, G., WINEMILLER, K. O., ZUANON, J. 2019. Unexpected fish diversity gradients in the Amazon basin. Sci. Adv. 5, eaav8681.). More than half of its linear extension comprises small-order streams (Macedo and Castello 2015MACEDO, M. & CASTELLO, L. 2015. State of the Amazon: Freshwater Connectivity and Ecosystem Health. (D. Oliveira, C.C. Maretti & S. Charity, eds). Brasília, Brazil, WWF Living Amazon Initiative. 136pp.) that have an extraordinarily diverse fish fauna (Bührnheim & Cox-Fernandes 2003BÜHRNHEIM, C.M. & COX-FERNANDES, C. 2003. Structure of Fish Assemblages in Amazonian Rain-Forest Streams: Effects of Habitats and Locality. Copeia 2003(2): 255-262., Mendonça et al. 2005MENDONÇA, F.P., MAGNUSSON, W.E. & ZUANON, J. 2005. Relationships Between Habitat Characteristics and Fish Assemblages in Small Streams of Central Amazônia. Copeia 4: 750-763., de Oliveira et al. 2009OLIVEIRA, R.R., ROCHA M.S., ANJOS, M.B., ZUANON, J. & L.H.R. PY-DANIEL. 2009. Fish fauna of small streams of the Catua-Ipixuna Extractive Reserve, State of Amazonas, Brazil. Check List 5(2):154-172., Barros et al. 2011BARROS, D.F., ZUANON, J., MENDONÇA, F.P., ESPÍRITO-SANTO H.M.V., GALUCH, A.V. & ALBERNAZ, A.L.M. 2011. The fish fauna of streams in the Madeira-Purus. Check List 7: 768-773., Silva-Oliveira et al. 2016SILVA-OLIVEIRA C., CANTO A.L.C. & RIBEIRO F.R.V. 2016. Stream ichthyofauna of the Tapajós National Forest, Pará, Brazil. ZooKeys 580: 125-144., Leitão et al. 2017LEITÃO, R.P., ZUANON J., MOUILLOT, D., LEAL, C.G., HUGHES, R.M., KAUFMANN, P.R., VILLÉGER, S., POMPEU, P.S., KASPER, D., DE PAULA, F.R., FERRAZ, S.F.B. & GARDNER, T.A. 2017. Disentangling the pathways of land use impacts on the functional structure of fish assemblages in Amazon streams. Ecography 125:336-342.).

Fish fauna studies in the Amazon basin are widespread for upland rainforest streams but still incipient for savannah landscapes. This vegetation has a restricted spatial distribution representing only 3-4% of the basin area (Pires and Prance 1985PIRES, J.M. & PRANCE, G.T. 1985. The vegetation types of the Brazilian Amazon. In Key environments: Amazonia (G.T. PRANCE & T.E. LOVEJOY, eds). Pergamon Press, Oxford, p. 109-145.) and, as expected, many endemic taxa (Barbosa et al. 2003BARBOSA, R.I., XAUD, M.R., SILVA G.N. & CATTÂNEO A.C. 2003. Forest fires in Roraima, Brazilian Amazonia. International Forest Fire News 28: 51-66., Plotkin & Riding 2011PLOTKIN R.L. & RIDING S. 2011. Biogeography of the Llanos de Moxos: natural and anthropogenic determinants. Geografica Helvetica 3:183-192., De Carvalho & Mustikin 2017DE CARVALHO, W.D. & MUSTIN K. 2017. The highly threatened and little known Amazonian savannahs. Nature Ecology & Evolution 1(4): 0100.). The savannah and its surrounding forest fragments along the lower Tapajós River are drained by small streams that are used for recreation by local populations and are threatened by urban growth. Therefore, it is urgent to conduct studies about these streams and their biota. Our goal here is to present the first fish fauna list for small streams in a savannah and forest fragments landscape, which is based on a standardized and well-tested active capture method used for small Amazonian streams.

Material and Methods

1. Study area

Sampling sites were selected in a landscape comprising forest fragments and savannah vegetation connected to a continuous forest (Amaral et al. 2017AMARAL, I.L., MAGNUSSON, W.E., MATOS, F.D.A., ALBERNAZ, A.N.K., FEITOSA, Y.O. & GUILLAUMENT, J.L. 2017. Disentangling structural patterns of natural forest fragments in a savanna matrix in the eastern Brazilian Amazon. Acta Amazonica 47(2): 111-122.), in the lower Tapajós River basin (Figure 1 and Table 1). The combined sampling sites represent a drainage area of 17,847 ha. Drainage flows into a main lake, Verde Lake, and smaller streams flow to the right margin of the Tapajós River. The drainage is poorly developed or branched, with sparse streams in the landscape (Figure 1 and 2). The streams have clear water, a natural condition that is attractive to local tourism (Fróis pers. obs. and Figure 1).

Figure 1
Study area and the 13 sample sites where the ichthyofauna was collected in small-order streams of savannah and forest fragments landscape in the lower Tapajós River (ADC is the code for Alter do Chão - the district name). The bigger streams at the study site are Jatuarana and Sonrisal, both third order streams. The image on the left represents elevation information from the SRTM radar, and on the right represents the vegetation in the study area (savannah = brown spots surrounding the ADC 01, 02, 03, 09, 10 and 11 samples). (SRTM 1 Arc-Second Global; Geographical Projection Lat-Long; Planimetric Datum WGS84; Database: IBGE/ANA).

Figure 2
Stretches from stream samples located in a savannah and forest fragments landscape in the lower Tapajós River basin.

The average annual precipitation is 1,991 mm and there are two well-defined climatic seasons: a dry period (precipitation <100 mm/month) between July and November and a rainy period (precipitation >100 mm) between December and June. The mean annual temperature is 25.9ºC, with a variation of 1.8 °C (Climate-data.org 2019CLIMATE DATA.ORG. Clima de Alter do Chão. http://climate-data.org/location/317494. Access 20 march 2019.
http://climate-data.org/location/317494... ). The vegetation is formed by semideciduous forest (IBGE 2012IBGE. 2012. Manual Técnico da Vegetação Brasileira: Sistema fitogeográfico. Inventário das formações florestais e campestres Técnicas e manejo de coleções botânicas. Procedimentos para mapeamentos. Nº 1, 2ª Edição ampliada. Centro de Documentação e Disseminação de Informações, Rio de Janeiro, 271p.) surrounded by savannah. This savannah contains two dominant grass species, a shrub vegetation dominated by Melastomataceae and Myrtaceae species, and sparse, short- to medium-sized trees (Magnusson et al. 2008MAGNUSSON, W.E., LIMA, A.P., ALBERNAZ, A.L.K.M., SANAIOTTI, T.M. & GUILLAUMET, J.L. 2008. Composição florística e cobertura vegetal das savanas na região de Alter do Chão, Santarém - PA. Revista Brasileira de Botânica 31(1): 165-177.).

2. Data collection

We sampled the ichthyofauna in 13 first- and second-order streams (50-meter stretch per stream) using the standardized active capture method described in Mendonça et al. (2005)MENDONÇA, F.P., MAGNUSSON, W.E. & ZUANON, J. 2005. Relationships Between Habitat Characteristics and Fish Assemblages in Small Streams of Central Amazônia. Copeia 4: 750-763.. We carried out the first sampling in October 2006 at the peak of the dry season. We sampled the streams two more times in 2018, once in the rainy season (March to May) and once in the dry season (October).

The 50-meter stretches were blocked with fine-mesh nets (5 mm stretched mesh size) and the fish were sampled with two people/two hours of effort using seine net and hand nets. Collected specimens were euthanized with a 2% clove oil solution (Fernandes et. al 2017FERNANDES, I.M., BASTOS, Y.F., BARRETO, D.S., LOURENÇO, L.S. & PENHA, J.M. 2017. The efficacy of clove oil as an anaesthetic and in euthanasia procedure for small-sized tropical fishes. Braz. J. Biol. 77(3): 444-450.), fixed in a 10% formalin solution, transferred to 70% alcohol, and are deposited in the INPA Fish Collection (Table S1). We identified specimens using the literature (Géry 1977GÉRY, J. 1977. Characoids of the world. T.F.H. Publications. Inc. Ltd. Neptune City, New Jersey, 672 pp., Vari 1982VARI, R.P. 1982. Systematics of the neotropical characoid genus Curimatopsis (Pisces: Characoidei). Smithsonian Contributions to Zoology 373 (1-3): 1-28., Kullander 1986KULLANDER, S. O. 1986. The cichlid fishes of the Amazon river drainage of Peru. Swedish Museum of Natural History, Stockholm, 431p.). The taxonomic classification follows Fricke et al. (2020)FRICKE, R., ESCHMEYER, W. N. & R. VAN DER LAAN. 2020. ESCHMEYER'S CATALOG OF FISHES: GENERA, SPECIES, REFERENCES. http://researcharchive.calacademy.org/research/ichthyology/catalog/fishcatmain.asp (last access 01 October 2020).
http://researcharchive.calacademy.org/re... . We evaluated the sampling effort from an abundance matrix and a first-order jackknife richness estimator (Krebs 1999KREBS, C. J. 1999. Ecological Methodology. Addison Wesley Educational Publishers, Menlo Park.), with 1,000 randomizations and considering each stream stretch as an independent sample unit, using the software ESTIMATES version 9 (Cowell 2013COLWELL, R.K. 2013. EstimateS: Statistical estimation of species richness and shared species from samples. Version 9. User's guide and application. http://purl.oclc.org/estimates.
http://purl.oclc.org/estimates... ). We calculated the fauna dissimilarity among sample units using the Jaccard distance for qualitative data (presence/absence) and Bray-Curtis distance for quantitative data, which is summarized as dendrograms drawn in PAST 3.20 (Hammer et al. 2001HAMMER, Ø., HARPER, D.A.T. & RYAN, P.D. 2001. Past: Paleontological Statistics Software Package for Education and Data Analysis. Palaeontologia Electronica 4(1): 1-9.).

Results

We sampled 6,094 individuals, belonging to 43 species, eight orders and 22 families (Table 2). The streams had clear (mean turbidity = 1.3 NTU, range = 0.02-6.2) and acidic waters (mean pH = 4.5, range = 3.4-5.3), with low conductivity (mean = 15.4 µS/cm, range = 6.4-46.4) and relatively high temperature (mean = 27.3 °C, range = 25.2-29.4). Characiformes was the richest and most abundant order, with 19 species and 4,094 individuals, followed by Cichliformes with eight species and 542 individuals. Gymnotiformes had one more species than Siluriformes, but the latter had three times more individuals. Cyprinodontiformes was represented by two species and Gobiiformes, Perciformes and Synbranchiformes by one species each (Table 3). The most abundant species were Copella nattereri (1,181 individuals, 19.4% of total), Iguanodectes variatus (1,070, 17.4%) and Laimosemion dibaphus (728, 11.8%), which together accounted for almost half of the total individuals (2,958, 48.7%) (Figures 3 and 4). The most frequent species were Aequidens epae, Helogenes marmoratus and L. dibaphus, collected in 11 of the 13 sampled streams.

Figure 3
Photo of the most representative orders collected in small-order streams of savannah and forest fragments landscape in the lower Tapajós River basin. Scale: 1 cm. Characiformes: (a) Acestrorhynchus falcirostris; (b) Curimatopsis evelynae; (c) Copella nattereri; (d) Moenkhausia copei; (e) Hemigrammus stictus; Perciformes: (f) Monocirrhus polyacanthus; Cichliformes: (g) Acarichthys heckelii, (h) Acaronia nassa; (i) Apistogramma gephyra; (j) Hypselecara coryphaenoides; Gymnotiformes: (k) Hypopygus lepturus and Siluriformes; (l) Helogenes marmoratus. Color in alcohol.

Figure 4
Species distribution using a relative abundance rank for 13 streams located in small-order streams of savannah and forest fragments landscape in the lower Tapajós River basin.

Thirty-two species were recorded during the first sampling (Dry season/2006), 29 species during the second sampling (Rainy season/2018) and 30 species during the third sampling (Dry/2018). Of the 43 species collected, 18 (41,9%) were recorded during all sampling periods. Species abundance did not vary among periods, except for Copella callolepis (Dry/2006: 14 individuals; Rainy/2018: 80; Dry/2018: 476), C. nattereri (677; 282; 201), Hyphessobrycon heterorhabdus (155; 11; 180) and Helogenes marmoratus (164; 44; 99) (Table 2). Some species distributions were related to a specific sample site or to a set of sample sites that were hydrographically related. Copella callolepis was present in a set of streams that drain into the southern border of Verde Lake (ADC 05, 06, 07, 08, 11 and 13). Its congeneric C. nattereri was present in the other sites (ADC 02, 03, 04, 09 and 10) and was never recorded in the same sample site. Elachocarax junki and Pygopristis denticulata were only sampled in the ADC 03 stream.

Species richness varied among streams. Sample sites ADC 02 and 10 had the greatest species richness (S = 19 and 28, respectively). Some of these species were considered occasional due to the variation between sampling periods (Table 2). In contrast, ADC 12 had the lowest species richness (3) (Table 1).

A faunistic analysis of similarity showed that stream ADC 12 was the most divergent sample followed by ADC 02 and 10, for both qualitative and quantitative data. A set of five streams formed a group with higher similarity (ADC 05, 06, 07, 08, 11 and 13) for the qualitative data, suggesting a geographical clustering for this subset of the whole fish assemblage (Figure 5).

Figure 5
Similarity analysis for fish fauna sampled in small-order streams of savannah and forest fragments landscape in the lower Tapajós River basin. (A) Similarity based on Bray-Curtis distance (abundance); (B) Based on Jaccard distance (presence/absence) (Cophen. corr. 0.92 and 0.96, UPGMA method) for all sampling periods and for all 13 sampled streams.

Discussion

This inventory is the first to focus exclusively on small streams within a savannah and forest fragments landscape and our results reinforce the importance of spatial heterogeneity to the regional fish community composition. The species richness recorded (43) is considered high given the relatively small number of streams sampled (13) and that only active capture methods were used. Many studies of headwater streams in the Amazon basin, with different sampling efforts and capture methods, show comparable species richness numbers (Araújo-Lima et al. 1999ARAÚJO-LIMA, C. A. R. M., L. F. JIMÉNEZ, R. S. OLIVEIRA, P. C. ETEROVICK, U. MENDONZA, & A. JEROZOLIMKI. 1999. Relação entre o número de espécies de peixes, complexidade de hábitat e ordem do riacho nas cabeceiras de um tributário do rio Urubu, Amazônia Central. Acta Limnologica Brasiliensia 11(2): 127-135., Bührnheim & Cox-Fernandes 2001BÜHRNHEIM, C. M. & COX-FERNANDES, C. 2001. Low seasonal variation of fish assemblages in Amazonian rain forest streams. Ichthyological Exploration of Freshwaters 12(1): 65-78., Mendonça et al. 2005MENDONÇA, F.P., MAGNUSSON, W.E. & ZUANON, J. 2005. Relationships Between Habitat Characteristics and Fish Assemblages in Small Streams of Central Amazônia. Copeia 4: 750-763., Espírito-Santo et al. 2009ESPÍRITO-SANTO, H.M.V., MAGNUSSON, W.E., ZUANON, J., MENDONÇA, F. P. & LANDEIRO, V.L. 2009. Seasonal variation in the composition of fish assemblages in small Amazonian forest streams: evidence for predictable changes. Freshwater Biology 54(3): 536-548., Oliveira et al. 2009OLIVEIRA, R.R., ROCHA M.S., ANJOS, M.B., ZUANON, J. & L.H.R. PY-DANIEL. 2009. Fish fauna of small streams of the Catua-Ipixuna Extractive Reserve, State of Amazonas, Brazil. Check List 5(2):154-172.).

The first-order jackknife estimator showed that 73% of the expected species were sampled (58.69 ±7.65). This efficiency of the sampling effort used here was satisfactory for active collection sampling in first- and second-order streams. For sections with two times the sample length (100 m), Anjos et al. (2007)ANJOS, M. B. & ZUANON, J. 2007. Sampling effort and fish species richness in small terra firme forest streams of central Amazonia, Brazil. Neotropical Ichthyology 5:45-52. using similar collection methods and sampling effort obtained 71.4 to 94.1% of the estimated species richness in central Amazonia.

The recorded fish fauna is characterized by typical inhabitants of Amazonian upland forest streams (Santos & Ferreira 1999SANTOS, G.M. & FERREIRA, E.J.G. 1999. Peixes da bacia amazônica. In: Estudos ecológicos de comunidades de peixes tropicais (R. LOWE-MCCONNELL, ed). São Paulo, EDUSP, p.345-373.). There was a dominance of species that occupy the uppermost (surface) strata of the water column and were observed in habitats with stronger currents, such as Iguanodectes variatus (Iguanodectidae) and Hyphessobrycon heterorhabdus (Characidae), as well as species present in slow-flowing backwaters, such as Copella spp. (Lebiasinidae). In addition, abundant species associated with deeper and slow-flowing environments, such as Aequidens epae, Apistogramma gephyra (Cichlidae) and Helogenes marmoratus (Cetopsidae), were also recorded in the present inventory, of which the latter is a common inhabitant of leaf-packed habitats under moderate to swift currents (Vari & Ortega 1986VARI, R.P. & H. ORTEGA. 1986. The catfishes of the neotropical Family Helogenidae (Ostariophysi: Siluroidei). Smith. Contrib. Zool. (442): 20., Carvalho et. al 2013CARVALHO, L.N., FIDELIS, L., ARRUDA, R., GALUCH, A. & ZUANON, J. 2013. Second floor, please: the fish fauna of floating litter banks in Amazonian streams and rivers. Neotropical Ichthyology 11(1): 85-94.).

In the Neotropics, dominance in fish community composition (in descending order of richness) is generally by Characiformes and Siluriformes, together accounting for about 80% of the species (Lowe-McConnell 1999LOWE-MCCONNELL, R. H. 1964. The fishes of Rupununi savanna district of British Guiana, South America. I. Ecological groupings of species and effects of the seasonal cycle on the fish. Journal of the Linnean Society 45(304):103-144.). Fish community composition studies of headwater streams throughout Amazonia show similar results regarding the distribution of species richness among taxonomic groups; usually Characiformes and Siluriformes are the most representative orders (Sabino & Zuanon 1998SABINO, J. & ZUANON. J. 1998. A stream fish assemblage in Central Amazonia: distribution, activity patterns and feeding behavior. Ichthyological Exploration of Freshwaters 8(3):201-210., Mendonça et al. 2005MENDONÇA, F.P., MAGNUSSON, W.E. & ZUANON, J. 2005. Relationships Between Habitat Characteristics and Fish Assemblages in Small Streams of Central Amazônia. Copeia 4: 750-763., Araújo et al. 2009ARAÚJO, T.R., CELLA-RIBEIRO, A., DORIA, C.R.C. & TORRENTE-VILARA, G.T. 2009. Ichthyofauna of the stream in the Madeira River. Biota Neotropica (9)3: 021-029.). However, in the present inventory, Characiformes and Cichliformes were the richest, which is similar to the pattern observed by Silva-Oliveira et al. (2016)SILVA-OLIVEIRA C., CANTO A.L.C. & RIBEIRO F.R.V. 2016. Stream ichthyofauna of the Tapajós National Forest, Pará, Brazil. ZooKeys 580: 125-144. for small, non-flooded rainforest streams in the lower Tapajós River basin. Characidae and Cichlidae were the most representative and, in both inventories, Loricariidae was absent. This pattern may be related to the natural characteristics lacustrine (ria-lake) environments originating from the Tapajós River estuary and favor a greater diversity of cichlids, similar to the distribution pattern observed for the ichthyofauna in streams in Caxiuanã, along the lower Anapu River (Montag et al. 2008MONTAG, L.F.A., FREITAS, T.M.S., WOSIACKI, W.B. & BARTHEM, R.B. 2018. Os peixes da Floresta Nacional de Caxiuanã (municípios de Melgaço e Portel, Pará - Brasil). Bol. Mus. Para. Emílio Goeldi. Ciências Naturais. 3(1):11-34.).

The fish community changes over time and space, with changes between streams in the same hydrographic basin or in spatially close basins, suggest that the fish assemblages can be influenced by local environmental characteristics of the streams in a small-scale context (Mendonça et al. 2005MENDONÇA, F.P., MAGNUSSON, W.E. & ZUANON, J. 2005. Relationships Between Habitat Characteristics and Fish Assemblages in Small Streams of Central Amazônia. Copeia 4: 750-763.). Changes in fish assemblage's composition during sampling periods, as observed for the abundances of C. nattereri (677; 282; 201), Hyphessobrycon heterorabdus (155; 11; 180) and Helogenes marmoratus (164; 44; 99), may be related to seasonality (Espírito-Santo et al. 2009ESPÍRITO-SANTO, H.M.V., MAGNUSSON, W.E., ZUANON, J., MENDONÇA, F. P. & LANDEIRO, V.L. 2009. Seasonal variation in the composition of fish assemblages in small Amazonian forest streams: evidence for predictable changes. Freshwater Biology 54(3): 536-548.). Small streams do not present predictable variations in water level but respond quickly to local rainfall that affect streams' structural conditions and water physicochemical parameters in shorter and more frequent periods (Walker 1995WALKER I. 1995. Amazonian streams and small rivers. In: Limnology in Brazil (J.G. TUNDISI, C.E.M. BICUDO & T.M. TUNDISI, eds). Brazilian Academy of Sciences, p. 167-193.).

Our data show low similarity with the fish fauna from first- to third- order upland rainforest streams in the Tapajós National Forest (TNF) (Silva-Oliveira et al. 2016SILVA-OLIVEIRA C., CANTO A.L.C. & RIBEIRO F.R.V. 2016. Stream ichthyofauna of the Tapajós National Forest, Pará, Brazil. ZooKeys 580: 125-144.), which is located on the same margin of the Tapajós River and nearly one hundred km away from our study area. We recorded only 18 species in common with the TNF inventory (23% similarity - Jaccard distance) among the 43 species collected in the present study and 53 species collected by Silva-Oliveira et al. (2016)SILVA-OLIVEIRA C., CANTO A.L.C. & RIBEIRO F.R.V. 2016. Stream ichthyofauna of the Tapajós National Forest, Pará, Brazil. ZooKeys 580: 125-144..

The low species richness for stream ADC 12 could be related to local features, given its partially degraded condition resulting from direct human use for recreation, featuring large and shallow pools covered by litter and low dissolved oxygen concentration (1.4 mg/L). At this site, we collected juveniles of Hoplias malabaricus, Laimosemion dibaphus and Synbranchus madeirae, which are capable of surviving in these conditions due to their respiratory adaptations (Lowe-McConnell 1964LOWE-MCCONNELL, R. H. 1964. The fishes of Rupununi savanna district of British Guiana, South America. I. Ecological groupings of species and effects of the seasonal cycle on the fish. Journal of the Linnean Society 45(304):103-144., Val et al. 1998VAL, A.L., SILVA, M.N.P. & ALMEIDA-VAL, V.M.F. 1998. Hypoxia adaptation in fish of the Amazon: a never-ending task. South African Journal of Zoology 33:2.).

We identified a subset of streams that are remarkably similar in fish assemblage composition when compared to the other sampled streams. These streams are hydrographically connected at the southern border of Verde Lake and drain into the forest fragments vegetation, sharing characteristics that may function as environmental filters for fish species (Mendonça et al. 2005MENDONÇA, F.P., MAGNUSSON, W.E. & ZUANON, J. 2005. Relationships Between Habitat Characteristics and Fish Assemblages in Small Streams of Central Amazônia. Copeia 4: 750-763.). Among them, ADC 05 is the least similar, which had one species (Nanostomus marginatus) exclusive to this sample site. Furthermore, this group of streams shares the presence of Copella callolepis, whereas its congener C. nattereri was recorded only to the northern portion of the drainage. This separation may be due to the lake acting as a barrier or the result of different biological interactions occurring at the opposite portions of the lake's shore.

High species richness values observed in the ADC 02 and ADC 10 streams could be related to lake proximity, being more affected by the river flood pulse and therefore having a distinct fish fauna (Walker 1990WALKER I. 1995. Amazonian streams and small rivers. In: Limnology in Brazil (J.G. TUNDISI, C.E.M. BICUDO & T.M. TUNDISI, eds). Brazilian Academy of Sciences, p. 167-193.). In the same stretches, we recorded Acestrorhynchus falcirostris and Mesonauta festivus juveniles, which shows that these streams can function as refuges and growth sites for lake-dwelling species (Goulding 1980GOULDING, M. 1980. The fishes and the forest: explorations in Amazonian natural history. Los Angeles, University of California Press, 280p., Meyer et al. 2007MEYER, J.L., D.L. STRAYER, J.B., WALLACE, S.L., EGGERT, HELFMAN, G.S. & LEONARD, N.E. 2007. The contribution of headwater streams to biodiversity in river networks. Journal of the American Water Resources Association 43(1): 86-103.). These two streams have a greater fluctuation in species number among periods, with the occurrence of non-resident species in the channel that use the habitat during the rainy season, such as Moenkhausia copei, Hemigrammus analis, H. levis, Acarichthys heckelii and Fluviphylax simplex. This result highlights the contribution of small streams for maintaining the regional diversity of the fish fauna.

Species richness in this inventory is numerically comparable but taxonomically distinct from that of streams in other forested, non-floodable areas throughout Amazonia and represents an important component of the regional fish fauna diversity. Therefore, this study is also important because it improves our knowledge about the fish fauna of small-order streams of mixed savannah-forest fragments landscape. This landscape is spatially restricted and poorly known in Amazonia in regard to fish communities and their relationships with the environment. In addition, the presence of species exclusive to some sampling sites, even after three sampling periods, confirms the need to implement riverscape protection strategies that consider a greater number of headwater streams, which will ensure the maintenance of hydrological connectivity (Jézéquel et al. 2020JÉZÉQUEL, C. TEDESCO, P.A., DARWALL, W., DIAS, M.S., FREDERICO, R.G., HIDALGO, M., HUGUENY, B., MALDONADO-OCAMPO, J., MARTENS, K., ORTEGA, H., TORRENTE-VILARA, G., ZUANON, J. & OBERDORFF, T. 2020. Freshwater fish diversity hotspots for conservation priorities in the Amazon Basin. Conservation Biology 00(0):1-10.) to safeguard the local fauna and regional diversity.

Acknowledgments

This is contribution # 67 of the Projeto Igarapés. We thank to Coordenação de Aperfeiçoamento de Pessoal de Nível Superior do Brasil (CAPES) for the scholarship (RF) and to scientific literature free access from Portal de Periódicos CAPES/MEC. We also thank to Nathan Smith for reviewing manuscript' English.

Funding -This work was financially supported by: Programa Nacional de Cooperação Acadêmica na Amazônia (PROCAD-AM/CAPES, edital N. 21/2018, processo Nº 88887.200472/2018-00) for supporting INPA's fish collection technical visit, and to the PELD/POPA (Conselho Nacional de Pesquisas do Brasil - CNPq: N. 441443/2016-8) and to Projeto Biodiversidade, hidrologia e qualidade das águas de Igarapés na cidade de Santarém-PA (MCTI/CNPQ/Universal 14/2014: N. 458297/2014-3) and JZ received a productivity grant from CNPq (#313183/2014-7).

Supplementary Material

The following online material is available for this article: Table S1 - Deposited material.

References

Publication Dates

History