The Amazon ecoregion is the largest tropical forest in the world and plays an important role in maintaining species diversity by providing food resources, shelter, different microclimates, and ecological processes, which also helps to maintain the structural integrity of habitats (Junk et al., 2007). South of this ecoregion is the transitional area with the Brazilian Savannah (Cerrado) ecoregion, an area that has been highly altered by the expansion of the agricultural frontier (Arc of Deforestation), which has deeply changed the region’s natural characteristics (Brando et al., 2013; Nepstad et al., 2014) and contributed to the loss of physical and biotic integrity in aquatic ecosystems (Cunha et al., 2015; Ilha et al., 2019). The removal of forest cover, especially of the riparian vegetation, facilitates erosion processes and the carrying of sediment, nutrients, and pollutants to water systems (Allan, 2004; Nessimian et al., 2008). Forest loss also changes the flux of organic matter, channel morphology, substrate characteristics, and water physico-chemical characteristics, reducing the integrity and quality of habitats and modifying their distributions (Cunha et al., 2015; Leal et al., 2016; Juen et al., 2016). The changes in landscape and habitat quality also lead to biodiversity loss, both in taxonomic diversity (Montag et al., 2019) and functional diversity (Benone et al., 2020; Leão et al., 2020).
Quantifying the response of stream organisms to environmental change using classical measures of species diversity and abundance (taxonomic diversity) together with a functional trait approach provides the most comprehensive understanding of changes in stream function and integrity (Petchey et al, 2007; Batalha et al., 2010; Luiza-Andrade et al., 2017). The functional approach refers to the morphological, behavioral, and physiological traits related to a species ability to exploit environmental resources, which can be used to understand how local communities are structured (MacArthur, Levins, 1967; Petchey et al., 2007; Leitão et al., 2018). Functional diversity becomes a link between species diversity and environmental change (Cadotte et al., 2011), and can be measured using indexes with various facets (e.g., Villéger et al., 2008). Those indexes describe the distribution of species and their abundances inside the community functional space, which is defined by the functional traits used. Each index covers a facet of functional diversity (richness, evenness, or functional divergence), making them complementary (Mason et al., 2005).
Several taxonomic groups have been used in studies considering functional diversity in response to environmental change in terrestrial ecosystems (e.g., ants, beetles, and birds) and in aquatic ecosystems (e.g., macroinvertebrates, zooplankton, and fish) (Batalha et al., 2010; Barragán et al., 2011; Moreira et al., 2016; Luiza-Andrade et al., 2017). Due to the sensitivity of some fish species to environmental impacts, the taxonomic and functional composition of their communities can respond in different ways, depending on the type, size, time of impact, and the type of ecosystem in which they live (Burress, 2015; Prudente et al., 2017; Teresa, Casatti, 2017; Leitão et al., 2018).
In Amazonian streams, deforestation affects fish assemblages and alters functional identity (Leal et al., 2017; Brejão et al., 2018). Fish species are distributed among various trophic levels and require a wide range of resources for their success in different environments, from rivers to temporary ponds (Brejão et al., 2013). Environments with less riparian forest cover tend to have more homogeneous environments, favoring generalist fish species with feeding strategies that allow them to occur in habitats with anthropogenic changes (Vieira et al., 2015; Zeni et al., 2017).
Although several studies have examined these relationships in the Amazon Forest biome (Leal et al., 2017; Brejão et al., 2018), we understand little about these relationships in the transitional area between the Amazon and the Savannah ecoregion with accelerated deforestation (Trigueiro et al., 2020). We assessed the taxonomic and functional structure of fish assemblages from streams in transitional areas between the Amazon and the Savannah to assess three main questions: (a) How does physical habitat integrity in areas with less riparian vegetation affect species richness, abundance, and composition of fish assemblages?; (b) How does environmental change in areas with less riparian vegetation and low physical habitat integrity affect the functional diversity of fish assemblages?; and (c) How does suppression of riparian vegetation affect stream habitats, and what functional characteristics are related to the physical habitat integrity of streams? Our operational hypothesis is that sites with riparian vegetation suppression will be characterized with low physical habitat integrity, resulting in altered fish fauna composition and functional diversity. Conversion of riparian vegetation should also have a negative impact on species with specific functional traits, especially those groups that depend on the riparian vegetation (Casatti et al., 2012). These changes will decrease species richness, abundance, functional divergence, functional richness, and functional evenness of fish assemblages, since reduced habitat integrity reduces important resources for the species, mainly those used as food (Leitão et al., 2018).
Material and methods
Study area. Nine streams (Fig. 1) were sampled in August 2017 on the Tanguro farm (Municipality of Querência, state of Mato Grosso, Brazil). Streams were selected to cover the variability of riparian and stream characteristics of the study area and included three forested watershed (i.e., reference sites) and six agricultural watersheds (i.e., altered sites). All streams with agricultural activities had a low degree of riparian forest, while the forested sites had extensive and well-preserved riparian vegetation. Therefore, to represent areas with less forest cover, we selected streams with less riparian forest and that were distant from each other in order to avoid spatial autocorrelation. If the impacts are strong, on a local scale, a small number of streams is capable of capturing the effects of vegetation loss on the fish fauna, as observed in other studies (Teresa et al., 2015; Virgilio et al., 2018, Ilha et al., 2018, 2019).
FIGURE 1 | Location of the nine streams sampled in the dry season of 2017, Tanguro Farm, state of Mato Grosso, Brazil.
The Tanguro farm is located in the Savannah-Amazon transitional area, in the upper Xingu River basin, which is part of the larger Amazon River basin. The native vegetation consists of perennial seasonal canopy forests transitioning from rainforests in the north to woody savannahs in the south of the Xingu Basin (Velasquez et al., 2010). According to the 1927 Köppen, Geiger climate classification (Peel et al., 2007), the region’s climate is the Aw type, with average annual precipitation of 1800 mm and an annual average temperature above 25°C (Rocha et al., 2014).
Forest cover. Using the ArcHydro tool, from ArcGis 10.1 software (Esri, 2012), with digital SRTM models of 30-meter resolution acquired from the NASA (National Aeronautics and Space Administration) homepage (https://earthexplorer.usgs.gov/), we delimited the catchments, their respective drainage systems, and the buffer zone. We evaluated land use within the riparian forest in a buffer zone measuring 60 m (30 m of land on each side of the drainage system) in the total catchment of each stream. We used this measure following the Brazilian Forest Code (Federal Law nº 12651/12), which designates riparian forests as Permanent Protection Areas (APP) in fluvial systems up to 10 m wide (Brasil, 2012). Land-use classes were identified using Digital Image Processing (PDI) of a set of Landsat 8 satellite images (USGS, 2018) using the PCI Geomatica and Ecognition programs.
The images were from the same year as the biological and physical habitat sampling (2017). Images were acquired in a geographic coordinate system projected in the geodesic DATUM SIRGAS UTM Zone 21. Next, we calculated the Kappa index according to Landis, Koch (1977), which reflects the quality of the supervised classification, ranging from 0 (low quality) to 1 (excellent quality). This index uses the classification made by the observer. In this image, the index places coordinate points where each land-use was considered. For example, where forest was considered, it seeks points that represent forest, and so forth for other land use classes. These points are compared with classified points to test the veracity of the information by counting how many forest points match the classified forest point.
The identified land use classes in riparian zone were: (a) forest, comprising areas occupied by rainforest; (b) secondary forest, comprising vegetation resulting from natural successional processes after total or partial suppression of the primary vegetation by anthropogenic activities or natural causes; and (c) agriculture areas, comprising mechanized or family farms and the presence of herbaceous plants, in addition to any indicators of homogenization seen in the satellite images (see Fig. S1). Even though it is in a forest-savannah transition area, the riparian vegetation was classified as forest due to its density of cover, a riparian vegetation characteristic.
Habitat. Environmental characterization of the sampled streams was carried out in the dry season, starting in the morning and extending into the afternoon, which is appropriate for our study due to the adequacy of the collection method (Frissell et al., 1986; Jaramillo-Villa, Caramaschi, 2008). In each stream, we delimited a 150 m stretch, which was then divided into 11 transects labeled from “A” (downstream) to “K” (upstream), obtaining 10 equidistant cross-sections of 15 m.
Environmental characterization (Kaufmann et al., 1999; Peck et al., 2006) consisted of 238 metrics including channel morphology, substrate, discharge and substrate, channel unit (water flow types), slope, sinuosity, riparian vegetation cover, woody debris tally, instream fish cover, and human impact (Tab. S2). We also measured water physical-chemical parameters including temperature, pH, conductivity, and dissolved oxygen (OD) at three sections (A, F and K) using a multiparameter probe. Habitat measurements were made within all cross-channel transects and longitudinal sections in each of the nine streams, and values are reported as averages and standard deviations for each stream.
Fish fauna. Fish sampling started in the morning and continued throughout the afternoon. Each stream was sampled only once. The sampling effort was six hours in each stream, with three being the number of collectors in each cross-section. Fish specimens were collected using circular 55 cm-diameter hand nets (sieves) and 40 cm-diameter dip nets (“puçá”in Portuguese), both with a 1 mm metallic mesh. We applied both methods simultaneously, always using two hand nets and dip nets in each section. The collected specimens were fixed in a 10% formalin solution and then transferred to 70% ethanol after 48 hours. Fish identification was carried out with the help of specialists and identification keys (Géry, 1977; Kullander, 1986). Species with taxonomic uncertainties were typed as morphospecies, taking care to standardize the nomenclature. The collected material was deposited in the Museu Paraense Emílio Goeldi (MPEG), Belém, state of Pará, Brazil. The list of catalog numbers for the species are listed in Tab. S6.
Functional traits. Fish specimens were classified into 11 trophic groups (Brejão et al., 2013) according to the feeding strategies of each species combined with their spatial distribution (stream margin, main channel, and water column). We also classified species into two life-history groups: 1) equilibrium strategy, with high juvenile survival rate, parental care, and large eggs, and 2) opportunistic strategy, characterized by small adult size, short generation period, high reproductive effort, and extended breeding seasons (Zeug, Winemiller, 2007). Feeding and reproductive strategies of individuals allow us to understand which factors determine the presence of a given species in a site, favoring not only their survival, but their perpetuation. Environmental changes modify these natural factors, causing species to leave the site looking for new resources and driving other species to occupy available resources the new trophic niche available (Zeug, Winemiller, 2007; Uieda, Pinto, 2011; Montag et al., 2019). Data for species were taken from the literature, and in the absence of information for a species, we used data for the genus (see S3).
Data Analyses. We first evaluated if there was spatial autocorrelation among the sampling units by calculating the fluvial distances among sampling sites. Using the ArcHydro tool, from ArcGis 10.1 software (Esri, 2012), with digital SRTM models of 30 meters resolution acquired from the NASA homepage (https://earthexplorer.usgs.gov/), we delimited the catchments and their respective drainage systems in order to verify if there was pseudo-replication among the data sampled in the streams. Next, a Principal Coordinates of Neighbor Matrices (PCNM) analysis was performed to transform the fluvial distances among the sites into a Euclidean distance matrix (Dray et al., 2006). The species abundance matrix and the spatial matrix generated by the PCNM were used to perform a Redundancy Analysis (RDA). The RDA result showed there was no correlation among the sampling sites (df = 2; F = 0.94; p = 0.552).
Given that land use variables are often autocorrelated, we performed a Principal Component Analysis (PCA), where the percentage of each land-use variable was transformed in a Euclidean distance matrix. Then, using the Broken-stick criterion, we selected the axes with the highest explanation power to represent forest cover (Legendre, Legendre, 2012). PCA was performed using the ‘princomp’ function of the Vegan package (Oksanen et al., 2017), in the Rsoftware(R Development Core Team, 2013).
Development of the Physical Integrity Index (PII). To calculate the Physical Integrity Index (PII) (Barbour et al., 1996), we used metrics from the physical habitat assessment in the three forested reference sites, and six agricultural sites. First, variables with more than 80% of values equal to 0 were removed. Second, we selected variables with reduced or no overlap between forested and agricultural watersheds based on Box-and-Whisker plots as indicators of the degree of habitat disturbance. Then, we used Student’s t-test (p <0.05), with the assumptions of normality and homoscedasticity, to verify whether the distributions were statistically different between both test and reference treatments (Zar, 2010). Metrics that did not show a statistically significant distribution were excluded. Metrics considered sensitive to environmental disturbances were submitted to a Spearman correlation analysis between pairs of metrics (rS ≥ 0.70 and p < 0.05). Thus, when two metrics were correlated, only the one with the greatest relevance related to habitat characteristics was selected to determine the structure of aquatic communities (Baptista et al., 2007).
The criteria for scoring the index were based on Barbour et al. (1996), using the three reference forest sites as a baseline. Metrics that decrease with impact intensification obtained a positive score, with the 25th percentile as the divisor of the category with the highest score, called the “upper limit”. For metrics that increase with impact intensification, with a negative score, the 75th percentile was used as the divisor of the category with the highest score.
The upper score represents 75% or more of the conditions found in reference sites and the lower score corresponds to the values that did not enter the distribution of these areas. The first case is given a score of five; in the second case, a score of one; intermediate conditions, values corresponding to 25% of the conditions found in reference sites, received a score of three. The index was expressed numerically by measuring average score values for each stream. We tested the hypothesis that forest cover reduction decreases the level of physical condition using linear regressions (Zar, 2010). For the analyses, the assumptions of homogeneity and normality of the residuals were analyzed. As predictor variables, we used the PCA axis selected by the Broken-Stick criterion, which represents forest cover, and as response variables, we used the Physical Integrity Index (PII) values for each stream.
Taxonomic diversity. To test the hypothesis that lower forest cover levels and lower physical conditions will decrease fish species richness and abundance, we performed simple linear regressions, where the assumptions of normality and homogeneity were reached (Zar, 2010). For these analyses, the predictive variables were the stream physical condition (represented by the PII) and forest cover (represented by the first PCA axis in the landcover analysis). We chose the latter because it represents the continuous gradient of forest cover, without stream categorization. Species richness and total abundance were the response variables.
To evaluate how fish species composition is distributed along the gradient of forest cover, we performed a Principal Coordinate Analysis (PCoA), using a Bray-Curtis distance matrix, where the abundance data was Hellinger-transformed. The PCoA was performed using the Vegan package (Oksanen et al., 2017), in the R software (R Development Core Team, 2013).
Functional diversity. We calculated three descriptive indexes of functional diversity (Villéger et al., 2008) (Tab. S4) using a categorical matrix with functional traits of life history and trophic groups together with the species abundance matrix:
i) Functional Richness (FRic) represents the volume occupied by the community in the functional space composed by community traits, allowing us to quantify the variation of each functional trait in each community (Mouchet et al., 2010). Decreased FRic values after disturbance indicate that disturbance acts as an environmental filter, selecting species with similar functional traits and that occupy a more restricted portion of the functional volume (Pakeman et al, 2011; Mouillot et al., 2013); ii) Functional Evenness (FEve) estimates the consistency with which species abundances are distributed in the functional space and can represent the efficiency with which available resources are used (Mouchet et al., 2010). With this index, we can infer if species are nested or dispersed evenly in the functional space. Under the influence of an environmental filter, coexisting species tend to be functionally similar, showing higher species grouping and consequently low heterogeneity in filling the functional space, resulting in low FEve values (Mouillot et al., 2013), and the iii) Functional Divergence (FDiv) assesses how much the taxa differ in the functional space. Higher FDiv values indicate a high level of niche differentiation between species, and thus low interspecific competition for resources. This index ranges from zero (when species are highly abundant and closer to the assemblage centroid in the occupied volume) to one (when highly abundant species are distant from the assemblage centroid) (Mouillot et al., 2013). All indexes were calculated using the FD package, using the dbFD function, which uses Principal Coordinate Analysis (PCoA) to scale samples on principal axes according to the Gower distance. The resulting axes are used as traits to calculate the indices. The indexes were calculated using R programming software (Laliberté, Legendre 2010; R Development Core Team, 2011; Laliberté et al., 2014).
We performed linear regressions (Zar, 2010), following the assumptions of normality and homogeneity of the residuals, for each descriptive index of functional diversity. We tested the hypothesis that lower values of physical condition and percentage of forest cover will lead to low values of richness, divergence and functional evenness of fish assemblages. We used the values obtained from the PII and the first PCA axis for land cover selected by the Broken-stick criterion as predictive variables. Additionally, we created ordination graphs, with the PCA axis representing the x axis and the abundance of trophic groups representing the y axis, to verify the organization of trophic groups related to forest cover loss. All analyses were conducted in the R programing environment (R Development Core Team, 2011).
Forest cover. The PCA analysis of the supervised classification of land use in the riparian buffer organized the sites according to the amount of forest cover on the first axis (84% of the variation; Fig. 2; Tab. 1). In relation to the Kappa index, which evaluates the quality of the classification of land-uses, we verified that the agricultural environments had excellent quality of classifications (1.0), followed by secondary forest (0.6) and forest (0.3). Even the value of 0.3 is classified by the literature as a good quality of interpretation (Niculescu et al., 2018); this value demonstrates the differences between high and low canopy that makes interpretation difficult, even in high resolution images.
FIGURE 2 | Principal component analysis performed with the percentages of land use in 60 m riparian buffers in the catchment of nine streams located in the Tanguro Farm, Municipality of Querência, state of Mato GrossoT. Sites 1, 2, and 3 overlapped, and were rearranged for better visualization. Sgrid: Site.
TABLE 1 | Land-use classes identified in riparian buffers in the catchment of nine streams in the Tanguro Farm, Municipality of Querência, state of Mato Grosso, and their correlations to Axis I of the PCA.
|Axis I||Axis II|
|% of explanation||84%||0.89%|
Habitat. We used five metrics to construct the stream Physical Integrity Index (PII): 1) mean depth; 2) width and depth ratio in the section; 3) volume of woody debris in the stream/m2 – size class 2 (all woody debris larger than 0.3 m diameter); 4) mean herbaceous ground cover; and 5) cultivation proximity index (Tab. 2; Fig. 3).The hypothesis that loss of riparian vegetation cover negatively affects the physical condition of streams was corroborated (r² = 0.70, p = 0.004) (Fig. 4). In general, the more forested streams in this study had lower mean depth (t = 2.42, p = 0.04), higher width and depth ratio in the section (t = 2.94, p = 0.02), lower cultivation proximity index (t = 2.95, p= 0.02), higher volume of wood debris in the stream (t = 2.64, p = 0.03), and lower mean herbaceous ground cover (t = 2.41, p = 0.04) compared to streams with less forest cover (Fig. 3).
FIGURE 3 | Box-and-whiskers plots of the five metrics comprising the Physical Integrity Index.
FIGURE 4 | Relationship between forest cover and the physical integrity index of streams located in the Tanguro Farm, Municipality of Querência, state of Mato Grosso. Forest cover is represented by the Axis I of the PCA performed with land use variables obtained from a 60 m buffer. Sites 1, 2 and 3 overlapped, and were rearranged for better visualization. Sgrid: Site.
TABLE 2 | Values of the metrics selected to calculate the scores of the Physical Integrity Index in the Amazon, state of Mato Grosso. PII: Physical Integrity Index; MEAN_TD: mean thalweg depth (cm); WDR_SEC: width and depth ratio in the section; VOL_WDS: volume of woody debris in the stream /m2 –size class 2; MEAN_HERBS: mean ground cover herbs; CULTIV_PI: cultivation proximity index.
Fish fauna. We collected 2,943 specimens, comprised of 29 species, 27 genera, and six orders (Tab. 3). Characiformes was the most representative order in terms of abundance, with 56% of individuals. The most abundant species were Moenkhausia phaeonota Fink, 1979 (27%), followed by Hemigrammus sp1. (24%), and Melanorivulus megaroni (Costa, 2010) (15%). The most representative (in terms of abundance) trophic functional groups were diurnal channel drift feeders (29%), diurnal backwater drift feeders (24%), and diurnal surface pickers (23%). For the life-history groups, 78% of individuals were opportunistic strategists.
TABLE 3 | Species list classified according to their trophic functional groups and life history. Species found in nine streams sampled in 2017, in the Tanguro Farm, Municipality of Querência, state of Mato Grosso. NIB: Nibblers; DBDF: Diurnal backwater drift feeders; DCDF: Diurnal channel drift feeders; ASP: Ambush and stalking predators; DSP: Diurnal surface pickers; NIP: Nocturnal invertebrate pickers; PB: Pickers and browsers; CNBP: Crepuscular to nocturnal bottom predators; GRU: Grubbers; GRA: Grazers; CNDF: Crepuscular to nocturnal drift feeders; OS: Opportunistic strategy; ES: Equilibrium strategy.
|Taxon||Species code||Total N||Trophic group||Life-history|
|Leporinus britskii Feitosa, Santos & Birindelli, 2011||sp1||2||NIB||OS|
|Moenkhausia oligolepis (Günther, 1864)||sp4||11||DCDF||OS|
|Moenkhausia phaeonota (Fink, 1979)||sp5||789||DCDF||OS|
|Characidium zebra (Eigenmann, 1909)||sp7||61||ASP||OS|
|Copella arnoldi (Regan, 1912)||sp9||159||DSP||OS|
|Melanorivulus megaroni (Costa, 2010)||sp10||453||DSP||OS|
|Gymnotus carapo Linnaeus, 1758||sp12||79||NIP||ES|
|Gymnotus coropinae Hoedeman, 1962||sp13||3||NIP||ES|
|Hypopomus brevirostris (Steindachner, 1868)||sp14||47||NIP||ES|
|Gymnorhamphichthys rondoni (Miranda Ribeiro, 1920)||sp15||32||NIP||ES|
|Hypopygus lepturus Hoedeman, 1962||sp16||74||NIP||ES|
|Eigenmannia trilineata (Lopes & Castello, 1966)||sp17||54||NIP||ES|
|Sternopygus macrurus (Bloch & Schneider, 1801)||sp18||4||NIP||ES|
|Aequidens tetramerus (Heckel, 1840)||sp19||178||PB||ES|
|Crenicichla inpa (Ploeg, 1991)||sp20||7||ASP||ES|
|Tatia aff. aulopygia (Kner, 1858)||sp21||8||CNBP||ES|
|Helogenes marmoratus Günther, 1863||sp23||52||CNDF||ES|
|Rhamdia aff. quelen (Quoy & Gaimard, 1824)||sp26||1||CNBP||ES|
|Synbranchus marmoratus Bloch, 1795||sp29||7||CNBP||ES|
Taxonomic Diversity. The hypotheses that lower forest cover in the riparian zone correlates with lower stream fish abundance (r² = 0.175, p = 0.262) and species richness (r² = 0.041, p = 0.602) were not supported. Additionally, we found no correlation between stream physical condition (PII) and abundance (r² = 0.149, p = 0.304) or richness (r² = 0.003; p = 0.891) of fish assemblages (Fig. S5). The first two axes of the PCoA explained 62% of the variation in the data (Fig. 5; Tab. S6). The results from the PCoA show that species were grouped according to different levels of forest cover as represented by the color gradient, with lighter colors representing sample points with less forest cover.
Functional diversity. As hypothesized, we found a positive correlation between forest cover in the riparian zone and functional richness (r² = 0.479, p = 0.03) (Fig. 6). However, contrary to our hypothesis we did not find a positive correlation with functional evenness (r² = 0.111, p = 0.379) or divergence (r² = 0.211, p = 0.214). Additionally, our hypothesis that reduced habitat integrity would correlate with lower values of functional richness (r² = 0.037, p = 0.618), functional divergence (r² = 0.238, p= 0.183), and functional evenness (r² = 0.262, p = 0.676) of fish assemblages was not supported.
The graph based on the distribution of functional groups in response to forest cover loss showed that groups such as ambush and stalking predators, diurnal surface pickers (opportunistic life history strategists), and pickers and browsers (equilibrium life history strategists) are associated more with environments with low forest cover values, while groups comprised of crepuscular to nocturnal drift feeders are more associated with environments containing high forest cover (Fig. 7).
FIGURE 5 | Relationship between species composition and the first PCA axis, ordinated by a Principal Coordinate Analysis. Lighter colors represent sample points with less forest cover. The species corresponding to the codes are in Tab. 2.
FIGURE 6 | Relationship between forest cover and the functional richness index observed in streams located in the Tanguro Farm, municipality of Querência, state of Mato Grosso. Forest cover is represented by the Axis I of the PCA performed with land use variables obtained from a 60 m buffer. Sites 1, 2 and 3 overlapped and were rearranged for better visualization.
FIGURE 7 | Distribution of functional groups related to forest cover loss in Amazonian streams, state of Mato Grosso. OS: Opportunistic strategy; ES: Equilibrium strategy. Forest cover values vary from green (more forest cover) to yellow (less forest cover).
Our results indicate that forest cover loss in the 60 m riparian buffer resulted in lower physical habitat quality (indicated by the Physical Integrity Index – PII) and functional richness. However, at the scale of our study, forest cover loss did not affect taxonomic richness, abundance, divergence, and functional evenness of fish species. The physical condition of the streams also did not influence the abundance, richness, functional richness, functional divergence, and functional evenness of the fish fauna.
Several authors have studied the effects of deforestation on fish fauna diversity in different environmental contexts (Ferreira et al., 2018; Ilha et al., 2018, 2019; Leitão et al., 2018; Montag et al., 2019). The results found in these studies tend to vary due to different factors such as the degree of impact, type of ecosystem, local characteristics, fauna diversity, and others. Our results diverged somewhat from previous findings, perhaps because the area we studied has a high percentage of preserved vegetation along the river network and a lower percentage of agricultural areas, which may help maintain certain natural characteristics of the ecosystem and mitigate the effects deforestation in the area (Paula et al., 2021). This abundance of preserved vegetation may also be helping the community to recover after disturbances from agricultural practices by retaining resource availability, especially habitat and nutrient resources that favor specialist species (Karr, 1981; Teresa, Casatti, 2010).
We observed that even with the presence of the forest extension in the sampled sites, it is still possible to detect changes in their physical condition. The mean herbaceous ground cover was higher and the volume of woody debris was lower at the streams in agricultural watersheds (i.e., altered sites). Deforestation likely makes these sites susceptible to the proliferation of herbaceous plants due to increased solar incidenceand lowered competition for resources; furthermore, it reduces the amountof allochthonous material in streams, as well as the supply of woody debris that are essential elements for communities of small water bodies that use them as shelter (Roth et al., 1996). The variable width and depth ratio also differed between reference and altered sites, becoming narrower and deeper in the altered sites, due to the removal of natural structures in the stream channel and resulting in slow water flow (Neill et al., 2006; Deegan et al., 2011).
Changes in the prevalence of certain functional groups between sites occurred possibly due to changes in the physical characteristics of streams. We observed that opportunistic life history strategist species were highly represented altered sites, as they can explore various environments. Groups such as crepuscular to nocturnal drift feeders, grubbers, and nibblers had a significant reduction across the forest gradient. Thus, the ecosystem functional volume is reduced since the available niches are underutilized, or contribute less efficiently toward ecosystem functions, such as the exchange of energy and matter or substrate bioturbation (Mason et al., 2005; Leduc et al., 2015).
The most representative functional group (diurnal channel drift feeders) was relatively homogeneous along the forest cover gradient. This distributional pattern may have occurred because this group was mostly composed of species from the Characidae family, which occupy a variety of habitats and have various feeding strategies (Brejão et al., 2013). Nocturnal invertebrate pickers showed little variation along the gradient, possibly because they belong to the order Gymnotiformes, which possess diverse feeding strategies and can make use of different habitat types (Albert, Crampton, 2006). Trophic plasticity suggests that each species can feed on many different prey types, which is not necessarily true for gymnotiforms. Gymnotus Linnaeus, 1758 can be piscivorous, but Gymnorhamphichthys Ellis, 1912 mostly feed on larval invertebrates.
Similar to other studies, grazers, represented by the Loricariidae family, were associated with environments that suffered environmental change (Bojsen, Barriga, 2002). We observed an increase of species from this family in areas with low forest cover. It is possible that additional light, resulting from reduced forest cover, increased the availability of periphyton on aquatic plants, pluricellular algae (epiphyton), and sediment (epilithon) (Sládeèková, 1962), which are important food resources for this group.
The presence of an ambush and stalking predator Characidium zebra Eigenmann, 1909, in sites with lower forest cover may be associated with prey availability, such as aquatic insect larvae (Cetra et al., 2011), which are more available in these types of environments (Juen et al., 2014; Marques et al., 2021). Diurnal surface pickers were also more abundant in environments with less forest cover. This group is represented by individuals of the Rivulidae family, which possess opportunistic feeding habits and are commonly found in shallow waters and deforested streams (Volcan et al., 2011; Ilha et al., 2019). Pickers and browsers showed a clear increase in abundance as the percentage of forest cover decreased. In this study, this group was mostly represented by the Cichlidae family, which, is comprised of some tolerant and generalist species that can present higher species richness in deforested environments and have been associated with sites that have suffered environmental degradation (Bojsen, Barriga, 2002; Burress, 2015; Ferreira et al., 2018).
Crepuscular to nocturnal drift feeders preferred forested streams; in this group, there was a predominance of Helogenes marmoratus Günther, 1863, a species that seeks shelter in the microhabitats between roots and wood in the riparian vegetation as well as in leaf banks, where they hide during the day (Zuanon et al., 2015). Crepuscular to nocturnal bottom predators had a high number of individuals in sites with reduced habitat integrity. The most representative individuals of this group belong to the Heptapteridae family, which has been recorded inhabiting various substrate types, e.g., leaf banks, sand, and groundwater (Zuanon et al., 2015), indicating their capacity to inhabit different environments.
Species with opportunistic reproductive strategies preferred environments with low forest cover values, suggesting that they can colonize disturbed or recently available habitats quickly. However, species with equilibrium strategies had high abundance in environments with intermediate forest levels and were inferred to be adaptive in habitats where resources are limited (Zeug, Winemiller, 2007).
The Amazon-Savannah transitional area is a region rich in biodiversity; however, due to the accelerated advance of deforestation, fish diversity and stream ecosystem functions are at risk. Our study was able to detect the effects of deforestation on the fish fauna, and, even with a low number of samples, it highlights the importance of complying with the Brazilian Forest Code (Brasil, 2012), which requires the creation of Permanent Protection Areas (APP) along the shoreline of any natural watercourse. Despite the observed decrease in functional richness, our results suggest that the presence of forest in the riparian zone, even if it is secondary forest, can buffer the effects of adjacent deforested areas on the taxonomic and functional diversity of the fish fauna, avoiding even greater losses to species diversity. However, we believe that additional studies in the area, with a larger number of samples, can increase the level of confidence in the results presented here.
We thank the Ecology and Conservation Laboratory at the Universidade Federal do Pará (UFPA) for all the support throughout this research, the Instituto of Pesquisa na Amazônia (IPAM) for all the logistic support during the field trips, and the Amaggi group for making it possible to do this research on their property. We also thank the following institutions: Conselho Nacional de Desenvolvimento Científico e Tecnológico – CNPq (PELD 441703/2016–0; site TANG); the NSF DEB (Division of Environmental Biology of the National Science Foundation) 1660034 and 1739724; the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – CAPES (for a master’s scholarship 88887.137921/2017-00 granted to PVF), doctoral scholarship granted to CM (process number: 88882.445579/2019–01), and post-doctoral scholarship granted to PI; the Fundações Estaduais de Amparo à Pesquisa – FAPs; and the British Council – BC – Fundo Newton.
Albert JS, Crampton WGR. Diversity and phylogeny of Neotropical electric fishes (Gymnotiformes). In: Bullock TH, Hopkins CD, Popper AN, Fay RR, editors. Electroreception. New York: Springer; 2006. p.360–409. http://dx.doi.org/10.1007/0-387-28275-0_13
Allan JD. Influence of land use and landscape setting on the ecological status of rivers. Limnetica. 2004; 23(3–4):187–98. Available from: http://www-personal.umich.edu/~dallan/pdfs/limnetica.pdf
Baptista DF, Buss DF, Egler M, Giovanelli A, Silveira P, Nessimian JL. A multimetric index based on benthic macroinvertebrates for evaluation of Atlantic Forest streams at Rio de Janeiro State, Brazil. Hydrobiologia. 2007; 575(83):83–94. http://dx.doi.org/10.1007/s10750-006-0286-x
Barbour MT, Diamond JD, Yoder C. Biological assessment strategies: applications and limitations. In: Grothe DR, Dickson KL, Reed DK, editors. Whole-effluent toxicity testing: An evaluation of methods and predictability of receiving system responses. Florida: Society of Environmental Toxicology and Chemistry Press; 1996. p.245–70.
Barragán F, Moreno CE, Escobar F, Halffter G, Navarrete D. Negative impacts of human land use on dung beetle functional diversity. PloS ONE. 2011; 6(3):e17976. http://dx.doi.org/10.1371/journal.pone.0017976
Batalha MA, Cianciaruso MV, Motta-Junior JC. Consequences of simulated loss of open Cerrado areas to bird functional diversity. Nat Conserv. 2010; 8(1):34–40. http://dx.doi.org/10.4322/natcon.00801005
Benone NL, Soares BE, Lobato CMC, Seabra LB, Bauman D, Montag LFA. How modified landscapes filter rare species and modulate the regional pool of ecological traits? Hydrobiologia. 2020; 1–16. http://dx.doi.org/10.1007/s10750-020-04405-9
Bojsen B, Barriga R. Effects of deforestation on fish community structure in Ecuadorian Amazon streams. Freshw Biol. 2002; 47(11):2246–60. https://doi.org/10.1046/j.1365-2427.2002.00956.x
Brando PM, Coe MT, DeFries R, Azevedo AA. Ecology, economy and management of an agroindustrial frontier landscape in the southeast Amazon. Philoso Trans R Soc Lond B Biol Sci. 2013; 368(1619):20120152.
Brasil. Lei No 12.651 de 25 de maio de 2012 [Internet]. Available from: www.planalto.gov.br/ccivil_03/_Ato2011-2014/2012/Lei/L12651.htm
Brejão GL, Gerhard P, Zuanon J. Functional trophic composition of the ichthyofauna of forest streams in eastern Brazilian Amazon. Neotrop Ichthyol. 2013; 11(2):361–73. http://dx.doi.org/10.1590/S1679-62252013005000006
Brejão GL, Hoeinghaus DJ, Pérez-Mayorga MA, Ferraz SFB, Casatti L. Threshold responses of Amazonian stream fishes to timing and extent of deforestation. Conserv Biol. 2018; 32(4):860–71. http://dx.doi.org/10.1111/cobi.13061
Burress ED. Cichlid fishes as models of ecological diversification: patterns, mechanisms, and consequences. Hydrobiologia. 2015; 748(1):7–27. http://dx.doi.org/10.1007/s10750-014-1960-z
Cadotte MW, Carscadden K, Mirotchnick N. Beyond species: Functional diversity and the maintenance of ecological processes and services. J Appl Ecol. 2011; 48(5):1079–87. http://dx.doi.org/10.1111/j.1365-2664.2011.02048.x
Casatti L, Teresa FB, Gonçalves-Souza T, Bessa E, Manzotti AR, Gonçalves CS, Zeni JO. From forests to cattail: how does the riparian zone influence stream fish? Neotrop Ichthyol. 2012; 10(1):205–14. https://doi.org/10.1590/S1679-62252012000100020
Cetra M, Rondineli G, Souza UP. Resource sharing by two nectobenthic freshwater fish species of streams in the Cachoeira river basin (BA). Biota Neotrop. 2011; 11(2):87–95. Available from: www.biotaneotropica.org.br/v11n2/en/abstract?article+bn01911022011
Cunha EJ, Montag LFA, Juen L. Oil palm crops effects on environmental integrity of Amazonian streams and Heteropteran (Hemiptera) species diversity. Ecol Indic. 2015; 52:422–29. https://doi.org/10.1016/j.ecolind.2014.12.024
Deegan LA, Neill C, Haupert CL, Ballester MVR, Krusche AV, Victoria RL et al. Amazon deforestation alters small stream structure, nitrogen biogeochemistry and connectivity to larger rivers. Biogeochemistry. 2011; 105(1–3):53–74. https://doi.org/10.1007/s10533-010-9540-4
Dray S, Legendre P, Peres-Neto PR. Spatial modelling: A comprehensive framework for principal coordinate analysis of neighbor matrices (PCNM). Ecol Modell. 2006; 196(3–4):483–93. https://doi.org/10.1016/j.ecolmodel.2006.02.015
Environmental Systems Research Institute (ESRI). ArcGIS 10.1 for Desktop [Internet]. Redlands: Environmental Systems Research Institute Inc; 2012.
Ferreira MC, Begot TO, Prudente BS, Juen L, Montag LFA. Effects of oil palm plantations on habitat structure and fish assemblages in Amazon streams. Environ Biol Fish. 2018; 101(4):547–62. https://doi.org/10.1007/s10641-018-0716-4
Frissell CA, Liss WJ, Warren CE, Hurley MD. A hierarchical framework for stream habitat classification: viewing streams in a watershed context. Environ Manage. 1986; 10(2):199–214. https://doi.org/10.1007/BF01867358
Géry J. Characoids of the World. Neptune City: T.F.H. Publications; 1977.
Ilha P, Rosso S, Schiesari L. Effects of deforestation on headwater stream fish assemblages in the Upper Xingu River Basin, Southeastern Amazonia. Neotrop Ichthyol. 2019; 17(1):e180099. http://dx.doi.org/10.1590/1982-0224-20180099
Ilha P, Schiesar L, Yanagawa FI, Jankowski K, Navas CA. Deforestation and stream warming affect body size of Amazonian fishes. PLoS ONE. 2018; 13(5):e0196560. https://doi.org/10.1371/journal.pone.0196560
Jaramillo-Villa U, Caramaschi ÉP. Índices de integridade biótica usando peixes de água doce: Uso nas regiões tropical e subtropical. Oecol Bras. 2008; 12(13):442–62. Available from: https://dialnet.unirioja.es/descarga/articulo/2882875.pdf
Juen L, Cunha EJ, Carvalho FG, Ferreira MC, Begot TO, Andrade AL et al. Effects of oil palm plantations on the habitat structure and biota of streams in eastern Amazon. River Res Appl. 2016; 32(10):2081–94. https://doi.org/10.1002/rra.3050
Juen L, Oliveira-Junior JMB, Shimano Y, Mendes TP, Cabette HSR. Composição e riqueza de Odonata (Insecta) em riachos com diferentes níveis de conservação em um ecótone Cerrado-Floresta Amazônica. Acta Amazon. 2014; 44(2):175–84. https://doi.org/10.1590/S0044-59672014000200008
Junk WJ, Soares MGM, Bayley PB. Freshwater fishes of the Amazon River basin: Their biodiversity, fisheries, and habitats. Aquat Ecosyst Health Manag. 2007; 10(2):153–73. https://doi.org/10.1080/14634980701351023
Karr JR. Assessment of biotic integrity using fish communities. Fisheries (Bethesda). 1981; 6(6):21–27. Kaufmann PR, Levine P, Robison EG, Seeliger C, Peck DV. Quantifying physical habitat in wadeable streams [Internet]. Washington: U.S. Environmental Protection Agency; 1999. Available from: https://archive.epa.gov/emap/archive-emap/web/pdf/phyhab.pdf
Kullander SO. Cichlid fishes of the Amazon River drainage of Peru. Stockholm: Swedish Museum of Natural History; 1986. Available from: http://svenkullander.se/publications/Kullander_Peru_1986.pdf
Laliberté E, Legendre P. A distance-based framework for measuring functional diversity from multiple traits. Ecology. 2010; 91(1):299–305. https://doi.org/10.1890/08-2244.1
Laliberté E, Zemunik G, Turner BL. Environmental filtering explains variation in plant diversity along resource gradients. Science. 2014; 345(6204):1602–05. https://doi.org/10.1126/science.1256330
Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977; 33(1):159–74. https://doi.org/10.2307/2529310
Leal CG, Pompeu PS, Gardner TA, Leitão RP, Hughes RM, Kaufmann PR et al. Multi-scale assessment of human-induced changes to Amazonian instream habitats. Landsc Ecol. 2016; 31(8):1725–45. http://dx.doi.org/10.1007/s10980-016-0358-x
Leal CG, Barlow J, Gardner TA, Hughes RM, Leitão RP, Nally RM et al. Is environmental legislation conserving tropical stream faunas? A large-scale assessment of local, riparian and catchment-scale influences on Amazonian fish. J Appl Ecol. 2017; 55(3):1312–16. https://doi.org/10.1111/1365-2664.13028
Leão H, Siqueira T, Torres NR, Montag LFA. Ecological uniqueness of fish communities from streams in modified landscapes of Eastern Amazonia. Ecol
Indic. 2020; 111:106039. https://doi.org/10.1016/j.ecolind.2019.106039
Leduc AOHC, Silva EM, Rosenfeld JS. Effects of species vs. functional diversity: Understanding the roles of complementarity and competition on ecosystem function in a tropical stream fish assemblage. Ecol Indic. 2015; 48:627–35. https://doi.org/10.1016/j.ecolind.2014.09.027
Legendre P, Legendre L. Numerical Ecology. Oxford: Elsevier; 2012.
Leitão RP, Zuanon J, Mouillot D, Leal CG, Hughes RM, Kaufmann PR et al. Disentangling the pathways of land use impacts on the functional structure of fish assemblages in Amazon streams. Ecography. 2018; 41(1):219–32. https://dx.doi.org/10.1111/ecog.02845
Luiza-Andrade A, Montag LFA, Juen L. Functional diversity in studies of aquatic macroinvertebrates community. Scientometrics. 2017; 111(3):1643–56. https://doi.org/10.1007/s11192-017-2315-0
Macarthur R, Levins R. The limiting similarity, convergence, and divergence of coexisting species. Am Nat. 1967; 101(921):377–85. Available from: http://max2.ese.u-psud.fr/epc/conservation/PDFs/HIPE/MacArthur1967.pdf
Marques NCS, Jankowski KJ, Macedo MN, Juen L, Luiza-Andrade A, Deegan LA. Riparian forests buffer the negative effects of cropland on macroinvertebrate diversity in lowland Amazonian streams. Hydrobiologia. 2021; 848:3503–20. https://doi.org/10.1007/s10750-021-04604-y
Mason NWH, Mouillot D, Lee WG, Wilson JB. Functional richness, functional evenness and functional divergence: the primary components of functional diversity. Oikos. 2005; 111(1):112–18. https://doi.org/10.1111/J.0030-1299.2005.13886.X
Montag LFA, Winemiller KO, Keppeler FW, Leão H, Benone NL, Torres NR et al. Land cover, riparian zones and instream habitat influence stream fish assemblages in the eastern Amazon. Ecol Freshw Fish. 2019; 28(2):317–29. https://doi.org/10.1111/eff.12455
Moreira FWA, Leite MGP, Fujaco MAG, Mendonça FPC, Campos LP, Eskinazi-Sant’Anna EM. Assessing the impacts of mining activities on zooplankton functional diversity. Acta Limnol Bras. 2016; 28(107):1–19. https://dx.doi.org/10.1590/S2179-975X0816
Mouchet MA, Villéger S, Mason NWH, Mouillot D. Functional diversity measures: An overview of their redundancy and their ability to discriminate community assembly rules. Funct Ecol. 2010; 24(4):867–76. https://doi.org/10.1111/j.1365-2435.2010.01695.x
Mouillot D, Graham NAJ, Villéger S, Mason NWH, Bellwood DR. A functional approach reveals community responses to disturbances. Trends Ecol Evol. 2013; 28(3):167–77. https://doi.org/10.1016/j.tree.2012.10.004
Neill C, Deegan LA, Thomas SM, Haupert CL, Krusche AV, Ballester VM et al. Deforestation alters the hydraulic and biogeochemical characteristics of small lowland Amazonian streams. Hydrol Process. 2006; 20(12):2563–80. https://doi.org/10.1002/hyp.6216
Nepstad D, McGrath D, Stickler C, Alencar A, Azevedo A, Swette B et al. Slowing Amazon deforestation through public policy and interventions in beef and soy supply chains. Science. 2014; 344(6188):1118–23. https://doi.org/10.1126/science.1248525
Nessimian JL, Venticinque EM, Zuanon J, Marco P, Jr., Gordo M, Fidelis L et al. Land use, habitat integrity, and aquatic insect assemblages in Central Amazonian streams. Hydrobiologia. 2008; 614(117):117–31. https://doi.org/10.1007/s10750-008-9441-x
Niculescu S, Billey A, Talab-Ou-Ali H, Jr. Random forest classification using Sentinel-1 and Sentinel-2 series for vegetation monitoring in the Pays de Brest (France) [Internet]. Berlin: SPIE Remote Sensing. Remote Sensing for Agriculture, Ecosystems, and Hydrology XX. 2018; 10783:1078305. https://doi.org/10.1117/12.2325546
Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, Mcglinn D et al. Vegan: Community Ecology Package [Internet]. R package; 2017. Available from: http://www.R-project.org
Pakeman RJ, Lennon JJ, Brooker RW. Trait assembly in plant assemblages and its modulation by productivity and disturbance. Oecologia. 2011; 167(1):209–18. https://doi.org/10.1007/s00442-011-1980-6
Paula FR, Leal CG, Leitão RP, Ferraz SFB, Pompeu PS, Zuanon JAS et al. The role of secondary riparian forests for conserving fish assemblages in eastern Amazon streams. Hydrobiologia. 2021. https://doi.org/10.1007/s10750-020-04507-4
Peck DV, Herlihy AT, Hill BH, Hughes RM, Kaufmann PR, Klemm DJ et al. Environmental monitoring and assessment program: Surface waters western pilot study: Field operations manual for wadeable streams. EPA 620/ R-06/003. Washington: Environmental Protection Agency; 2006.
Peel MC, Finlayson BL, McMahon TA. Updated world map of the Köppen-Geiger climate classification. Hydrol Earth Syst Sci. 2007; 11:1633–44. Available from: https://hal.archives-ouvertes.fr/hal-00305098/document
Petchey OL, Evans KL, Fishburn IS, Gaston KJ. Low functional diversity and no redundancy in British avian assemblages. J Anim Ecol. 2007; 76(5):977–85. https://doi.org/10.1111/j.1365-2656.2007.01271.x
Prudente BS, Pompeu PS, Juen L, Montag LFA. Effects of reduced-impact logging on physical habitat and fish assemblages in streams of Eastern Amazonia. Freshw Biol. 2017; 62(2):303–16. https://doi.org/10.1111/fwb.12868
R Development Core Team. R: A language and environment for statistical computing [Internet]. Vienna: R Foundation for Statistical Computing; 2013. Available from: https://www.R-project.org/
R Development Core Team. R: A language and environment for statistical computing [Internet]. Vienna: R Foundation for Statistical Computing; 2011. Available from: https://www.R-project.org/
Rocha W, Metcalfe DB, Doughty CE, Brando P, Silvério D, Halladay K et al. Ecosystem productivity and carbon cycling in intact and annually burnt forest at the dry southern limit of the Amazon rainforest (Mato Grosso, Brazil). Plant Ecol Divers. 2014; 7(1–2):25–40. https://doi.org/10.1080/17550874.2013.798368
Roth NE, Allan JD, Erickson DL. Landscape influences on stream biotic integrity assessed at multiple spatial scales. Landsc Ecol. 1996; 11:1141–56. https://doi.org/10.1007/BF02447513
Sládeèková A. Limnological investigation methods for the periphyton (“Aufwuchs”) Community. Bot Rev. 1962; 28:286–350. https://doi.org/10.1007/BF02860817
Teresa FB, Casatti L. Importância da vegetação ripária em região intensamente desmatada no sudeste do Brasil: um estudo com peixes de riacho. Panam J Aquat Sci. 2010; 5(3):444–53. Available from: https://hdl.handle.net/11449/22556
Teresa FB, Casatti L. Trait-based metrics as bioindicators: Responses of stream fish assemblages to a gradient of environmental degradation. Ecol Indic. 2017; 75:249–58. https://doi.org/10.1016/j.ecolind.2016.12.041
Teresa FB, Casatti L, Cianciaruso MV. Functional differentiation between fish assemblages from forested and deforested streams. Neotrop Ichthyol. 2015; 13(2):361–70. https://dx.doi.org/10.1590/1982-0224-20130229
Trigueiro WR, Nabout JC, Tessarolo G. Uncovering the spatial variability of recent deforestation drivers in the Brazilian Cerrado. J Environ Manage. 2020; 275:111243. https://dx.doi.org/10.1016/j.jenvman.2020.111243
Uieda VS, Pinto TLF. Feeding selectivity of ichthyofauna in a tropical stream: Space-time variations in trophic plasticity. Community Ecol. 2011; 12(1):31–39. https://doi.org/10.1556/ComEc.12.2011.1.5
U. S. Geological Survey (USGS). Earth Explorer. [Internet]. United States. 2018. Available from: https://earthexplorer.usgs.gov/
Velasquez C, Alves HQ, Bernasconi P. Fique por dentro: a bacia do rio Xingu em Mato Grosso. São Paulo: Instituto Socioambiental; 2010. Available from: https://www.socioambiental.org/sites/blog.socioambiental.org/files/publicacoes/10395.pdf
Vieira TB, Dias-Silva K, Pacífico ES. Effects of riparian vegetation integrity on fish and heteroptera communities. Appl Ecol Environ Res. 2015; 13(1):53–65. https://doi.org/10.15666/aeer/1301_053065
Villéger S, Mason NWH, Mouillot D. New multidimensional functional diversity indices for a multifaceted framework in functional ecology. Ecology. 2008; 89(8):2290–301. https://doi.org/10.1890/07-1206.1
Virgilio LR, Ramalho WP, Silva JCBS, Suçuarana MS, Brito CH, Vieira LJS. Does riparian vegetation affect fish assemblage? A longitudinal gradient analysis in three Amazonian streams. Acta Sci Biol Sci. 2018; 40(1):e42562. https://doi.org/10.4025/actascibiolsci.v40i1.42562
Volcan MV, Fonseca AP, Robaldo RB. Reproduction of the threatened Annual Killifish Austrolebias nigrofasciatus (Cyprinodontiformes: Rivulidae), confined in a natural environment. J Threat Taxa. 2011; 3(6):1864–67. https://doi.org/10.11609/JoTT.o2575.1864-7
Zar JH. Biostatistical Analysis. Upper Saddle River: Prentice-Hall; 2010.
Zeni JO, Hoeinghaus DJ, Casatti L. Effects of pasture conversion to sugarcane for biofuel production on stream fish assemblages in tropical agroecosystems. Freshw Biol. 2017; 62(12):2026–38. https://doi.org/10.1111/fwb.13047
Zeug SC, Winemiller KO. Ecological correlates of fish reproductive activity in floodplain rivers: a life-history-based approach. Can J Fish Aquat Sci. 2007; 64(10):1291–301. https://doi.org/10.1139/F07-094
Zuanon J, Mendonça FP, Espírito-Santo HMV, Dias MS, Galuch AV, Akama A. Guia de peixes da Reserva Ducke – Amazônia Central. Manaus: Editora INPA; 2015. Available from: https://www.researchgate.net/publication/299459227_Guia_de_Peixes_da_Reserva_Ducke
 Programa de Pós-graduação em Ecologia, Universidade Federal do Pará, Rua Augusto Corrêa, 01, Guamá, 66075-110 Belém, PA, Brazil. (PVF) firstname.lastname@example.org (corresponding author), (CM) email@example.com.
 Laboratório de Ecologia e Conservação, Instituto de Ciências Biológicas, Rua Augusto Corrêa, 01, Guamá, 66075-110 Belém, PA, Brazil. (LFAM) firstname.lastname@example.org, (ATN) email@example.com.
 Instituto de Pesquisa Ambiental da Amazônia, Rua Horizontina, 104, 78640-000 Canarana, MT, Brazil. firstname.lastname@example.org.
 Programa de Pós-graduação em Ecologia Aquática e Pesca – Universidade Federal do Pará, Rua Augusto Corrêa, 01, Guamá, 66075-110 Belém, PA, Brazil. email@example.com.
 Woods Climate Science Center, Woods Hole Road, 149, 02540-1644 Falmouth, MA, United State. firstname.lastname@example.org.
 Laboratório de Ecologia – Universidade Federal do Pará, Rua Coronel José Porfírio, 2515, São Sebastião, 68372-040 Altamira, PA, Brazil. email@example.com.
Pâmela V. Freitas: Data curation, Formal analysis, Investigation, Methodology, Software, Supervision, Validation, Visualization, Writing-original draf, Writing-review and editing.
Luciano F. A. Montag: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing-original draf, Writing-review and editing.
Paulo Ilha: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing-original draf, Writing-review and editing.
Naiara R. Torres: Formal analysis, Methodology, Visualization, Writing-review and editing.
Calebe Maia: Conceptualization, Data curation, Formal analysis, Methodology, Software, Visualization, Writing-review and editing.
Linda Deegan: Conceptualization, Funding acquisition, Investigation, Project administration, Resources, Supervision, Visualization, Writing-review and editing.
Anderson Thiago Nascimento: Visualization, Writing-review and editing.
Karina Dias Silva: Conceptualization, Formal analysis, Investigation, Methodology, Software, Supervision, Validation, Visualization, Writing-original draf, Writing-review and editing.
CThe sampling and transportation of the biological material were authorized by the Biodiversity Authorization and Information System (SISBIO), under the Permanent License number 8878–1 to collect zoological material. The biological material is filed under the Animal Use Ethics Commission – CEUA/UFPA nº 8293020418 (ID 000954), following the rules issued by the National Council for the Control of Animal Experimentation (CONCEA).
The authors declare no competing interests.
How to cite this article
Freitas PV, Montag LFA, Ilha P, Torres NR, Maia C, Deegan L, Nascimento AT, Silva KD. Local effects of deforestation on stream fish assemblages in the Amazon-Savannah transitional area. Neotrop Ichthyol. 2021; 19(3):e210098. https://doi.org/10.1590/1982-0224-2021-0098
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
Creative Commons CC-BY 4.0
© 2021 The Authors.
Diversity and Distributions Published by SBI
August 29, 2021 by Priscila Camelier
Submitted December 2, 2020
Epub October 18, 2021