Amazon floodplain fish communities: Habitat ...

Biological Conservation 195 (2016) 118–127

Contents lists available at ScienceDirect

Biological Conservation journal homepage: www.elsevier.com/locate/bioc

Discussion

Amazon floodplain fish communities: Habitat connectivity and conservation in a rapidly deteriorating environment Lawrence E. Hurd a,⁎, Raniere G.C. Sousa b, Flávia K. Siqueira-Souza c, Gregory J. Cooper d, James R. Kahne,f, Carlos E.C. Freitas c a

Department of Biology, Washington and Lee University, Lexington, VA 24450, USA Departmamento de Ciências Pesqueiras, Universidade Federal do Rondônia, Rua da Paz, 4376. Presidente Médici, RO 76.916-000, Brazil Departmamento de Ciências Pesqueiras, Universidade Federal do Amazonas, Av. Gen. Rodrigo Otávio, 3000, 69077-000 Manaus, AM, Brazil d Department of Philosophy and Environmental Studies Program, Washington and Lee University, Lexington, VA 24450, USA e Environmental Studies Program and Department of Economics, Washington and Lee University, Lexington, VA 24450, USA f Universidade Federal do Amazonas, Brazil b c

a r t i c l e

i n f o

Article history: Received 13 July 2015 Received in revised form 29 December 2015 Accepted 2 January 2016 Available online xxxx Keywords: Amazon Basin Fish species diversity Floodplain lake communities Habitat connectivity Metapopulations

a b s t r a c t The Amazon River Basin contains the world's highest fish species diversity, with a hydrologic cycle that creates a patchy distribution of floodplain lakes at low water and affords dispersal and colonization opportunities through reconnected lakes, rivers, and flooded forests during high water. This connectivity is increasingly threatened by dam construction and droughts caused by climate change. Although the metapopulation framework has not been widely applied to freshwater ecosystems, it should represent a fruitful approach to conservation of important fish stocks and species diversity in Amazonian floodplains. Our examination of the evidence for metapopulation structure reveals that: (1) Although many economically important migratory species are not currently metapopulations (either demographically or genetically), connectivity is crucial to their life histories and anthropogenic stresses may induce metapopulation structure in these species; (2) Some large migratory pimelodid catfish with homing behavior to natal headwater streams appear to be the most spatially expansive metapopulations in existence among freshwater fish; (3) Non-migratory species are less well studied, but some (perhaps many) such species already exist as metapopulations and are vulnerable to disruptions in patterns of connectivity. Connectivity plays a crucial role in each of these cases, so the most promising conservation strategies involve: (1) reduction in dam building; (2) establishment of large enough protected areas to incorporate high β diversity and maintain patterns of connectivity during anomalous low water events; (3) implementation of governmentally facilitated community-based fishing agreements to curb overexploitation and monitor sustainable population levels and connectivity in protected areas. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction The Amazon River Basin, which encompasses the world's largest remaining tropical rainforest, has the highest diversity of fish species of any region in the world, with more than 2000 species identified so far (Reis et al., 2003). Much of this high fish diversity can be attributed to the physical complexity of the system, including variation in water depth, temperature, acidity, velocity, sediment and nutrient loads, and relative isolation of floodplain lakes from the river and each other (Freitas et al., 2010b). These factors have promoted adaptive radiation

⁎ Corresponding author. E-mail addresses: [email protected] (L.E. Hurd), [email protected] (R.G.C. Sousa), fl[email protected] (F.K. Siqueira-Souza), [email protected] (G.J. Cooper), [email protected] (J.R. Kahn), [email protected] (C.E.C. Freitas).

http://dx.doi.org/10.1016/j.biocon.2016.01.005 0006-3207/© 2016 Elsevier Ltd. All rights reserved.

into a wide variety of local environments with different selective regimes: large river channels, floodplain lakes and streams, and seasonally flooded forests. The Amazon Basin is quite old, initially formed near the end of the Cretaceous, with diversification of major Amazonian freshwater fish lineages such as characins, cichlids and catfish occurring during the Paleogene, between 65 and 23 million years ago (Hoorn et al., 2010). The threats to biodiversity caused by the loss of tropical ecosystems are frequently documented and discussed (e.g., Malhi et al., 2008; Bradshaw et al., 2009), but the emphasis has generally been on the terrestrial component of these systems. Aquatic ecosystems within tropical regions are faced with a number of anthropogenic threats: overfishing (Batista et al., 1998), invasive alien species (Latini and Petrere, 2004; Pelicice and Agostinho, 2008a), hydroelectric dam-building (Finer and Jenkins, 2012; Araújo et al., 2013), deforestation (Phillips et al., 2009, Lobόn-Cerviá et al., 2015) and droughts caused by both deforestation

L.E. Hurd et al. / Biological Conservation 195 (2016) 118–127

and global warming (Marengo et al., 2008, Freitas et al., 2013b, Nazareno and Laurance, 2015). Globally, freshwater ecosystems cover less than 1% of the earth's surface but account for a substantial fraction (about 6%) of total described species, and therefore are worthy of greater attention in general (Dudgeon et al., 2006). Regionally, the aquatic portion of the Amazon Basin covers more than 1 million km2, and produces 18% of the earth's river discharge (Castello et al., 2013). The fish component of this ecosystem represents significant biodiversity, comprising about 7% of the 28,000 known fish species (Hickman et al., 2014). Further, many species have been exploited by human inhabitants of the region as important sources of food for centuries prior to European colonization (Verissimo, 1895). Fish inhabiting floodplain lakes in the Amazon Basin are subjected to an annual hydrologic cycle alternating between high and low water seasons. These fish depend on temporal variation in accessible connectivity among lakes, rivers, and seasonally flooded forests to complete their life cycles (Fernandes et al., 2009). The movement of fish species among these habitats at different stages of the hydrologic cycle suggests a system within which some species might be distributed over a region as localized populations that experience a limited degree of crossmigration. In contrast, other species may experience the region as a more-or-less continuous habitat in which the rising water that washes over the floodplain during the annual high water stage prevents isolation. Rare species make up a large fraction of total species richness in any large collection (Preston, 1948), including floodplain lake fish (Yamamoto et al., 2014) and so may be particularly vulnerable to extinction if connections among habitat patches are severed by dams, or the period of lake isolation from the river is extended by droughts. The patchy distribution of floodplain lakes during the low water season, and the network of connections that exist among them (the main river channel and seasonally flooded forests), suggest that regional variations in topography, water chemistry and flow may be at least as important as local lake conditions in determining floodplain lake community structure. Because of this, metapopulation dynamics, involving interacting local populations within a larger region of floodplain habitat, may be applicable to floodplain fish. Here we examine this proposition as a foundation for conservation strategies in an environment where connectivity is directly threatened by dams and climate change. We begin by explaining the dynamic structure of the physical environment in the Amazonian floodplain ecosystem, then introduce the background of metapopulation theory and discuss its applicability to fish species, and finally discuss how this conceptual framework relates to conservation efforts aimed at preserving species diversity and sustainably exploiting fish stocks. Specifically, we address three questions: (1) What criteria should we use to judge whether a species is distributed as a metapopulation, and are there fish species in the Amazon Basin that currently meet these criteria? (2) How do we expect the changing aquatic environment of the floodplains to affect species that are, or are not, currently structured as metapopulations? (3) How does the metapopulation point of view contribute to designing effective strategies for conservation of fish populations and preservation of species diversity? 2. Structure of the physical environment 2.1. The hydrologic cycle and habitat connectivity The Amazon River Basin is physically heterogeneous, being divided in terms of drainage systems of different geological origins and physical characteristics. The principal systems in this region are whitewater and blackwater. Whitewater rivers such as the Solimões, Madeira, and Purus Rivers carry a high load of sediments and nutrients from pre-Andean headwaters. These rivers are therefore highly productive, with a nearly neutral pH, and support high fish species diversity and productive fish stocks (Lowe-McConnell, 1999; Freitas et al., 2010b). In contrast, blackwater rivers such as the Urubu, Uatumã and Negro originate from the

119

old plateaus of Guyana in Northern South America. They carry a lighter sediment load, are more acidic and less productive than whitewater rivers, but have equivalent numbers of fish species (Goulding et al., 1988; Freitas et al., 2010b). The hydrologic cycle, or “flood pulse” (Junk et al., 1989) in this vast system is characterized by dramatic changes in water level. It consists of four seasons, rising water, high water, receding water, and low water, each of which have different characteristics with regard to interhabitat connectivity (Fig. 1). Depending on specific location within

Fig. 1. A portion of a river and its adjacent floodplains, including lakes and connection channels, taking into account the season of the hydrologic cycle: (A) rising water season: the connectivity is increasing so that the water flows from the river to the lakes, increasing their size. The timing of lake connection depends on topography and the distance between the river and lake. Some lakes will be connected only in high water season (dashed lines); (B) high water (flood) season: connectivity is highest and the number of connected lakes at the floodplain will vary with the intensity of the flood pulse. Some lakes could be connected just in years of highest floods. Other lakes could be completely connected as a single aquatic environment; (C) receding water season: connectivity is diminishing as the water flows from lakes to the river, reducing their size. This process is faster than rising water. Some fish species could be retained in the lakes and die during years of extreme drought; (D) low water season: connectivity is lowest and the degree of isolation is a function of the lake position in the floodplain and drought intensity. Extreme drought years could cause a complete disconnection of some floodplain lakes.

120


the basin, shifts in the water level occur at approximately 5–8 month intervals (Bittencourt and Amadio, 2007), and the normal amplitude of change in depth over the cycle can be from 2 to 18 m (Richey et al., 1986). The flood pulse is generally considered to be the main structuring force for fish species in terms of distribution and population density (Winemiller and Jepsen, 1998; Garcez and Freitas, 2007), thus providing supportive evidence for the increasing focus on regional drivers of local community diversity (Ricklefs, 2008). The area and depth of floodplain lakes, as well as their degree and duration of isolation, are highly variable among lakes, years, and seasons of the cycle (Fig. 1). There is also high spatial heterogeneity at the regional scale in floodplain lakes (Fig. 2). Island lakes have different fish assemblages than coastal lakes, which in turn differ in species composition from mainland lakes (Freitas et al., 2013a). Water transparency, important to visually-oriented predators, is similarly variable over time and among lakes (Rodriguez and Lewis, 1994). Oxygen concentrations also may decline over time. However, even hypoxic environments may be refugia from predators or competition for some fish species (Junk et al., 1983). The intensity and timing of the flood pulse is controlled by several factors, including those that act on a global scale. The cyclical phenomenon of warming in the Pacific Ocean, termed El Niño, is related to severe drought in the Amazon Basin (Melack and Coe, 2013). Alternatively, La Niña is associated with strong floods (Ronchail et al., 2002). These events and the warming of the Tropical North Atlantic Ocean have been used to explain extreme climatic events in this region (Marengo et al., 2008; Oliver et al., 2009). Over the past three decades, there has been a trend toward increased lengths of dry periods (Fu

et al., 2013), and climate models generally predict longer and more intense droughts in the future (Salazar et al., 2007; Cox et al., 2008). A marked increase in duration of the low water season caused by climate change is likely to depress biodiversity, because of both a loss of connectivity among aquatic environments and a loss or delay of access to resources from the flooded forest. Allochthanous input of plant and animal material directly from the forest at high water is an important source of sustenance for many floodplain fish species, such as the commercially important characid frugivore, Colossoma macropomum (Goulding, 1980; Waldhoff et al., 1996; Correa and Winemiller, 2014). A severe drought in the Central Amazon Basin during 2005 caused measurable changes in fish community structure, some persisting over the subsequent two years (Freitas et al., 2013b). 2.2. Regional and local environmental forces The relative importance of physical v. biological factors to fish assemblages in floodplain lakes changes over the hydrologic cycle. The flood pulse has the potential to homogenize regional fish species distributions (Hoeinghaus et al., 2003; Thomaz et al., 2007; Freitas et al., 2010a). However, researchers have observed non-random assortment of fish assemblages, particularly during the low water season (e.g, Petry et al., 2003; Arrington and Winemiller, 2006; Fernandes et al., 2009). As the water recedes, and during the subsequent dry season, lakes get shallower and smaller in area, so that local biotic forces including density-dependent interactions such as competition and predation become more important (Rodriguez and Lewis, 1997; Winemiller and Jepsen, 1998; Thomé-Souza and Chao, 2004). Piscivorous predators,

Fig. 2. Different types of Amazonian floodplain lakes based on their proximity to the main river channel. M = mainland (upland) lake that is farthest from the main river channel, and may lose connectivity with the river and adjacent floodplain lakes during low water season; I = island lake, which shrinks in area during dry season but remains connected to the river through surrounding flooded forests; C = coastal lake, which remains connected to the river through narrow channels during low water.


including human fishers, can reduce or eliminate prey species in diminished volumes of water, and competitive exclusion may occur within guilds of fish as resources decline over a protracted dry period, so species assemblages can become different among lakes. As water levels rise, the intensity of biotic interactions can be reduced or eliminated preventing extinction of prey or subordinate competitor species. One density-dependent interaction that may be important in structuring floodplain fish communities is parasitism. Known taxa of fish parasites include protozoans, nematodes, arthropods, trematodes and acanthocephalans (Thatcher, 2006), which disperse among individual fish in a population through proximity or predation, depending on the life cycle of the parasites. The powerful influence of parasites on structure, dynamics, and behavior of host populations has been documented from a wide variety of taxa, including snails (Curtis and Hurd, 1983), mice (Scott, 1987), birds (Dobson and Hudson, 1992), and fish (Poulin, 2000). It has been suggested that the success of invasive alien species may even be facilitated by the lack of their normal specialist parasites in the invaded habitat (Torchin et al., 2003). Further, there is a potential for transmission of novel parasites and pathogens from alien species that may be inimical to native Amazonian fish (Thatcher, 2006), particularly when they are already stressed from overcrowding during low water. An important consequence of the annual hydrologic cycle is the constant shift in available food resources. The main carbon source at the base of the food chain is allochthanous (Correa and Winemiller, 2014). During high water the flooded forest provides plant material such as leaves and fruit, and animal food such as terrestrial insects, which are scarce or absent at low water. It is therefore not surprising that so many fish species are opportunistic omnivores (Jepson and Winemiller, 2002; Röpke et al., 2014). Most of the information used to categorize fish into trophic levels has been gleaned from stomach content analysis (e.g., Goulding, 1980; de Mérona and Rankin-de-Mérona, 2004), but examination of fish diets based on fatty acids and stable isotopes showed that many species switch diets between plant and animal sources according to differences in availability over time and space (Mortillaro et al., 2015). Given the patterns of connectivity in the system, and the annual reshuffling of these patterns via the flood pulse, it is clear that regional forces are important to floodplain fish community structure. This sets the stage for examination of the potential for metapopulation dynamics as a descriptor of floodplain fish populations, and for exploring the relevance of this concept to conservation of fish populations and preservation of species diversity. 3. Floodplain metapopulations? 3.1. Theoretical underpinnings Metapopulation theory has deep historical roots beginning with Andrewartha and Birch (1954) and the concept of “spreading of risk” (den Boer, 1968). The core idea involves representing the landscape in terms of a series of spatially discrete patches, each suitable for colonization by local populations of a regionally distributed species. This regional metapopulation is stabilized by a balance between extinction of local populations with colonization from other local populations. Levins (1969) modeled this process by analogizing the colonization and extinction of habitat patches to the birth and death of organisms. Thus, a metapopulation is a population of populations. The metapopulation is most often defined in terms of both local populations of organisms and the habitat patches suitable for occupancy by those organisms (Hanski and Simberloff, 1997). While suitable habitat patches clearly need to be part of the picture when analyzing metapopulation dynamics, habitat patches are not organisms. To avoid confusion we will distinguish between the metahabitat as a set of habitat patches suitable for occupancy, and the metapopulation as the

121

actual assemblage of local populations that occupy some subset of the metahabitat. Thus, the spatial distribution of a metahabitat is normally relatively constant, whereas the distribution of the metapopulation can change over time. The classical Levins metapopulation model embodies several presuppositions. First, the local populations are sufficiently discrete that demographic trajectories are largely determined by local processes. Second, the persistence of local populations is short relative to the persistence of the regional population. Third, population dynamics (rates of birth, death, immigration and extinction) among the various local populations is asynchronous so that they would not all be expected to go extinct at the same time. Fourth, dispersal rates are high enough to facilitate recolonization of unoccupied patches but not so high that they turn the metapopulation into a single, patchily distributed population. Many fish populations in the Amazon Basin do not look like strong candidates for the classical model. Migration, reproduction, and feeding of virtually all fish species are synchronized to stages of the hydrologic cycle. However, the annual flood pulse potentially homogenizes species distributions across the floodplain during high water (Thomaz et al., 2007). Thus, the conditions of local population distinctness, limited dispersal, and asynchronicity appear to be mostly violated in this system. In fact, classical metapopulations appear to be relatively rare in nature (Harrison and Taylor, 1997) and ecologists have broadened the notion to include any system that involves patchy distributions of local populations in space with some exchange of individuals. Several different models have been recognized (Hanski and Simberloff, 1997) and they can be set out along a continuum reflecting changes in rates of dispersal (Harrison and Taylor, 1997). At one extreme is the patchy population, which is a limiting case because dispersal rates are so high that the regional population, despite its patchy distribution, behaves as a single entity both genetically and demographically. A limiting case at the other extreme is the nonequilibrium model, which describes a situation so disconnected that demographic trajectories in the subpopulations are determined entirely by local processes (Harrison and Taylor, 1997). If local population growth rates are negative and/or populations are small then these local populations will be threatened with extinction and, if the phenomenon is widespread, the regional population will be in jeopardy as well. The limiting cases that form the ends of this continuum are spatially structured regional populations, but they are not metapopulations. Genuine metapopulations require intermediate levels of dispersal (Fig. 3). Some metapopulations embody a source-sink population structure. As originally defined by Pulliam (1988) a set of local populations connected by dispersal has a source-sink structure when the finite rate of increase at equilibrium is positive in some populations (λ N0) and negative (λ b0) in others. The general idea is that dispersal from overproducing local populations into underproducing local populations will prevent extinction of the latter, a phenomenon known as the “rescue effect” (Brown and Kodric-Brown, 1977). However, there are certain limitations to the original idea (Runge et al., 2006). The equilibrium assumption is likely to fail for many systems, and the model gains wider applicability if sources and sinks are defined in terms of rates of growth that may be variable but that exhibit tendencies over a period of time, i.e., sources tend to exhibit positive growth and sinks negative growth, even though sources may occasionally behave like sinks, and vice versa (Liu et al., 2012). Another limitation is that the original model ignored emigration from sink populations, implicitly treating such emigration as mortality. However, emigration from sink populations can have important impacts on overall metapopulation dynamics (Runge et al., 2006; Liu et al., 2012). While refinements continue in the literature, the core idea remains an important perspective on metapopulation structure. The existence of source-sink dynamics, and of metapopulation structure more generally, is important from a management perspective. To what extent do these models apply to Amazon fish populations?

122


Fig. 3. (A) Four floodplain lake types during the high water season of the hydrologic cycle. Increased drought reduces water level for longer periods during low water, and reduces connectivity during high water. Reliability of fish interpopulation connectivity among lakes and between lakes and the river decreases from lake type 1 (regionally open, patchy populations) to type 4 (locally isolated, closed populations). (B) Intermediate connectedness is expected to promote metapopulation dynamics (solid line) compared to a completely open and panmictic regional population (type 1) and completely closed local populations (type 4). The probability of local extinction (dashed line) increases from left to right on the x-axis: As droughts extend isolation of lakes during low water and the metahabitat decreases as a result, lakes change (moving left to right on x-axis), such that presently completely connected populations (type 1) become subject to metapopulation dynamics, and those populations now distributed as metapopulations (types 2–3) become closed (type 4), and may face extirpation. This representation is consistent with a different depiction of the continuum from open patchy to closed nonequilibrium local populations provided by Harrison and Taylor (1997: Fig. 2).

3.2. Life history considerations A number of Amazonian fish, especially those of economic importance, exhibit various forms of migratory behavior (Ribeiro and Petrere, 1990; Barthem et al., 1991; Araujo-Lima and Goulding, 1997; Araujo-Lima and Ruffino, 2003; Granado-Lorencio et al., 2005). These and other studies report that migratory characiforms, e.g., Colossoma (Charicidae), Brycon (Charicidae),and Prochilodus (Prochilodontidae), spend most of the high water season feeding in the flooded forest in either whitewater or blackwater rivers, where they develop significant fat reserves. During receding water they migrate to nutrient-rich whitewater main river channels (de Lima and Araujo-Lima, 2004). During rising water they spawn along the margins of the main river channel with larvae and juveniles passively dispersing downstream until they enter the floodplain nursery habitats, after which spent adults disperse back into the flooded forest to begin the cycle again (Goulding, 1980; Cox-Fernandes, 1997; Araujo-Lima and Ruffino, 2003). Several migratory characins such as Prochilodus spp. (Oliveira et al., 2009), Piaractus mesopotamicus (Iervolino et al., 2010) and C. macropomum (Santos et al., 2007) appear to form large panmictic populations. However, microsatellite analysis indicated a more constrained “stepping stone” migratory pattern among local populations of C. macropomum rather than total panmixia, leaving open the question of some form of metapopulation structure (Aldea-Guevara et al., 2013). Several species in the genus Brycon also are genetically fragmented, in these cases by anthropogenic effects (Oliveira et al., 2009). In contrast to the relatively short migrations of the characins, the catfish Brachyplatystoma rousseauxii (Pimelodidae) has the longest migration known for a freshwater fish (Vásquez et al., 2009). These fish breed in the Andean and pre-Andean headwaters, and larvae drift down river through five countries (approximately 5500 km) to the estuary of the Amazon. After maturation adults move back upriver, and Randomly Amplified Polymorphic DNA (RAPD) analysis indicates

that they may return to their natal headwaters (Batista and AlvesGomes, 2006), which means that each tributary may produce a distinct genetic local population, the sum of which could constitute a migratory metapopulation of enormous geographic range. The regional population of another long-distance migratory pimelodid, Pseudoplatystoma corruscans, consists of six genetically distinct populations in the La Plata Basin of Brazil, based on microsatellite data (Pereira et al., 2009). Thus, even though long-range migrant species may move together with members of different headwater origins, evidence suggests that they exist as separate local populations within the matrix of a regional metapopulation. Other long distance migratory catfish appear to lack this homing behavior. Mitochondrial DNA (mtDNA) evidence indicates that the commercially important pimelodids Brachyplatystoma vaillantii and Brachyplatystoma flavicans form large regional panmictic populations (Santos et al., 2007). The lack of genetic structure in migratory populations is perhaps unsurprising given the opportunities for gene flow provided by their migratory behaviors. The situation is less clear for nonmigratory species, where we have fewer studies to draw upon. A population of Cichla monoculus (Cichlidae) in the lower Solimões revealed both highly structured local populations and low levels of heterozygosity within each population using RAPD analysis, suggesting a metapopulation structure (Dos Santos et al., 2012). More sensitive techniques using both mtDNA and nuclear DNA were applied to a different species in this genus, C. pleiozona, in the Upper Madeira Basin (Carvajal-Vallejos et al., 2010). These results were mixed, although they found generally higher levels of gene flow than the previous study and the intriguing possibility of sex-biased dispersal. More recently, a study of yet another species in this genus, C. temensis used both mtDNA and microsatellite data to study population genetic structure across a broad distribution of the species in both the Amazon and Orinoco basins (Willis et al., 2015). This study also revealed low vagility and limited gene exchange among local populations, suggesting a potential metapopulation


123

structure as with C. monoculus. A study of Osteoglossum ferreirai (Osteoglossidae) using both nuclear and mtDNA described low levels of genetic variation within local populations and evidence of inbreeding and/or nonrandom mating (Olivares et al., 2013). This suggests the existence of a potentially unstable metapopulation structure for this species. The above examples use genetic information to suggest patterns of connectivity that have a bearing on the metapopulation status of various fish species. This is a popular strategy because actual dispersal is hard to measure, especially at large spatial scales. However it is important to remember that demographic population structure and genetic population structure are distinct; patterns in one domain need not reflect patterns in the other (Lowe and Allendorf, 2010), which is the basis for the familiar distinction between management units (MUs) and evolutionarily significant units (ESUs) (Moritz, 1994). For example, the long-distance migratory catfish that exhibit homing behavior to natal headwater streams exist as genetically distinct subpopulations (Batista and Alves-Gomes, 2006). If selective pressures differ among these habitats, their genetic contributions to the metapopulation will not be equal. Thus, a level of harvest that might seem sustainable from the demographic (numbers, size, reproductive condition, etc.) point of view when local populations merge might actually be driving some local populations toward extinction. This would reduce the overall genetic diversity of the metapopulation (Policansky and Magnuson, 1998). Another example is a study of Prochilodus mariae (Prochilodontidae) in Venezuelan rivers using a combination of genetic and otolith studies to investigate its spatial ecology (Collins et al., 2013). Otolith data reveal the existence of distinct local populations in four of their six study sites even though microsatellite analysis indicates that the regional population is essentially panmictic. In this case the evidence suggests that although there is enough exchange of individuals to prevent genetic differentiation, demographic considerations indicate that levels of dispersal among local populations are insufficient to compensate for the impacts of severe overfishing or habitat destruction. Therefore, although management decisions based on genetic information alone might suggest that there is no need to pay attention to spatial patterns in fishing pressure or habitat degradation, these fish must be managed as demographically distinct units. The migratory characins represent an important group of Amazon fishes, both in terms of regional fisheries and ecosystem dynamics, and P. mariae appears to be fairly representative of the group from a life history perspective. It remains an open, and important, question whether the panmixia that appears to characterize many characin species obscures an underlying demographic population structure that could be important from a management perspective.

important determinant of community structure early in the high water season, and the local physical environment was more important later in that same season (Fernandes et al., 2014). Although not many studies have explicitly taken the metacommunity approach, there have been numerous studies of community structure and community assembly in floodplain habitats, many of which are at least implicitly metacommunity studies. These studies reveal that community assemblages differ among different types of floodplain lakes (e.g., Freitas et al., 2013a). Furthermore, contrary to earlier evidence (Goulding et al., 1988), community assembly can be nonrandom during high water (Fernandes et al., 2014) and often is nonrandom at low water (Petry et al., 2003; Arrington and Winemiller, 2006; Fernandes et al., 2009). Nonrandom assembly appears to be due to habitat affinities and, as the season progresses, to biotic interactions. The heterogeneity in fish community composition has important implications for how species diversity in this region must be considered. The measurement of species diversity in communities is a matter of scale: species richness in a single lake that is located on a floodplain region containing many lakes is α diversity; the differences in species composition among lakes in the region is β diversity; and these two measures taken together constitute γ diversity on the regional scale (Whittaker, 1972). For Amazonian floodplain lakes the contribution of β diversity to γ can be as large as, or larger than, that of α (Freitas et al., 2013b). This means that no single lake is likely to contain most of the diversity to be found within a floodplain region. The high importance of β diversity in this ecosystem therefore requires consideration of species diversity on both local and regional scales. Several conclusions emerge from the focus on the ecology of regional processes. First, many of the migratory species probably exist as patchy populations. There is evidence that some of these species might be headed in the direction of a metapopulation structure as a consequence of anthropogenic activities — mainly the disruption of connectivity through dams and droughts. Second, it appears likely that there are many metapopulations in existence already, especially among nonmigratory species. Third, β diversity is highly important in this freshwater ecosystem. Finally, metapopulation structure exists along a continuum from open patchy populations to closed nonequilibrium populations (Fig. 3). One of the most pressing conservation concerns for this system is that disruptions of connectivity will impact important regional processes in ways that move populations to the right along this continuum.

3.3. Metacommunities and species diversity

Dam construction is both an immediate and long-term threat to connectivity of habitats and migratory patterns of fish species (Sá-Oliveira et al., 2015a, 2015b), as well as to the timing and magnitude of the flood pulse. Currently there are 18 hydroelectric dams operating in the Amazon Basin, eight under construction, and 53 planned. In addition, another 62 sites have been identified with potential for hydroelectric facilities (Kahn et al., 2014). Dam construction also exacerbates another well-known anthropogenic threat to Amazonian biodiversity, deforestation. The projected inundated area associated with planned dams is roughly three times the total amount of upland deforestation, mainly clearing for agriculture, which occurred in the Brazilian Amazon in 2013 (Kahn et al., 2014). We know of no studies that specifically examine the effect of deforestation on floodplain fish species. However, extensive deforestation of floodplain forests would interrupt allochthanous food sources at high water, increase siltation from terrestrial runoff and alter riparian flow (Castello et al., 2013). Dams are especially harmful to species with long-distance migratory patterns, and many of the economically valuable species such as the pimelodid catfish described earlier, fall into this category. Existing dams have already been shown to disrupt migratory patterns

Emphasis on regional processes also has led to the emergence of metacommunity ecology. A metacommunity is a set of local communities of potentially interacting species linked by dispersal (Holyoak et al., 2005). Metacommunity dynamics is a function of the interaction between regional (primarily dispersal) and local (both biotic and abiotic) forces. The significance of regional processes, coupled with the patchiness of floodplain lakes, suggest that the metacommunity framework should be fruitful. Most work on metacommunities has consisted of the theoretical elaboration of a set of standard models. However, the homogenizing influence of the flood pulse raises questions about the extent to which any of these standard models really fit the system. Perhaps there is some yet to be articulated hybrid (or set of hybrids) that can be made to work, but the formal metacommunity approach is of limited utility in understanding floodplain community dynamics at present. A second, more empirical, approach to metacommunities involves the investigation of patterns of species presence/absence along environmental gradients (Pressley et al., 2010). Application of this approach to the floodplain system revealed that dispersal was an

4. Conserving fish stocks and preserving diversity 4.1. Damming in the Amazon Basin

124


(Barthem et al., 1991; Pringle et al., 2000) and the ambitious construction program described above will obviously result in further disruptions. Ameliorative strategies such as fish ladders have proven to be of limited effectiveness or to have negative effects (Pelicice and Agostinho, 2008b; Agostinho et al., 2013; Pelicice et al., 2014). The problems are not only in the upstream direction. Downstream dispersal of larvae and juveniles also is disrupted, and even when fish make it past the dam mortality is high (Pelicice et al., 2014).

of carrying more seeds per individual, distributing them farther and over more years than smaller fish, the reduction in numbers of large individuals by overfishing of wild stocks undoubtedly has a negative impact on regional plant recruitment dynamics (Horn et al., 2011; Costa-Pereira and Galetti, 2015). Clearly, the other anthropogenic factors that threaten survival and connectivity in fish communities such as dam-building, deforestation and climate change, also can reduce terrestrial plant diversity by disrupting ichthyochory (Horn et al., 2011).

4.2. Climate change and regional drought 4.3. Floodplain fisheries management Global climate change affects nearly everything about the environment directly or indirectly, and we are far from being able to predict its effects on the great majority of species. Kokko and Lopez-Sepulcre (2006) have noted that the biology of dispersal will require greater attention if we are to understand the effects of climate change on animal species. For floodplain fish, the main problem is likely to be drought caused by a combination of deforestation and global warming. Drought has been documented to have a number of deleterious effects on floodplain ecosystems (Lake, 2011), and can be expected to decrease the spatial distribution of the metahabitat, normally a relative constant, for many fish species. A principal effect of drought in floodplains, as with dams, is to interrupt connectedness for open patchy populations, for metapopulations that depend on interhabitat migration, and for lake residents that depend on access to the forest and relief from abiotic deterioration and density-dependent mortality. Increased drought during low water should reduce connectivity and cause a shift in probabilities of extinction (Fig. 3), as open populations that depend on timely migration and dispersal are forced into metapopulation status, and those already distributed as metapopulations become more isolated. The larger piscivorous fish such as tucunaré, or smaller opportunistic predators such as piranha, may dominate isolated lakes assuming they can withstand the deteriorating physical environment. To the extent that conservation measures involve the creation of preserves it is especially important to take into consideration the impending increased frequency of droughts when designing preserves, maximizing the inclusion of lakes and multi-habitat floodplain areas that have the highest probability of maintaining connectivity during low-water events. Obviously, the ultimate conservation strategy is to ameliorate climate change, which is likely to be a long, slow process compared to the rate of environmental degradation. More immediately, it is important to incentivize conservation policy by quantifying the economic value of goods and services delivered by Amazon Basin floodplain ecosystems. The most significant resource is the yield from Amazonian fisheries, both for subsistence consumption and commercialization. For example, per capita fish consumption in the Amazon is very high: 5.8 times the world average for rural populations and 2.5 times the world average for those living in cities (Castello et al., 2013). Given the important role of allochthanous nutrient input in much of the system (Waldhoff et al., 1996), the fishery resource specifically, and ecosystem processes more generally, depend on maintaining the integrity of the flooded forest. River access to important forest resources (e.g. açai fruit) also is important (Castello et al., 2013). Significant ecosystem services include the role of forested riparian corridors in maintaining water quality and tempering the extremes of the flood pulse. Forests are also important for maintaining the hydrologic cycle and, of course, carbon sequestration (Castello et al., 2013). A service of floodplain ecosystems that is provided directly by fish such as C. macropomum that consume fruit from terrestrial plants is the dispersal of seeds that helps maintain regional plant species diversity, a process known as ichthyochory (Horn et al., 2011). This relationship between terrestrial plant reproduction and fish apparently is ancient (Gottsberger, 1978). In fact, it is possible that fish were the first vertebrate seed dispersers (Horn et al., 2011). Since large-bodied frugivorous fish (mainly Characiformes and Siluriformes) are capable

The recommendation for preservation of species diversity, given the high importance of β diversity mentioned earlier, is to create protected preserves that encompass as large a number and variety of floodplain lakes as possible, at least in whitewater systems. It is possible that α diversity is a larger component of regional species richness in blackwater systems, so that preserves would require fewer lakes to represent the bulk of species richness there, but more evidence is needed before we can be sure (Goulding et al., 1988). Preserving diversity should also function to protect most commercially valuable fish populations, but many such populations also face pressure from overfishing. Overexploitation is problematic because, by definition, it is unsustainable. Further, if a floodplain population manifests source-sink dynamics, then it is vital to avoid overexploitation of sources that may be necessary to supply immigrants to sinks. However, sources and sinks can be difficult to identify and they can flip roles over time, i.e., sources can become sinks and vice versa (Falcy and Danielson, 2011). In his original presentation, Pulliam (1988) suggested that understanding the dynamics of source-sink systems required knowledge of juvenile and adult mortality, reproductive success, and dispersal for the relevant local populations, a tall order for most floodplain species. There are many standard tools for managing fish populations such as closed seasons, size restrictions, harvest quotas and gear restrictions. Unfortunately, the effectiveness of these regulatory instruments depends on a consistent regime of enforcement. The vastness and isolation of the region, combined with the lack of resources for managing agencies, results in inadequate enforcement and thus poor compliance. One alternative option, which has met with some degree of success, is to implement community-based management systems (de Castro and McGrath, 2003; Lima, 1999; McGrath, 2012). Under these arrangements local communities monitor fish populations and develop rules governing the use of lakes in their neighborhoods. Since the 1990s there has been an effort to incorporate these community management systems into a more formal state-based management framework as a way to overcome some of the institutional shortcomings of the community approach alone (de Castro and McGrath, 2003; McGrath, 2012). The effectiveness of community-based management schemes has not been extensively researched. Almeida et al. (2009) conducted a pairwise investigation of nine fishing communities in the lower Amazon that had community based fishing arrangements. Each community was paired with a similar community that lacked such arrangements. The authors found a significant increase (48%) in catch per unit effort among the communities that had fishing arrangements and they attributed this to higher fish densities in the managed systems. Another example is Mamirauá, a sustainable development reserve along the middle Solimões for which there is evidence for the recovery of both Arapaima gigas and C. macropomum (McGrath, 2012; Aldea-Guevara et al., 2013). These successes were the result of adopting sustainable levels of exploitation by exclusion of commercial fishing, adoption of restrictions on fishing gear and catch, monitoring and enforcement by local community members, informal assessment of fish densities, and the designation of certain areas as preserves (Almeida et al., 2009; Aldea-Guevara et al., 2013). Despite these successes, the community-based fishing approach still faces a number of institutional and ecological impediments. Ecological impediments include a lack of understanding of the potential role of


metapopulation structure, including possible source-sink dynamics, in the managed region. Mamirauá differs importantly from many other community-based management systems because it is formally designated as a preserve. The designation of protected areas (PAs) has long been a popular strategy for the conservation of terrestrial biodiversity and it is rapidly gaining prominence for marine systems as well. However, PAs have generally not been a significant component of efforts to conserve freshwater biodiversity. It has been suggested that the traditional concept of PAs is a poor fit for freshwater rivers (Abell et al., 2007). However, PAs represent a promising approach to accommodating the regional processes that shape floodplain fish communities. Historically A. gigas (Arapaimidae) existed as one large panmictic population along the entire length of Amazon varzea habitat, but overexploitation in the center of the species' range has created a genetic bottleneck there at the origin of genetic diversity for the population (Hrbek et al., 2005). They recommended creating protected areas near both ends of the species' range, where among-population genetic diversity is higher than in the genetically depleted center. They also recommended establishing a source-sink metapopulation structure by creating a protected area in the center of the range, so that the center will not be a sink for source populations in preserves at the edges of the range. The source habitats would serve as a repository for genetic diversity and also generate harvestable surplus in adjacent exploitable sink habitats (see also Iervolino et al., 2010). There is evidence that a large number of other economically important fish species, together with important non-fish varzea inhabitants such as caiman, turtles and manatee, have a genetic structure similar to A. gigas that would also benefit from a system of strategically placed PAs (Hrbek et al., 2007). The general problem with establishing PAs in the Amazon Basin is lack of data on population dynamics and species distributions. Using newly available spatial and remote-sensing data sets, Thieme et al. (2007) developed a conservation planning framework for data-poor areas where this crucial biological information is lacking. Their framework, developed for the Madre de Dios River in the Amazon Basin, includes stream and floodplain classification, connectivity analysis, and an assessment of intactness for the various subbasins. This represents a “coarse filter,” but it can be updated as relevant biological information is acquired. If we wait for the data to come in it will be too late, so it makes sense to use the information that we do have as best we can. This framework offers a promising way to do so. Acknowledgments LEH was funded by a fellowship from CAPES (Brazil), and by the Herwick Professorship Endowment at Washington and Lee University. JRK, LEH and CG also were supported by Lenfest research grants from Washington and Lee University. JRK received additional support from the Herndon Professorship Endowment at Washington and Lee University, and from FAPEAM (Brazil). CECF was funded by a grant from CNPq (Brazil). CECF and JRK were also funded by CNPq grant number 302430/2012 (Brazil). References Abell, R., Allan, J.D., Lehner, B., 2007. Unlocking the potential of protected areas for freshwaters. Biol. Conserv. 134, 48–63. Agostinho, A.A., Agostinho, C.S., Pelicice, F.M., Marquis, E.E., 2013. Fish ladders: safe fish passage or hotspot for predation? Neotropical Ichthyol. 10, 687–696. Aldea-Guevara, M.I., Hargrove, J., Austin, J.D., 2013. Diversity and gene flow in a migratory frugivorous fish: implications for Amazonian habitat connectivity. Conserv. Genet. 14, 935–942. Almeida, O.T., Lorenzen, K., McGrath, D.G., 2009. Fishing agreements in the lower Amazon: for gain and restraint. Fish. Manag. Ecol. 16, 61–67. Andrewartha, H.G., Birch, L.C., 1954. The Distribution and Abundance of Animals. University of Chicago Press, Chicago IL. Araújo, E.S., Marques, E.E., Freitas, I.S., Neuberger, A.L., Fernandes, R., Pelicice, F.M., 2013. Changes in distance decay relationships after river regulation: similarity among fish

125

assemblages in a large Amazonian river. Ecol. Freshw. Fish http://dx.doi.org/10. 1111/eff.12054. Araujo-Lima, C.A.R.M., Goulding, M., 1997. So Fruitful a Fish: Ecology, Conservation, and Aquaculture of the Amazon's Tambaqui. Columbia University Press, NY. Araujo-Lima, C.A.R.M., Ruffino, M.L., 2003. Migratory fishes of the Brazilian Amazon. In: Carolsfeld, J., Harvey, B., Ross, C., Baer, A. (Eds.), Migratory Fishes of South America: Biology, Fisheries and Conservation Status. International Development and Research Centre, World Fisheries Trust, Ottawa, pp. 235–307. Arrington, D.A., Winemiller, K.O., 2006. Habitat affinity, the seasonal flood pulse, and community assembly in the littoral zone of a Neotropical floodplain river. J. N. Am. Benthol. Soc. 25, 126–141. Barthem, R.B., Ribeiro, M.C.L.B., Petrere Jr., M., 1991. Life strategies of some long distance migratory catfish in relation to hydroelectric dams in the Amazon Basin. Biol. Conserv. 55, 339–345. Batista, J.S., Alves-Gomes, J., 2006. Phylogeography of Brachyplatystoma rousseauxii (Siluriformes: Pimelodidae) in the Amazon Basin offers preliminary evidence for the first case of “homing” for an Amazonian migratory catfish. Genet. Mol. Res. 5, 723–740. Batista, V.S., Inhamuns, A.J., Freitas, C.E.C., Freire-Brasil, D., 1998. Characterization of the fishery in river communities in the low-Solimões/high-Amazon region. Fish. Manag. Ecol. 5, 419–435. Bittencourt, M.M., Amadio, S.A., 2007. Rapidity proposal to identify the hydrological periods in the Solimões–Amazonas River floodplain areas near to Manaus. Acta Amazon. 37, 303–308. den Boer, P.J., 1968. Spreading of risk and stabilization of animal numbers. Acta Biotheor. 18, 165–194. Bradshaw, C.J.A., Sodhi, N.S., Brook, B.W., 2009. Tropical turmoil: a biodiversity tragedy in progress. Front. Ecol. Environ. 7, 79–87. Brown, J.H., Kodric-Brown, A., 1977. Turnover rates in insular biogeography: effect of immigration on extinction. Ecology 58, 445–449. Carvajal-Vallejos, F.M., Duponchelle, F., Ballivian, J.P.T., Hubert, N., Rodríguez, J.N., Berrebi, P., Cornejo, S.S., Renno, J.-F., 2010. Population genetic structure of Cichla pleiozona (Perciformes: Chichlidae) in the Uppere Maera Basin (Bolivian Amazon): sex-biased dispersal? Mol. Phylogenet. Evol. 57, 1334–1340. Castello, L., McGrath, D.G., Hess, L.L., Coe, M.T., Lefevre, P.A., Petry, P., Macedo, M.N., Renó, V.F., Arantes, C.C., 2013. The vulnerability of Amazon freshwater ecosystems. Conserv. Lett. 6, 217–229. de Castro, F., McGrath, D.G., 2003. Moving toward sustainability in the local mangement of floodplain lake fisheries in the Brazilian Amazon. Hum. Organ. 62, 123–133. Collins, S.M., Bickford, N., McIntyre, P.B., Coulon, A., Ulseth, A.J., Taphorn, D.C., Flecker, A.S., 2013. Population structure of a neotropical migratory fish: constrasting perspectives from genetics and otolith microchemistry. Trans. Am. Fish. Soc. 142, 1192–1201. Correa, S.B., Winemiller, K.O., 2014. Niche partitioning among frugivorous fishes in response to fluctuating resources in the Amazonian floodplain forest. Ecology 95, 210–224. Costa-Pereira, R., Galetti, M., 2015. Frugivore downsizing and the collapse of seed dispersal by fish. Biol. Conserv. 191, 809–811. Cox, P.M., Harris, P.P., Huntingford, C., Betts, R.A., Collins, M., Jones, C.D., Jupp, T.E., Marengo, J.A., Nobre, C.A., 2008. Increasing risk of Amazonian drought due to decreasing aerosol pollution. Nature 453, 212–215. Cox-Fernandes, C., 1997. Lateral migrations of fishes in Amazon floodplain. Ecol. Freshw. Fish 6, 36–44. Curtis, L.A., Hurd, L.E., 1983. Age, sex, and parasites: spatial heterogeneity in a sandflat population of Ilyanassa obsoleta. Ecology 64, 819–828. Dobson, A.P., Hudson, P.J., 1992. Regulation and stability of a free-living host–parasite system: Trichostrongylus tenuis in red grouse. II. population models. J. Anim. Ecol. 61, 487–498. Dos Santos, C.H.D.A., De Sousa, C.F.S., Paula-Silva, M.dN., Val, A.L., De Almeida-Val, V.M.F., 2012. Genetic diversity in Cichla monoculus (Spix and Agassiz, 1931) populations: implications for management and conservation. Am. J. Environ. Sci. 8, 35–41. Dudgeon, D.A., Arthington, A.H., Gessner, M.O., Ka- wabata, Z.-I., Knowler, D.J., Leveque, C., Naiman, R.J., Prieur-Richard, A.H., Soto, D., Stiassny, M.J.L., Sullivan, C.A., 2006. Freshwater biodiversity: importance, threats, status and conservation challenges. Biol. Rev. 81, 163–182. Falcy, M.R., Danielson, B.J., 2011. When sinks rescue sources in dynamic environments. In: Liu, J., Hull, V., Morzillo, A.T., Wiens, J.A. (Eds.), Sources, Sinks and SustainabilityCambridge Studies in Landscape Ecology. Cambridge University Press, New York, pp. 139–154. Fernandes, R., Gomes, L.C., Pelicice, F.M., Agostinho, A.A., 2009. Temporal organization of fish assemblages in floodplain lagoons: the role of hydrological connectivity. Environ. Biol. Fish 85, 99–108. Fernandes, I.M., Henriques-Silva, R., Penha, J., Zuanon, J.P., Neto, P.R., 2014. Spatiotemporal dynamics in a seasonal metacommunity structure is predictable: the case of floodplain-fish communities. Ecography 37, 464–475. Finer, M., Jenkins, C.N., 2012. Proliferation of hydroelectric dams in the Andean Amazon and its implications for Andes-Amazon connectivity. PLoS One 7, 1–9. Freitas, C.E.C., Siqueira-Souza, F.K., Florentino, A.C.R., Hurd, L.E., 2013a. The importance of spatial scales to analysis of fish diversity in Amazonian floodplain lakes and implications for conservation. Ecol. Freshw. Fish http://dx.doi.org/10.1111/eff.12099. Freitas, C.E.C., Siqueira-Souza, F.K., Guimarães, A.R., Santos, F.A., Santos, I.L.A., 2010a. Interconnectedness during high water maintains similarity between fish assemblages of island floodplain lakes in the Amazon Basin. Zoologia 27, 931–938. Freitas, C.E.C., Siqueira-Souza, F.K., Humston, R., Hurd, L.E., 2013b. An initial assessment of drought sensitivity in Amazonian fish communities. Hydrobiologia 705, 159–171. Freitas, C.E.C., Siqueira-Souza, F.K., Prado, K.L.L., Yamamoto, K.C., Hurd, L.E., 2010b. Factors determining fish species diversity in Amazonian floodplain lakes. In: Rojas, N., Prieto,

126


R. (Eds.), Amazon Basin: Plant Life, Wildlife and Environment. Nova Science Publishers, New York, pp. 41–76. Fu, R., Yin, L., Li, W., Arias, P.A., Dickinson, R.E., Huang, L., Chakraborty, S., Fernandes, K., Liebmann, B., Fisher, R., Myeni, R.B., 2013. Increased dry-season length over southern Amazonia in recent decades and its implication for future climate projection. PNAS 110, 18110–18115. Garcez, R.C.S., Freitas, C.E.C., 2007. The influence of flood pulse on fish communities of floodplain canals in the Middle Solimões River, Brazil. Neotropical Ichthyol. 6, 249–255. Gottsberger, G., 1978. Seed dispersal by fish in the inundated regions of Humaita, Amazonia. Biotropica 10, 170–183. Goulding, M., 1980. The Fishes and the Forest. University of California Press, Berkeley. Goulding, M., Carvalho, M.L., Ferreira, E.G., 1988. Rio Negro, Rich Life in Poor Water. SPB Academic PUblishing, The Hague, The Netherlands. Granado-Lorencio, G., Araujo Lima, C.R.M., Lobón-Cerviá, J., 2005. Abundance-distribution relationships in fish assembly of the Amazonas floodplain lakes. Ecography 28, 515–520. Hanski, I., Simberloff, D., 1997. The metapopulation approach, its history, conceptual domain and application to conservation. Metapopulation Biology: Ecology, Genetics and Evolution. Academic Press, San Diego, pp. 5–26. Harrison, S.T., Taylor, A.D., 1997. Empirical evidence for metapopulation dynamics. In: Hanski, I., Gilpin, M. (Eds.), Metapopulation Biology: Ecology, Genetics and Evolution. Academic Press, San Diego, pp. 27–42. Hickman, C.P., Roberts, L.S., Keen, S.L., Eisenhour, D.J., Larson, A., I'Anson, H., 2014. Integrated Principles of Zoology. 16th ed. McGraw-Hill, NY. Hoeinghaus, D.J., Layman, C.A., Arrington, D.A., Winemiller, K.O., 2003. Spatiotemporal variation in fish assemblage structure in tropical floodplain creeks. Environ. Biol. Fish 67, 379–387. Holyoak, M., Leibold, M., Mouquet, M., Holt, R.D., Hoopes, M.F., 2005. Metacommunities: a framework for large-scale community ecology. In: Holyoak, M., Leibold, M.A., Holt, R.D. (Eds.), Metacommunities: Spatial Dynamics and Ecological Communities. University of Chicago Press, pp. 1–31. Hoorn, C., Wesselingh, F.P., ter Steege, H., Bermudez, M.A., Mora, A., Sevink, J., Sanmartin, I., Sanchez-Meseguer, A., Anderson, C.L., Figueiredo, J.P., Jaramillo, C., Riff, D., Negri, F.R., Hooghiemstra, H., Lundberg, J., Stadler, T., Särkinen, T., Antonelli, A., 2010. Amazonia through time: Andean uplift, climate change, landscape evolution, and biodiversity. Science 330, 927–931. Horn, M.H., Correa, S.B., Parolin, P., Pollux, P.J.A., Anderson, J.T., Lucas, C., Widmann, P., Tjiu, A., Galetti, M., Goulding, M., 2011. Seed dispersal by fishes in tropical and temperate fresh waters: the growing evidence. Acta Oecol. 37, 561–577. Hrbek, T., Crossa, M., Farias, I.P., 2007. Conservations strategies for Arapaima gigas (Schinz, 1822) and the Amazonian várzea ecosystem. Braz. J. Biol. 67, 909–917. Hrbek, T., Farias, I.P., Crossa, M., Sampaio, I., Porto, J.I.R., Meyer, A., 2005. Population genetic analysis of Arapaima gigas, one of the largest freshwater fishes of the Amazon Basin: implications for its conservation. Anim. Conserv. 8, 297–308. Iervolino, F., de Resende, E.K., Hilsdorf, A.W.S., 2010. The lack of genetic differentiation of pacu (Piaractus mesopotamicus) populations in the Upper-Paraguay Basin revealed by the mitochondrial DNA D-loop region: implications for fishery management. Fish. Res. 101, 27–31. Jepson, D.B., Winemiller, K.O., 2002. Structure of tropical river food webs revealed by sable isotope ratios. Oikos 96, 46–55. Junk, W.J., Bayley, P.B., Sparks, R.E., 1989. The flood pulse concept in river-floodplain systems. In: Dodge, D.P. (Ed.), Proceedings of the International Large River Symposium (LARS). Canadian Special Publication of Fishery and Aquatic Science 106, pp. 110–127. Junk, W.J., Soares, F.M., Carvalho, F.M., 1983. Distribution of fish species in a lake of the Amazon River floodplain near Manaus (Lago Carmaleão), with special reference to extreme oxygen conditions. Amazoniana 7, 397–431. Kahn, J.A., Feitas, C.E.C., Petrere, M., 2014. False shades of green: the case of Brazilian Amazonian hydropower. Energies 7, 6063–6082. Kokko, H., Lopez-Sepulcre, A., 2006. From individual dispersal to species ranges: perspectives for a changing world. Science 313, 789–791. Lake, P.S., 2011. Drought and Aquatic Ecosystems: Effects and Responses. Wiley-Blackwell, UK. Latini, A.O., Petrere, M., 2004. Reduction of a native fish fauna by alien species: an example from Brazilian freshwater tropical lakes. Fish. Manag. Ecol. 11, 71–79. Levins, R., 1969. Some demographic and genetic consequences of environmental heterogeneity for biological control. Bull. Entomol. Soc. Am. 15, 237–240. Lima, D.M., 1999. Equity, sustainable development and biodiversity preservation: some questions on the ecological partnership in the Brazilian Amazon. In: Padoch, C., Ayres, M., Pinedo-Vasquez, M., Henderson, A. (Eds.), Conservation and Development of Amazonian Varzea. New York Botanical Garden Press, New York, pp. 247–263. de Lima, Á.C., Araujo-Lima, C.A.R.M., 2004. The distributions of larval and juvenile fishes in Amazonian rivers of different nutrient status. Freshw. Biol. 49, 787–800. Liu, J., Hull, V., Morzillo, A.T., Wiens, J.A., 2012. Sources, Sinks and Sustainability. Cambridge University Press. Lobόn-Cerviá, J., Hess, L.L., Melack, K.M., Araujo-Lima, C.A.R.M., 2015. The importance of forest cover for fish richness and abundance on the Amazon floodplain. Hydrobiologia 750, 245–255. Lowe, W.H., Allendorf, F.W., 2010. What can genetics tell us about population connectivity? Mol. Ecol. 19, 3038–3051. Lowe-McConnell, R., 1999. Estudos Ecológicos em Comunidades de Peixes Tropicais. EDUSP, São Paulo – SP (524 pp.). Malhi, Y., Roberts, J.T., Betts, R.A., Killeen, T.J., Li, W., Nobre, C.A., 2008. Climate change, deforestation, and the fate of the Amazon. Science 319, 169–172.

Marengo, J.A., Nobre, C.A., Tomasella, J., Oyama, M.D., de Oliveira, G.S., de Oliveira, R., Carmago, H., Alves, L.M., Brown, I.F., 2008. The drought of Amazonia in 2005. J. Climatol. 21, 495–516. McGrath, D.G., 2012. Case analyses on experiences of formalization of informal sectors. Development of a formal co-management system for floodplain fisheries in the Lower Amazon Region of Brazil. Cent. Int. For. Res. 1–54. Melack, J.M., Coe, M.T., 2013. Climate change and the floodplain lakes of the Amazon Basin. In: Goldman, C.R., Kumagai, M., Robarts, R.D. (Eds.), Climatic Change and Global Warming of Inland Waters: Impacts and Mitigation for Ecosystems and Societies. John Wiley, Chichester, UK, pp. 296–310. de Mérona, B., Rankin-de-Mérona, J., 2004. Food resource partitioning in a fish community of the central Amazon floodplain. Neotropical Ichthyol. 2, 75–84. Moritz, C., 1994. Defining ‘evolutionary significant units’ for conservation. Trends Ecol. Evol. 9, 373–375. Mortillaro, J.M., Pouilly, M., Wach, M., Freitas, C.E.C., Abril, G., Meziane, T., 2015. Trophic opportunism of central Amazon floodplain fish. Freshw. Biol. http://dx.doi.org/10. 1111/fwb.12598. Nazareno, A.G., Laurance, W.F., 2015. Brazil's drought: beware deforestation. Science 347, 1427. Olivares, A.M., Hrbek, T., Escobar, M.D., Caballerao, S., 2013. Population structure of the black arowana (Osteoglossum ferreirai) in Brazil and Columbia: implications for its management. Conserv. Genet. 14, 695–703. Oliveira, C., Foresti, F., Hilsdorf, W.S., 2009. Genetics of neotropical fish: from chromosomes to populations. Fish Physiol. Biochem. 35, 81–100. Oliver, L.P., et al., 2009. Drought sensitivity of the Amazon rainforest. Science 323, 1344–1347. Pelicice, F.M., Agostinho, A.A., 2008a. Fish fauna destruction after introduction of a nonnative predator (Cichla kelberi) in a neotropical reservoir. Biol. Invasions http://dx. doi.org/10.1007/s10530-008-9358-3. Pelicice, F.M., Agostinho, A.A., 2008b. Fish-passage facilities as ecological traps in large neotropical rivers. Conserv. Biol. 22, 180–188. Pelicice, F.M., Pomppeu, P.S., Agostinho, A.A., 2014. Large reservoirs as ecological barriers to downstream movements of neotropical migratory fish. Fish Fish. http://dx.doi.org/ 10.1111/faf.12089. Pereira, L.G.H., Foresti, F., Oliviera, C., 2009. Genetic structure of the migratory catfish Pseudoplatystoma corruscans (Siluriformes: Pimelodidae) suggests homing behavior. Ecol. Freshw. Fish 18, 215–225. Petry, P., Bayley, P.B., Markle, D.F., 2003. Relationships between fish assemblages, macrophytes and environmental gradients in the Amazon River floodplain. J. Fish Biol. 63, 547–579. Phillips, O.E., et al., 2009. Drought sensitivity of the Amazon rainforest. Science 37, 637–669. Policansky, D., Magnuson, J.J., 1998. Genetics, metapopulations, and ecosystem management of fisheries. Genet. Fish. Manag. 8, S119–S123. Poulin, R., 2000. Variation in the intraspecific relationship between fish length and intensity of parasitic infection: biological and statistical causes. J. Fish Biol. 56, 123–137. Pressley, S.J., Higgins, C.L., Willig, M.R., 2010. A comprehensive framework for the evaluation of metacommunity structure. Oikos 119, 908–917. Preston, F.W., 1948. The commonness, and rarity, of species. Ecology 29, 254–283. Pringle, C.M., Freeman, M.C., Freeman, B.J., 2000. Hydrologic alterations on riverine macrobiota in the new world: tropical-temperate comparisons. Bioscience 50, 807–823. Pulliam, H.R., 1988. Sources, sinks, and population regulation. Am. Nat. 132, 652–661. Reis, R.E., Kullander, S.O., Ferraris Jr., C.J., 2003. Check List of the Freshwater Fishes of South and Central America. Edipucrs, Porto Alegre. Ribeiro, M.C.L.d.B., Petrere Jr., M.,., 1990. Fisheries ecology and management of the jaraqui (Semaprochilodus taeniurus, S. insignis) in central Amazonia. Regul. Rivers Res. Manag. 5, 195–215. Richey, J.E., Meade, R.H., Salati, E., Devol, A.H., Nordin Jr., C.F., dos Santos, U., 1986. Water discharge and suspended sediment concentrations in the Amazon River, 1982–1984. Water Resour. Res. 22, 756–764. Ricklefs, R.E., 2008. Disintegration of the ecological community. Am. Nat. 172, 741–750. Rodriguez, M.A., Lewis Jr., W.M., 1994. Regulation and stability of fish assemblages of neotropical floodplain lakes. Oecologia 99, 166–180. Rodriguez, M.A., Lewis Jr., W.M., 1997. Structure of fish assemblages along environmental gradients in floodplain lakes of the Orinoco River. Ecol. Monogr. 67, 109–128. Ronchail, J., Cpchonneau, G., Molinier, M., Guyot, J.-L., Chaves, A.G.dM., Guimara, V., Oliveira, E., 2002. Int. J. Climatol. 22, 1663–1686. Röpke, C.P., Ferreira, E., Zuanon, J., 2014. Seasonal changes in the use of feeding resources by fish in stands of aquatic macriphytes in an Amazonian floodplain, Brazil. Environ. Biol. Fish 97, 401–414. Runge, J.P., Runge, M.C., Nichols, J.D., 2006. The role of local populations within a landscape context: defining and classifying sources and sinks. Am. Nat. 167, 925–938. Salazar, L.F., Nobre, C.A., Oyama, M.D., 2007. Climate change consequences on the biome distribution in tropical South America. Geophys. Res. Lett. 34, L0970. Santos, M.C.F., Ruffino, M.L., Farias, I.P., 2007. High levels of genetic variability and panmixia of the tambaqui Colossoma macropomum (Cuvier, 1816) in the main channel of the Amazon River. J. Fish Biol. 71, 33–44. Sá-Oliveira, J.C., Hawes, J.E., Isaac-Nahum, V.J., Peres, C.A., 2015a. Upstream and downstream responses of fish assemblages to an Eastern Amazonian hydroelectric dam. Freshw. Biol. http://dx.doi.org/10.1111/fwb.12628. Sá-Oliveira, J.C., Isaac, V.J., Ferrarim, S.F., 2015b. Fish community structure as an indicator of the long-term effects of the damming of an Amazonian river. Environ. Biol. Fish 98, 273–286. Scott, M.E., 1987. Regulation of mouse colony abundance by Heligmosomoides polygyrus (Nematoda). Parasitology 95, 111–129.

L.E. Hurd et al. / Biological Conservation 195 (2016) 118–127 Thatcher, V.E., 2006. Amazon Fish Parasites. second ed. Pensoft Publishers, Sofia, Bulgaria. Thieme, M., Lehner, B., Abell, R., Hamilton, S.K., Kellndorfer, J., Powell, G., Riveros, J.C., 2007. Freshwater conservation planning in data-poor areas: an example from a remote Amazonian basin (Madre de Dios River, Peru and Bolivia). Biol. Conserv. 135, 484–501. Thomaz, S.M., Bini, L.M., Bozelli, R.L., 2007. Floods increase similarity among aquatic habitats in river-floodplain systems. Hydrobiologia 579, 1–13. Thomé-Souza, M.J.F., Chao, N.L., 2004. Spatial and temporal variation of benthic fish assemblages during the extreme drought of 1997–98 (El Niño) in the middle Rio Negro, Amazonia, Brazil. Neotropical Ichthyol. 2, 127–136. Torchin, M.E., Lafferty, K.D., Dobson, A.P., McKenzie, V.J., Kuris, A.M., 2003. Introduced species and their missing parasites. Nature 421, 628–630. Vásquez, A.G., Alonso, J.-C., Carvajal, F., Moreau, J., Nuñez, J., Renno, J.-F., Tello, S., Montreuil, V., Duponchelle, F., 2009. Life-history characteristics of the large Amazonian migratory catfish Brachyplatystoma rousseauxii in the Iquitos region, Peru. J. Fish Biol. 75, 2527–2551.

127

Verissimo, J., 1895. A Pesca na Amazônia. Livraria Clássica de Alves, Rio de Janeiro. Waldhoff, D., Sant-Paul, U., Furch, B., 1996. Value of fruits and seeds from the floodplain forests of central Amazonia as food resource for fish. Ecotropica 2, 143–156. Whittaker, R.H., 1972. Evolution and measurement of species diversity. Taxon 21, 213–215. Willis, S.C., Winemiller, K.O., Montaña, C.G., Macrander, J., Reiss, P., Farias, I.P., Orti, G., 2015. Population genetics of the speckled peacock bass (Cichla temensis). South America's most important inland sport fishery. Conserv. Genet. http://dx.doi.org/10. 1007/s10592-015-0744-y. Winemiller, K.O., Jepsen, D.B., 1998. Effects of seasonality and fish movement on tropical river food webs. J. Fish Biol. 53, 267–296. Yamamoto, K.C., Freitas, C.E.C., Zuanon, J., Hurd, L.E., 2014. Fish diversity and species composition in small-scale artificial reefs in Amazonian floodplain lakes: refugia for rare species? Ecol. Eng. 67, 165–170.