Assessing ecological communities in the wake of ongoing land use change in the Atlantic Forest and Pantanal of Brazil

Abstract

Abstract. Globally, land use change (LUC) is responsible for 20% of anthropogenic CO2 emissions, widespread losses of ecosystem services and biodiversity, and erosion of ecosystem resilience to disturbance and climate change. Although LUC has peaked in developed northern temperate countries, it is ongoing in wilderness-rich equatorial countries, where economic and governmental incentives continue to drive expansion of agricultural development and natural resource extraction. In Brazil, vast tracts of tropical forest and other types of natural vegetation cover have been (and are being) replaced by crops and planted exotic (non-native) pasture, the latter for livestock rearing. I investigated how a range of ecological communities are being affected by ongoing LUC in two Neotropical biomes, the Atlantic Forest and Pantanal of Brazil. My objectives were to: (1) help understand the consequences of LUC, (2) provide baselines for current and future monitoring studies, and (3) help prioritize conservation actions. The Atlantic Forest, which originally covered 1.5 million km2 of eastern Brazil, is a biodiversity hotspot with a range of tropical forest formations and a high proportion of endemic species. Five-hundred years of LUC, including coffee and sugarcane cultivation, timber extraction, wood charcoal production for industry, and urban development, shrank forested area to 11% of original coverage, creating a highly-fragmented landscape of mostly small (<100 ha) forest patches within an agriculturally-dominated matrix. In a fragmented region of inland seasonal forest, called the Planalto or interior Atlantic Forest, I investigated the state of headwater stream fish communities 80 years after much of the original vegetation was cleared for coffee plantations. The Pantanal, which covers 150,500 km2 in central-western Brazil, southeastern Bolivia and northeastern Paraguay, is a highly-seasonal floodplain comprised of tropical forest, savanna, and wetland formations supporting large populations of fishes, waterbirds, and charismatic wildlife species. Since the late 1800s, LUC in the Pantanal has centered around extensive cattle ranching and more recently (since the 1960s) the expansion of grazing lands through conversions of native vegetation to planted exotic pasture. In the southern Pantanal, I investigated how cattle activity and ranching practices are affecting: (1) the ii macroinvertebrate and waterbird communities of enrichment-prone alkali soda lakes, and (2) the mammalian and avian forest communities that aggregate at fruiting trees. Because historical records of biodiversity were lacking and focal species of conservation importance were largely unidentified in the study regions, community-level assessments were ideal starting points for evaluating the consequences of LUC. The assessments related abundance trends of multiple species to major environmental gradients, allowing me to characterize communities in relation to LUC and identify indicator species with strong responses to LUC. Analyses included ordinations that described species compositional trends along LUC gradients, univariate and multivariate comparisons of community composition between categories representing different levels of LUC, and indicator species analyses that identified species characteristic of particular LUC categories. Despite the wide array of taxa and environments investigated, there were a number of common communitylevel responses to increasing intensity of LUC: (1) turnover and loss of species and guilds, (2) loss of environmental and biotic heterogeneity, (3) dominance by enrichment- and disturbance-tolerant species, (4) bottom-up trophic cascading, and in one case (5) an ecosystem regime shift. Applying results from the assessments of ecological communities and environments (Chapters 2–4, summarized below), I presented conservation priorities aimed at curbing impacts from LUC and maintaining biodiversity and ecosystem services. A key challenge will be convincing stakeholders, which are mostly private landowners, to change and improve their land use practices. To that end, successful approaches will need to balance environmental and economic sustainability. An effective conservation strategy will combine: (1) dissemination of basic research results demonstrating impacts from LUC, (2) additional research on sustainable management practices, (3) a landowner outreach program demonstrating the economic and long-term environmental benefits of adopting sustainable land use practices, and (4) an environmental education program that introduces sustainable practices to future landowners, land managers, and rural laborers. iii Chapter summaries of community-level responses to LUC. In streams of a highly fragmented Atlantic Forest landscape (Chapter 2), accelerated erosion and channel aggradation in deforested agricultural basins shifted the stream food web from one based on forest-generated detritus to one based on riparian macrophytes and epiphytic algae. This in turn caused a compositional shift in the fish community from detritivorous and insectivorous species, which were characteristic of forested streams, to highly-abundant herbivorous and omnivorous species characteristic of marsh-dominated agricultural matrix streams. These changes in agricultural matrix streams were accompanied by increased dominance of fish species tolerant of aggraded stream-channel conditions, and loss of longitudinal environmental and biotic heterogeneity among sites, the latter caused by accelerated sedimentation, burial of natural barriers to fish movements, and upstream colonization by disturbance-tolerant species. In this fragmented landscape, prevention of additional deforestation is a top priority, and protection and restoration of riparian forests will be essential for maintaining the biodiversity and ecosystem services of Atlantic Forest streams. In Pantanal forests, long- and short-term impacts from cattle activities, as well as unsustainable ranching practices, reduced the heterogeneity and availability of fruit sources, which in turn reduced the diversity and abundance of mammalian and avian frugivores (Chapter 4). As frugivore diversity and abundance declined, the seed dispersal services provided by frugivores also declined, creating a negative feedback loop that is expected to further reduce the diversity and availability of fruit sources and frugivores. If current trends continue, frugivory in Pantanal forests will be dominated by cattle, and forest vegetation will be comprised of a species-poor homogenized assemblage of cattle-dispersed plants. To maintain the diversity and unique interactions that characterize the fruiting trees and frugivores of the Pantanal, it will be necessary to preserve standing forests, rather than convert them to exotic pasturelands, and promote novel cattle management practices that confine cattle activities to pasture environments, alleviating impacts to forest and aquatic (see below) environments. In alkali soda lakes of the Pantanal, nutrient and organic matter enrichment associated with high levels of cattle activity caused an ecosystem regime shift from a high water-transparency macrophyte- and benthic iv algae-dominated state to a high biogenic-turbidity phytoplankton-dominated state (Chapter 3). The shift to a phytoplankton-dominated state was associated with increasingly extreme physicochemical conditions, increased water-column primary productivity and respiration, a loss of depth-related environmental and biotic heterogeneity, and impoverishment of macroinvertebrate and waterbird communities. An array of trophic pathways that supported a diversity of macroinvertebrate and waterbird species in soda lakes exposed to low levels of cattle activity was replaced in high cattle-activity soda lakes by 1 dominant phytoplankton-based trophic pathway that supported 1 superabundant macroinvertebrate species and a few abundant waterbird species. In these high cattle-activity lakes, negative feedback loops associated with nocturnal oxygen depletion, as well as bioturbation by superabundant macroinvertebrate and waterbird populations, stimulated nutrient releases from sediments, exacerbating enriched conditions. Long-term recovery of nutrient-enriched cattle-impacted soda lakes will require management practices that limit cattle activity near soda lakes, as well as preservation of lake-side forests, which maintain regional hydrologic balance. In addition to promoting novel cattle management practices (as mentioned above for Pantanal forests), providing incentives for establishment of private reserves and ecotourism operations will help maintain the unique biotas of alkali soda lakes in the Pantanal.