Understanding Biodiversity Loss

IALE Landscape Research and Management papers

Understanding Biodiversity Loss

An overview on Forest Fragmentation in South America

Edited by Maria Jose Pacha, Sandra Luque, Leonardo Galetto, Louis Iverson

ISSN: 1570-6532

SANDRA LUQUE, MSc PhD Research Director at Cemagref, France. Landscape ecologist working in the development of rapid assessment methods for biodiversity evaluation and comprehensive landscape monitoring and modelling. At the present she serves as elected vice-president for IALE (International Association for Landscape Ecology). She is serving in the board of the IALE-IUFRO Working Group on Forest Landscape Ecology. For 18 years she has been working on remote sensing/GIS and landscape ecology issues in relation to change detection, forest ecology, biodiversity indicators, and biodiversity habitat quality models. At the present, she is the co-ordinator of the scientific group of experts on forest fragmentation and biodiversity loss in South-America. She is merited coordinator of international and national level research projects. MARIA JOSE PACHA MSc PhD working for the NGO Fundacion Vida Silvestre, Argentina. Her background is in Vegetation Ecology with links at the landscape level. She has worked in applied conservation projects in Argentina and the United Kingdom in Nature Reserves and National Parks. Also her research interests are about linking social and natural aspects of conservation. She has worked as project coordinator of the Atlantic Forest Programme of FVSA and at the present she coordinates a WWF UNESCO project on financing mechanisms for protected areas in South America. LEONARDO GALETTO PhD Researcher from CONICET (National Research Council of Argentina) and Professor at the Universidad Nacional de Córdoba (Argentina), working with plant reproductive ecology and forest fragmentation for the last 18 years. He is particularly interested about the changes on animal-plant interactions due to habitat loss, studying different processes as pollination, dispersion and herbivory. He also has worked in ethnobotany and the relationship of rural people perception of local resources and their availability. At present supervises doctoral and postdoctoral students working with different approaches in the Chaco forest. He is Editor of Kurtziana (a regional journal of Botany) and involved as Vice-President with the Asociación Argentina de Ecología LOUIS R. IVERSON PhD Research Landscape Ecologist for United States Forest Service in Delaware, Ohio. Vice-president for International Association for Landscape Ecology and book review editor for the journal, Landscape Ecology. His current research concerns potential changes in tree species following climate change in the United States, the use of fire and thinning to restore oak-hickory forest communities, and modelling the advance of the emerald ash borer, an insect killing ash trees in the central United States.

IALE electronic publication series

Understanding Biodiversity Loss: An overview on Forest Fragmentation in South America

Edited by Maria Jose Pacha, Sandra Luque, Leonardo Galetto, Louis Iverson

Target audience: The publication is aimed at advanced undergraduate and graduate students, researchers and teachers, professional landscape ecologists, policy makers and practitioners with a special interest in South American forest issues.

Aim: to demonstrate the contribution that Landscape Ecology can make to forest management and the understanding of forest fragmentation and biodiversity loss. This publication is specially targeted to describe and determine causes of forest fragmentation in South America. The papers that contribute to the series present study cases from Argentina, Chile and Southern Brazil. The different forests that are covered in this issue are: Atlantic Forest of Argentina and Brazil, Gran Chaco (Argentina, Paraguay and Bolivia), Valdivian Forest (south of Argentina and Brazil) and Yungas rainforest (Argentina and Bolivia).

Cite as:

Pacha, M.J., Luque, S., Galetto, L. and Iverson, L. (2007) Understanding biodiversity loss: an overview of forest fragmentation in South America. IALE Landscape Research and Management papers. International Association of Landscape Ecology

Preface This publication is the result of a series of papers presented during the workshop organized in Bariloche, Argentina: “Understanding Biodiversity Loss: A Workshop on Forest Fragmentation in South America “ (26 – 30 June 2006). The workshop allowed an assessment of the situation for the region and provided an analysis of the state of the art on the subject and an identification of gaps in research. More importantly the activity, funded by MEDD, France (Ministry of ecology and sustainable development); allowed an assembly of experts working on the evaluation of temperate and subtropical forests within the region (South America). The workshop was focused within the framework of the “Paris Declaration for the biodiversity" (Paris Conference, January 2005). As the Declaration states, we aim at bringing researchers together from developing countries and reinforcing the links between North and South in order to work towards an improved protection of biodiversity. The overall goal of creating a network of experts working more precisely on forest fragmentation, biodiversity loss and conservation issues targeted two main issues: i) improve the knowledge and the relevance of the indicators that can be developed and used in relation to forest biodiversity loss in South America. ii) facilitate building capacity not only in monitoring and evaluating forest fragmentation but also on forest restoration to mitigate the existing trends on biodiversity loss for the region. Sandra Luque

S Luque

Introduction Overview of Biodiversity Loss in South America S. Luque & M.J. Pacha In response to global concern over the rapid loss of the world’s biodiversity, the 6th Conference of the Parties of the Convention on Biological Diversity (CBD) adopted a global target to reduce the rate of biodiversity loss by 2010 (CBD 2002). This target, which was later endorsed by the World Summit on Sustainable Development (United Nations 2002), has also been adopted by a number of regional scale policies and processes. The European Union Sustainable Development Strategy (2001a) and various other European Union policies (EC 1998, 2001b, c) set similar or even more ambitious biodiversity goals. The Pan-European Ministerial ‘Environment for Europe’ process adopted a resolution on halting the loss of biodiversity by 2010 (UN/ECE 2003). This widespread adoption of targets for reducing the rate of biodiversity loss has highlighted a need for indicators that will allow policy makers to track progress towards these ambitious goals. Recognising this need, the Convention of the parties (CoP) of the CBD identified a series of biodiversity indicators for immediate testing (UNEP 2004). Such indicators are needed at national, regional and global levels. In June 2004, the Environment Council of the EU adopted a set of 15 headline indicators for biodiversity to evaluate progress towards the 2010 target (Council of the European Union 2004). This set of indicators was recommended by the EU Biodiversity Expert Group and its Ad Hoc Working Group on Indicators, Monitoring and Assessment, and the Malahide stakeholder conference (Anonymous 2004). Both the CBD decision and the European documents recommend, among other indicators for immediate testing, indicators of trends in abundance and distribution of selected species. Species trend indicators are considered a sensitive measure of biodiversity change (Balmford et al. 2003; Ten Brink et al. 1991; Ten Brink 2000), and one such approach, composite species trend indicators, has been increasingly applied. In addition to the global-scale Living Planet Index (Loh 2002, and this volume) there are several instances of the successful implementation of such indicators, principally at national scales (Jenkins et al. 2004). The UK Headline indicator of wild bird populations (Gregory 2003a) is one example. The European Bird Census Council (EBCC) has used a similar approach to develop the Pan-European Common Bird Index for farmland and forest birds (Gregory 2003b; Gregory et al. 2004). Another set of indicators is directly related to forest biodiversity and, in particular, to forest cover loss. Valid indicators for this target area are poorly developed. During the International Conference on Biodiversity (Paris, January 2005), the workshop on “Biodiversity Indicators and the 2010 target: scientific challenges in meeting and assessing progress towards the 2010 biodiversity targets and related goals”; identified forest fragmentation as a key indicator to be added to the list. However, consensus and work is needed in the application of the indicator as a tool for monitoring forest status within the Action 6 framework.

We need also to reach a consensus on the use of the indicator according to general guidelines. These guidelines, as established in the CBD 2010 targets, need to be set in order to develop suitable indicators for informing the general public on biodiversity trends. The indicators should match the set of requirements as listed in the CBD general guidelines and principles for developing national-level biodiversity monitoring programmes and indicators (UNEP 2003a). These principles require that an indicator be, among other characteristics: policy and biodiversity relevant; scientifically sound; broadly accepted; affordable to produce and update; sensitive; representative; flexible; and amenable to aggregation. Within this context, we intend to build up this network in order to reinforce the local capacity of different actors and to coordinate actions to prepare integrated projects at the international level that can have an impact at the global level. The goal is to work on native forest that has a particular important biodiversity value and that is been neglected until now in International projects. It is only with a proactive co-operation between North and South that we will be able to reach the targets set up by the CBD 2010 and reinforced during the Biodiversity Conference in Paris (January 2005). At present, we have an exchange group but not an actual official network due to the lack of funding (FRAGFORNET (http://sympa.lyon.cemagref.fr/wws/info/fragfornet)). But in the light of our efforts we hope to be able to develop a network based on this core group of experts that work on forest fragmentation, biodiversity loss and conservation issues. The aims are to improve the knowledge and the relevance of the indicators that can be developed, in particular for temperate and subtropical forested habitats in South America. We organize our network of contacts on the basis of decisions taken at the time of the seventh conference of the parts of the CBD (UNEP/CBD/COP/7/21), which relates to the biological diversity of forests (VII/I). The general objective is to build a flexible network in which national and regional actions can be developed according to the priorities and nationals’ interests in order to implement an effective program for the protection of biodiversity in the future.

The critical importance to help reduce the rate of biodiversity loss by 2010 in South America Biological diversity, the variety of all forms of life on Earth, plays a critical role in meeting human needs directly while also maintaining the ecological processes upon which our survival depends (BSP 1996). By any standard of measure, the Latin American and Caribbean (LAC) region is the repository of some of the world’s richest biodiversity, containing 40% of Earth’s plant and animal species (Global Environment Outlook 2000). Nine of the 25 most biodiverse countries are located in the LAC region (Caldecott et al. 1994). Of the 229 terrestrial ecoregions (geographically distinct assemblages of natural communities that share a large majority of species, dynamics and environmental conditions) designated in the region by the World Wildlife Fund (WWF), 57 are considered to be highest priority for conservation at the regional scale. Although South America still maintains vast areas of intact tropical and temperate forest, the region’s biodiversity is facing significant and growing

threats, including increased rates of deforestation. One of the problems to monitor, manage and restore biodiversity is the unequal distribution of funding (Castro and Locker 2000). In order to effectively mitigate these threats, practitioners and donors in the conservation community must work together with host countries to improve the conservation of the region’s biodiversity. The biological importance of the native forest1 Most of the native forests that are represented in this publication are amongst the most threatened ecoregions around the world and are included in the Global 2000 list of the Worldwide Fund for Nature (WWF) which includes representations of all major habitat types in each major biogeographic unit of the world and aims to prioritize conservation actions worldwide. They are located in the southern part of South America. Chile and Argentina together harbour the largest temperate rainforest area of South America, and more than half of the temperate forests in the Southern Hemisphere (Donoso, 1993; Wilcox, 1996). These forests are classified as temperate rainforests because of their geographical location outside the tropics, and because they experience high rainfall and low temperatures in winter. Similar forests are found in Tasmania, New Zealand and the Pacific Northwest in North America. Forests in South America are important as they store vast quantities of carbon that contribute to global climate regulation, flood control, water purification and soil nutrient cycling, as well as providing habitat for a high diversity of species that contribute to the genetic material for valuable new products and a foundation for the resilience of natural systems. First, the Valdivian Rainforest Ecoregion is located from 36º S through 48º S, and extends from southern Chile and adjacent Argentina This ecoregion includes terrestrial, freshwater and coastal marine ecosystems and is a worldwide reserve due to its unique biodiversity and high biogeographic and ecological significance, covering a total area of 10.5 million ha. These forests have a high diversity of trees and shrub species (over 35% of tree genera are endemic), and are the habitat of 60 endemic bird and 38 endemic mammal species. Reports indicate 23% of endemism for reptiles, 30% for birds, 33% for mammals, 50% for fish and 76% for amphibians. Forests in southern Chile include Fitzroya cupressoides, the second longest living tree worldwide that may live over 3600 years. These forests are home to over 900 vascular plant species, including 60 tree species, of which over 90% are endemic (Arroyo et al., 1995). Due to their special biodiversity assemblages, the Valdivian forests provide important ecosystem services that are the basis for several relevant economic activities, including the conservation of biological diversity of aquatic ecosystems, water production (quantity and quality), salmon farming (accounting for over 1 billion US dollars of annual exports, which represent 80% of the exports from the Lake Region in southern Chile), sport fishing and ecotourism.

1

In collaboration with Adriana Rovere, Cristian Echeverria and Daniel Somma

Secondly, the Gran Chaco Americano is the largest dry forest in South America and the continent’s most extensive forested region after Amazonia. It occupies territories in Argentina, Paraguay, Bolivia, and Brazil. Extending from tropical latitudes (18°S) to subtropical zones (31°S), the Chaco shows strong climatic gradients, generating (together with geological and topographic characteristics) a range of environments: wide plains, swamps, and dry or seasonally flooded savannas, marshes, salt flats, and a variety of forests and scrublands. This landscape diversity translates into a high diversity of animal and plant species that makes the Chaco a key area for biodiversity conservation. The Chaco also provides important resources for the local population and industries. Because of the fragility of the Gran Chaco’s natural resources, the irreversibility of many of the changes that have already taken place in the region, and the human pressures currently underway, urgent steps are needed to reconcile productive sector interests with the sustainable development and conservation of the region’s natural resources through a long-term vision (TNC et al., 2005). Thirdly, the Yungas rainforest in the Northwest of Argentina and South of Bolivia. is considered the most extensive biodiversity hotspot in Argentina. There are about 5,000,000 hectares that host 60% of the total bird species of Argentina and some key mammals like the jaguar (Panthera onca) and the Andean taruca (Hippocamelus antisensis). The latter is a mountain deer nominated as a “national natural heritage”. At the same time, this subtropical rainforest acts as an important carbon sink, an important water producer and the watershed shield for an extended irrigated region. The Yungas conditions are changing rapidly, as indicated by an estimated 1,250,000 hectares converted to agriculture between 1975 and 1988 (no later figures are available for the whole region). Synergy will be strengthened in the near future with ongoing local initiatives such as the Yungas biosphere reserve project and the local reforestation program in old degraded forests. The forth important native forest that is described in this publication is the Atlantic Forest which is one of the most endangered and biologically diverse biomes on the Earth. It is one of the Conservation International's “Hot Spots; it is a WWF Global 200 Ecoregion; it overlaps significantly with Birdlife International's Endemic Bird Areas of the World: and it is one of the IUCN/WWF Centers of Plant Diversity. Unfortunately, with high human population densities and deep soils, only 7% of the original forest cover remains intact. Fifteen different ecoregions can be recognized within the Atlantic Forest biome. The Upper Parana Atlantic Forest constitutes the southwestern portion of the Atlantic Forest ecoregion, and extends from the western slopes of the Serra do Mar in Brazil to eastern Paraguay and the Misiones Province in northeast Argentina. A continuous subtropical semideciduous forest originally covered this entire area with a high diversity of animal and plant species. Deforestation in the Upper Parana Atlantic Forest has been most severe in Brazil, where only 3% of the original forest remains intact, most of it in protected areas. Paraguay has approximately 10% of its original forest cover remaining, though with a very high deforestation rate. In contrast, approximately 50% of the original cover remains intact in the Argentine province of Misiones with the bulk located in the almost continuous

block in the northcentral portion of the province. This is the largest remaining forest block of the Atlantic Forest, and still contains the original set of large vertebrates including top predators such as harpy eagles, crested eagles, jaguars, pumas, and ocelots, and large herbivores such as tapirs, brocket deer, and peccaries (Di Bitetti et al 2003).

Pressures on the native forests All the native forests that have been described are suffering from intensive pressures that are threatening the biodiversity and persistence of the ecoregions in the short and long term. In the Valdivian forests, the main threats are degradation due to inadequate logging practices, and destruction due to conversion to agriculture and fast-growing plantations, as well as human-set fires. Current estimates indicate that only 10-20% of the native forests are adequately managed. Most of them are either destroyed or degraded, leading to important negative reductions on the biological diversity of terrestrial ecosystems, as well as on the ecosystem services which depend on this diversity. The Gran Chaco and the Yungas are suffering irrational logging for soybean and cane plantations that bring huge revenues to the international market. Large areas are being cleared and small subsistence farming and communities are losing not only their ecosystem but also their identity and livelihood. On the other hand, the greatest threat to biodiversity in the Upper Parana Atlantic Forest ecoregion is the extreme degree of forest fragmentation and degradation due to agricultural expansion. Large-scale soybean production, pine plantations, and pasture for cattle ranching, as well as small-scale tobacco, yerba-mate plantations and subsistence agriculture are all contributing to forest fragmentation across the ecoregion. Other causes of forest conversion and degradation include squatting by landless people, the construction of infrastructure (dams, roads, etc.), illegal and unsustainable hunting of wildlife, and unsustainable exploitation of the native forest. However, despite the high degree of forest fragmentation across the ecoregion, there are still excellent opportunities for conserving remaining large forest blocks, as is the case with the Misiones. As a result of all the pressures on these native forests, they are being degraded, fragmented and large areas of forest are being lost.

Forest fragmentation within the context of this publication Fragmentation is simply the disruption of continuity (Lord and Norton 1990). When defined in this manner, the concept of fragmentation can be applied to any domain in which continuity is important to the functioning of ecosystems (Lord and Norton 1990). In a restricted way, fragmentation occurs when a

large expanse of habitat is transformed into a number of smaller patches of smaller total area, isolated from each other by a matrix of habitats unlike the original (Wilcove et al. ,1986). The fragmentation of natural habitats is usually a result of the expansion of land use that accompanies human population growth. As fragmentation proceeds, average fragment size and total fragment area decreases and insularity of fragments increases (Moore 1962; Webb and Haskins 1980; Burgess and Sharpe 1981). Habitat fragmentation and forest loss have been recognized as a major threat to ecosystems worldwide (Armenteras et al. 2003; Dale and Pearson, 1997; Iida and Nakashizuka, 1995; Noss, 2001). These two processes may have negative effects on biodiversity, by increasing isolation of habitats (Debinski and Holt, 2000), endangering species, and modifying species’ population dynamics (Watson et al., 2004). Fragmentation may also have negative effects on species richness by reducing the probability of successful dispersal and establishment (Gigord et al., 1999; Luque et al., 1994; Luque 2000) as well as by reducing the capacity of a patch of habitat to sustain a resident population (Iida and Nakashizuka, 1995). For example, fragmentation of the Maulino temperate forest in central Chile has affected the abundance of bird richness (Vergara and Simonetti, 2004) and regeneration of shade-tolerant species (Bustamante and Castor, 1998), and has also favoured the invasion of alien species (Bustamante et al., 2003). The ecological consequences of fragmentation can differ depending on the pattern or spatial configuration imposed on a landscape and how this varies both temporally and spatially (Armenteras et al., 2003; Ite and Adams, 1998). Some studies have shown that the spatial configuration of the landscape and community structure may significantly affect species richness at different scales (Steiner and Köhler, 2003). Other authors emphasise the need to incorporate the spatial configuration and connectivity attributes at a landscape level in order to protect the ecological integrity of species assemblages (Herrmann et al., 2005; Piessens et al., 2005). The dynamics of populations inhabiting terrestrial habitat fragments have received considerable research attention, including studies of birds, mammals, invertebrates, and plants (see Herkert 1994 and references therein). Perhaps the most extensively studied system thus far is the breeding birds of eastern North American deciduous forest, where several researchers have shown that habitat fragmentation adversely affects many forest birds species (Herkert 1994; Robbins 1979, Leck et al. 1981, 1988, Askins and Philbrick 1987, Johnston and Hagan 1992). Although there is general agreement on the effects of fragmentation on breeding birds within forest habitats, the mechanisms that account for these trends are not clear (Lynch 1987, Martin 1988). There is a need for studies that provide a quantitative treatment of landscape pattern changes and dynamics to better understand the widespread population decline of several species in fragmented landscapes. In this sense, this publication provides case studies that serve as examples of the type of research that needs to be integrated in collaborative projects. In order to better understand fragmentation, we need to be able to compare different study sites and species information to target the many unresolved questions that exist within the subject, as has been pointed out by several authors (Farigh 2003, Bissonette & Storch 2002).

About this publication The contributions for this publication are grouped considering three main topics that are considered relevant in these ecoregions such as the need for science-based information on the status of these native forests and to suggest practical tools for their conservation, restoration and management. In this sense all the contributions use different concepts of landscape ecology and they show the growing importance of this discipline in integrating different concepts of remote sensing, geographical information systems, description of the natural world and incorporation of socio economic and cultural aspects. In the first part, Multitemporal changes and forest status, the authors describe changes in forest cover over a period of time using mainly remote sensing techniques. They evaluate the composition and landscape structure of the remaining forest fragments. The first four contributions are from the Gran Chaco region: Torrella et al. describes the changes that have occurred in nearly 50 years in a particular area covering 72.000 ha and have analysed the surface reduction and the forest fragmentation using several landscape indices. Guinzburg et al. used satellite images to analyse the effect of agriculture expansion in the last ten years in an area of approximately 4 million ha. Parmunchi et al. concentrated on the last 5 years and identified landscape pattern changes in the Chaco region along a precipitation gradient beween 1,100 and 600 mm annual rainfall isolines. Menghi and Sueldo focused on the mountainous Chaco region in Central Argentina and described the landscape spatial structure and habitat diversity of Polylepis astralis and Lithraea ternifolia forests in high mountain areas. Echeverria et al. analysed land-use change and spatial patterns in the Valdivian forest over twenty years using landscape indicators and satellite images. Finally, Vibrans et al. presented a forest inventory and characterised the phytosociology of a section of the Atlantic Forest in the Santa Catarina state in Brazil. In the second part of the publication, Ecological consequences of forest fragmentation, three contributions aim at exploring some aspects on how fragmentation can affect interactions among species and ecological processes. Vidella et al. analysed the effect of fragmentation on herbivory, comparing fragmented and continuous patches of the Chaco forests in Central Argentina. Marchelli et al. concentrated more on reproductive characteristics and evaluated the degree of connectivity between populations of three tree species of the Valdivian Forests of Argentina through pollen flow studies. Finally in this section, Echeverria et al. described the effect of landuse change on the water regime in ñadi soils in the Valdivian Forest of Chile. The third part, Landscape Ecology for conservation, management and restoration, has contributions of different authors that want to show how the landscape approach can be used to provide recommendations for better management of natural resources based on local needs. Lencinas et al. focused on analysing how alternative forest management methods can improve the species richness of Nothofagus pumilo forests in Patagonia, taking into consideration a landscape level approach. On the other hand,

Bachmann et al. introduced different land-use planning strategies in the Yungas region of Argentina, aimed at improving biodiversity conservation. Lara et al. presented the case of experimental plantation of Fitzorya cupressoides, aimed at genetic conservation of the species in southern Chile. Finally, Pacha et al. showed how landscape ecology can serve as a useful tool to identify conservation priorities and help to plan actions for conservation at the species, landscape and ecoregional level in the Upper Parana Atlantic Forest. We hope that these contributions can foster further research and actions to conserve the temperate forests of South America and their rich ecosystems.

References Aguilar, R., and L. Galetto. 2004. Effects of forest fragmentation on male and female reproductive success in Cestrum parqui (Solanaceae). Oecologia 138: 513-520. Aguilar, R; L Ashworth; L Galetto, MA Aizen. 2006. Plant reproductive susceptibility to habitat fragmentation: review and synthesis through a meta-analysis. Ecology Letters 9: 968-980. Anonymous. 2004. Final message from Malahide. Halting the loss of biodiversity. Priority objectives and targets for 2010. http://www.eu2004.ie/templates/document_file.asp?id=17810 (downloaded 1 Oct. 2004). Armenteras, D., F. Gast, H. Villareal. 2003. Andean forest fragmentation and the representativeness of protected natural areas in the eastern Andes, Colombia. Biological Conservation 113: 245-256 Armesto, J.J, Rozzi, R., Smith-Ramírez, C., Arroyo, M. T. K. 1998. Conservation Targets in South American Temperate Forests. Science, Vol 282, Issue 5392, 1271-1272 Arroyo, M. K. , Cavieres, L., Peñaloza , Riveros, M., A. M. Faggi. 1995. Relaciones fitogeográficas y patrones regionales de riqueza de especies en la flora del bosque lluvioso templado de Sudamérica. In: Ecología de los Bosques Nativos de Chile. J. J. Armesto, C. Villagrán, M. K. Arroyo (Eds). Santiago. Chile. pp 71-92. Askins, R.A., and M.J. Philbrick. 1987. Effects of changes in regional forest abundance on the decline and recovery of a forest bird community. Wilson Bulletin 99:7-21. Balmford, A., Green, R.E., and Jenkins, M. 2003. Measuring the changing state of nature. Trends in Ecology and Evolution 18: 326-330. Biodiversity Support Program, Conservation International, The Nature Conservancy, World Resources Institute and World Wildlife Fund. 1995. A regional analysis of geographic priorities for biodiversity conservation in Latin America and the Caribbean. Washington, D.C.: Biodiversity Support Program. Biodiversity Support Program. 2000. Final proceedings of donors workshop on funding for biodiversity conservation and ecoregional planning in the Latin American and Caribbean region, December 13-14. Washington, D.C. : World Bank, USAID, BSP. Bissonette J. A. and Storch I. 2002. Fragmentation: Is the Message Clear? Conservation Ecology 6(2): 14. URL: http://www.consecol.org/vol6/iss2/art14 Burgess, R. L. and D. M. Sharpe, eds. 1981. Forest island dynamics in man-dominated landscapes. Springer Verlag, New York.

Bustamante, R., Serey, I., Pickett, S. T., 2003. Forest fragmentation, plant regeneration and invasion processes in Central Chile. In: Bradshaw, G., Marquet, P. (Eds.), How landscapes change: Human disturbance and ecosystem fragmentation in the Americas. Springer- Verlag, pp. 145 - 160 Bustamante, R., Castor, C., 1998. The decline of an endangered ecosystem: The Ruil (Nothofagus alessandrii) forest in Central Chile. Biodiversity and Conservation 7, 1607-1626. Caldecott, J.O., M.D. Jenkins, T. Johnson and B. Groombridge. 1994. Priorities for conserving global species richness and endemism. In World Conservation Monitoring Centre, Biodiversity Series No. 3, ed. N. Mark Collins, 17. Cambridge, UK: World Conservation Press. Retrieved March 17, 2000 from the World Wide Web: http://www.nih.gov/fic/opportunities/countries.html. Castro, G. and I. Locker. 2000. Mapping conservation investments: An assessment of biodiversity funding in Latin America and the Caribbean. Washington, D.C.: Biodiversity Support Program. Castro, G. and I. Locker. 2000. Mapping conservation investments: An assessment of biodiversity funding in Latin America and the Caribbean. Washington, D.C.: Biodiversity Support Program. Castro, G. and I. Locker. 2000. Mapping conservation investments: An assessment of biodiversity funding in Latin America and the Caribbean. Washington, D.C.: Biodiversity Support Program. CBD. 2002. Decision VI/26 Strategic Plan for the Convention on Biological Diversity. (https://www.biodiv.org/decisions/default.aspx?lg=0&dec=VI/26 Accessed 1 October 2004). ——. 1996. Biodiversity: Facts on the foundation of life. Washington, D.C.: Biodiversity Support Program. Council of the European Union. 2004. Halting the loss of biodiversity by 2010 – Council conclusions. 10997/04, unofficial information note. (http://www.ibot.cas.cz/biop/texts/Council_Conclusions_30_June_2004%5B1%5D.d oc Accessed 1 October 2004). DALE, V.H., S. M. PEARSON. 1997. Quantifying habitat fragmentation due to land use change in Amazonia. In: Laurance, W., R. Bierregaard (eds.). Tropical Forest Remnants. The University of Chicago Press, Chicago, pp. 400-414. Debinski, D., Holt, R., 2000. A survey and overview of habitat fragmentation experiments. Conservation Biology 14, 342-355. Di Bitetti, M.S., G. Placci and L.A. Dietz, 2003. A biodiversity vision for the Upper Parana Atlantic Forest ecoregion: designing a biodiversity conservation landscape and setting priorities for conservation action. World Wildlife Fund, Washington D.C., 145 p. Donoso, C., 1993. Bosques templados de Chile y Argentina: variación, estructura y dinámica. Universitaria, Santiago, Chile. EC. 1998. Communication of the European Commission to the Council and to the Parliament on a European Community Biodiversity Strategy. Brussels. (http://europa.eu.int/comm/environment/docum/9842en.pdf Accessed 1 October 2004) EC. 2001a. A Sustainable Europe for a Better World: A European Union Strategy for Sustainable Development. COM(2001)264 final – Brussels. (http://www.europa.eu.int/eur-lex/en/com/cnc/2001/com2001_0264en01.pdf Accessed 1 October 2004) EC. 2001b. Communication from the Commission to the Council and the European Parliament - Biodiversity Action Plans in the areas of Conservation of Natural Resources, Agriculture, Fisheries, and Development and Economic Co-operation.

Brussels. (http://biodiversity-chm.eea.eu.int/convention/cbd_ec/strategy/BAP_html Accessed 1 October 2004). EC. 2001c. Environment 2010: Our future, Our choice – the Sixth Environment Action Programme. COM(2001) 31 final – Brussels. Fahrig L. 2003. Effects of habitat fragmentation on biodiversity Annu. Rev. Ecol. Evol. Syst. 34:487–515 Galetto, L., Aguilar R., Musicante M., Astegiano J., Ferreras A., Jausoro M., Torres C., Ashworth L., Eynard C. 2007. Ecología Austral 17: 67-80. Gigord, L., Picot, F., Shykoff, J. 1999. Effects of habitat fragmentation on Dombeya acutangula (Sterculiaceae), a native tree on La Réunion (Indian Ocean). Biological Conservation 88, 43-51. Gregory, R.D., Eaton, M.A., Noble, D.G., Robinson, J.A., Parsons, M., Baker, H., Austin, G. and Hilton, G.M. 2003a. The state of the UK’s birds 2002. The RSPB, BTO, WWT and JNCC, Sandy, UK. Gregory, R.D., Vorisek, P., Van Strien, A.J., Eaton, M. and Wotton, S.R. 2003b. From bird monitoring to policy-relevant indicators. A report to the European Topic Centre on Nature Protection and Biodiversity. Herkert, J.M. 1994. The effects of habitat fragmentation on midwestern grassland bird communities. Ecological Applications 4(3):461-471. Herrmann, H., Babbitt, K., Baber, M., Gongalton, R., 2005. Effects of landscape characteristics on amphibian distribution in a forest-dominated landscape. Biological Conservation 123, 139-149. Iida, S., T. Nakashizuka. 1995. Forest fragmentation and its effect on species diversity in suburban coppice forests in Japan. Forest Ecology & Management 73: 197-210 Ite, U. E., Adams, W. M., 1998. Forest conversion, conservation and forestry in Cross River State, Nigeria. Applied Geography 18, 301-314 Jenkins, M, Kapos, V. & Loh, J. 2004. Rising to the biodiversity challenge. Draft discussion paper for CBD CoP7. World Bank, UNEP-WCMC and WWF. Johnston, D.W., and J.M. Hagan III. 1992. An analysis of long-term breeding bird censuses from eastern deciduous forest. Pp. 75-84 in Ecology and Conservation of Neotropical Landbirds. J.M. Hagan III and D.W. Johnston (eds.). Smithsonian Institute, Washington, D.C., USA. Leck, C.F., B.G. Murry, Jr., and J. Swinebroad. 1981. Changes in breeding bird population at HutchesonMemorial Forest since 1958. Hutcheson Memorial Forest Bulletin 6:8-14. Leck, C.F., B.G. Murry, Jr., and J. Swinebroad. 1988. Long-term changes in the breeding bird populations of a New Jersey forest. Biological Conservation 46:145-157. Loh, J. 2002. Living Planet Report 2002. World Wide Fund for Nature, Gland, Switzerland. Lord, J.M. and D.A. Norton. 1990. Scale and the spatial concept of fragmentation. Conservation Biology 4:197-202. Luque, S, R. G. Lathrop Jr, and J. A. Bognar. 1994. Temporal and Spatial Changes in the New Jersey Pine Barrens Landscape. Landscape Ecology 9(4):287-300. Luque, S. 2000. Evaluating Temporal Changes Using Multispectral Scanner and Thematic Mapper Data on the Landscape of a Natural Reserve: The New Jersey Pine Barrens, a Case Study. International Journal of Remote Sensing Special Issue Remote Sensing and Landscape Ecology: Landscape Patterns and Landscape Change 21(13&14):2589-2611. Lynch, J.F. 1987. Responses of breeding bird communities of forest fragmentation. Pp. 123140 in Nature Conservation: the Role of Remnants of Native Vegetation. D.A. Saunders, G.W. Arnold, A.A. Burbidge, and A.J.M. Hopkins (eds.). Surrey Beatty and Sons, Sydney, Australia.

Martin. T.E. 1988. Habitat and area effects on forest bird assemblages: is nest predation an influence? Ecology 69:74-84. Moore, N.W. 1962. The heaths of Dorset and their Conservation. Journal of Ecology 50:369391. Noss, R.F. 2001. Forest Fragmentation in the Southern Rocky Mountains. Landscape Ecology 16: 371-372. Piessens, K., Honnay, O., Hermy, M., 2005. The role of fragmented area and isolation in the conservation of heathland species. Biological Conservation 122, 61-69. Robbins, C.S. 1979. Effects of forest fragmentation on bird populations. Pp. 198-212 in Proceedings of the Worshop on Management of North-Central and Northeastern Forest for Non-game Birds. R.M. Degraaf and N. Tilghman (eds.). United States Forest Service General Technical Report NC-51. Steiner, N., Köhler, W., 2003. Effects of landscape patterns on species richness – a modelling approach. Agriculture Ecosystems & Environment 2086, 1-9. Ten Brink, B.J.E, Hosper, H. and Colijn, F. 1991. A Quantitative Method for Description and Assessment of Ecosystems: the AMOEBA-approach. Marine Pollution Bulletin 3: 65-70. Ten Brink, B.J.E., 2000. Biodiversity indicators for the OECD Environmental Outlook and Strategy; A feasibility study. RIVM report 402001014, Bilthoven. TNC, FVSA, DeSdel Chaco, WCS, 2005. Evaluación Ecorregional del Gran Chaco Americano / Gran Chaco Americano Ecoregional Assessment. Fundación Vida Silvestre Argentina, Buenos Aires. UN/ECE. 2003. Declaration by the environment Ministers of the region of the United Nations Economic Commission for Europe (UNECE). Fifth Ministerial Conference "Environment for Europe" Kiev, Ukraine, 21-23 May 2003. ECE/CEP/94/Rev.1. (http://www.rusrec.ru/homepage/databases/int_law/ece.cep.94.rev.1.e.pdf Accessed 1 October 2004) UNEP, 2004. Strategic Plan: future evaluation of progress. UNEP/CBD/COP/VII/30. Montreal.

Part I Multitemporal changes and forest status 1. Análisis multitemporal de la fragmentación y reducción del Bosque de Tres Quebrachos S. A. Torrella, R. G. Ginzburg y J. M. Adámoli

19

2. Cuantificación y análisis regional de la expansión agropecuaria en el Chaco Argentino R. G. Ginzburg, S. A. Torrella y J. M. Adámoli

28

3. Landscape changes due to native forest loss along a precipitation gradient in the Chaco region, Argentina M. G i, J. Bono, M. Stamati, C. Montenegro, M. Brouver, E. Manghi and M. Strada

39

4. Landscape mosaic, habitat structure and fragmentation of native forests at Córdoba mountain areas (Argentina central). M. Menghi and R. del Sueldo

50

5. Patterns of land use change and forest fragmentation in the temperate forests in southern Chile C. Echeverría,, D. A. Coomes, A. C. Newton, A. Lara and J. M. Rey-Benayas

63

6. Forest floristic inventory of Mixed Ombrophilous Forest and Deciduous Forest of Santa Catarina State, Southern Brazil: preliminary results. C. Vibrans; A. Uhlmann; L. Sevegnani; M. Marcolin; N. Nakajima,C. R. Grippa, E. Brogni & M. Braga Godoy

74

Part II Ecological consequences of forest fragmentation 1. Habitat fragmentation effects on insect herbivory in Chaco Serrano woodlands M. Videla, L. Cagnolo, G. Valladares, A. Salvo, S. Fenoglio

86

2. Extensive pollen flow may counteract the effects of landscape fragmentation P. Marchelli, A.C. Moreno & L.A. Gallo

94

3. Effects of forest loss on soil water regime in the temperate landscape in southern Chile C. Echeverría, O. Thiers, A. Lara

102

Part III Landscape Ecology for conservation, management and restoration 1. Mitigation of biodiversity loss in Nothofagus pumilio managed forests of South Patagonia M.V. Lencinas, G. Martínez Pastur, E. Gallo, A. Moretto, C. Busso & P. Peri

112

2. Estrategias de ordenamiento territorial y conservación de la naturaleza en la Ecoregión de las Yungas (noroeste de Argentina) L. I. Bachmann, C. L. Daniele, A. G. Frassetto

121

133 3. Ecological restoration of Fitzroya cupressoides, a long-lived conifer in southern Chile A. Lara, C. Echeverría, O. Thiers, F. Bustos, E. Huss, B. Escoba 4. Conservation approaches in the Atlantic Forest of Argentina: from eco-region to single-species M. J. Pacha, M. S. Di Bitetti, G. Placci, E. Carabelli1, A. Paviolo, C. D. De Angelo , M. Jaramillo

142

Part I

Multitemporal changes and forest status

Análisis multitemporal de la fragmentación y reducción del Bosque de Tres Quebrachos S. A. Torrella1, R. G. Ginzburg1 y J. M. Adámoli 1y2 1

Laboratorio de Ecología Regional, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires. Ciudad Universitaria, Pab. II, 4º Piso (1428), Ciudad de Buenos Aires, Argentina. Tel.: 054-011-4576-3300 int. 214. e-mail: [email protected] 2 Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET).

Abstract Since the beginning of the 20th century, the Chaco region in Argentina has been the scenario of consecutive agricultural frontier increases. This expansion has been produced disregarding any territorial planning criteria, thus jeopardizing the conservation of many Chaco’s environments, including the object of this study: the Tres Quebrachos Forest. The landscape structure, interpreted as an agricultural matrix of remaining patches of the original Tres Quebrachos Forest, was analyzed in an area of 71.975 hectares in the surroundings of Charata, Chaco Province. Based on aerial photographies and satellite images, the forest patches were digitalized and thematic maps were generated for the years 1957, 1975, 1988 & 2002. Several landscape indices were used to quantify, among others, two key issues: a) the surface reduction (the forest’s surface decreased from 25.898 hectares in 1957 to 14.917 in 2002), and b) the forests fragmentation (163 to 303 patches during the same period). The 1988-2002 period showed the most tangible variations for many of the studied indicators. This analysis provides essential information to design conservation strategies for these forests.

Resumen Desde principios del siglo XX la región chaqueña en la República Argentina ha sido escenario de sucesivos avances de la frontera agrícola. Esta expansión, realizada al margen de cualquier criterio de planificación territorial, compromete seriamente la conservación de muchos ambientes chaqueños, entre ellos, el Bosque de Tres Quebrachos, objeto de este estudio. En un área de 71.975 hectáreas alrededor de Charata, Provincia de Chaco, se analizó la estructura del paisaje, interpretada como una matriz agrícola con parches remanentes del Bosque de Tres Quebrachos original. Sobre la base de fotografías aéreas e imágenes satelitales, se digitalizaron los parches de bosque y se generaron mapas temáticos para los años 1957, 1975, 1988 y 2002. Se utilizaron diversos índices de paisaje, que permitieron cuantificar, dos aspectos clave: a) la reducción en superficie (los bosques pasaron de 25.898 hectáreas en 1957 a 14.917 en 2002) y b) la fragmentación de los bosques (pasaron de 163 a 303 parches en ese período). El período 1988-2002 presentó las variaciones más marcadas para muchos de los índices estudiados. Este análisis provee información esencial para diseñar estrategias de conservación para estos bosques.

Understanding biodiverity loss: an overview on forest fragmentation in South America 19

Palabras clave Chaco Subhúmedo Central; Bosque de tres quebrachos; Expansión agrícola; Fragmentación de bosques; Deforestación. Introducción El Bosque de Tres Quebrachos es el único ecosistema donde coexisten el quebracho blanco (Aspidosperma quebracho blanco), el quebracho colorado chaqueño (Schinopsis balansae) y el quebracho colorado santiagueño (Schinopsis lorentzii). La distribución espacial de este bosque en la República Argentina coincide con la subregión ecológica del Chaco Subhúmedo Central, que abarca el centro-oeste de la Provincia del Chaco, este de Santiago del Estero y noroeste de Santa Fe (Morello y Adámoli, 1968 y 1974). Dentro de esta subregión, el Bosque de Tres Quebrachos ocupa aproximadamente un 50 % de la superficie, y se manifiesta donde los suelos son profundos, de buena fertilidad natural y sin mayores limitaciones por inundaciones, salinidad ni sequías, es decir que este bosque se corresponde con los mejores suelos agrícolas de la zona, también llamada “el óptimo de los tres quebrachos”. (Morello y Adámoli, 1974). Las áreas ocupadas por el Bosque de Tres Quebrachos han sido transformadas para la agricultura desde las primeras décadas del siglo XX, por colonias de pequeños productores agrícolas. En un análisis que contempló la expansión agrícola en la región chaqueña (ver “Cuantificación y análisis regional de la expansión agropecuaria en el chaco argentino” en esta publicación), la subregión ecológica del Chaco Subhúmedo Central concentró casi el 50 % de la expansión total (aunque ocupa tan sólo algo más del 10 % del área), y para el año 2002 se hallaba ocupada por agricultura cerca de la mitad de su superficie. En el presente trabajo se cuantifica la sensible reducción del área y se evalúa la intensa fragmentación de los parches del Bosque de Tres Quebrachos. Asimismo, se alerta acerca de los riesgos de desaparición de este bosque, por la intensidad del avance agropecuario y por la falta de criterios de planificación que compatibilicen los intereses de la producción, con los de la conservación de los bienes y servicios ambientales en ecosistemas únicos. Materiales y Métodos El área de estudio abarca 71.975 hectáreas alrededor de la Ciudad de Charata en la Provincia del Chaco, República Argentina (Figura 1), en la zona denominada “el óptimo de los tres quebrachos” (Morello y Adámoli, 1974). Se analizó la evolución (1957-2002) de la estructura del paisaje, en particular los parches remanentes del Bosque de Tres Quebrachos dispuestos en la matriz agrícola. Se trabajó con fotografías aéreas del año 1957 (escala 1:40.000) las que fueron escaneadas y luego ensambladas para obtener un mosaico fotográfico del área de estudio. Se utilizaron tres imágenes satelitales, una Landsat 2 de 1975 (path-row 245-79), una Landsat 5 (228-79) del año 1988 y una Landsat 7 (228/79) del año 2002. Se georreferenció la imagen satelital Landsat 7 del 2002, y sobre esta base se registraron las demás imágenes y el mosaico fotográfico. A partir de la interpretación visual de las imágenes satelitales y el mosaico fotográfico, se procedió al mapeo de los fragmentos de bosque, visualizando las escenas en pantalla a escala 1:150.000, utilizando el programa Arc View Gis 3.2.


Se aplicaron diversos índices de paisaje que permitieron cuantificar, la fragmentación y reducción de superficie que sufrieron estos bosques a lo largo de 45 años. Para el cálculo de superficies y la estadística de los parches de bosque digitalizados se empleó la extensión Patch Analyst del mismo programa. Figura 1. Situación geográfica del área de estudio (1.a-República Argentina; 1.b-Provincia del Chaco; 1.c-Área de estudio en torno a la Ciudad de Charata). 1.c

1.a

Ruta Provincial 94

1.b

Resultados Los resultados muestran la importante reducción de superficie y la fuerte fragmentación que sufrieron estos bosques durante el período analizado (tabla 1 y figura 2). La superficie total pasó de 25.898 ha en 1957 a 14.917 ha en 2002 (figura 3.a). O sea que en 45 años se perdieron 10.981 hectáreas, lo que significó una pérdida del 42,4 % de la superficie de bosques, los que sólo ocupan 20,72 % de la zona estudiada. Tabla 1. Índices de paisaje calculados. Año

Superficie total (ha)

Número de parches

163

Tamaño medio de parche (ha) 156,96

Mediana del tamaño de parche (ha) 10,63

Desvío estándar del tamaño de parche (ha) 830,41

1957

25.898,48

1975 1988

24.143,88 22.154,06

197 227

122,56 97,59

12,15 10,85

706,32 460,79

2002

14.916,74

303

49,23

12,64

126,53


Tabla 1. Índices de paisaje calculados (continuación). Año

Borde total (m)

Borde medio de parche (m)

Densidad de borde (m/ha)

1957 1975 1988 2002

1.195.779 1.214.357 1.183.251 1.091.352

7.247,15 6.164,25 5.212,56 3.601,82

46,17 50,30 53,41 73,16

Media de Índice perímetro/área de (m/ha) forma medio 481,98 1,72 399,20 1,67 371,95 1,63 279,25 1,61

Índice de forma medio pesado por área 5,91 5,83 4,56 2,58

Figura 2. Evolución de los parches remanentes del Bosque de Tres Quebrachos (gris) dispuestos en la matriz agrícola (blanco), entre los años 1957 y 2002.

2.a

2.b

2.c

2.d

La cantidad de parches aumentó en todos los períodos analizados, pasando de 163 en 1957 a 303 en 2002 (figura 3.b), es decir que la reducción de superficie fue acompañada por un intenso proceso de fragmentación. La intensidad de la reducción fue tal, que más allá de esta fragmentación, entre 1988 y 2002 hubo 35 parches de bosque que fueron eliminados completamente. La fragmentación afecto al tamaño medio de parche, el cual se redujo 68,64 % durante todas las etapas, cayendo de 157 ha en 1957 a 49 ha en 2002 (figura 3.c). Sin embargo, la mediana del tamaño de parche no se modificó sustancialmente, lo que muestra la conjunción observada del aumento del número de parches y la disminución de sus tamaños,


o lo que es lo mismo el fraccionamiento de los parches de mayor tamaño (este fraccionamiento de los parches más grandes se ve reflejado también al analizar la disminución conjunta del tamaño medio de parche y del desvío estándar del tamaño de parche, el cual cayó de 830 ha en 1957 a 127 ha en 2002). Figura 3. Índices de paisaje calculados para los parches de Bosque de Tres Quebrachos. Numero de parches

Superficie total de bosques (ha) 310

28,000.00

290

26,000.00

270

24,000.00

250

22,000.00

230

20,000.00

210

18,000.00

190

16,000.00

170

14,000.00

150

1957

3.a

1975

1988

2002

3.b

1957

1975

1988

2002

Densidad de borde (m/ha)

Tamaño medio de parche (ha) 80.00

160.00

75.00

140.00

70.00

120.00

65.00 60.00

100.00

55.00

80.00

50.00

60.00

45.00

40.00

40.00

1957

3.c

1975

1988

2002

3.d

Media de la relacion perímetro área (m/ha)

1957

1975

1988

2002

Índice de forma medio pesado por área 6.00

520.00

5.50 470.00

5.00 4.50

420.00

4.00

370.00

3.50 320.00

3.00

270.00

2.50

3.e

1957

1975

1988

2002

3.f

1957

1975

1988

2002

Tanto el borde total como el borde medio del parche disminuyeron en el transcurso de los 45 años estudiados, lo que es consecuencia por un lado, de la reducción sufrida en su


superficie (en cada período la superficie total de bosque fue menor y a su vez los parches fueron de menor tamaño) y por el otro de la simplificación de sus formas. Esto mostraría que, si bien, dado el aumento del número de parches se esperaría que aumente el borde total, el proceso de reducción de la superficie predominó sobre el de fragmentación. El índice de densidad de borde (figura 3.d), que indica la cantidad de metros de borde de bosque que hay por hectárea de bosque, aumentó 58,46 %, trepando de 46 a 73 m/ha entre 1957 y 2002. Esto significa que aunque la cantidad de borde total de bosque se redujo en el área de estudio, su relación con la cantidad neta de bosque remanente, aumentó en todos los períodos. Para los parches de bosque la relación media de perímetro/área disminuyó en todos los períodos, mostrando que el perímetro de cada parche es cada vez menor respecto de su superficie (figura 3.e). Considerando que tanto el borde medio de parche como el tamaño medio de parche disminuyeron en todos los años analizados, se llega a la conclusión de que en los parches predominó la simplificación de los bordes sobre la pérdida de superficie (la disminución en la cantidad de borde/parche fue superior a la disminución del área de estos parches). El proceso de simplificación de los bordes se ve claramente al analizar que el índice de forma medio presentó valores decrecientes para todos los años (esta variable mide la complejidad de los parches en comparación con la forma de un círculo; cuanto más irregulares son los parches, mayor es el valor del índice). En tanto que al ponderar el índice y considerar la superficie de cada parche (índice de forma pesado por área) la reducción fue sensiblemente mayor, indicando la tendencia de los bordes de los parches a formas más regulares (figura 3.f). Es de notar que ambos procesos, de reducción y fragmentación, se produjeron simultáneamente. Durante el período de 45 años considerado, la cantidad de parches de bosque y la superficie que ellos ocupan aumentó en las clases menores a 1000 hectáreas, mientras que se redujo para los parches mayores (tabla 2 y figura 4). Cabe aclarar, por si quedan dudas, que el aumento de superficie ocupada por parches pequeños no se dio por regeneración de bosques, sino debido a la fragmentación de los parches de mayor tamaño. Tabla 2. Distribución de los parches de bosque en clases de superficies. Año

Clase

1957 Nº Superf. parches (ha)




< 10 ha 10-100 ha 100-1000 ha > 1000 ha

79 61 18 5

88 83 23 3

108 93 23 3

128 140 34 1

301,48 1.857,99 5.902,07 17.836,94

354,61 2.612,77 7.534,23 13.642,27

468,17 2.997,04 8.396,26 10.292,59

597,63 4.747,00 8.110,00 1.462,12


Figura 4. Superficie (4.a) y número (4.b) de parches de bosque distribuidos en clases de superficies. Superficie (ha) por clase de tamaño de

Número de parches por clase de tamaño

parche

160

18,000

< 10ha

16,000

140

14,000

10 a 100 ha

12,000 10,000 8,000

100 a 1000 ha

6,000 4,000

>1000 ha

2,000

100

10 a 100 ha

80 60 40

100 a 1000 ha

20

>1000 ha

0 1957

0

4.a

< 10ha

120

1957

1975

1988

2002

1975

1988

2002

4.b

Por último, el período comprendido entre los años 1988 y 2002 presentó las variaciones más marcadas para muchos de los índices estudiados: la superficie total distribuida en parches mayores a 1.000 hectáreas cayó de 10.293 ha a 1.462 ha; la pérdida de superficie de bosque alcanzó a 65,9% (figura 5) y la densidad de borde aumentó 37%. Figura 5. Pérdida de bosques durante los períodos 1957-1975, 1975-1988 y 1998-2002.

Conclusión Los resultados obtenidos muestran la significativa reducción de superficie y la intensa fragmentación que ha sufrido el Bosque de Tres Quebrachos, proceso que se ha acelerado en el último período. Dado el alto valor de conservación de este tipo de bosque, es imperioso tomar medidas de protección y manejo para asegurar su preservación.


Un criterio básico de conservación indica que debería preservarse al menos una parte de cada tipo de bosque natural que exista (Burkart, 1999), ya que cada tipo de bosque contiene un elenco diferente de especies de plantas y animales, con sus propias interacciones, lo que es fundamental para la conservación. La autoperpetuación del bosque depende de la interacción entre manchones en diversos estados de desarrollo o sucesión. A su vez, las interacciones entre el área protegida y su entorno influyen en el funcionamiento de los ecosistemas protegidos, por lo cual además de proteger un área, es necesario planificar los usos en el entorno, con normas de manejo adecuadas y sustentables (Morello y Matteucci, 1999). De acuerdo con diversos especialistas, el mínimo a conservar debería estar entre 15 y 30 % (Mackinnon et al., 1986; Reid y Miller, 1989), entre los cuales sugieren 5 a 10% de protección estricta, como mínimo, y 10 a 20 % adicional de áreas no estrictas (un 10 % bajo protección estricta sería insuficiente, sin amplias áreas de amortiguación de bosques manejados). El Bosque de Tres Quebrachos presenta tal nivel de fragmentación, sobreexplotación y ritmo de deforestación, que si no se adoptan medidas urgentes, en pocos años más es posible que ya no queden masas disponibles con número, tamaño y conectividad mínimas como para asegurar su protección y el cumplimiento de los servicios ecológicos que este ambiente presta. Debido a la falta de tierras fiscales en el área, la única posibilidad real de conservación consistiría en integrar una red de áreas protegidas en propiedades privadas, para poder conservar muestras representativas de la diversidad ecológica de este tipo de bosques. Entre los especialistas en biología de la conservación, existe consenso acerca de que las curvas especie-área permiten predecir la proporción de especies que se extinguirán en una región con base en la cantidad de hábitat que se pierde (Raven, 1987, 1988a, b; Myers, 1988a; Simberloff, 1986; Lovejoy, 1980; Wilson, 1988; Reid y Miller, 1989). La tasa de extinción de especies en base a diversos escenarios de deforestación, no tiene una relación lineal, ya que para una pérdida de 11 % de la superficie se prevé una pérdida de 2 % de especies, mientras que con 44,8 % de pérdida de superficie, las pérdidas de especies llegarían a 35 % (Reid, 1992). Los trabajos basados en aplicaciones de la teoría de biogeografía de islas (MacArthur y Wilson, 1967), demuestran que en bosques tropicales, cuando un hábitat pierde el 90% de su extensión, con el tiempo se extingue la mitad de sus especies (Myers, 1988b). Sin embargo, (Wilson, 1988) destaca que estas proyecciones son conservadoras, porque aún cuando una porción de especies sobreviva, probablemente hayan sufrido una significativa reducción en la variación genética entre sus miembros, debido a la pérdida de genes que se dio junto con la disminución del número de individuos. De esta forma, cuando un bosque se reduce, por ejemplo de 10.000 a 1.000 ha, algunas extinciones de especies son inmediatas; otras especies seguirán existiendo pero en poblaciones que se habrán reducido de forma muy peligrosa para su viabilidad futura. Las pérdidas estimadas en el Bosque de Tres Quebrachos, son del orden del 80 % de la superficie original. Los cambios climáticos podrían exacerbar esta pérdida potencial. Para detener la pérdida de especies, es necesario por un lado disminuir la tasa de deforestación y por el otro, racionalizar el uso sustentable del bosque y proteger los hábitats claves (con alta riqueza de especies y endemismos). El Bosque de Tres Quebrachos se está perdiendo y fragmentando aceleradamente. En un escenario en el que la frontera agrícola continúa avanzando con intensidad, donde no hay ningún tipo de áreas protegidas y que presenta una notoria escasez de tierras fiscales,


las perspectivas de conservación de las especies y servicios ambientales de este ecosistema tan particular, está muy comprometida. Por ello surge la necesidad de implementar con urgencia acciones concretas que permitan controlar el patrón espacial y la localización de los fragmentos remanentes, asegurando la existencia de áreas relativamente grandes de hábitats naturales y semi-naturales, para reducir la pérdida de especies. Agradecimientos: Silvia Matteucci, Patricia Kandus, Martha Gazzano. Bibliografía Burkart, R. (1999) Conservación de la biodiversidad en bosques naturales productivos del subtrópico argentino. Matteucci, Solbrig, Morello y Halffter (editores). Biodiversidad y uso de la tierra. Conceptos y ejemplos de Latinoamérica. EUDEBA, Buenos Aires, 589 pp. Lovejoy, T.E. (1980) A projection of species extinctions, in Council on Environmental Quality (CEQ), The Global 2000 Report to the President, Vol. 2. CEQ, Washington, DC, pp. 328-31. MacArthur, R.H. y Wilson, E.O. (1967) The Theory of Island Biogeography. Princeton University Press, Princeton, RI. Mackinnon, J.; Mackinnon, K.; Child, G. y Thorsell, J. (comps). (1986). Managing protected areas in the Tropics. IUCN/UNEP, Gland (Suiza), 290 pp. Morello, J. y Adámoli, J. (1968) Las grandes unidades de vegetación y ambiente del Chaco argentino. Primera parte: Objetivos y metodología. Serie fitogeográfica Nº 10, INTA, Buenos Aires. 125 pp. Morello, J. y Adámoli, J. (1974) Las grandes unidades de vegetación y ambiente del Chaco argentino. Segunda parte: Vegetación y ambiente de la provincia del Chaco. Serie fitogeográfica Nº 13, INTA, Buenos Aires. 130 pp. Morello, J. y Matteucci, S. (1999) Biodiversidad y fragmentación de los bosques en la Argentina. Matteucci, Solbrig, Morello y Halffter (editores). Biodiversidad y uso de la tierra. Conceptos y ejemplos de Latinoamérica. EUDEBA, Buenos Aires, 589 pp. Myers, N. (1988a) Threatened biotas: 'hotspots' in tropical forests. Environmentalist, 8(3), 120. Myers, N. (1988b) Tropical forests and their species - going, going...?. E.O. Wilson and F.M. Peter, eds. Biodiversity. National Academy Press, Washington, D.C., pp. 28-35 Raven, P. H. (1987) The scope of the plant conservation problem world-wide. D. Bramwell, O. Hamann, V. Heywood, and H. Synge, eds. Botanic Gardens and the World Conservation Strategy. Academic Press, London, pp. 19-29. Raven, P. H. (1988a) Biological resources and global stability. S. Kawano, J.H. Connell, and T. Hidaka, eds. Evolution and Coadaptation in Biotic Communities. University of Tokyo Press, Tokyo, pp. 3-27. Raven, P.H. (1988b) Our diminishing tropical forests. E.O. Wilson and F.M. Peter, eds. Biodiversity. National Academy Press, Washington, D.C., pp. 19-22. Reid, W. R. (1992) How many species will there be? T. C. Whitmore and J. A. Sayer, eds. Tropical deforestation and species extinction. New York. Reid, W. y Miller, K. (1989) Keeping options alive. The scientific basis of conserving biodiversity. World Resources Institute, Washington, D.C. Simberloff, D. (1986) Are we on the verge of a mass extinction in tropical rain forests? D.K. Elliott, ed. Dynamics of Extinction. New York, NY, pp. 165-180. Wilson, E.O. (1988) The current state of biological diversity. E.O. Wilson and F.M. Peter, eds. Biodiversity. National Academy Press, Washington, D.C.


Cuantificación y análisis regional de la expansión agropecuaria en el Chaco Argentino R. G. Ginzburg1, S. A. Torrella1 y J. M. Adámoli1,2 1

Laboratorio de Ecología Regional, Departamento de Ecología, Genética y Evolución, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires. Ciudad Universitaria, Pab. II, 4º Piso (1428), Ciudad de Buenos Aires, Argentina. Tel.: 054-011-4576-3300 int. 214. e-mail: [email protected] 2 Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET).

Resumen El aumento en la producción agrícola registrado en el país en los últimos 15 años trajo aparejadas importantes implicancias ambientales. En la región chaqueña se generó una fuerte presión sobre el uso de la tierra, que desencadenó un intenso proceso de deforestación. Mediante el uso de imágenes satelitales se detectó que entre 1992 y 2002 las áreas transformadas en el zona de estudio tuvieron un aumento del 60%, de 3.014.107 a 4.816.502 hectáreas; es decir que se han perdido 1.802.395 ha de ambientes nativos, constituidos principalmente por bosques. Esta expansión tiene a la soja como su principal motor, y todo indica que la demanda de esta oleaginosa se mantendría o incrementaría en los próximos años. A esto hay que agregar la presión generada por el inminente desarrollo del mercado de los biocombustibles. Al desarrollarse este proceso sin una efectiva regulación por parte del estado, se compromete tanto la conservación de la biodiversidad, como también la sustentabilidad de los emprendimientos productivos. Palabras clave Región chaqueña; Expansión agrícola; Chaco subhúmedo central; Bosque de tres quebrachos.

Abstract The increase of crop production in Argentina during the last 15 years entailed strong environmental implications. In Chaco region a strong pressure on land use appeared and unleashed an intense deforestation process. The use of satellite images detected in the 1992 – 2002 period, an increase of 60% in land conversion: from 3.014.107 to 4.816.502 hectares; this implies that 1.802.395 ha of native environments, mainly forests, have been lost. This expansion is leaded mainly by soy, and all signs indicate that the demand for this oleaginous will stay constant or even be increased during the next years. The pressure generated by the imminent development of the biofuel market must also be taken into account. As this process has been developed in absence of an effective State regulation, both biodiversity conservation and productive endeavor sustainability have been jeopardized.

Key words Chaco region; agricultural expansion; central sub-humid Chaco, Tres quebrachos Forest


Introducción En los últimos años se dio en la Argentina un significativo aumento en la producción agrícola; la superficie dedicada a cultivos anuales se incrementó en el período 1988-2002 en 5.500.000 hectáreas, pasando de 13.800.000 a 19.300.000 ha (INDEC). En este proceso cumple un papel fundamental el cultivo de la soja, que en el mismo período incrementó su superficie de 4.670.000 a 11.639.240 ha sembradas en el país, constituyendo en la actualidad cerca del 50% de la producción total del sector (SAGPyA). Los indicadores internacionales señalan que la demanda de soja no sólo se mantendría en los próximos años sino que además aumentaría sensiblemente, tanto en lo referido a las fuertes demandas de granos, aceites y harinas, como por la inminente expansión del uso de biocombustibles a nivel nacional e internacional. La producción de soja ha aumentado en las distintas regiones del país, empezando por la región pampeana donde se expandió, en gran medida sobre otros cultivos -principalmente maíz-, así como sobre terrenos antes dedicados a la ganadería (Azcuy Ameghino y León, 2005). En la región chaqueña el proceso tuvo distintos matices; la soja desplazó a cultivos tradicionales, como el algodón en la Provincia del Chaco, pero motorizó además una importante expansión de la agricultura en general: Entre 1995 y 2005, la superficie sembrada con cultivos anuales en Salta, Chaco y Santiago del Estero aumentó de 1.800.000 a 3.100.000 ha, y la sembrada con soja de 420.000 a 1.760.000 ha (ODSMA - OEA, en preparación).Al mismo tiempo, la región recibió a buena parte de los emprendimientos ganaderos desplazados de la región pampeana por la mayor rentabilidad de la actividad agrícola. Esta combinación constituyó una fuerte presión sobre los bosques nativos, que terminó desencadenando un proceso de desmontes generalizados. Sólo entre 1998 y 2002 se desmontaron 306.000 ha en Santiago del Estero, 194.000 en Salta y 118.000 en Chaco (Montenegro et al., 2004), convirtiendo a la región en la de mayor tasa de deforestación del país. La expansión agropecuaria en la región chaqueña generó opiniones encontradas: por una parte se celebra la incorporación de nuevas áreas productivas al mapa agrícola del país y los ingresos económicos que ello implica. Por otra parte se alerta sobre los riesgos que conlleva el modelo adoptado, en consonancia con la creciente preocupación por los temas de sostenibilidad agraria (INTA-INDEC, 1994; Viglizzo, 2001). En este sentido se destaca que el proceso se lleva adelante sin que exista un plan de ordenamiento territorial, por lo que se permite que el avance se realice sobre zonas en las que no está garantizada la sustentabilidad de la producción, ya sea por condiciones edáficas o climáticas (Adámoli et al., 2004; Adámoli, 2005; Grau et al., 2005). Tampoco se están contemplando los riesgos ambientales de los procesos de expansión, como la pérdida de biodiversidad (Torrella et al., 2003), la simplificación del paisaje (Forman y Godron, 1985), o la conectividad entre ambientes. El fenómeno también tuvo implicancias relevantes en el ámbito social, los productores y pobladores tradicionales de la región no se vieron incluidos mayoritariamente en el nuevo modelo, ni percibieron los beneficios económicos que este generó (Reboratti, 2005); también se alerta sobre la preocupante concentración de la renta y la disminución del trabajo rural (Barsky y Gelman, 2001). El Gran Chaco Americano es una vasta planicie de más de 1.000.000 km2, de los cuales el 60% están en Argentina. Presenta en su extensión una gran variabilidad climática, acompañada por distintas formaciones fisonómicas, entre las que se destacan distintos tipos de bosques, caracterizados por la dominancia de especies del género Schinopsis; sabanas


y pastizales (Morello y Adámoli, 1974; Prado, 1993). Su alta biodiversidad y grado de intervención antrópica la sitúan en el contexto regional y nacional como una de las áreas prioritarias para la conservación (Dinerstein et al, 1995; Bertonatti y Corcuera, 2000). En el presente trabajo se describe y analiza la configuración espacial del proceso de expansión agropecuaria en la región chaqueña, en el período 1992-2002, mediante la interpretación de imágenes satelitales y el uso de sistemas de información geográfica, en una aproximación tendiente a desentrañar sus implicancias ambientales en el nivel ecorregional. Materiales y Métodos Se definió como área de estudio a parte del Chaco argentino donde se desarrolla con mayor intensidad el proceso descrito de expansión agropecuaria. El área, con una superficie total de 41.748.513 hectáreas, quedó comprendida por el Chaco salteño, el norte de la Provincia de Santa Fe, y las Provincias del Chaco, Formosa y Santiago del Estero (a excepción de su extremo SE) (Figura 1). Sobre imágenes satelitales se identificaron visualmente y se mapearon, trabajando a escala 1:250.000, todas las parcelas en las que la cobertura vegetal original ha sido sustituida por cultivos, tanto agrícolas como pasturas (en adelante “áreas transformadas”). Esta digitalización se realizó con el programa Arcview 3.2. Se utilizaron mosaicos “Mr Sid” compuestos a partir de imágenes Landsat de acceso libre en internet (http://glcfapp.umiacs.umd.edu:8080/esdi/index.jsp); para abarcar la totalidad del área de estudio fueron necesarios los mosaicos de las ubicaciones 20-20, 20-25, 21-20 y 21-25. Se utilizó una serie de mosaicos del período 1986-1992 que en el texto y las tablas se indica como 1992 y otra del período 1999-2002, que se indica como 2002. La ventaja de poder disponer de mosaicos de imágenes tiene la limitación de la amplitud temporal de las series. No obstante, el volumen de información generada, las tendencias detectadas y la localización espacial de los procesos, consideramos que son válidos para la toma de decisiones y la planificación en esta escala de trabajo. En la enorme mayoría de los casos la diferenciación entre la vegetación nativa y cultivos es inequívoca, pero en algunos potreros de ganadería extensiva sobre campos naturales, pueden generarse confusiones, porque ciertos tipos de manejo pueden presentar un patrón similar al de las parcelas cultivadas. En estos casos la identificación de las parcelas se hizo ampliando sensiblemente la escala de la imagen, para mejorar la definición. Este tipo de errores no son intrínsecos de la metodología, ya que inclusive una clasificación automática sin una exhaustiva verificación a campo también puede presentarlos, incluso en mayor medida. Para analizar el proceso de expansión en las distintas zonas climáticas, se dividió la región chaqueña a partir de un análisis bibliográfico de datos de precipitación anual (Galmarini y Raffo del Campo, 1964; Bianchi 1981; Bruniard 1987). Así quedaron definidas las siguientes zonas (figura 1): Chaco Árido: menos de 500 mm, marginal en el área de estudio definida para este trabajo. Chaco Semiárido: 750 a 500 mm, la de mayor extensión territorial. Chaco Subhúmedo: 750 a 900 mm, presenta una faja muy angosta en el borde oeste de la región chaqueña (subhúmedo occidental), y una faja más ancha en la frontera entre Santiago del Ester, Chaco y Santa Fe (subhúmedo central).


Chaco Húmedo: más de 900 mm, se extiende por el este de las provincias de Formosa, Chaco y Santa Fe. Figura 1. Área de estudio, división política y zonas climáticas.

Estas zonas no son estables en el tiempo, ya que pueden desplazarse, en función de ciclos plurianuales secos o húmedos. Tomando como referencia a la isohieta de 750 mm en la Provincia del Chaco, durante un ciclo húmedo esa isohieta se desplaza hacia el Oeste, mientras que en un ciclo seco se desplaza hacia el Este (en el borde occidental de la región, los desplazamientos son en sentido inverso). Así, queda determinada una faja de variabilidad climática. En los últimos 25 años hubo un sensible desplazamiento de las isohietas hacia el Oeste, mientras que en los 5 años más recientes, hay evidencias de una tendencia más seca. Esto indicaría un proceso de reversibilidad climática. Las áreas con riesgo de reversión fueron definidas como aquellas en las que en una situación normal están dentro de la zona del subhúmedo, pero en un ciclo seco quedan incluidas en la zona correspondiente al subhúmedo seco a semiárido. Se identificaron así dos zonas con riesgo de reversión, en las áreas de contacto entre el Chaco Semiárido, y ambas porciones del Chaco Subhúmedo.


Resultados Los resultados obtenidos indican que para la primera serie temporal analizada (1992), las áreas transformadas cubrían 3.014.107 ha, o sea el 7,22 % de la superficie total estudiada. Para el año 2002 este valor trepó hasta el 11,54 % (4.816.502 ha). Las áreas transformadas tuvieron una expansión del 59,8 %, es decir que en este período fue sustituida la cobertura vegetal nativa, constituida principalmente por bosques, en 1.802.395 ha. Como se observa en la figura 2, estas áreas no tienen una distribución homogénea en el área de estudio, sino que se presentan agrupadas en núcleos de diferentes características. Figura 2. Superficies donde la cobertura vegetal nativa ha sido eliminada.

Para visualizar más claramente los núcleos en los que se concentran las áreas transformadas, se dividió el área de estudio en hexágonos regulares de 10.000 hectáreas. Las áreas transformadas dentro de cada hexágono, fueron expresadas como % de cada polígono. En la figura 3 se muestran de esta manera las áreas transformadas para 1992 y 2002. Se distinguen claramente los principales núcleos agrícolas de la región: El grueso de los cultivos se localiza en el núcleo del centro-sur de la región, correspondiendo básicamente a la zona climática del “Chaco Subhúmedo Central”. Este núcleo se presenta dividido por una depresión salobre. La mayor parte del núcleo se localiza en torno al límite entre las provincias de Chaco y Santiago del Estero. La porción sur de este núcleo se encuentra en el límite entre el sudeste de la provincia de Santiago del Estero y el noroeste de Santa Fe.


En la porción más húmeda de la región, en el este, predominan los suelos inundables. Por tal motivo, la agricultura sólo se expresa en los terrenos topográficamente más altos: - En el NE del área de estudio, puede observarse que las áreas cultivadas presentan un alineamiento en sentido NO-SE, acompañando el trazado de los albardones de los ríos. -

En el extremo SE del área, la agricultura se localiza en la llamada “dorsal oriental” en el NE de Santa Fe y sur de la provincia de Chaco, de orientación N-S.

El núcleo occidental de la agricultura está formado a su vez por varios centros, que se disponen de norte a sur recostados sobre las primeras estribaciones andinas. Hay un par de núcleos en pleno Chaco Semiárido (ver figura 5a), que corresponden a zonas de irrigación: el más grande y evidente, ubicado en Santiago del Estero, en el área de riego del río Dulce, y el otro es el de la localidad de Joaquín V. González en la provincia de Salta, sobre el área de riego del río Juramento. Figura 3. Áreas transformadas para 1992 (a) y 2002 (b), expresadas como porcentaje de hexágonos de 10.000 ha.

a

b

En la figura 4 se representa, también como porcentaje de la superficie de los hexágonos, la expansión agropecuaria, medida como la diferencia entre la superficie transformada en 2002 y la transformada en 1992. Se observa que en el período estudiado la expansión no fue homogénea ni se dio en todos los núcleos, sino que también estuvo concentrada en ciertos puntos. Pueden advertirse además diferentes patrones de “expansión”: en algunos casos hubo una “intensificación” o expansión interna, entendiendo por esto que dentro de un núcleo hay más superficie transformada, pero que el núcleo no expandió sus límites; mientras que en otros casos se produjo un expansión propiamente dicha, aparecen nuevas áreas transformadas en las periferias de los núcleos, que expanden de hecho sus límites.


Figura 4. Áreas de expansión agropecuaria entre 1992 y 2002, expresada como porcentaje de hexágonos de 10.000 ha.

La principal zona de expansión resulta entonces de una combinación de estos procesos en las dos zonas climáticas del Chaco Subhúmedo (Figuras 4 y 5a). En el límite de Santiago del Estero con Chaco y Santa Fe, se observa por una parte una intensificación de la agricultura que se refleja en los tonos más oscuros de los hexágonos, pero también se ve un avance de la frontera agropecuaria hacia el Oeste. En el Subhúmedo Occidental, los mayores valores de la expansión se registran en el Norte. La frontera Oeste de Santiago del Estero registra poca expansión, reflejando que ya estaba fuertemente ocupada en 1992, salvo en el límite Sur con Catamarca. Dentro del Chaco Semiárido, hay algunos puntos importantes en Salta, cerca de la triple frontera con Chaco y Santiago del Estero. La información disponible indica que se trata de grandes campos ganaderos. Puede observarse que el área de irrigación de Santiago del Estero permanece con pocos cambios, con una ligera expansión hacia el Este. Contrariamente a lo que podría pensarse, las menores tasas de expansión se registraron en el Chaco Húmedo. Las áreas de cultivos sobre albardones en Chaco y Formosa prácticamente no tuvieron crecimiento, mientras que el área de la Dorsal Oriental de Santa Fe, tuvo un pequeño incremento en intensidad, pero no expansión efectiva. Zonas de riesgo climático


Como se explicó, a partir del mapa de zonas climáticas se identificaron dos fajas críticas en cuanto la variabilidad de sus precipitaciones (Figura 5b). Estas fajas históricamente (registros climáticos de largo plazo), formaron parte de la zona semiárida, pero debido a los desvíos de precipitaciones de los últimos 25 años, pasaron a formar parte de la zona subhúmeda. Sin embargo, en los últimos 5 años hay fuertes evidencias de una reversibilidad hacia las condiciones semiáridas antecedentes. La angosta faja occidental corresponde a la transición entre el Chaco Semiárido y el Chaco Subhúmedo Occidental, que recibe precipitaciones determinadas por la cercanía con las primeras estribaciones andinas. La faja oriental corresponde a la transición entre el Chaco Semiárido y el Chaco Subhúmedo Central, cuyas precipitaciones están originadas por las corrientes húmedas que ingresan al continente desde el Este. Figura 5. Áreas transformadas según zonas climáticas (a) y zonas de riesgo climático (b).

a

b

Los resultados en cuanto a superficies transformadas en cada una de las zonas y períodos que se muestran en la tabla 1, son sumamente preocupantes. Hay casi 1.300.000 ha en zonas de riesgo climático (la cuarta parte del total de las áreas transformadas), de las cuales aproximadamente la mitad corresponden a desmontes realizados a partir del año 1992. Teniendo en cuenta que en la gran mayoría de estos campos no se hacen rotaciones con ganadería, ni siquiera rotaciones con gramíneas como maíz y sorgo, y que los suelos supuestamente cultivados en siembra directa no tienen cobertura superficial y presentan evidencias de procesos erosivos (ODSMA-OEA, en preparación; experiencia personal), existe un riesgo muy fuerte de que se dispare un proceso de desertificación de consecuencias muy graves.


Tabla 1. Superficies agropecuarias en zonas de riesgo climático. Zona de riesgo climático Occidental Oriental Total

Superficie Sup. Sup. total (ha) transformada transformada 1992 (ha) 2002 (ha) 1.471.957 358.271 519.798 4.310.011 315.863 769.738 5.781.969 674.134 1.289.537

Diferencia Expansión (ha)

161.527 453.876 615.403

45,09 % 143,69 % 91,29 %

Perspectivas de conservación

Finalmente se analizó la ubicación geográfica de las áreas naturales protegidas (considerando las de la administración nacional y provincial, los Sitios Ramsar y las Reservas de Biosfera). Los resultados son preocupantes, puesto que virtualmente no existen áreas protegidas en las zonas con un desarrollo agrícola histórico ni en aquellas en las que se concentra la expansión actual (Figura 6). Resulta evidente entonces que esta red de áreas protegidas no es suficiente para garantizar la conservación de los ambientes nativos frente a la expansión agropecuaria. Peor aún, recientemente el gobierno de la provincia de Salta ha desafectado y vendido parte de la Reserva de Pizarro, única reserva vinculada con el núcleo de expansión de Las Lajitas, los lotes fiscales 32 y 33. Figura 6. Sistema de áreas naturales protegidas en el área de estudio; se observa que es totalmente periférico a las zonas con desarrollo agropecuario.


Discusión y conclusiones La expansión agropecuaria en la región chaqueña tiene en la soja a su principal motor, ya sea por el cultivo de esta oleaginosa en la región, o por el asentamiento de emprendimientos ganaderos que se ven desplazados por la mayor rentabilidad de la agricultura en las zonas más favorables, como la región pampeana. Los indicadores internacionales señalan en forma consistente que la demanda de soja no es un fenómeno pasajero, ya que no sólo se mantendría en un largo plazo para granos, harinas y aceites, sino que además hay fuertes indicios de que se incrementaría significativamente. Hay que considerar además el compromiso de muchos países (Estados Unidos, la Unión Europea, China, etc.) de alcanzar en el corto plazo distintas metas de incorporación de biocombustibles (biodiesel y bioetanol) como corte en las fuentes de energía fósil. Argentina aparece como uno de los países con potencial para aumentar significativamente su producción, no sólo por mejoras en los rendimientos, sino también por expansión de su frontera agrícola. Considerando sólo el mercado interno, de acuerdo a la ley 26.093 a partir del año 2010 la totalidad de los combustibles expendidos en el país deberán tener un 5% de biocombustibles en su formulación. Para el biodiesel, la demanda sería de unos 600.000 metros cúbicos, que de producirse a partir de soja implicarían unas 650.960 tn de aceite. Esta cantidad puede ser obtenida por la reducción de las exportaciones, o bien incrementando el área sembrada en una superficie de 1.815.176 ha. (ODSMA - OEA, en preparación) es decir, un área similar a la expansión registrada en este trabajo. En la región todavía existen vastas superficies disponibles potencialmente para el uso agropecuario, especialmente en las fajas indicadas como zonas de riesgo climático, y su continuidad hacia el semiárido. Hasta el momento esta expansión se ha dado en forma espontánea (sino anárquica) sin una regulación efectiva por parte del gobierno nacional ni de los gobiernos provinciales. Por lo general, cuando las normas regulatorias existen, los gobiernos no cuentan con la infraestructura necesaria o con la decisión política para garantizar su cumplimiento. Si bien es cierto que existen algunas experiencias y esfuerzos para avanzar en esta tarea, muchas de ellas impulsadas por organizaciones no gubernamentales, y también por organismos del estado como la Administración de Parques Nacionales o el Instituto Nacional de Tecnología Agropecuaria, éstas todavía no se han plasmado en acciones concretas sobre el terreno. De hecho el proceso de expansión estuvo regido por las leyes del mercado y la oferta de tierras, con todos los riesgos que eso implica. El escenario planteado pone en evidencia que es imprescindible la implementación de un programa de ordenamiento territorial a escala regional. Desde el punto de vista ambiental, es necesario señalar que hay serios riesgos de desertificación, así como riesgos de pérdida de ecosistemas únicos, como es el caso del Quebrachal de tres quebrachos, que ya ha sufrido una reducción muy drástica en su superficie y conectividad (ver “Análisis multitemporal de la fragmentación y reducción del bosque de tres quebrachos” en esta misma publicación). Pero también es necesario que se prevea y regule la expansión de aquí en adelante, de forma de establecer prioridades ambientales, garantizando la conservación y la conectividad de los elementos más relevantes o amenazados, así como también el cumplimiento de los servicios ecológicos que prestan estos ambientes. Como se señaló al comienzo de este trabajo, el crecimiento agropecuario es una gran oportunidad para el país y para la región chaqueña en particular. Por eso mismo, se impone un mínimo de racionalidad para evitar perjuicios para los productores, para el ambiente y en definitiva para el país.


Bibliografía Adámoli, J.; Ginzburg, R.; Torrella, S.; Herrera, P. (2004) Expansión de la frontera agrícola en la región chaqueña: el ordenamiento territorial como herramienta para la sustentabilidad. Gerencia ambiental 11 (112): 810-823. Adámoli, j. (2005) “Problemas ambientales de la agricultura en la región chaqueña” A. Brown, U. Martinez Ortiz, M. Acerbi y J. Corcuera (Eds). La Situación ambiental Argentina 2005, Fundación Vida Silvestre Argentina, Buenos Aires. Pp 436-442. Azcuy Ameghino, E. y Leon, A. (2005) La sojización, contradicciones, intereses y debates. Revista interdisciplinaria de estudios agrarios 23: 133-157. Bertonatti, C. y Corcuera, J. (2000) Situación Ambiental Argentina 2000, Fundación Vida Silvestre Argentina, Buenos Aires. Bianchi, A. (1981) Las precipitaciones en el Noroeste Argentino. Instituto Nacional de Tecnología Agropecuaria, Salta. Barsky, O. Y Gelman, J. (2001) Historia del agro argentino. Grijalbo-Mondadori, Buenos Aires Bruniard, E. (1987) Atlas geográfico de la Provincia del Chaco I El medio natural. Instituto de Geografía N° 5, Facultad de Humanidades. Universidad Nacional del Noreste. Dinerstein, E., Olson, D.M., Graham, D.J., Webster, A.L., Primm, S.A., Bookbinder, M.P., Ledec, G., (1995) Una Evaluación del Estado de Conservación de las Ecorregiones Terrestres de América Latina y el Caribe. Banco Mundial, Washington DC. Forman, R. Y Godron, M. (1985) Landscape ecology. Wiley and Sons, Editors. Galmarini, A. y Raffo del Campo, J. (1964) Rasgos fundamentales que caracterizan el clima de la región chaqueña. CONADE N° 9. 178 pp. Grau, R.; Gasparri, I.; Aide, M. (2005) Agriculture expansion and deforestation in seasonally dry forests of north-west Argentina. Environmental conservation, 32 (2):140-148. INDEC. Instituto Nacional de Estadísticas y Censos. Censos Nacionales agropecuarios. www.indec.gov.ar/nuevaweb/cuadros/11/d060104.xls INTA-INDEC (1994) Desarrollo Agropecuario sustentable. Eds. L. Verde y E. Viglizzo. Buenos Aires Montenegro, C.; Gasparri, I.; Manghi, E.; Strada, M.; Bono, J. y Parmuchi, G. (2004) Informe sobre deforestación en Argentina. Secretaría de Ambiente y Desarrollo Sustentable, Dirección de Bosques, Ministerio de Salud y Ambiente. 8pp Morello, J. y Adámoli, J. (1974) Las Grandes Unidades de Vegetación y Ambiente del Chaco Argentino. Segunda parte: Vegetación y ambiente de la Provincia del Chaco. Serie Fitogeográfica, 13. INTA, Buenos Aires. ODSMA - OEA (en preparación) “Evaluación regional del impacto de sustentabilidad de la cadena productiva de la soja”. Análisis ambiental. Prado, D.E. (1993) What is the Gran Chaco Vegetation in South América? I A review. Contribution to the study of flora and vegetation of the Chaco. V. Candolle, 48: 145-172. Reboratti, C. (2005) “Efectos sociales de los cambios en la agricultura”. Ciencia Hoy 15 (87), pp 52-61. SAGPyA. Estimaciones agrícolas. http://www.sagpya.mecon.gov.ar Torrella, S., Herrera, P. y Adámoli J. (2003) “Sostenibilidad de la expansión agraria en la región chaqueña: condiciones favorables y factores limitantes” 3ras Jornadas Interdisciplinarias de Estudios Agrarios y Agroindustriales. Buenos Aires. Viglizzo, E. (2001) “La trampa de Malthus: agricultura, competitividad y medio ambiente en el siglo XXI”. Editorial Universiaria de Buenos Aires, Buenos Aires


Landscape changes due to native forest loss along a precipitation gradient in the Chaco region, Argentina M. G. Parmuchi, J. Bono, M. Stamati, C. Montenegro, M. Brouver, E. Manghi and M. Strada Unidad de Manejo del Sistema de Evaluación Forestal (UMSEF) – Dirección de Bosques – Secretaría de Ambiente y Desarrollo Sustentable de la Nación (Forest Evaluation System Management Unit - Native Forest Division - Secretariat for the Environment and Sustainable Development), San Martín 451 – (1004) Ciudad de Buenos Aires – Argentina [email protected], www.medioambiente.gov.ar/umsef

Abstract From 1998 to 2002, an important landscape transformation occurred in the Chaco region as a consequence of the expansion of the agricultural frontier. It produced not only a loss of forest ecosystems but also an increase in fragmentation and a decrease in connectivity. Therefore, we evaluated landscape pattern changes in the Chaco region in Argentina along a precipitation gradient. We analysed three transects in the east-west direction, from the 1,100 to the 600 mm annual rain isolines, and established 6 sample units in each transect. We calculated landscape indexes for 1998 and 2002 using land coverage obtained by visual interpretation of satellite images. Land classes were Forest Land, Other Wooded Land and Other Land. Different landscape patterns were evident showing association with natural and anthropic factors. Deforestation is present in some samples situated between 700-900 mm but there is not a clear pattern associated with the rain gradient. In areas with high deforestation, indexes allow us to detect more obvious landscape changes. ____________________________________________________________________________

Introduction In Argentina, native forests cover approximately 30.300.000 ha and are the most biodiverse systems in the country. During the last years, the agricultural expansion mainly for the culture of soybean produced a significant conversion of native forest into anthropic ecosystems (UMSEF, 2006). One of the most affected zones is the Chaco region, which is situated in northern Argentina and are dominated by xerophitic forests that cover almost 21.700.000 ha. It is the second largest native forest region in South America after the Amazon rainforest (Eva et al., 2004) and is considered internationally as a key area in terms of biodiversity conservation and a key area for the production of timber and non-timber goods and services (FVSA, 2006). However, the region lost 920.000 ha of forest between 1998 and 2002 with an annual rate of deforestation of –0.9 % (UMSEF, 2006), a process that continues at present. As a consequence, the region has suffered landscape structural changes that have resulted in decreased connectivity among patches, putting species at risk of extinction and modifying population dynamics.


However, spatial changes are not homogeneous across the region since they respond to natural and anthropic variables that influence their probability of occurrence. Water availability determines whether or not land can be used for agriculture. For instance, those areas that receive less than 600 mm annual rain are not suitable for soybean crops in the Chaco region. Moreover, there is a rainfall gradient that historically defined areas under agricultural use (primarily cotton and wheat) in the region (Morello et al., 2005). In relation to anthropic variables, proximity to roads is considered an important factor in the land cover change analysis since it determines accessibility to forest and likelihood of replacement (Geist and Lambin, 2002). It is important to take into account the variation in rain quantity and distribution because climate change, added to economic interests in promoting deforestation to convert land to agriculture, and technological improvements, could modify more intensively forest landscape and consequently their goods and services. The development of policies and programs for conservation, restoration and sustainable use depends on management and land planning strategies based on information about forest status, changes and responses. In this context, the Native Forest Division (Dirección de Bosques) of the Secretariat for Environment and Sustainable Development of Argentina (Secretaría de Ambiente y Desarrollo Sustentable) is in charge of monitoring status of native forests through analysis of deforestation, fragmentation and their causes. For this, we consider the landscape ecology frame since its theories and methodological tools allow us to build an integrated approach (Forman and Godron, 1986; Gustafson, 1998; Turner et al., 2001; Lawrence et al.; 2000; Vos et al., 2001). The objective of this work is to analyse landscape pattern status and changes in the Chaco region in Argentina along a precipitation gradient. Study Area The Chaco region comprises Humid, Semiarid, Arid and Low Mountain subregions according to climatic and geomorphologic criteria (Figure 1). The Humid subregion is characterized by annual rainfalls that vary from 750 to 1300 mm. The most important tree species are Schinopsis balansae (quebracho colorado chaqueño), Aspidosperma quebracho-blanco (quebracho blanco), Astronium balansae (urunday), Ziziphus mistol (mistol), Phyllostylon rhamnoides (palo amarillo) and several species of the genus Prosopis. The Semiarid subregion is characterized by annual rainfalls from 750 to 1300 mm and is dominated by xerophyllous open forests where the main tree species are Schinopsis lorentzii (quebracho colorado santiagueño), A. quebracho blanco, Z. mistol, Bulnesia sarmientoi (palo santo) and several species of the genus Prosopis. Forests alternate with wetlands, prairies and palms (Cabrera, 1976; Red Agroforestal Chaco Argentina, 1999). This study included the provinces of Chaco, Formosa, Santiago del Estero, Salta and Tucumán in Argentine.


Figure 1. A) Study area, transects and samples, B) Republic of Argentina in South America and C) Study Area in the Chaco region in Argentina.

Methods In this study we used the land cover classification carried out by UMSEF for 1998 and 2002 (UMSEF, 2002, 2003a, 2003b, 2004a, 2004b, 2005) through visual interpretation of Landsat 5 TM and 7 ETM images, at a scale of 1:50.000 with a minimum map unit of 10 ha. Land cover classes are Forest Land (FL), Other Wooded Land (OWL) and Other Land (OL) (Table 1). Land cover classes are based on crown cover and physiognomic features, according to Forest Resources Assessment (FRA-FAO, 2000) classification and adapted to Argentinean characteristics.

Table 1. Land Cover Classes Definition. Land Cover Class Forest Land

Definition Land with tree crown cover of more than 20 percent. The trees should be able to reach a minimum height of 7 meters at maturity in situ. It may consist of closed forest formation where trees of various storey and undergrowth cover a high proportion of the ground.

Other Wooded Land

Land with tree crown cover of 5-20 percent able to reach a height of 7 m at maturity in situ or crown cover of more than 20 percent of trees not able to reach a height of 7 m at maturity in situ or with shrub cover of more than 20 percent.

Other Land

Land not classified as forest land or other wooded land as defined above. It includes agricultural land, meadows and pastures, built-on areas, barren land, among others.


In order to study the landscape pattern change, we set up three 500 km-long transects (north: N, center: C and south: S) in a west-east direction considering the rainfall gradient and in each transect established six equidistant circular samples of 100,000 ha (Figure 1). In each sample we calculated the following landscape indexes (Forman and Godron, 1986; Forman, 1995) for 1998 and 2002: • • • • • • •

FL, OWL and OL Area to document the presence and representativity of each class Number of FL, OWL y OL Patches to characterize landscape configuration, in particular spatial heterogeneity Mean Size & Largest FL Patches to characterize landscape configuration, in particular continuity Mean Distance to Nearest FL Neighbor (MDNN) to characterize patch isolation Deforested Area in the period 1998-2002 Road Density to characterize accessibility Annual Deforestation Rate (r) to characterize the rate of native forest loss in relation to forest area at the beginning of the period (Puyravaud, 2003).

In this study we used Arcview GIS 3.2. Results & Discussion Figure 2 shows samples for 2002 situated in the three selected transects. Each one presents a particular landscape pattern although in some cases samples share characteristics along the precipitation gradient. Samples in the east (number 6), that receive more than 1000 mm of annual rain, have an OL matrix with several FL small irregular patches while samples in the 700 to 900 mm isohyets shows a similar pattern but their forest patches present regular shapes (C1 and S4). Moreover, we distinguish a set of samples with less than 600 mm of precipitation with a FL matrix and OL or OWL perforations or gap formation (N2 and C2).

Figure 2. 2002 Samples of the North (N), Centre (C) and South (S) transects along a precipitation gradient (PP).


In relation to deforestation between 1998 and 2002, we notice that samples C1 and S4 have the highest values (more than 6,000 ha) and also high annual deforestation rate (near -6.7 %) due to the small area covered by FL in 1998 (Fig. 3 and 4). They are located in areas with annual rainfalls above 700 mm. On the other hand, S3 has a large deforested area but lower annual deforestation rate (-1.7 %) because it preserves a large forest area in 1998.

Figure 3. Distribution of deforested areas in the different samples during 1998-2002 period along a precipitation gradient.

Figure 4. Deforested Areas (ha) and Annual deforestation rate.


The number of FL patches in samples C1, S4 and C4 increased dramatically as a consequence of a forest fragmentation process in these areas. On the other hand, the number of FL patches in samples in the east and those with precipitation less than 600 mm do not change during the period (Figure 5).

Figure 5. Number of Forest Land Patches in 1998 and 2002.

Figure 6 illustrates different examples of typical and contrasting landscape pattern in the study area (N2, C1 and S6). N2 has low deforestation (conversion to cattle range) since it is placed in the area with the lowest precipitation and poor water access for irrigation. It is characterized by one large FL patch that covers almost the whole area and little variation between years, which causes a low matrix perforation. C1 is located in places where there are not water restraints to agriculture and thus presents high deforestation. This process causes an increase of fragmentation that is evident through new FL patches smaller than 250 ha in 2002 and shrinkage of high FL patches (1,000-10,000 ha). Since C1 is part of an area that was historically designated for agriculture, it presents an OL matrix and several small and irregular FL patches. Finally, S6 is situated in the east where annual rainfalls are above 1,000 mm and presents an OL matrix that corresponds to wetlands. They are designated for cattle range and are not useful for agricultural uses, and so they do not suffer deforestation or spatial changes.


Figure 6. Examples of different land patterns.

We distinguish two cases when analysing changes in mean distances between FL patches and their nearest neighbor between 1998 and 2002: a decrease in samples N1 and S4 and an increase in C1, C3, S2 and S3 (Table 2). We detect that there are new patches as a result of dissection of the forest and also a patch size decrease during the period in most samples. The index changes according to the dominance of one of these processes. When fragmentation is the dominant process, the index decreases because there are new FL patches and less distance between them. Contrarily, when the main process is shrinkage, the index increases since there is the same number of patches but they are more isolated. Other samples such as C4 and S1 do not show changes in the index although they have changes between years but they are offsetting.

Table 2. Mean Distance to nearest neighbour (MDNN) Forest Land (FL) Patch in 1998 and 2002 (PP: precipitation ranges). MDNN (m) 1998 Sample N1 N2 N3 N4 N5

PP (mm) 800-700 1000 800-700 1000 800-700 1000

368.1 309.3 0.0 291.9 265.6 184.3 517.3 663.4 634.4 227.7 468.5 470.2 261.9

368.9 370.0 0.0 305.6 264.5 184.3 517.3 660.4 679.3 279.3 456.2 470.2 260.3

We are not able to calculate the index for samples N2, N3 and C2 because they are composed of only one FL patch that covers almost the whole sample (Table 3). However, N2 and N3 present some deforested areas which results in a matrix perforation and then in the decrease of the largest FL patch size (Table 3). Although C1 and S4 have a similar landscape pattern (a large OL matrix alternated with FL patches) and present a decrease in mean FL size due to high deforestation, both samples behave differently in relation to mean OL size (Table 3). In C1, forest fragmentation and reduction seem to be the dominant processes and they lead to an increase of OL matrix size without an outstanding change in the number of OL patches (from 17 to 13), increasing mean OL size. On the other hand, deforestation in S4 mainly produces a perforation of remnant FL patches causing an increase of the number of OL patches (from 19 to 34) and consequently a decrease in mean OL size. Moreover, these samples show a reduction of approximately 36% in the largest FL patch size (Table 3).

Table 3. Mean and Largest Patch Size in 1998 and 2002 (FL: Forest land, OWL: Other wooded land, OL: Other land, PP: precipitation ranges).

Sample N1 N2 N3 N4 N5 N6 C1 C2 C3 C4 C5 C6 S1 S2

PP (mm) 800-700 1000 800-700 1000 800-700 1000

2516 244 176 69

150 116 193 89

806 3787 8895 10413

2266 175 166 69

150 116 191 89

729 2302 9005 10412

63242 5823 5706 853

59560 3691 5706 853

The south-centre of the study area shows a high road density that is related to a historical agricultural use (S4) while the west-centre is characterized by a low road density (N3 and C3). Isolated areas have low probability of having important native forest loss. However, we do not find a direct relation between road density and deforestation (Figure 6). In general, roads are non-pavement and samples that have more pavement or consolidated roads do not show more deforestation.

Figure 6. Road Length & Deforestation (a) and Road types (b).


Conclusions Along transects different landscape patterns are evident showing association with natural and anthropic factors that are not at present necessarily associated with the rain gradient. In the East, where precipitation exceeds 900 mm, there is a natural pattern characterized by an OL matrix (mainly wetlands) and a high density of small FL patches that do not vary in time. In the zones with precipitation lower than 700 mm, in general there is a FL matrix that presents OL perforations due to cattle range or old dry riverbeds. Between 700 and 900 mm, the most common pattern is an OL matrix with small geometric FL patches as a consequence of human pressure mainly for the development of agriculture that is evident during the analyzed period. Deforestation is an evident process in some samples situated between 700-900 mm but there is not a clear pattern associated with the rain gradient although it is well-known that the 600 mm isohyet is a threshold for the culture of soybean. Although it is expected that areas with precipitation above 700 mm show high deforestation, some samples present little or no change. In general, this behavior is due to the lack of infrastructure and the presence of wetlands. On the other hand, some deforested areas appear in zones with precipitation below 600 mm. They correspond to areas where shrubs have been eliminated for cattle grazing or those areas that have access to irrigation channels. In areas with high deforestation, indexes allow us to detect more obvious landscape changes, that are reflected in an increase in FL patches fragmentation and reduction, affecting their connectivity. Acknowledgements To Ph. D. Andrew J. R. Gillespie for his useful suggestions. References Cabrera, A. L. (1976) Regiones Fitogeográficas Argentinas. Enciclopedia Argentina de Agricultura y Jardinería. 2ª Edición. Tomo II. Fascículo I. Acme S.A.C.I., Buenos Aires. Argentina. 85 pp. Eva, H. D; Belward, A. S; De Miranda, E. E; Di Bella, C. M; Gond, V; Huber, O; Jones, S; Sgrenzaroli M. & Fritz S. (2004) A land cover map of South America. Global Change Biology 10: 731–744 Food and Agricultural Organization (FAO). Evaluación de los Recursos Forestales al Año 2000. Forman, R. T. T. & Godron, M. (1986) Landscape Ecology. John Wiley & Sons. New York. 619 pp. Forman, R.T.T. (1995) Land Mosaics: The Ecology of Landscapes and Regions, Cambridge University Press, United Kingdom, 632 pp. Fundación Vida Silvestre Argentina (FVSA). (2006) La situación ambiental Argentina. 587 pp.


Geist, H. J. & Lambin, E. F. (2002) Proximate causes and underlying driving forces of tropical deforestation. Bioscience, 52:143-150. Laurance, W. F; Vasconcelos, H. L. & Lovejoy, T. E. (2000) Forest loss and fragmentation in the Amazon: implications for wildlife conservation. Oryx, 34 (1): 39 – 45. Morello, J; Pengue, W. & Rodríguez, A. (2005) Etapa de uso de los recursos y desmantelamiento de la biota del Chaco. Fronteras, 4:1-18. Puyravaud, J. P. (2003) Standardizing the calculation of the annual rate of deforestation. Forest Ecology and Management, 117: 593-596. Red Agroforestal Chaco Argentina. (1999) Estudio Integral de la Región del Parque Chaqueño. Gerencia Técnica Bosques Nativos. Secretaría de Desarrollo Sustentable y Política Ambiental. Ministerio de Desarrollo Social y Medio Ambiente. Informe Esencial. Unidad del Sistema de Evaluación Forestal (UMSEF). (2002) Cartografía y Superficie de Bosque Nativo de Argentina. Dirección de Bosques, Secretaría de Ambiente y Desarrollo Sustentable, Ministerio de Salud y Ambiente. Buenos Aires, Argentina. 32 pp. Unidad del Sistema de Evaluación Forestal (UMSEF). (2003a) Mapa forestal provincia del Chaco. Actualización Año 2002. Dirección de Bosques, Secretaría de Ambiente y Desarrollo Sustentable, Ministerio de Salud y Ambiente. Buenos Aires, Argentina. Unidad del Sistema de Evaluación Forestal (UMSEF). (2003b) Mapa forestal provincia de Salta. Actualización Año 2002. Dirección de Bosques, Secretaría de Ambiente y Desarrollo Sustentable, Ministerio de Salud y Ambiente. Buenos Aires, Argentina. Unidad del Sistema de Evaluación Forestal (UMSEF). (2004a) Mapa forestal provincia de Tucumán. Actualización Año 2002. Dirección de Bosques, Secretaría de Ambiente y Desarrollo Sustentable, Ministerio de Salud y Ambiente. Buenos Aires, Argentina. Unidad del Sistema de Evaluación Forestal (UMSEF). (2004b) Mapa forestal provincia de Santiago del Estero. Actualización Año 2002. Dirección de Bosques, Secretaría de Ambiente y Desarrollo Sustentable, Ministerio de Salud y Ambiente. Buenos Aires, Argentina. Unidad del Sistema de Evaluación Forestal (UMSEF). (2005) Mapa forestal provincia de Formosa. Actualización Año 2002. Dirección de Bosques, Secretaría de Ambiente y Desarrollo Sustentable, Ministerio de Salud y Ambiente. Buenos Aires, Argentina. Unidad del Sistema de Evaluación Forestal (UMSEF). (2006) Monitoreo del Bosque Nativo de Argentina – Período 1998-2002. I Jornadas-Taller Nacionales de Protección y Manejo Sustentable del Bosque Nativo. Entre Ríos, Argentina. Vos, C. C; Verboom, J; Opdam, P. F. M; & Ter Braak, C. J. F. (2001) Toward Ecologically Scaled Landscape Indices. The American Naturalist, 183 (1): 24 - 40.


Landscape mosaic, habitat structure and fragmentation of native forests at Córdoba mountain areas (Argentina central). M. Menghi1 and R. del Sueldo CONICET1, Centro de Ecología y Recursos Naturales Renovables (CERNAR), Edificio de Investigaciones Biológicas , Facultad de Ciencias Exactas, Físicas y Naturales- Universidad Nacional de Córdoba. Ciudad Universitaria, X5016GCA, Córdoba, Argentina. email:[email protected] __________________________________________________________________________ Abstract The growing of human pressures over mountainous areas that directly affects native ecosystems, in particular forests, is a major issue of world wide concern. In Central Argentina, the mountain region at Córdoba, native biodiversity is under serious threat due to exotics plantations, as well as urbanization that are expanding without an integrated plan that looks after a sustainable use of natural resources, including the visual landscape. The present work, focused mainly on the diagnosis and hypothesis exploration, it presents first results obtained from a current landscape spatial structure analysis, as well as habitat diversity and status study. Native woods of Polylepis australis “tabaquillo” at upper areas, and of Lithraea ternifolia “molle” toward lower ones, have both shown huge retraction and fragmentation into small isolated patches, as well as habitat alteration, which appear to be mainly related to traditional wood cutting, grazing livestock and burnt. Toward middle and lower sectors of the studied basin, the Pinus plantation caused high retraction and fragmentation of the Hetherotalamus alienus “romerillo” shrub-land. Recent intensification of land use increased the areas with complete decline of native woody strata. The interplay of Pinus plantation and/or urbanization, among others pressures, are increasing the rate of wood habitat loss and subsequently transforming the landscape. _____________________________________________________________________ key words: elevation gradient, spatial variation, habitat status, boundary type, diversity

Introduction Topography and rock substrate are major sources of landscape heterogeneity at Córdoba mountains (central Argentina). Different geological parent materials, their associated landforms and soil conditions, combined with climate are determinants of local and regional plant spatial patterns along an elevation gradient from 650 to 2950 m. a. s. l. (Luti et al.1979; Menghi et al., 1989; Acosta et al., 1992, among others). Besides natural conditions, present and past human activities have also affected plant spatial patterns, with the increasing trend to substitute the native plant cover. The historical human-nature interaction through livestock rearing and wood-cutting based on native resources, as well as fire, were major sources of vegetation shaping up to 1950s. Related to those disturbances, the presence and abundance of some components were altered but the broad structure of native ecosystems has remained relatively unchanged from a physiognomic point of view, with exception of woods. Toward lower mountainous areas, wood cutting led to overgrowth of the less productive under-story spiny shrub species,. At the opposed elevation end, the same disturbance have mainly promoted the conversion of

Understanding biodiverity loss: an overview on forest fragmentation in South America 1 50

woodlands to shrubby growth forms of the same tree species, as well as the wood habitat loss against the growing of secondary grasslands or rock outcrop areas. In agreement with recent global trends related to the advance of agriculture, tree plantation and urbanization over mountainous areas, the human impact over ecosystem and landscape components of Córdoba mountains has been growing in both, intensity and extension. The conversion of wood habitat to open grasslands or its replacement by exotic wood or by bare or paved soil (roads and residences) are pointed out. The expanding of Pinus spp. plantations, incipient 30 years ago (Menghi and Luti, 1982), has increased the loss rate of natives habitat as well as the disruption of extensive patches into small isolated ones, with notable impact on species diversity and on visual landscape. The process, which has already started, is complex ought to the natural complexity of ecosystems as well as to the presence of actors and social elements interacting at different spatio-temporal scales. In addition, there are very few integrated studies done aiming to support the development of resources use programs. Present work, focused mainly to the diagnosis and hypothesis exploration, intended to stand out landscape ecological elements and interactions that are predominant for large geographical areas (Forman and Godron, 1986; Turner and Ruscher, 1988; Wiens et al., 1989; Turner, 1990; Turner and Gardner, 1991; Gardner and O´Neill, 1991; Matteucci, 1998, among others) as well as to gain insights into those proper of more detailed scales. In the present study, we analyzed current landscape spatial patterns, habitat status, the habitat and boundary variations along the elevation gradient, with emphasis on P. australis “tabaquillo” and of L. ternifolia “molle” native woods. Study area The study was carried out at “Los Reartes” water-basin river (31°50'S/64°50'W) (central Argentine) (Fig. 1). It involves an area of 77.622 ha varying in geomorphology, rock substrate and soil along an elevation gradient from 650 to 2400 m. a. s. l. (Ambrosino, 2000). From a bio-geographical point of view, the basin is included in the Chaco province (Cabrera, 1976). The potential native plant cover is represented by more or less continuous belts with domain of wood of L. ternifolia “molle” (from 900 to1100 m. a. s. l.), of shrub-land of Hetherotalamus alienus “romerillo” and of grasses, from lower to upper sectors respectively (Luti et al., 1979). At the upper sector, the grassland domain could have a coarse-grained spatial pattern including patches of woodland and shrub-land of P. australis “tabaquillo” (over 1500 m. a. s. l.). Historically, the economic activities were livestock rearing and wood cutting based on native plant resources. Currently, the Pinus plantation is displacing native vegetation over extensive areas (del Sueldo, 2004), and activities like tourism and urbanization are experimenting a sustainable growth. Methodology The landscape spatial structure analysis followed two main steps (Fig. 1 A, B). It was based on a subset of a Landsat TM image (pixel of 30 x 30 m) (Path Row 229/ 83) (Comisión Nacional de Actividades Espaciales, CONAE) dated September 2002, when the contrast between native (semi-deciduous) and exotic (evergreen) woods was enhanced. Georeferencing (to the Gauss/Kruger projection), geometric and radiometric corrections were carried out through a digital elevation model based on topographic maps N° 3166-36-3 and N° 3166-36-4 (scale 1:50.000) (Military Geographic Institute) (Fig. 1 A). Seventeen control points at sites easily identifiable in the satellite image, were obtained with a Magellan 2000 XL GPS.


Land cover classes were separated by a physiognomic criteria based on authors previous experience, as well as on current field observations and image resolution, taking into account a minimum mapping unit of 3 x 3 pixels (0.81 ha). A supervised classification, based on a false color composition (bands 4, 3, 2) and the maximum probability algorithm (ENVI 3.5) (Chuvieco, 1996; Johnston, 1998), was carried out. The first step included the analysis of all the basin (77.622 ha) (Fig. 1A) at mosaic and patch levels. Broad and clear land cover types at the resolution level of analysis were considered: wood of L. ternifolia “molle”; wood of Pinus spp. “pino”, agriculture, urban, artificial lake, shrub-land of H. alienus “romerillo”; shrub-bland of Acacia spp. “espinillo”, open grasslands. The upland grassland domain was classified like a mosaic ought to the mix of numerous small patches of “tabaquillo” (trees or shrubs), as well as of areas locally dominated by rock outcrop, by tall tussock grasses or by lawns, difficult to discriminate over extensive areas. The second step (Fig 1 B) was focused on an area of 21.000 ha including extensive Pinus plantations, as well as on analysis at mosaic, patch (Fig. 1 B a) and boundary (Fig. 1 B b, c) levels. Sub-areas of 1.800 ha (N=12), three at each of four sectors (upper, middle upper, middle low, lower) along the elevation gradient (Fig. 1B, a) were selected, excluding the valley and piedmont. This analysis considered the same land cover types of first step, plus the local habitat variation of the grassland belt, previously not discriminated. The variety and number of boundaries (sensu Rescia et al. 1994) were registered within fifteen circles of 240 m of radio, regularly distributed at each area of 1800 ha (N= 180) (Fig. 1 B b, c). Figure 1. Localization of the study area and simplified scheme of the methodological procedure.

Location of "Los Reartes" river basin

Landscape analysis Based on the land cover maps obtained, the spatial structure at mosaic, habitat and patch levels (Fragstat 3.3) (Mc Garigal et al., 2002) was analyzed.


Each habitat was analyzed considering its total area, number, patches density and mean area, area of largest patch. From an existing data base and current field observations (visual inspection and systematic sa mpling along interception lines of 60 m in length), the native habitat was also analyzed at plant community level. The number of strata; the maximum strata height, the plant cover; the floristic composition and the character (native or exotic) of dominant species were considered. The habitat and boundary diversities of landscape mosaic, as well as their spatial variation along the elevation gradient, were calculated through the Shannon index (H´ = - Σ pi log pi) (Magurran, 1988). The richness (r) was the variety of the considered element (habitat, or boundary type). The relative abundance (pi) of each element was calculated from the percentage of area (habitat) or the number (boundary), represented in their respective totals within the analyzed area. Results and discussion The mosaic has shown predominance of rural landscape with a native plant cover matrix, which represents the 77 % of the total area. The matrix is made by grasslands, shrublands and woodlands varying in their conservation degree, with irregular edges and gradual transitions between adjoining habitat (Fig. 2, Table 1). It is suggested that this type of contact could function like a biological corridor (Taylor et al., 1993) and, from the high frequency (81%) detected all over the studied area (Table 4), it could provide connectivity to native ecosystems. Figure 2. Map of land cover types of "Los Reartes" river basin


Map references: M[G+Pol+R]: Mosaic of grassland (G), wood of Polylepis australis "tabaquillo" (Pol) and rock outcrop (R) (from 1.400 to 2.400 m a. s. l.). Grassland patch (G) (30%). It includes tall tussock grasses (Deyeuxia hieronymi, Festuca lilloi, F. hieronymi, Stipa pseudohichu), short sized grasses (Stipa eriostachya, S. flexibarbata, Agrostis montevidensis, Eragrostis lugens, montevidensis, Poa annua, Carex fuscula) and lawns (Alchemilla pinnata, peruviana, Gentiana achalensis).

Poa stuckertii, neesiana, S. Piptochaetium Muhlembergia

Polylepis patch (Pol) (10%), Woodland/shrubland of Polylepis australis "tabaquillo" (7 - 9 m) , Maytenus boaria "maiten", Berberis spp, Escallonia cordobensis, Heterothalamus alienus, Cassia hookeriana, Colletia spinosissima, Pernettya poeppigii, Baccharis spp., and ferns. Rock patch (R) (60%). Predominance of rock outcrop with Stevia achalensis, Solanum incisum, Berberis hieronymi, Satureja odora, Blechnum penna-marina, Elaphoglossum gayanum y Polypodiun gilliesii, and grasses. G: Grassland (from 1.000 to 1.800 m a. s. l.). The floristic composition could vary with altitude and the conservation degree. Deyeuxia and Festuca species are dominant upwards, and Stipa filiculmis, S. trichotoma, S. tenuissima, Paspalum dilatatum, P. notatum downwards. Schizachyrium spicatum, S. imberbe, Sporobolus indicus, Aristida spegazzini, among others, are common on less fertile and/or degraded soils. Het: Hetherotalamus shrubland (from 1.100 to 1.400 m a. s. l.). It is a low sized shrubland (1-2 m) with predominance of Heterothalamus alienus “romerillo”, Eupatorium buuniifolium, Baccharis spp., over a grassland matrix. Emergent L. ternifolia "molle" trees, and/or spiny shrubs can be frequent downwards. Aca: Acacia shrub-land (from 900 to 1.100 m a. s. l.). It is tall sized (2-5 m) shrub-land, closed or open, with predominance of spiny shrubs of Acacia spp. “espinillo”, Schinus spp. "moradillo", grasses and forbs. Isolated trees of L. ternifolia "molle" and Fagara coco "coco" can be observed accompanied by Colletia spinosissima, Condalia microphylla, Baccharis articulata, Geoffroea decorticans, Lippia turbinata, Lycium spp., among others. Lit: Lithraea ternifolia "molle" woodland (from 900 to 1.100 m a. s. l.). The well preserved wood has a tree stratum ( 7 m) made by L. ternifolia, Fagara coco “coco”; Ruprechtia apetala "manzano del campo", accompanied by shrubs (Acacia spp., Schinus spp, Bacharis spp.), grasses and forbs. Towards lower areas the tree stratum could include Aspidosperma quebracho-blanco "quebracho blanco"; Prosopis spp. "algarrobo", Celtis tala "tala", Geoffroea decorticans "chañar", Jodinia rhombifolia "sombra de toro". Pin: wood of Pinus spp. "pino". Plantation of Pinus elliottii, P. taeda and P. radiata (P. insignis). Agr: Cultures. Zea maiz "maíz"; Solanum tuberosum "papa"; Sorghum spp., Avena spp., Eragrostis curvula "pasto llorón"; Secale cereale “centeno argentino”. Urb: Urban. Villages, towns and secondary residences. Lak: Artificial lake Los Molinos.

With exception of the complex mosaic ( M[G+ Pol + R] ) and the lake (Lak), most of the landscape units have shown numerous patches with averages areas ranging from 0.64 to


6.62 ha, as well as small large patch areas (Table 1). The "romerillo" shrubland (Het) was an exception showing a large patch which represented the 17.84 % of basin surface. It is pointed out, the current small area of well preserved L. ternifolia "molle" (Lit) wood (3148 ha), which includes only one patch of 415 ha, two of 100 ha each one, and numerous small patches of less than 1 ha. This wood area represents the 19 % of its potential. From the analysis by elevation sector (Table 2) emerged that the mosaic (M[G+ Pol + R]) (Fig. 2) is made by 30 % of grasslands (G), 10 % of "tabaquillo" woodland/shrub-land (Pol) and 60 % of rock outcrop. The detailed analysis of 5400 ha (Fig. 1B), revealed the presence of 282 ha of "tabaquillo" wood (Pol) fragmented in 1016 patches with an average area of 0.26 ha ± 0.45, or 0.14 ha ± 0.8, depending on the altitude (upper, or middle-upper sectors of the basin). The large part was related to ravines. With respect to grassland habitat (G), the mapped area (Fig. 2) involves different community types and processes along the elevation gradient. The 30% of (G) reported for the mosaic at upper sector, includes semi-natural climax and well preserved grasslands and lawns (C), as well as degraded ones (del Sueldo, 2004). The contrasting proportion of (G) and (C) areas detected (Table 2), could be partially explained by the advance of degraded lawns against productive grasslands lost by burnt and overgrazing. In contrast, a large part of grassland habitat (G) mapped towards lower sectors (Fig. 2), is made by secondary communities including many exotic species, which are promoted by “molle” wood or shrub-land alteration. Table 1. Spatial structure of land-cover types detected on "Los Reartes" river basin

Land-cover types 1. M[G+ Pol + R] 2. G 3. Het 4. Pin 5. Aca 6. Lit 7. Agr 8. Urb 9. Lak

Ta (ha) (%) 19.252 (27) 4.060 (6) 15.547 (22) 7.670 (11) 13.080 (18) 3.148 (4) 7.295 (10) 701 (1) 647 (1)

Np 85 3.731 2.349 1.695 3.201 2.812 1.400 1.091 1

Lpi (%) 24.76 0.61 17.84 2.27 9.64 0.53 7.65 0.42 0.83

Map (ha) 226.5 1.09 6.62 4.52 4.08 1.12 5.21 0.64 323.73

References: (M[G + Pol + R]: mosaic of grassland, Polylepis woodland and rock outcrop; G: grassland; Het: Hetherotalamus shrubland; Pin: Pinus plantation; Aca: Acacia shrubland; Lit: Lithraea woodland; Agr: agriculture; Urb: urban; Lak: artificial lake. Ta: total area ; Np: number of patches; Lpi: large patch index; Map: mean patch area.

The rest of basin surface (23 %) is covered by exotic and/or artificial habitat (agriculture, Pinus plantation, lake, urban) (Tables 1), which are more frequent towards lower areas (Fig. 2). This type of habitat has in common sharp and straight edge, as well as a simplified ecosystem structure. The landscape mosaic contrast is enhanced by the boundaries sharpness and the different nature of adjoining habitats (earth/water; native/exotic; natural/build, among others). Concerning organism movements, the described spatial structure, could be resistant to desirable native species, and at the same time permeable for opportunistic native and/or exotic ones.


Table 2. Spatial structure of land-cover types at each sector of the elevation gradient of "Los Reartes" river basin Land-cover types / Sectors Upper 1. M[G + R] 2. C 3. Pol 4. G 5. Het 6. Pin

Ta (ha) (%)

Np

Lpi (%)

Map (ha) (SD)

444 (23.7) 385 (21.6) 235 (13.0) 198 (11.0) 446 (24.7) 107 (6.0)

946 929 674 733 879 361

2.61 7.98 0.43 0.26 4.08 0.19

0.47 (2.40) 0.44 (5.03) 0.26 (0.45) 0.27 (0.37) 0.70 (3.42) 0.10 (0.20)

274 (15.3) 165 (9.18) 52 (2.86) 354 (19.7) 780 (43.4) 175 (9.71)

746 772 342 729 551 149

0.17 0.18 0.04 1.45 19.2 0.64

0.37 (0.40) 0.21 (0.24) 0.14 (0.08) 0.47 (1.54) 1.53 (17.8) 0.88 (1.29)

Middle Low 1. G 2. Het 3. Pin

587 (32.5) 791 (44.0) 422 (23.4)

322 399 311

23.5 21.7 5.38

0.50 (3.30) 7.43 (65.4) 1.47 (6.56)

Lower 1. G 2. Het 3. Pin 4. Aca 5. Lit 6. Agr

255 (14.3) 552 (31.0) 296 (16.5) 555 (31.0) 109 (6.0) 33 (1.8)

777 842 198 790 345 41

1.20 6.27 7.44 7.40 0.48 0.06

0.32 (1.07) 0.65 (4.88) 1.26 (8.88) 1.03 (9.05) 0.10 (0.87) 0.53 (0.51)

Middle Upper

1. 2. 3. 4. 5. 6.

M[G + R] C Pol G Het Pin

References: (M[G + R]: mosaic of grassland and rock outcrop; C: lawn “césped”; Pol: Polylepis woodland G: grassland; Het: Hetherotalamus shrubland; Pin: Pinus plantation; Aca: Acacia shrubland; Lit: Lithraea woodland; Agr: agriculture). Ta: total area ; Np: number of patches; Lpi: large patch index; Map: mean patch area. SD standard deviation.

The native habitat retraction and fragmentation detected (Tables 1, 2), involved decrement of areas with and “interior effect” and increments of the length of boundaries and of the variety of contacts (Table 4) and therefore of "edge effects" . The dynamic and direction of the biotic transition promoted (Peters et al., 2006), as well as the regulating factors, vary along the elevation gradient. These factors would be mainly related to local biotic-abiotic interactions upwards, and to the intensity of human impact toward lower sectors. With respect to species changes, it was observed that the selective logging and/or complete decline of “molle” trees led to the overgrowth of under-story spiny shrubs with predominance of Acacia spp. (Aca) (Fig. 2, Table 1), which would be promoted by competition process for light. At present, this secondary shrub-land covers the large part (13.300 ha) of the original area of the “molle” wood maintaining gradual transitions with the


remaining well preserved patches. These disturbed areas are being notably invaded by woody exotic species (Ligustrum lucidum “ligustro”, Pyracantha coccinea “crateus”, Gleditsia triacanthos “espina corona”, Melia azedarach "paraiso"; Prunus persica “durazno”, etc.), which have references of high aggressiveness against native plant species related to different disturbed native ecosystems (Vitouseck et al., 1997; Aragón and Morales, 2003; Casavecchia, 2005; Aimar et al., 2006). This process seems to agree, at least in part, with the uni-directional transition hypothesis (Peter et al., 2006), and it is suggested that the spontaneous recovering of the original tree species could be limited. At the upper sector of the basin, man disturbance has mainly promoted the conversion of “tabaquillo” wood (Pol) to shrubby growth forms of the same species. Also a directional process was observed, causing woody habitat retraction against the growth of secondary grasslands, or eroded and rock outcrop areas. The exotics invasion would be minor at the upland grazed grasslands (Diaz et al., 1994), and, up to present, also at disturbed “tabaquillo” woods. That problem increases at areas higly transformed by man activity, like rural towns (Villa Alpina, La Cumbrecita) and roadsides. The Pinus plantation (Pin), the most extensive exotic habitat (7670 ha) at the studied basin (Table 1), has replaced the native shrub-land of Hetherotalamus “romerillo” (Het) in the 40 % of its original area, semi-natural grasslands of Festuca and Stipa related to moderated relief and elevation, and a minor proportion of "molle" wood areas already altered (Fig. 2). The Pinus patches have extensive nucleus involving physical conditions adverse for native plant species, as well as 1.200 km of straight edges, sharp and about 20 m height. This type of edge itself, could influence organism movement (native or exotic) and function like a shelterbelt with microclimate effects over adjoining habitats (Ryszkowski, 1992). Landscape diversity and dynamism The values obtained of habitat diversity (H´1.85), equity (E 0.80) and dominance (0.34) suggest spatial pattern heterogeneity at the regional scale, with local spatial variation among the sectors analyzed along the elevation gradient (Table 3). Table 3. Spatial variation of habitat diversity along the elevation gradient of "Los Reartes" river basin.

Parameters

Upper

Richness (r) Diversity (H’) Equity (E)

6 2.34 0.94

Sectors Middle Middle Low Upper 6 3 2.18 1.53 0.84 0.97

Lower 6 2.20 0.85

The boundary analysis (Table 4) pointed out a similar spatial pattern derived from the diversity index along the elevation gradient, the predominance of native/native boundaries and a clear trend of native/exotic contacts increase in proportion (from 16,7 to 50%) and frequency ( from 3.1 to 21.4) towards lower areas.


Table 4. Spatial variation of boundary types, abundance and diversity index (H´) along the elevation gradient of "Los Reartes" river basin.

Boundary types M[G + R] / Het M[G + R] / C M[G + R] / G M[G + R] / Pol Pol / Het Pol / C Pol / G Pol / Pin C/G C / Pin Het / C Het / G Het / Pin G / Pin Het / Agr Het / Lit Het / Aca Aca / G Aca / Pin Aca / Agr Aca / Lit G / Agr G / Lit Total number of boundaries (N) Number of boundaries type (n) Proportion of exotic/native boundaries (%) Frequency of exotic/native boundaries (%) Diversity (H’)

Sectors Upper Middle Upper 82 33 112 68 77 34 42 14 150 58 18 21 113 32 1* 99 40 1* 74 47 221 186 10* 14* 21* 10*

1019 12 16.7 3.1 3.14

559 14 28.6 4.7 2.92

Middle Low

Lower

177 46* 59*

144 26* 50* 1* 7 121 141 42* 1* 32 5* 16

282 3 66.7 37.2 1.26

586 12 50.0 21.4 2.11

References: M[G + R]: mosaic of grassland and rock outcrop; Pol: Polylepis woodland; C: lawn “césped”; G: grassland; Het: Hetherotalamus shrubland; Pin: Pinus plantation; Aca: Acacia shrubland; Lit: Lithraea woodland; Agr: agriculture). The abundance values of contacts between natives and exotics land-cover units are enhanced *.

The structural landscape fragmentation detected would be mainly related to the interaction of natural factors upwards, and to man and natural factors downwards. Uplands, the natural spatial variability in landforms like extensive areas with rock outcrops, ravines, undulated plains, among others, have shown to be a strong determinant of a cellular plant spatial pattern (Acosta et al., 1992). That pattern is concordant with the richness of habitat types and high frequency of sharp native/native boundaries detected (Table 4). According to the stationary transition hypothesis (Petter et al., 2006), the major coarsegrained spatial structure, typical of the landscape under consideration, could be relatively stable at regional scale, ought to strong control of ecosystem inherent abiotic constraints. Nevertheless, the human impacts, continuous over long periods, has been promoting boundary fluctuations and patch area changes, which have currently landscape evidences of


high concern. The retraction of well preserved wood (Pol) to ravines, and of grassland (G) areas against degraded lawns, as well as the increment of contacts with poor soil and/or vegetation conditions (ex. C/R; G/C, Pol /C, Pol/G, etc.) (Table 4) are mainly promoted by human pressures. At middle parts of the elevation gradient traditional and new land-uses interactions could enhance landscape dynamics and plant diversity changes at areas relatively stable up to present. The disruption of extensive "romerillo" shrub-land (Het) areas, and of its ecotone with the grassland domain, by Pinus plantations, increased the boundary types and number (Table 4) as well as potential edge effects on the ecosystem (Table 4). The high frequency Het/G contacts, could increase the risk of shrubs invasion on overgrazed and/or burnt grasslands. Furthermore, the Pin/G and Pin/Het contacts could have shelterbelt effects on the adjoining native habitats and influence species composition and productivity. Towards the valley and piedmont, the landscape transformation is closely related to man activities (roads, villages, agriculture, plantation, selective cutting, open grasslands, etc) involving drastic ecosystem changes (soil loss and pavement, alteration of seed bank, elimination of native species and introduction of exotic ones, retraction and isolation of native patches, etc.). These processes appear irreversible, and the "molle" wood seriously endangered.

Concluding remarks The descriptions of plant spatial pattern and physiognomy carried out, agree with previous studies at neighbor mountainous areas along the elevation gradient (Menghi and Luti, 1982; Menghi and Intile, 1992) and at upper sector (Cingolani et al., 2004). The present study, also contributed information on the native habitats alteration degree at landscape level, and on boundary types and frequency. With respect to native woods, the main factors determining current habitat status of “tabaquillo” wood (Pol), are still matter of research and discussion. On one part, there are field observations which support that ravines physical factors (microclimate and relief) would protect the adult and young trees against livestock grazing and fire, and at the same time could provide proper sites for seed germination and seedling growing (Renison et al., 2002). On the other hand, according to habitat models (currently in process), the current wood surface would be quite below its potential area, thus suggesting retraction promoted by human activities. Present results, contributed some evidence for the second hypothesis. Several indicators of burnt trees and/or wood cutting where detected at grassland and at Pol/G boundary areas. This transition, also suggested that the neighbor wood patch is particularly exposed to burnt by fire propagation, ought to frequent use of fire to manage grasslands. At the opposed elevation end, the “molle” habitat alteration is clearly related to past and present human impacts. References for neighbor areas (Convenio UNC-CONEA 1976-1989; Menghi and Luti, 1982; Menghi and Cabido, 1986) pointed out wood retraction when the Pinus plantation activity was incipient at Córdoba mountain areas. Although present results have also shown that the "pino" plantation was not the major cause of loss of the "molle" wood habitat, this fact could be changing. The deforestation process with urbanization goal is also growing at different Córdoba mountain areas, with alarming references (Gavier and Bucher, 2004). The native wood habitat loss, dramatic by itself, involves huge impact on wood ecosystem services like the native visual landscape, animal and plant diversity, water-shed and soil


protection, microclimate, among others, which are the bio-geophysical support for many human activities developed at Córdoba mountain and plain surrounding areas. The diversity and intensity of human actions, and the associated landscape transformation, was historically limited by physical factors at Córdoba mountain areas. This situation is currently changing due to recent access to powerful technology and external capital, and to local and regional people migrations looking for job opportunities and new life styles. The development of infrastructure and services (secure roads, hotels, filling stations, etc.) observed at areas considered of low accessibility up to few years ago, also represents a new focus of human impact on natives ecosystem and diversity. Finally, we conclude that the present study, focused mainly on the diagnosis and hypothesis exploration, therefore help to point out a complex mosaic of native and exotic habitats with multiple edge effects, which are not static. The dynamism and direction of the biotic transitions related to the detected boundary types would affect in different ways the flow of organisms and material (Peters et al., 2006). In the same way, the landscape spatial pattern dynamics described here needs more research. Acknowledgments This study was partially supported by the Secretary of Science and Technology (SECyTUNC projects 2003-2004) and by the “Instituto de Altos Estudios Espaciales Mario Gulich (CONAE- UNC). We specially thanks to Dr. Marcelo Scavuzzo and to Lic. Mario Lamfri for their assessing and help with satellite image processing. Bibliografía Acosta, A.; Díaz, S.; Menghi, M. & Cabido, M. (1992) Patrones Comunitarios a Diferentes Escalas Espaciales en Pastizales de las Sierras de Córdoba, Argentina. Revista Chilena de Historia Natural 65: 195-207. Aimar, L.; Massuh, Y.; Ruiz de los Llanos, E; Ferreiro, G.; del Sueldo, R. & Menghi, M. (2006) La Reserva Natural de Fauna Laguna La Felipa (SE Córdoba, Argentina) No Sería Sustentable Bajo el Manejo Actual. Libro de Resúmenes. XXII Reunión Argentina de Ecología. Córdoba, Argentina. p. 327. Ambrosino, S.N. (2000) Caracterización Geológica y Geomorfológica de la Cuenca Embalse de Los Molinos, Córdoba. Trabajo Final de Grado, Escuela de Geología, Facultad de Ciencias Exactas, Físicas y Naturales, Universidad Nacional de Córdoba, Argentina. Aragón, R. & Morales, J.M. (2003) Species Composition and Invasion in NW Argentinian Secondary Forests: effects of land use history, environment and landscape. Journal of Vegetation Science 14: 195-204. Cabrera, A. (1976) Regiones Fitogeográficas Argentinas. Enciclopedia Argentina de Agricultura y Jardinería. Editorial ACME. Buenos Aires, Argentina. Casavecchia, C. (2005) Aplicación de Imágenes Landsat para Detección y Monitoreo de Bosques de Siempre Verde (Ligustrum lucidum) en la Sierra de San Javier, Tucumán, Argentina. Tesis de Licenciatura. Facultad de Ciencias Exactas, Físicas y Naturales, Universidad Nacional de Córdoba, Argentina. Cingolani, A.; Renison, D.; Zak, M. & Cabido, M. (2004) Mapping Vegetation in a Heterogeneous Mountain Rangeland Using Landsat Data: an alternative method to define and classify landcover units. Remote Sensing of Environment 92: 84-97


Convenio UNC-CONEA (1976-1989) Evaluación del Estado de la Cuenca de Alimentación del Embalse Rio III y Relación de los Recursos Naturales con las Actividades Humanas. Informes l979-l986. CERNAR-UNC. Chuvieco, E. (1996) Fundamentos de Teledetección Espacial. Tomo II. 3ª Edición Revisada. RIALP. Madrid, España. del Sueldo, R. (2004) Análisis de la Estructura Espacial del Paisaje de la Cuenca del Río Los Reartes, Valle de Calamuchita (Córdoba, Argentina). Tesis de Maestría, Centro de Ecología y Recursos Naturales Renovables, Facultad de Ciencias Exactas, Físicas y Naturales, Universidad Nacional de Córdoba, Argentina / Universidad Internacional de Andalucía, España. Diaz, S., Acosta, A. and Cabido,M. (1994). Community structure in montane grasslands of central Argentina in relation to land use. Journal of Vegetation Science (5)4:483-488, Forman, R.T.T. & Godron, M. (1986) Landscape Ecology. John Wiley and Sons, New York, USA. Gardner, R.H. & O’Neill, R.V. (1991) Pattern, Process and Predictability: The Use of Neutral Models for Landscape Analysis. M.G. Turner & H.G. Gardner (Eds.). Quantitative Methods in Landscape Ecology. Ecological Studies; Vol. 82. Springer-Verlag, New York, USA. pp. 289307. Gavier, G. & Bucher, E. (2004) Deforestación de las Sierras Chicas de Córdoba (Argentina) en el período 1970-1997. Academia Nacional de Ciencias. Córdoba, Argentina. Miscelánea N°. 101. Intile, A.C. (1989) Cartografía de la Vegetación y Análisis de Características Estructurales de Interés Paisajístico. Subcuenca del Río Anizacate, Córdoba. Seminario de Grado. Facultad de Ciencias Exactas, Físicas y Naturales, Universidad Nacional de Córdoba, Argentina. Johnston, C.A. (1998) Geographic Information Systems in Ecology. Methods in Ecology Series. Blackwell Science Ltd., Oxford, London. Luti, R.; Solís, M.A.B.; Galera, F.M.; Ferreyra, N.M.; Berzal, M.; Nores, M.; Herrera, M.A. & Barrera, J.C. (1979) Vegetación. J.B. Vázquez; R.A. Miatello & M.E. Roqué (Eds.), Geografía Física de la Provincia de Córdoba. Editorial Boldt, Buenos Aires, Argentina. pp. 297-368. Magurran, A.E. (1988) Ecological Diversity and its Measurement. Princeton University Press, Princeton, NJ, USA. Matteucci, S.D. (1998) La Creciente Importancia de los Estudios del Medio Ambiente. S.D. Matteucci & G.D. Buzai (Comp.). Sistemas Ambientales Complejos: Herramientas de Análisis Espacial. Editorial Universitaria de Buenos Aires EUDEBA. Buenos Aires, Argentina. pp. 19-30. McGarigal, K.; Cushman, S.A.; Neel, M.C. & Ene, E. (2002) FRAGSTATS: Spatial Pattern Analysis Program for Categorical Maps. Computer software program produced by the authors at the University of Massachusetts, Amherst. Available at the following web site: www.umass.edu/landeco/research/fragstats/fragstats.html


Menghi, M. & Luti, R. (1982) Mapa Fisonómico de Vegetación de la Cuenca de Alimentación del Embalse Río Tercero. Escala 1:250.000. Ecología Argentina 7: 185-194. Menghi, M. & Cabido, M. (1986) Un Método Fisonómico-Estructural para la Clasificación de la Vegetación de la Cuenca de Alimentación del Embalse Río Tercero, Córdoba (Argentina). Documents Phytosociologiques X (II):305-317. Menghi, M.; Cabido, M.; Peco, B. & Díaz Pineda, F. (1989) Grassland Heterogeneity in Relation to Lithology and Geomorphology in the Córdoba Mountains, Argentina. Vegetatio 84: 133-142. Menghi, M. and Intile,C. (1992). Landscape variability detected from two sampling procedures. Esc. 1:50.000. Landscape research 17(2):80-89. Peters, D.; Gosz, J.; Pockman, W.; Small, E.; Parmenter, R.; Collins, S. & Muldavin, E. (2006) Integrating Patch and Boundary Dynamics to Understand and Predict Biotic Transitions at Multiple Scales. Landscape Ecology 21: 19-33. Renison, D.; Cingolani, A.M. & Suárez, R. (2002) Efectos del Fuego Sobre un Bosquecillo de Polylepis australis (Rosaceae) en las Montañas de Córdoba, Argentina. Revista Chilena de Historia Natural 75: 719-727. Rescia, A.J.; Schmitz, M.F.; Martín de Agar, M.P.; de Pablo, C.L.; Atauri, J.A. & Díaz Pineda, F. (1994) Influence of Landscape Complexity and Land Management on Woody Plant Diversity in Northern Spain. Journal of Vegetation Science 5: 505-516. Ryszkowski, L. (1992) Energy and Material Flows Across Boundaries in Agricultural Landscapes. A.J. Hansen & F. di Castri (Eds.). Landscape Boundaries. Springer-Verlag, New York. pp. 270-284. Taylor, P. D.; Fahrig, L.; Henein, K. & Merrian, G. (1993) Connectivity is a Vital Element of Landscape Structure. Oikos 68: 571-573. Turner, M.G. & Ruscher, C.L. (1988) Changes in Landscape Patterns in Georgia, USA. Landscape Ecology 1(4): 227-240. Turner, M.G. (1990) Spatial and Temporal Analysis of Landscape Patterns. Landscape Ecology 4: 21-30. Turner, M.G. & Gardner, H.G. (1991) Quantitative Methods in Landscape Ecology. Ecological Studies; Vol. 82. Springer-Verlag, New York, USA. Vitousek, P.M.; D'antonio, CM.; Loope, L.; Rejmanek, M. & Westbrooks, R. (1997) Introduced Species: a significant component of human-caused global change. New Zealand Journal of Ecology 21(1):1-16. Wiens,g J.A. (1989) Spatial Scaling in Ecology. Functional Ecology 3: 385-397.


Patterns of land use change and forest fragmentation in the temperate forests in southern Chile C. Echeverría1, D. A. Coomes2, A. C. Newton3, A. Lara1 and J. M. Rey-Benayas4 1

Millennium Scientific Nucleus FORECOS – Instituto de Silvicultura, Universidad Austral de Chile. Casilla 567, Valdivia, Chile. e-mail: [email protected] 2 Department of Plant Sciences, University of Cambridge, UK. 3 School of Conservation Sciences, Bournemouth University, Poole, Dorset, UK. 4 Departamento de Ecología, Edificio de Ciencias, Universidad de Alcalá, Alcalá de Henares, Spain.

Abstract The spatial patterns of land use change and forest fragmentation were assessed using multitemporal land-cover maps derived from satellite imagery in southern Chile. There was a reduction in forest area of 24% between 1976 and 1999, equivalent to an annual forest loss rate of 1.1%. However, the highest rate is recorded for the period 1976-1985 with 1.6%yr-1. Most of the native forest was converted to shrublands which, in turn, were replaced by crops pasture and agricultural lands. The native forests became increasingly surrounded by arboreus shrublands, which reveal the forest logging around fragment edges. A decline in size and core area and an increase in isolation of forest fragments were observed through time. Results revealed that land-cover change is leading to a transitional landscape, where the matrix of native forests is being converted to a matrix of non-forest elements. Fragmentation was associated with forest logging and expansion of agricultural land. Key words: satellite imagery, landscape indices, forest matrix, forest loss, spatial configuration

Introduction Few studies of the consequences of fragmentation on landscape configuration have been made in temperate forests (Staus et al., 2002). Although previous studies have examined the consequences of fragmentation on biological processes within the remnant habitats (Willson et al., 1994; Bustamante and Castor, 1998), it is also important to quantify the extent of fragmentation at the landscape level in order to establish the population-level effects of fragmentation. Analysis of previous fragmentation studies revealed the need to analyze the matrix habitat to understand the dynamics of remnant fragments (Debinski and Holt, 2000). The interaction of patches with the surrounding matrix appears to be a determinant for species to be retained in isolated patches. If a native forest matrix shifts to a predominance of other land uses, then species dependent on a forest habitat are likely to be adversely affected (Formand and Godron, 1986). For instance, fragmentation of temperate forests in central Chile due to the expansion of a matrix of industrial pine plantations (Echeverría et al., 2006) has reduced the flow of biotic pollinators and dispersers among individuals of the endangered tree Nothofagus alessandrii (Bustamante and Castor, 1998). The purpose of this study is to assess the temporal and spatial changes of landscape elements in a forested matrix influenced by forest fragmentation. In particular, I examined the


spatial changes of land-cover types over time intervals using remotely sensed imagery. Forest fragmentation was assessed over time utilizing selected landscape indices. Method The study area corresponds to 503,287 ha located between 41° 30’ S, 73° W and 42° 20’S S, 74° W in southern Chile (Fig.1). The zone is characterised by a rainy temperate climate with an oceanic influence and without dry-periods (Di Castri and Hajek, 1976), with a mean annual precipitation of 2090 mm. The landscape is dominated by Valdivian temperate rain forests, surrounded by crops and pasture lands. Approximately 40% of the remaining forest occurs on acidic, shallow, poorly-drained soil referred to as ñadis, which is classified as Gleysol. A similar percentage of native forest is found in volcanic ash soil, and the remaining portion corresponds mainly to marine sediments. The forests are characterised by the presence of several broad-leaved evergreen tree species. In some sites, the long-lived conifers such as Fitzroya cupressoides and Pilgerodendron uvifera (both Cupressaceae) can also be found. Anthropogenic disturbance has led to widespread early successional stages of the forest, which are characterized by a high abundance of young trees. In the middle of the 20th century a significant area of native forests was cut down and burnt as a result of European settlements. Satellite data To analyze the spatial and temporal changes of the land use types, a set of three Landsat satellite scenes were acquired at different time intervals: 1976 (Multispectral Scanner), 1985 (Thematic mapper), and 1999 (Enhanced Thematic Mapper). The smallest patches (less than 5 pixels) were removed from all the images. It was necessary to correct the images geometrically, atmospherically and topographically before they could be used to assess changes in forest cover and fragmentation (Chuvieco, 1996; Rey-Benayas and Pope, 1995). Supervised, maximum likelihood classifications were performed on each of the three images to classify the land cover types using training locations, obtained from field surveys. Each land cover map was validated using ground-based data. Overall agreement of classification was 88.8% for the 1976 MSS, 89.6% for 1985 TM image, and 91.9% for the 1999 ETM+ image. Forest loss and landscape spatial pattern analysis The resulting categories of land cover were grouped into forest or non-forest categories to create a binary forest/non-forest map. This map was analyzed using ARC VIEW 3.2 software1 and its extension Arc View Spatial Analyst 2.0 for Windows to quantify land cover change and forest loss. The formula used to determine the annual rate of deforestation was (FAO, 1995):

 P =   

1 /( t2 −t1 )

A  A  2

1

 − 1 * 100  

(Equation 1)

where P is the percentage loss per year, A1 and A2 are the forest area at time t1 and t2 respectively. Next, landscape spatial indices were computed using FRAGSTATS (version 3) (McGarigal et al., 2002). The following indices were calculated: a) mean patch size (ha), b) patch density (number of patches per 100 hectares), c) mean proximity index (ratio between 1

ESRI 1996-2000. Environmental Systems Research Institute, Inc. 380 New York St., Redlands, CA92373-8100, USA.


the size and proximity of all patches whose edges are within 1 km-search radius of the focal patch), and d) total interior forest area (total patch size remaining after removing a specific buffer edge of 100m) (ha). Results Forest loss Approximately 23% of the native forests in 1976 had disappeared by 1999 (Fig. 1). During the whole study period, the annual forest loss was of 2,614 ha year-1, equivalent to 1.1%yr-1. Most of the forest loss was concentrated in the first nine years of the study period, at a deforestation rate of 1.6% yr-1, corresponding to 4,049 ha yr-1. In the second time interval, the rate decreased considerably to approximately 0.62% yr-1, equivalent to 1,341 ha yr-1. The proportion of the landscape represented by native forests decreased gradually across the time periods. In 1976, this land cover type covered 266.8 thousand hectares, equivalent to 53% of the total study area (Fig. 1). In 1985, native forests represented 45% of the total area (Fig. 3.2), and the large fragments had been divided into smaller patches (Fig. 1). Fourteen years later, the area of native forests declined to 206.7 thousand hectares, which corresponded to 41% of the total land area (Fig. 1). Approximately 16% of native forests (equivalent to 4,700 ha yr-1) in 1976 was transformed into shrublands and 8% (1,230 ha yr-1) were converted into crops and pasture and other land cover types in 1985. During the same time interval, 18% of shrublands was replaced by crops and pasture. Between 1985 and 1999, 18% (3,027 ha yr-1) of the native forests was converted to shrublands and 4% (equivalent to 660 ha yr-1) was cleared for crops and pasture. Of the total area of shrublands in 1985, 24% was converted to crops and pasture. Across the two study periods, 62% of the original forests remained as native forest and 29% was replaced by shrublands and 6% by crops and pasture. In both study periods, most of the area covered by crops and pasture was derived from shrublands. The rates of transition derived from the analysis of land cover change show the transition of a landscape dominated by native forests in 1976 (53%) to a landscape where non-forest land uses are becoming the dominant land cover type (60%). Trends in forest fragmentation Results showed an increasing proportion of the total area occupied by small patches over time. In 1976, 77% of the forest area was concentrated in patches of more than 10,000 ha. In 1985, patches larger than 10,000 ha had decreased to 47% of the total area, while the patches of less than 100 ha had increased to 20% of the total forest area (Fig. 2). In 1999, there was a considerable division of forest fragments which led to a 27% increase in the area occupied by patches less than 100 ha (Fig. 2). The mean size of forest patches decreased significantly from 47 ha in 1976 to 24 ha in 1985 (Table 1). This decline in the patch size during the first time interval was associated with an increase in the patch density for the same period (Table 1). Patch density showed an increasing trend overall through time, reaching its maximum value of 0.65 fragments per 100 ha in 1999 (Table 1). The main change in the mean proximity was observed from 1976 to 1985, when the value decreased to almost one-fifth of its initial value (Table 1). Between 1985 and 1999, the mean proximity also demonstrated a significant decline. During these periods, the neighborhood of forest patches rapidly became occupied by areas of a different land cover type, as native forest patches became further apart and less contiguous in distribution. Interior areas in 1976 significantly declined in 1985. However, interior area of


patches in 1985 did not present significant difference with those areas that were interior in 1999.


1976

1985

1999

CHILE ANTARTICA 90º 53º

Figure 1.Temporal variation of the major land cover types over 23 years in southern Chile.


Table 1. Temporal variation in landscape indices for the native forests in southern Chile. Kruskal-Wallis tests were applied to assess significant differences over time between indices estimated at the patch level Minimum and maximum values are given for mean patch size and mean proximity.

Landscape indices Mean patch size (ha) Patch density (n/100 ha) Mean proximity Total interior forest area (ha)

1976 47 (0.45 - 132,971.2) 0.36 19,350 (0.0 369,603.5) 143,428

χ 12

1985

1999

24 (0.45 - 49,767.4) 0.60 4,380 (0.0 - 152,583.1)

19 (0.45 - 42,785.3) 0.65 2,552 (0.0 - 120,135.5)

1976-1985 201.2 168.9

89,007

69,900

3.85

χ 12 1985-1999 0.12 18.5

*** *** *

0.45

n.s. *** n.s.


150,000

100,000

50,000

Total forest area (ha)

150,000

1

2

3

4

5

6

7

8

9

100,000

50,000

2,000-5,000

5,000-10,000

10,000-20,000

1,000-2,000

2,000-5,000

5,000-10,000

10,000-20,000

>100,000

1,000-2,000

500-1,000

20,000-100,000

500-1,000

100-500

0-100

150,000

100,000

50,000

-

>100,000

20,000-100,000

100-500

0-100

Fragment size (ha) Figure 2.Temporal variation of the forest fragment size for the years 1976, 1985 and 1999.


DISCUSSION AND CONCLUSIONS Deforestation and land use change In the last three decades, the study landscape has experienced forest loss at an annual rate higher than that estimated between 1995 and 1998 further north to the present study area, in which the annual forest loss reached 0.3% (CONAF et al., 1999). Deforestation in other temperate forests has proceeded at a similar rate. In western Oregon, almost 20% of the forests were cleared between 1972 and 1995, giving an annual deforestation rate of 0.5%-1.2% (Cohen et al., 2002). In the same region, between 1972 and 1988, deforestation by clearcutting reached 1.2% of the entire study area including wilderness areas (Spies et al., 1994). Changes in the native forest matrix The pattern of change reveals the transition of the landscape dominated by a large contiguous extent of native forest to a landscape characterized by smaller and more isolated forest fragments surrounded by arboreus shrublands, shrublands, crops and pasture lands. This pattern was associated with an increasing adjacency between arboreus shrublands and native forest as a result of logging in the borders of forest patches. In contrast to this, in central Chile, native forests were primarily surrounded by arboreus shrublands at the earliest stage of deforestation (Echeverría et al., 2006). With progressive deforestation, pine industrial plantations largely dominated the neighboring areas of native forest patches. The total area of crops and pasture land seemed to be relatively constant over the latter period, increasing by 9,000 ha in 14 years, while shrubland and other land cover types displayed much greater temporal variation. The results contrast with changes occurring in an agriculture landscape in southern Wisconsin, where the amount of natural vegetation (excluding forest) remained stable, and patches of forest appeared and disappeared at the average rate of about 4% per year (Forman and Godron, 1986). This stationary distribution (described as a shifting mosaic by Forman and Godron, 1986) was not found in the current study area. The landscape element that has the greatest area constitutes the matrix (Forman and Godron, 1986). Up to 1976, native forests predominated both in terms of relative area and percent of aggregation. However, the land-cover change analysis revealed a substantial modification of the landscape composition associated with a lower dominance of native forests in the landscape across time. This pattern has led to a transitional landscape characterised by a matrix of shrublands and crops and pasture in 1999. This change in the matrix may lead to changes in the functioning of the landscape (Forman and Godron, 1986; Bennet, 2003 ), as some ecological processes associated with native forests can be altered by the dominance of non-forest land cover types. This new less desirable matrix may reduce the movement of many species from fragment to fragment, isolating gene pools and reducing local genetics (Barnes, 2000). A matrix of open lands surrounding forest patches may alter species composition (Echeverria et al., in press), increase the susceptibility of windthrow of existing trees, and allow for the invasion of exotic species (Barnes, 2000). In Central Chile, the rapid expansion of exotic-species plantations in the last decades has led to a dominance of this land-cover type in the landscape. This matrix of plantations has caused an isolation of the remnant forest patches which has affected the regeneration of shade-tolerant plants depending on biotic pollinators and dispersers (Bustamante and Castor, 1998; Grez et al., 1998).


Spatial patterns of forest fragmentation The reduction and division of the forest habitat is one of the recognizable results of the process of fragmentation (Ingegnoli, 2002). Owing to this, the size-frequency distribution of remnants in fragmented landscapes is strongly skewed towards small blocks (Bennett, 2003). With progressive forest loss and fragmentation, large forest areas were divided into smaller patches, which led to a decline in the mean patch size and a change in the patch size distribution by 1999. A similar pattern was observed in the highlands of Chiapas, Mexico between 1976 and 1996, in which the mean patch size and the number of fragments greater than 10,000 ha of dense forest decreased over time while small fragments increased (Ochoa-Gaona, 2001). In Bolivia, the fragmentation of the tropical deciduous forest showed a similar trend in which the smallest classes of patch size increased from 1977 to 1998 (Steininger et al., 2001). The progressive increase in these types of patches may have some negative consequences on the occurrence of certain species. For instance, in the Atlantic forest of Brazil, the smallest fragments (less than 50 ha) were associated with a diminishing mammal community as these fragments were below the carrying capacity of these species (Pardini et al., 2005). The increase in the number of forest fragments also demonstrates that native forests of the study area were continuously affected by fragmentation over time. A similar situation was observed between 1975 and 1990 in central Chile, where the number of patches increased at higher densities from 0.93 to 1.65 (Echeverría et al., 2006) compared to 0.36 to 0.6 from 1976 to 1985 in southern Chile. In Limón, Costa Rica, the patch density reached to 0.4 in 1997 (Van Laake and Sanchez-Azofeifa, 2004), similar to the density recorded in 1976 by the present study. The decline in the amount of large forest fragments with interior areas reveals the loss of large fragments with high quality habitat due to fragmentation. A very similar trend was described in Costa Rica in 1997, using the same distance from the edge, where large-scale deforestation for agricultural expansion has reduced the core area representing 34.8% of the total forest (Van Laake and Sanchez-Azofeifa, 2004). This decrease in area of large forest patches in the study area may jeopardise the survival of some species dependent on this type of habitat. For instance, some studies conducted in Chile have demonstrated that the reduction of large forest fragments with core area may have some influence the population of rodents (Donoso et al., 2003) and birds of temperate forest (Wilson et al., 1994; Cornelius et al., 2000, Vergara and Simonetti, 2004). This study demonstrates that the progressive fragmentation and forest loss are associated with dramatic changes in the spatial structure of the temperate forest landscape in southern Chile.

References Bennett, A. (2003) Linkages in the landscape. The role of corridors and connectivity in wildlife conservation. IUCN, Gland, Switzerland and Cambridge, UK. 254 pp. Bustamante, R.; Castor, C. (1998) The decline of an endangered ecosystem: The Ruil (Nothofagus alessandrii) forest in Central Chile. Biodiversity and Conservation 7, 1607-1626. Chuvieco, E. (1996) Fundamentos de teledetección espacial. Ediciones RIALP, S. A. 3rd Edition. Madrid.


CONAF; CONAMA; BIRF; Universidad Austral de Chile; Pontificia Universidad Católica de Chile; Universidad Católica de Temuco (1999) Catastro y Evaluación de los Recursos Vegetacionales Nativos de Chile. Monitoreo de Cambios. Santiago, Chile. Cohen, W.; Spies, T.; Alig, R.; Oetter, D.; Maiersperger, T.; Fiorella, M. (2002) Characterizing 23 years (1972-95) of stand replacement disturbance in western Oregon forest with Landsat imagery. Ecosystems 5, 122-137. Cornelius, C.; Cofre, H. ; Marquet, P. (2000) Effects of Habitat Fragmentation on Bird Species in a Relict Temperate Forest in Semiarid Chile. Conservation Biology 14, 534-543. Debinski, D.; Holt, R. (2000) A survey and overview of habitat fragmentation experiments. Conservation Biology 14, 342-355. Di Castri, F.; Hajek, E. (1976) Bioclimatología de Chile. Ediciones Universidad Católica de Chile, Santiago. Donoso, D.; Grez, A.; Simonetti, J. (2003) Effects of Forest fragmentation on the granivory of differently sized seeds. Biological Conservation 115, 63-70. Echeverría, C.; Coomes, D.; Newton, A.; Salas, J.; Rey-Benayas, J. M.; Lara, A (2006) Rapid fragmentation and deforestation of Chilean Temperate Forests. Biological Conservation 130, 481-494. ESRI (1999) Environmental Systems Research Institute, Inc., Redlands, California, USA. FAO (1995) Forest resources assessment 1990. Global Synthesis. FAO, Rome. Forman, R.T.T.; Godron, M. (1986) Landscape Ecology. John Wiley & Sons, USA. 619 p. McGarigal, K.; Cushman, S. A.; Neel, M. C.; Ene, E. (2002) Fragstats: Spatial Pattern Analysis Program for Categorical Maps. Retrieved January 20, 2003, from University of Massachusetts, Landscape Ecology Program Web site: www.umass.edu/landeco/research/fragstats/fragstats.html. Ochoa-Gaona, S. (2001) Traditional land-cover systems and patterns of forest fragmentation in the highlands of Chiapas, Mexico. Environ. Manage. 27, 571-586. Pardini, R.; Marquez de Souza, S.; Braga-Neto, R.; Metzger, J.P. (2005) The role of Forest structure, fragments size and corridors in maintaining small mammal abundance and diversity in an Atlantic forest landscape. Biological Conservation 124, 253-266. Rey-Benayas, J. M.; Pope, K. (1995) Landscape ecology and diversity patterns in the seasonal tropics from Landsat TM imagery. Ecological Applications 5, 386-394. Spies, T.; Ripple, W.; Bradshaw, G. (1994) Dynamics and pattern of a managed coniferous forest landscape in Oregon. Ecological Applications 4, 555-568. Staus, N.; Strittholt, J.; Dellasala, D.; Robinson, R. (2002) Rate and patterns of forest disturbance in the Klamath-Siskiyou ecoregion, USA between 1972 and 1992. Landscape Ecology 17, 455-470. Steininger, M.; Tucker, C.; Ersts, P.; Killeen, T.; Villegas, Z.; Hecht, S. (2001) Clearance and fragmentation of tropical deciduous forest in the Tierras Bajas, Santa Cruz, Bolivia. Conservation Biology 15, 856-866. Van Laake, P.; Sánchez-Azofeifa, G.A. (2004) Focus on deforestation: zooming in on hot spots in highly fragmented ecosystems in Costa Rica. Agriculture, Ecosystems and Environment 102, 3–15.


Vergara, P.; Simonetti, J. (2004) Avian responses to fragmentation of the Maulino in central Chile. Oryx 38, 383-388. Willson, M.; De Santo, T. I.; Sabag, C.; Armesto, J. J. (1994) Avian communities of fragmented south-temperate rainforests in Chile. Conservation Biology 8, 508-520.


Forest floristic inventory of Mixed Ombrophilous Forest and Deciduous Forest of Santa Catarina State, Southern Brazil: preliminary results. A. C. Vibrans1; A. Uhlmann1; L. Sevegnani1; M. Marcolin1; N. Nakajima1, C. R. Grippa1 , E. Brogni2 & M. Braga Godoy1 1

Universidade Regional de Blumenau, Blumenau-SC, Brazil. Email: [email protected] 2 Fundação Municipal do Meio Ambiente de Itajaí-SC, Brazil.

Abstract Forest floristic inventory of Santa Catarina State in Soutjern Brazil aimes quantification of forest resources and increase of knowledge on vascular plant diversity. The study includes sampling of trees, shrubs, herbs and vascular epiphytes. Santa Catarina State’s surface is about 95,443 km², covered by four major vegetation types: Dense Tropical Humid Forests, Mixed Ombrophilous Forests with Araucaria angustifolia, highland grasslands and Seasonal Deciduous Forests. Forest land cover is about 30%, although extremely fragmentized, specially in the western part. This pilot study has been realized in Mixed Subtropical Forest and Seasonal Deciduous Forest during 2005, measuring 74 sample plots (2000 m² each) randomly distributed in the forest type’s area (41,380 km²). Myrtaceae, Fabaceae, Lauraceae are the most numerous families, followed by Asteraceae and Solanaceae. There are a restricted number of very frequent tree species and a large number (120) of species occurring in less than 10 of the total of 74 sample plots; 46 species are present in only one plot, showing vulnerability of species diversity. Multivariate analysis starting from a species density matrix was executed in order to aggregate similar sample plots from the floristic point of view and identify species groups which preferentially occur together. The analysis shows two well distinguished plot clusters, one of them dominated by typical mixed forest species; the other cluster clearly dominated by supposed to be typical seasonal deciduous forest represents, although located in geographical region until now attributed to mixed forests. Other studies are needed to confirm this floristic tendency which could refine existent vegetation maps.

Introduction The drastic reduction of the Forest cover of the state of Santa Catarina during the second half of the 20th century and the constant pressure exerted by the economic activities on the forest remains lead to the Secretary of Agriculture and Rural policies to formulate in 2003, the Project of the Forest floristic inventory of Santa Catarina State, according to the Resolutions of the Brazilian Environmental National Council, Brazil- CONAMA 278/2001 and 309/2002. The forest floristic inventory has as its main objective to catalogue and analyze the


floristic composition and the horizontal and vertical structure of the forest remains of Santa Catarina state. The main purpose of the project is the need to record arboreal diversity of the forests and to collect data on the plant endangered species. Mixed ombrophylous forest (IBGE, 1990) or araucária forest (KLEIN, 1978) includes the forest types on the plateaus in the western regions of Rio Grande do Sul, Santa Catarina, Paraná, São Paulo, Rio de Janeiro and Minas Gerais states and also the northeast of Argentina and southeast of Paraguay, originally covering 400 thousand km2, today remaining less than 5%. The original covering of Santa Catarina was about 2/3 of the whole area of the state, mainly over high altitudes (more than 500m a.s.l.), on the Serra Geral plateau up to the border of Argentina. Physiognomically this type of forest had the emergent stratum composed by Araucaria angustifolia (Bertol.) Kuntze (reaching up to 45m tall). The canopy is located about 10 m below and it is mainly composed by large broadleafed species which composition varies according to the region but includes the Lauraceae (Ocotea porosa (Nees) Barroso, O. odorifera (Vell.) Rohwer, O. puberula (Rich.) Nees, Persea major Kopp, Cryptocarya aschersoniana Mez, Nectandra lanceolata Nees), Myrtaceae (Eugenia involucrata DC, E. pyriformis Cambess., Campomanesia xanthocarpa O. Berg), Meliaceae (Cedrela fissilis Vell.), Dicksoniaceae (Dicksonia sellowiana Hook), Winteraceae (Drymis brasiliensis Miers), and others (GAPLAN, 1986). Lumber exploitation, agriculture, cattle-raising, reforestation with exotic species (Pinus spp.) as well as the expansion of urban areas cities, are factors, which had caused in the past and in the present drastic reduction and fragmentation of this peculiar forest type. Seasonal deciduous forest – in the western part of Santa Catarina State - Brazil covered the side hills and affluents of the Uruguay River penetrating towards the north (GAPLAN, 1986). Typical species of the seasonal forest co-occurred with Araucaria angustifolia and all the other species of mixed forest, leading to Klein’s conclusion that the major portion of the western region of Santa Catarina constitutes a transition area between two phytogeographical regions (Klein 1979). This kind of forest shows remarkable characteristic leaves-fall above 50%, among the species, which compose the canopy. The low temperatures mark the climatic seasonality with an annual average below 15º C in the winter and above 20º C in the summer. The leaves-fall of the species of this forest is determined by a canopy dominated by deciduous leguminous depicting Apuleia leiocarpa McBride and Parapiptadenia rigida (Benth.) Brenan. also recording Patagonula americana Kuntze, Holocalyx balansae Micheli, Syagrus romanzoffiana (Cham.) Glassman, Cedrela fissilis Vell., Tabebuia spp., Cordia trichotoma (Vell.) Steud., Diatenopterix sorbifolia Radlk, Luehea divaricata Mart., Ocotea puberula Ness, Nectandra megapotamica Mez, among dozen others (KLEIN, 1978). The presence of many lumber species of great economic value and the clearing of the areas for the agricultural and cattle-raising purposes has stimulated the raising of cycles of economic exploitation which has lead to the deforestation accounting for only 3% of the remaining forests that nowadays are extremely fragmented. Objective The main task of this work phase is the inventory of forest remnants of mixed ombrophilous forest, as well as the forests located in the area of natural grasslands and the


seasonal deciduous forests, according to the phytogeographical map of Santa Catarina, Brazil (Klein, 1978). Methods Sixty temporary sample units were implemented and measured in order to survey the actual state of the vegetation. Fourteen permanent sample units were installed aiming to follow the forest dynamics. The sample units were randomly distributed in the mixed forest remnants identified through visual interpretation of Landsat 2003 images. The sample units distribution took place by a raffling system with 15.023 fragments bigger than 10 ha identified in the image mentioned above (Figure 1). Within each raffled fragment the sample units were allocated randomly. Each sample unit includes the following component: adult trees, thin trees, natural regeneration, herbs, epiphytes and shrubs. Each unit was subdivided in order to make a more appropriate sample as described in Figure 2: K – Herbs, shrubs, and natural regeneration with DBH < 1 cm; surveyed by using a 2 x 2 sample unit (4 m2); L – Natural regeneration with 1 cm ≤ DBH < 10 cm; surveyed by using a 5 x 5m sample unit (25 m2); M – Thin trees with DBH ≥ 10 cm – subunits S1a of 10 x 25 m (250 m2); in this subunit epiphytes were inventoried by means of raffling 02 hostage trees; N – Adult trees DBH ≥ 20 cm – subunits S1b of 10 x 25 m (250 m2); O – Adult trees DBH ≥ 30 cm – subunits S2 of 10 m x 50 m (500 m2); P – Adult trees DBH ≥ 40 cm - subunits S3 of 10 m x 50 m (500 m2); Q – Adult trees DBH ≥ 50 cm - subunits S4 of 10 m x 50 m (500 m2).

In the permanent sample units, were measured all arboreal species above 10 cm DBH in all subunits (S1, S2, S3, S4).


Mixed ombrophilous forest

Seasonal deciduous forest

Permanent sample units (14) Temporary sample units (60)

Grassland

Figure 1: Phytogeographical map of Santa Catarina state (Klein, 1978) with 60 temporary sample units and 14 permanent sample units within the mixed ombrophilous forest and (originally) grassland areas.

The collecting data of each individual measured includes: perimeter at breast high (PBH), commercial high, total high and stem quality. All the individuals that were imprecisely identified in the field were collected and taken to the laboratory to be corrected identified. The identified specimens were included in the collection of herbarium Roberto Miguel Klein of Universidade Regional de Blumenau and Herbário Barbosa Rodrigues (HBR). S2

S1

S3

S4

S1a S1b 1