<%BANNER%>

The Importance of Private Lands for Conservation

Permanent Link: http://ufdc.ufl.edu/UFE0025132/00001

Material Information

Title: The Importance of Private Lands for Conservation Land Cover Change Analysis in The Caldenal Region, Argentina
Physical Description: 1 online resource (55 p.)
Language: english
Creator: Gonzalez Roglich, Mariano
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: Interdisciplinary Ecology -- Dissertations, Academic -- UF
Genre: Interdisciplinary Ecology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Protected areas are critical for the conservation of the world s biodiversity; however, parks isolated from their surroundings will not assure the maintenance of biodiversity and working ecosystems. New management practices assuring the conservation of biodiversity and the provision of goods and services need to be designed and promoted. Private lands have the potential to fulfill this dual role, but first we need to know which land use practices in private land really contribute to the maintenance of native biodiversity. In this study, a land use and land cover change analysis, between 1987, 1999 and 2008, was developed for the Caldenal region, Argentina to determine the effect of private game reserves on landscape scale processes of change. Game reserves were found preferentially located in the areas with highest proportion of forest cover. No differences were found in the rates of conversion to agriculture between game reserves and cattle ranches for any of the two periods analyzed. Differences were found in the rates of deforestation between the first and second period. While deforestation stopped in the game reserves and big cattle ranches, it increased in intensity in the small cattle ranches. This suggests that, for the period analyzed, the incorporation of fee hunting as a land use has not significantly changed management practices in the area, being deforestation more associated with the agricultural capabilities of the of the ranches, than to the land use type. Furthermore, the dependence of hunting activities on exotic species makes game reserves active agents of introduction of non-native species in this region. Further research on the role of this land use type on the introduction of species, and on the effect of these new species on the rest of the Caldenal ecosystem will allow for a more complete evaluation of the value of game reserves as conservation tools in central Argentina.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Mariano Gonzalez Roglich.
Thesis: Thesis (M.S.)--University of Florida, 2009.
Local: Adviser: Southworth, Jane.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0025132:00001

Permanent Link: http://ufdc.ufl.edu/UFE0025132/00001

Material Information

Title: The Importance of Private Lands for Conservation Land Cover Change Analysis in The Caldenal Region, Argentina
Physical Description: 1 online resource (55 p.)
Language: english
Creator: Gonzalez Roglich, Mariano
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: Interdisciplinary Ecology -- Dissertations, Academic -- UF
Genre: Interdisciplinary Ecology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Protected areas are critical for the conservation of the world s biodiversity; however, parks isolated from their surroundings will not assure the maintenance of biodiversity and working ecosystems. New management practices assuring the conservation of biodiversity and the provision of goods and services need to be designed and promoted. Private lands have the potential to fulfill this dual role, but first we need to know which land use practices in private land really contribute to the maintenance of native biodiversity. In this study, a land use and land cover change analysis, between 1987, 1999 and 2008, was developed for the Caldenal region, Argentina to determine the effect of private game reserves on landscape scale processes of change. Game reserves were found preferentially located in the areas with highest proportion of forest cover. No differences were found in the rates of conversion to agriculture between game reserves and cattle ranches for any of the two periods analyzed. Differences were found in the rates of deforestation between the first and second period. While deforestation stopped in the game reserves and big cattle ranches, it increased in intensity in the small cattle ranches. This suggests that, for the period analyzed, the incorporation of fee hunting as a land use has not significantly changed management practices in the area, being deforestation more associated with the agricultural capabilities of the of the ranches, than to the land use type. Furthermore, the dependence of hunting activities on exotic species makes game reserves active agents of introduction of non-native species in this region. Further research on the role of this land use type on the introduction of species, and on the effect of these new species on the rest of the Caldenal ecosystem will allow for a more complete evaluation of the value of game reserves as conservation tools in central Argentina.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Mariano Gonzalez Roglich.
Thesis: Thesis (M.S.)--University of Florida, 2009.
Local: Adviser: Southworth, Jane.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0025132:00001


This item has the following downloads:


Full Text

PAGE 1

THE IMPORTANCE OF PRIVAT E LAND S FOR CONSERVATION: LAND COVER CHANGE ANALYSIS IN THE CALDENAL REGION, ARGENTINA By MARIANO GONZALEZ ROGLICH A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2009 1

PAGE 2

2009 Mariano Gonzalez Roglich 2

PAGE 3

To m y family and friends 3

PAGE 4

ACKNOWL EDGMENTS This research was supported by the Fulbrigh t Program, Research Fellowship Program (CLP-WCS), Tropical Conservation and Deve lopment Program (UF), La Pampa Natural Resources Agency, La Pampa Ecology Agency and La Pampa Emergency Management Agency. I am deeply grateful to my advisor Jane S outhworth and committee members Lyn Branch and Brian Child for their support, encouragement and ideas while developing this project. The following individuals played a sign ificant role on this study, and w ithout their help, this research would not have been possible: Maite Betelu, Mari sa Urioste, Fabian Titarelli, Gustavo Romero, Diego Villarreal, Eugenia Estanga Mollica, Moni ca Castro, Marcelo Casanovas y Yanina Rubio. Field assistants Marina Cock a nd Juan Lagos deserve special recognition for their hard work and dedication. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4LIST OF TABLES................................................................................................................. ..........6LIST OF FIGURES.........................................................................................................................7 CHAPTER 1 INTRODUCTION................................................................................................................. .102 MATERIALS AND METHODS...........................................................................................17Study Area..............................................................................................................................17Remote Sensing......................................................................................................................18Image Pre-Processing......................................................................................................18Land Cover Classes.........................................................................................................19Image Classification........................................................................................................20Trajectory Analysis.........................................................................................................21Statistical Analysis..........................................................................................................2 33 RESULTS...................................................................................................................... .........26Land Cover Patterns...............................................................................................................26Preferential Location.......................................................................................................... ....26Conversion to Agriculture...................................................................................................... 27Rates of Change......................................................................................................................28Trajectory Analysis............................................................................................................ .....284 DISCUSSION................................................................................................................... ......31Processes of Land Cover Change...........................................................................................31Game Reserves and Conservation..........................................................................................34REFERENCES..............................................................................................................................46BIOGRAPHICAL SKETCH.........................................................................................................55 5

PAGE 6

LIST OF TABLES Table page 2-1 Satellite images used in for the analysis............................................................................43 2-2 Confusion matrix for the 2008 land-cover classification...................................................43 2-3 Trajectory analysis scheme...............................................................................................4 4 2-4 Statistical summary of the different zones analyzed..........................................................44 3-1 Annual rate of conversion to agriculture...........................................................................45 3-2 Annual rate of change of natural land cover classes..........................................................45 6

PAGE 7

LIST OF FI GURES Figure page 2-1 Relative location of the study area in La Pampa Province, Argentina..............................37 2-2 Natural gradients.......................................................................................................... ......37 2-3 Spatial distribution of the parcels analyzed.......................................................................38 3-1 Land cover maps of the central Caldenal region of 1987, 1999 and 2008........................38 3-2 Mean land cover for 1987, 1999 and 2008........................................................................39 3-3 Mean land-cover change traj ectories between 1987, 1999 and 2008................................40 3-4 Spatial distribution of natu ral changes between 1987 and 2008........................................41 3-5 Spatial distribution of agricu ltural changes between 1987 and 2008................................42 7

PAGE 8

Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science THE IMPORTANCE OF PRIVATE LAND S FOR CONSERVATION: LAND COVER CHANGE ANALYSIS IN THE CALDENAL REGION, ARGENTINA By Mariano Gonzalez Roglich December 2009 Chair: Jane Southworth Major: Interdisciplinary Ecology Protected areas are critical for the conservati on of the worlds biodiversity; however, parks isolated from their surroundings will not assure the maintenance of biodiversity and working ecosystems. New management practices assuri ng the conservation of biodiversity and the provision of goods and services need to be de signed and promoted. Pr ivate lands have the potential to fulfill this dual role, but first we need to know which land use practices in private land really contribute to the ma intenance of native biodiversity. In this study, a land use and land cover change analysis, between 1987, 1999 and 2008, was developed for the Caldenal region, Argentina to determine the effect of private game reserves on landscape scale processes of change. Game reserves were found preferentially located in the areas with highest proportion of forest cover. No differences were found in the rates of conversion to agriculture between game reserves and cattle ranches for any of the two periods analyzed. Differences were found in the rates of deforestation between the first and second period. While deforestation stopped in the game reserves and big cattle ranches, it increase d in intensity in the small cattle ranches. This suggests that, for the period anal yzed, the incorporation of f ee hunting as a land use has not significantly changed management practices in th e area, being deforestation more associated with the agricultural capabilities of the of the ra nches, than to the land use type. Furthermore, the 8

PAGE 9

dependence of hunting activities on exotic species m akes game reserves active agents of introduction of non-native species in this region. Further research on the role of this land use type on the introduction of species, and on the effect of these new species on the rest of the Caldenal ecosystem will allow for a more complete evaluation of the value of game reserves as conservation tools in central Argentina. 9

PAGE 10

CHAP TER 1 INTRODUCTION In the last few centuries humans have ge nerated an unprecedented situation of global environmental change, via population grow th, technology and economic development (Houghton, 1994; MEA, 2005). Loss of natural ecosys tems and species, increase in agricultural areas, soil degradation, emission of greenhouse gases and urbanization are some of the manifestations of this change that may risk the long term global sustainability (Fearnside, 2005; IPCC, 2007). This is why monitoring and unders tanding the magnitude and consequences of human impacts on the environment ha s been identified as a global research priority (Ojima et al., 1994; Lambin & Geist, 1999; Turner et al., 2007). Land cover and land use change is one of the main drivers of this process of global change (Foley et al., 2005), ther efore learning about the processes of land use change, its consequences on land-cover, and its final interactions with the social and environmental systems is critical for the implementation of adequate management measures. In order to preserve threatened ecosystems from degradation or replacement, various conservation approaches have been applied (B awa & Seidler, 1998; Fra nk et al., 2003; AzevedoRamos et al., 2006; Tole, 2006; Brockington, 2007) but among them protected areas have been the cornerstone of conservation strategies so far (Margules & Pressey, 2000). As defined by the IUCN (1994) there is a wide ra nge of protected area definiti ons, based on their objectives, management and relationship with human populati ons. However, a big debate is occurring among conservation practitioners, donors and NGOs about which conservation strategies are more ecological, social, and econom ically sustainable over time. Th e options are antagonistically presented as exclusionary conservation versus the community based appr oaches to conservation (Chapin et al., 2000; Wilshusen et al., 2002; Terborgh, 2004; Andrade, 2005; Hutton et al., 10

PAGE 11

2005). The exclusionary conservation approach is based on the assum ption of incompatibility of conserving functional ecosystems with people liv ing in them. The community based approaches are based on the opposite assumption, that people will better conserve natu ral resources if they can make sustainable utilization of them. The fact that by the year 2005 there were over 100,000 protected areas worldwide, covering over 12% of the lands surface, highlig hts the importance assigned to parks as conservation tools (Chape et al., 2005). Its effectiveness in maintaining forest cover inside their boundaries has been demonstrated in a variety of regions (Bruner et al., 2001; Sanchez-Azofeifa et al., 2003; Southworth et al., 2004b) though, degradation beyond those boundaries often compromises the ecological objectives of the prot ected areas (Soul et al ., 1999; Clerici et al., 2007), especially if they are too small or too isol ated (Sanchez-Azofeifa et al., 2002). Even when the extent of the protected area system has increa sed significantly in the last century, it is not enough to ensure the conservation of global biodiv ersity (Chape et al., 2005). New strategies need to be implemented to comple ment the role of protected areas. Biodiversity conservation can also be ach ieved under private management based on ecotourism, hunting or just becau se of the value assigned to na ture by the owners of the land (Langholz & Lassoie, 2001; Frank et al., 2003; Langholz & Kerley, 2006). Different mechanisms are being developed to favor conservation prac tices in private lands, such as conservation easements (Engel, 2007; Rissman et al., 2007; Rissman & Merenlender, 2008), payment for ecosystem services (Engel et al ., 2008; Klimek et al., 2008; Tall is et al., 2008), and taxes and subsidies modifying private behavior (Jenkins et al., 2004). The potential significance of private conservation initiatives is especia lly important in countries, like Ar gentina, in which most of the 11

PAGE 12

land is privately owned. However, in order to co rrectly im plement any of these strategies basic information on the real conservation valu e of different land uses is needed. The importance of wildlife utilization as a s ource of income in private landholdings has increased in recent years, as an alternative to traditional rural production systems (e.g. cattle ranching), providing incentives to wildlife a nd habitat conservation (Williams & Lathbury, 1996; Butler et al., 2005). Wildlife, as a land use, has be en proven effective in southern Africa from an economic and biodiversity conservation perspective (Bond et al., 2004). Trophy hunting is of major importance to conservation in Africa creating economic incentives for conservation over vast areas unsuitable for other activities (Lindsey et al., 2006b). Fee hunting is promoting habitat conservation in California, where landowners agr ee to develop habitat enhancement activities in order to obtain hunting benefits (Williams & Lathbury, 1996). However, no landscape scale land cover change assessment has been performed in the Caldenal savanna of central Argentina to determine the conservation value of private ga me reserves to maintain native ecosystems. Lack of agreement in the conservation community is not uncommon. Diff erent criteria have been proposed for deciding conservation pr iorities (Mittermeier et al., 1998; Kareiva & Marvier, 2003; Mittermeier et al ., 2003; Rodrigues et al., 2004; Ho ekstra et al., 2005; Brooks et al., 2006), and as already mentioned, there are di fferences in how to effectively assure the achievement of the conservation objectives. It is not surprising the n, that there is a the lack of standardized and comparable methods for asse ssing the conservation outcome of different management practices over time (Green et al., 2005b; Gaston et al., 2008). Three main groups of methodologies can be found in the literature analyz ing the effectiveness of conservation tools. The first one is based on comparison of demographi c data of target species in areas subject to conservation management over time (Stoner et al., 2007) or comparing data from similar areas 12

PAGE 13

protected and not protected (Car o, 2001; W eller, 2008). These types of studies would ideally be preferred because they directly measure the variab le of conservation interest, however the lack of reference data and the need of high cost data collection campaigns limit its implementation in most cases. The second group is based on survey data. Asking managers, local people and experts about the conservation outcomes of particul ar practices is used to assess the performance of different management practices (Bruner et al., 2001; Langholz & Lassoie, 2001), though the validity of the results have been questioned because of the partic ular motives of the respondents might bias the results in any particular direct ion (Vanclay, 2001). The third group of assessment procedures for determining the effectiveness of conservation practices is the use of surrogates (Caro & O'Doherty, 1999), and among them the ma intenance of habitat as a surrogate for biodiversity conservation (Lombard et al., 2003 ; Altmoos & Henle, 2007; Lindsay et al., 2008). The main assumption is that the maintenance of native habitats is directly related to the maintenance of the other components of local biod iversity, usually much more difficult or costly to assess. Limitations to this approach are importa nt, given that the existe nce of forest cover does not assure the conservation of local biodive rsity (Redford, 1992; Terborgh et al., 2008). However, these limitations do not invalidate the utility of habitat assessments as a first analysis of effectiveness of conservation practices, given the inevitable necessity of maintaining native habitats for any in situ conservation program. Remote sensing is a particularly effective tool for the evaluation of conservation strategies at the landscape level (Turner et al., 2003) becau se it offers important means for detecting and analyzing changes in landscapes over large temporal and spatial scales (Narumalani et al., 2004). Information on the distribution of species habi tats, and habitat changes over time, can be generated from satellite images over greater spatia l extents and more frequent time steps than 13

PAGE 14

possible with trad itional field studies (Nage ndra, 2001). The use of remote sensing for determining the effectiveness of protected areas is a commonly used procedure. Clerici et al (2007) used a trajectory analysis to determine the effectiveness of two biosphere reserves in West Africa maintaining savanna habitats within th eir boundaries, finding that even when the parks had been effective in maintaining forest cover, conversion in its periphery compromised the long term maintenance of species richness. Similar results were found for protected areas in other regions of the world like Eastern Africa (V ogt et al., 2006), Costa Rica (Sanchez-Azofeifa et al., 1999; Sanchez-Azofeifa et al., 2003), Honduras (Southwor th et al., 2004b) and Paraguay (Huang et al., 2007). Global remote sensing datasets have also been used to determine the performance of parks at broader scales. Joppa et al. (2008) used one of these global datasets to compare actual land cover in protected areas of the Amazon, Congo, Sout h American Atlantic Coast and West Africa. They identified two groups, one formed by regions with larger parks maintaining natural habitats in and outside its boundaries (Amazon and Congo) and those in which protected areas were smaller and with sharp boundaries due to conversion in its surroundings (South American Atlantic Coast and West Africa). However, the main tenance of forest cover has not been the only characteristic of protected areas analyzed thr ough remote sensing techniques. Wright et al. (2007) determined the effectiveness of protected areas globally in reducing fire frequency within their boundaries as compared to unprotected areas Fires were more fre quent outside protected areas by up to four orders of magnitude. These are just a some exampl es of the acceptance of land cover change analysis, through the use of remotely sensed data, as valid procedures to study the effect of different land use pr actices on land cover changes over time. 14

PAGE 15

The Caldena l region in central Argentina is a transitional savanna ecosystem at the margin of the Pampas grasslands, and is at the present time a reservoir of some of the best perennial grasslands in the country. However, currently on ly 0.1% of the region is under formal protection and the remainder is undergoing rapid conversio n for agriculture (Mendez, 2007b). With more than 40% of its area converted to agriculture in th e last century and a very scarce protected areas system, this region has been id entified as one of the endangere d crisis ecoregions worldwide (Hoekstra et al., 2005). In additi on to habitat conversion, much of the remaining Caldenal is highly degraded, where the open savanna character ized by scattered trees and dense cover of perennial grasses (Koutche & Carmelich, 1936) ha s been replaced by thick thorn scrub (Lell, 2004). As a result this system is one of the hi ghest conservation priori ties for the national government of Argentina (APN, 2007). In La Pampa province there ar e two publically owned protecte d areas in the Caldenal, but of those, only one preserves forest and grasslands ecosystems of relative magnitude, Parque Luro (76 km2). The identification of the role of diffe rent land uses in private landholdings in maintaining natural land covers is critical for the long term sust ainability of this threatened habitat and its biodiversity, particularly in a province in which 95% of the land is privately owned. Previous research has id entified private game reserves as land uses with potential for conservation of native resources in the Caldenal region (Gobb i, 1994).The objective of this research was to identify, using remote sensing data from 1987, 1999 and 2008, the importance of private game reserves for the conservation of native habitats in the Caldenal savanna of central Argentina. The processes of land cover change were analyzed at a landscape scale, comparing the land cover composition, the rate s of conversion to agriculture and the rates of change over time among game reserves, cattle ranches a nd the only protected area of the region. 15

PAGE 16

Understanding the ef fect of different private land us es in the Caldenal region will be critical for designing a regional conservation strategy complementing strict conservation in governmentally owned protected areas with working ecosystems in private lands. 16

PAGE 17

CHAP TER 2 MATERIALS AND METHODS Study Area The Caldenal is a semiarid savanna ecosystem of about 170,000 km2 located in central Argentina, primarily in La Pampa province. This xerophytic open forest system compromises the southern half of the Espinal ecoregion (Cabre ra, 1994; Olson et al., 2001) and is a transitional ecosystem between the Pampas grasslands, to the east, and the dry Monte ecoregions, to the west (Figure 2-1). It is dominated by Caldn ( Prosopis caldenia ) with understory of perennial grasses frequently interrupted by dunes, wetlands and lagoons (Cabrera, 1994). The climate is semiarid temperate with a mean annual temperature of 15C (max = 24C and min = 8C) and high diurnal and seasonal ranges due to low atmos pheric humidity. Major changes occurred in the area with the arrival of new se ttlers at the end of the 19th century and have intensified since then. Changes in fire regimes (Medina, 2007), introduction of species (e.g. red deer, Cervus elaphus ; wild boar, Sus scrofa ), extractive logging, overgrazing by li vestock, and replacement of natural systems for agriculture are the biggest threats the region has faced historically (Amieva, 1993; Mendez, 2007b). The area analyzed in this project (31,000 km2) is located in the central portion of the Caldenal region in La Pampa province. It was se lected based on the clumped spatial distribution of the game reserves and the pres ence of the only park in a Caldenal terrestrial environment in the province (Figure 2-1). This area is char acterized by steep northeast-southwest gradients (Figure 2-2). Rainfall decrease from 900 mm a year to 450 in only 200 km, being mainly concentrated in the summer. Soil types also follow the same gradient, finding the most fertile soils to the northeast of the province (Cano et al., 1980). The interaction of these varied environmental conditions highly influences the productivity of rural enterprises. Agrarian 17

PAGE 18

econom ic unit is the term used in local legislation to refer to the minimum amount of land a family needs to make a dignified living, ther efore the smallest possible size an individual property can legally be divided in (Pampa, 1973). The size of these units follow the same pattern described by natural gradients, finding the smallest pr operties to the northeast of the study area (3 km2) and the biggest to the southwest (25 km2, Figure 2-2). The economy of the region is almost exclusiv ely supported by the rural production. Cattle ranching is the main economic activity of the region followed by agricultu re (Cano et al., 1980; Mendez, 2007a). However, hunting and touris m are growing and complementary activities, focused on the attraction of international tour ists (Gobbi, 1994). Fee hunting is organized in private rural enterprises called game reserves ( cotos de caza in Spanish). Game reserves, as defined by local regulations, are pieces of land de dicated to hunting activ ities while assuring the sustainable utilization of wild life (SAA, 2004). Given the reduced diversity of big game native species in the region, game reserves have intr oduced a variety of exotic species. Red deer ( Cervus elaphus ), wild boar ( Sus scrofa ), antelope ( Antilope cervicapra ) and buffalo ( Bubalus bubalis ) are some of the species hunting enterprise s have actively introduced and manage to increase their game offer. Game reserves star ted formally in La Pampa in 1987, and since then 149 such reserves have been registered with the local government. Sixty one are still working, and of those, 56 are located in the Caldenal region covering over 3,222 km2 (4.4% of the ecoregion in the province). Remote Sensing Image Pre-Processing Landsat 5 TM images were obtained for th e end of the summer season for 1987, 1999 and 2008, three scenes were needed for each date to cover the study area (Tab le 2-1). Images were obtained from two sources: the Global Land Cover Facility (1987) and the Ecology Agency of 18

PAGE 19

La Pam pa (1999 and 2008). Images of 1987 were resampled to 30 m to match the spatial resolution of the images from the other source (the original spatial resolution was 28.5 m). All images were geometrically registered to a GIS layer of roads collected with a GPS during fieldwork (error < 0.5 pixel). Roads were used to rectify the images given the lack of high resolution cartography for the area and the abundance of road data collected during the training sample collection. Radiometric calibration was app lied to the images converting digital numbers to surface reflectance in order to correct for sensor gains and atmospheric distortion between dates (Green et al., 2005a) For each of the three dates, images were mosaicked to create a single image on which to perform the classification pr ocedures (image processing was done with ERDAS Imagine). Land Cover Classes In the classification process 6 land cover classe s were identified: 1) Agriculture: areas in which the land cover resulted from direct human influence through agricult ural practices. Three sub classes composed the agriculture class: active fields, bare soil, and fallow; 2) Grasslands: the original intention was to differentiate between native grasslands and planted pastures, but it was not possible, so this class includes both; 3) Shrublands: areas covered with woody vegetation usually having multiple permanent stems branching from or near the ground. This class includes two different units in terms of composition and ecological characteri stics: in the south west of the study area shrublands are composed mainly by Larrea sp, indicating the transition between the Caldenal and the Monte ecoregions. In the rest of the study area shrublands indicate sites in which Caldn (the dominant tree species in the region) is present in early successional stages, both colonizing areas of grassland or re-growing after forest fires; 4) Forest: areas covered by at least 10 percent tree crown cover (FAO, 2002); 5) Burned: areas that experienced fires up to approximately six months before image capture ; 6) Lagoon: this class includes both permanent 19

PAGE 20

and tem poral water bodies. In this document, I will refer to as naturally vegetated land covers to the group of grasslands, shrubla nds and forest. The term natura l is used to differentiate agricultural areas from the other ty pes of vegetation in the area, but it should not be interpreted as areas with lack of human influence. Image Classification During summer 2008, 452 training samples were collected in the study area registering information on vegetation structure and com position, among other variab les. In addition, 234 complementary samples were extracted from high resolution imagery of the area (Google Earth). Sixty five percent of the traini ng samples were used during the classification procedure and the remainder was used for the accuracy assessment. Different band combinations were tried for th e supervised classifi cation procedure using the maximum likelihood classifier. The best comb ination in terms of se parability among classes was: band2, band3, band4, band5, and band7 from the La ndsat images, together with a digital elevation model of the area (NASA, 2009), and two layers representing the X and Y coordinates. The digital elevation model and the X-Y coordi nates incorporated information in the model useful to differentiate land cover classes that we re not originally separable based on the spectral data alone. The accuracy of the classificati on was 85.6% (overall kappa statistics = 0.8164) (Jensen, 2005), a little over the usual threshold to accept individual classification (Campbell, 2002). A rule based approach, as described by Dani els (2006), was later incorporated to improve the quality of the final land cover map. Using data mining software (Compumine, 2009) rules were generated and applied in ERDAS Imagin e Knowledge Engineer. Rules were based on derived indices of Tasseled Ca p transformation and NDVI (Normalized difference vegetation Index); and soil data of the ar ea (soil map modified from Ca no et al. 1980). Tasseled Cap transformation is a method of data reduction wi th minimal information loss (Crist & Kauth, 20

PAGE 21

1986) widely used in vegetation m apping and land cover change analysis (Kiage et al., 2007; Chen & Rao, 2008; Jupiter & Marion, 2008; Karn ieli et al., 2008; Siwe & Koch, 2008). The tasseled cap transformation of Landsat thematic mapper (TM) consists of six multispectral features, with the first three bands accounting for the most variati on in the data set. These first three features have been labe led as brightness, greenness and wetness given their ability to describe those physical features of the land surface (Crist & Kauth, 1986; Jin & Sader, 2005). The NDVI is also a widely used index in land cove r change analysis studies given its ability to describe the active vegetation covers (Hansen et al., 2000; Loveland et al ., 2000; Southworth et al., 2004a; Alcaraz-Segura et al., 2009; De La Maza et al., 2009). It is computed as a ratio of the differences between near infrared and red bands to the addition of those bands (Jensen, 2005). The rule based approach generated a more accur ate and precise land cover map of the area for the year 2008 (accuracy = 88.5% and overall kappa statistics = 0.8525, Table 2-2). The same procedure was applied for 1987 and 1999 images. The lack of reference data for 1987 and 1999 prevented the assessment of classification accu racies for those dates. However, given the consistency in the remote sensing data used (all Landsat 5TM images from the same season) and of the classification procedure ap plied, I assumed they did not di ffer significantly from that of 2008. Visual examination of the products supported the assumption. Trajectory Analysis The original interest was not only quantifyi ng net changes among classes between the three dates analyzed, but understanding th e nature of the changes occurr ing in the area, as such two sets of trajectory maps (Petit et al., 2001; Southworth et al., 2004b) were created. The first consisted of two two-date trajectories, one for each time period (1987-1999 and 1999-2008), and the second on a single three-date trajectory su mmarizing changes occurred across the 21-year period. The number of possible comb inations of 6 land-cover classe s for the different dates made 21

PAGE 22

the in terpretation of trajectories difficult and unpr actical (36 possible trajec tories for the two-date analysis and 216 for the three-date analysis). In order to overcome this, a four level trajectory analysis scheme was developed (Table 2-3). The first level classified pixels as changed or unchanged according to the maintenance or not of the same land cover class for the three dates. The second level classified changed pixels acco rding to the nature of the change in: (1) conversion to agriculture; (2) regeneration from agriculture to a natural land cover; or (3) natural change if the land cover changed among grassland, shrubland and/or forest. Natural changes were described in this study as those changes in vegetation cover occurring between grassland, shrubs and forest. The term natural in this case again, refers to the lack of conversion to agriculture, and not to th e lack of human influence in the pro cess of change (e.g. fire is a main force of change in the Caldenal and is highly influenced by management practices). The third level classified the areas of natural change in: (1) opening (if the vegetatio n vertical structure in the final date was lower as compared to the in itial date for each period); (2) closing (if the vegetation structure is higher in th e final date as compared to the initial date for each period); (3) water increase; or (4) water de crease. The fourth level of th e classification has different meanings depending on the previous levels cla ssification. For areas conve rted to agriculture, level four indicates the origin al land cover being converted to agriculture, fo r areas being regenerated from agriculture level four i ndicates the natural la nd cover product of the regeneration, for areas where vegetation was ope ned or closed during the period level four indicates the final land cover type product of th e change, finally for th e areas of unchanged land covers level four indicates th e land cover type that remained unchanged during the period. This simplified version of the trajectories allowed fo r a more rapid and intuitive understanding of the changes occurring in the study area. For example, pixels identified as Ag riculture-Agriculture22

PAGE 23

Agriculture were classified as Unchanged-Agriculture; pixels id en tified as Forest-AgricultureAgriculture or Forest-Forest-Agri culture were classified as C hanged-Conversion to AgricultureForest indicating that the land c over had changed during the period of analysis, that the change was a conversion from a natural land cover to ag riculture, and that the original land cover was forest. Statistical Analysis For the analysis of the processes on land cover change and the determination of the rates of conversion at a parcel level a geographic information system of the study area was developed. A digitized cadastral map of La Pampa provin ce was provided by the La Pampa Ecology Agency (sensitive information was previously removed, providing only the exte nt and the cadastral coding at parcel level). The areas of analysis were: 1) Parque Luro Protected Area, from here on referred to as Park; 2) Game reserves: Records from the Natural Resource Agency were used to identify at a parcel level each of the game reserves registered in the province. Given the variability on the starting date of game reserv es, 2 subpopulations were defined: (a) those that have been working for a longer period (befor e March 1999), hereafter refereed as old game reserves (Old_GR); and (b) those working for a shorter time (after Ma rch 1999), and hereafter refereed as new game reserves (New_GR). Those ra nches previously registered as game reserves but no longer working were directly removed from the analysis to avoid confounding effects. In order to compare the land cover changes among the main land uses of the region, a sample of the parcels dedicated livestock production was extracte d. Given the differences in the size of the units (Old_GR mean size = 71.5 km2, New_GR mean size = 45.3 km2, and cattle ranches mean size = 10.9 km2) a random sample of ranches was not considered appropriate to compare with the park and game reserves. To co ntrol for the significantly smaller size of cattle ranches, 2 blocked random samples (N = 40 each one ) were extracted from all the parcels used 23

PAGE 24

f or that activity (Table 2-4). The first sample was extracted from the parcels with sizes bigger than 48.6 km2 (mean sample size of game reserves minus one standard error). This sample was comparable, in the average extent of the individu al parcels, to the game reserves and will be refereed as big cattle ranches (Big_CR). The second sample was selected from those cattle ranches with sizes smaller than 48.6 km2, this group will be refereed as small cattle ranches (Small_CR). Digitization errors in the boundaries of each of the parcels selected for the analysis were corrected using the 2008 image mosaic as re ference. Both the land cover change and the trajectory analysis were performe d at the parcel level for each of the 5 zones: Park, Old_GR, New_GR, Big_CR and Small_CR (F igure 2-3). The results of Park will be presented and discussed, however they were not be included in the statistical analysis given the presence of only one park in the study area. In order to determine the conservation signifi cance of private game reserves in the region the land cover composition in 1987 (before establishm ent) and its rates of change were analyzed and compared with that of cattle ranches. Differences in the initial proportion of the different land cover classes were tested using KruskalWallis test, using Bonfarroni correction when pairwise comparisons were needed (Kruskal & Wallis, 1952; Shaffer, 1995). For analyzing the effect of the different land uses on land cover ch anges a set of rates were calculated. The annual rate of change of each natural land cover was computed at the property level using the formula described by Dirzo & Garcia (1992): 1* 11/1 1 12 t t t changeA AA rt Where At1 is the extent in hectares of the la nd cover at the begi nning of the period; At2 is the extent of the land cover of inte rest at the end of the period; and t is the time in years for the 24

PAGE 25

period of analysis. This rate was computed on grassland, shrubl and and forest covers for GRs and CRs. Given the im portance assigned by previous studies to the proce sses of conversion in the area, the annual gross rate of conversion to agriculture was al so calculated using a modified version of the rate of change formula: t tnatural p converted conversionA A r/1 1_ _11 Where Aconverted_p is the extent in hectares converte d to agriculture during the period; Anatural_t1 is the extent in hectares of the natural land cove rs (grasslands, shrubla nds and forest) at the beginning of the period; and t is the time in years. Differences in rates of change and conversion between periods and land uses were tested using Kruskal-Wallis test (when pairwise comparisons were needed Bonfarroni correction was applied) (Kruskal & Wallis, 1952; Shaffer, 1995). 25

PAGE 26

CHAP TER 3 RESULTS Land Cover Patterns Three land cover maps resulted from the classification procedure, one for 1987, 1999 and 2008 (Figure 3-1). Across the three dates the basic land cover patter n is repeated. Agriculture is concentrated in the eastern sector of the study area while grasslands are re stricted to the western sector and to one of th e transversal valleys ( Valle Argentino ). Shrublands, as previously explained, comprise two different compositi onal communities. The community composed mainly of Larrea sp is restricted to the southwestern portion of the study area, indicating the transition to the Monte ecoregion. On the othe r hand, the shrublands composed by juvenile stages of Caldn (product of fire regeneration or colonization) are distri buted all over the area. Forest distribution shows a significant difference in the latitudinal gradient. While forest cover is evenly distributed in the northern sector (especially in 1987), it is re stricted to the slopes of the transversal valleys in the central and southern portion of the area analy zed. The distribution of burned areas in the study site grea tly varies between years, showi ng the random nature of fire in this region. Preferential Location The first prerequisite for a conservation area to exist is to be placed in areas where the land covers of conservation interest are located. To determine if game reserves fulfill this first prerequisite to be considered valuable for the conservation of the Caldenal, the relative proportions of the different land covers were compared amo ng Old_GR, New_GR, Big_CR and Small_CR for 1987, before game reserves were formally established. The mean area dedicated to agriculture was lowest in Old_GR (4.6%) as compared to New_GR (11.5%), Big_CR (12.3%) and Small_CR (56.5%) (Kruskal-W allis test, p<.0001). Agricultura l cover was statistically 26

PAGE 27

significan tly higher in Small_CR as compared to the other 3 zones (no significant differences were found between the first 3 zones). The mean area covered by grasslands was highest in Big_CR (24.7%), followed by Old_GR (23.7%), New_GR (16.4%) and Small_CR (12.4%) (Kruskal-Wallis test, p=0.0018). Gr assland cover was statistica lly significantly lower in Small_CR than in the other 3 zones (no significa nt differences were found between the first 3 zones). The mean area covered by shrubs was highest in Big_CR (22.6%), followed by New_GR (11.2%), Old_GR (10.6%) and with Small_CR ( 5.5%) presenting the lowest shrubs cover (Kruskal-Wallis test, p=0.0001), with the Small_CR s shrub cover being significantly lower than that from the other 3 zones (no significant di fferences were found between the first 3 zones). Based on the differences in the land cover of agri culture, grassland and sh rubs two groups were identified, on one side GRs and Big_CR, a nd on the other Small_CR. A different pattern emerges when forest cover is analyzed. Old_GR and New_GR presented the highest forest cover in 1987 (59.7% and 58.4% respectively), followed by Big_CR (39.0%) and Small_CR (24.6%) presenting the lowest (Kruskal-Wallis test, p=<. 0001). Three significantly different groups were identified relative to the forest cover. On one extrem e GRs showed the highest forest cover, in the middle Big_CR, and Small_CR presented the lowe st proportion of its area occupied by forest (Figure 3-2). Conversion to Agriculture In both periods the Park presented the lowest annual rates of conversion from natural land covers to agriculture (Table 3-1), followed in increasing order by Old_GR, Big_CR, New_GR and Small_CR. The first four zones presented m ean annual rates lower than 1% for both periods, much lower than that observed in Small_CR (8.4% for period 1 and 11.6% for period 2), although any of the differences between groups for each period were statistically significant. No significant differences were identified neither be tween period 1 and 2 for each of the zones; 27

PAGE 28

however an increase in the m ean annual rate of conversion to agricu lture in Small_CR might suggest an intensification of the process in these areas. Rates of Change Conversion to agriculture is one of the ch anges occurring in the caldenal, which is extremely important from the conservation pers pective. However, natural changes are much more common in terms of spatial extent (Figure 3-3-C). To accura tely address this processes the annual rate of change for each of the natural land covers was calculated (Table 3-2). In order to determine the effect of GRs on the land cover change dynamics, these rates were compared for each period among the different zones, finding no clearly distinct patterns between GRs and CRs. When the comparison was made between period 1 and 2 for each of the zones, most of the rates showed no significant differenc e, except from that of forest. The forest mean annual rate of change was negative for all GRs and CRs be tween 1987 and 1999 (-5.1% in Old_GR, -3.6% in New_GR, -4.7% in Big_CR, and -2.1% in Small_ CR) indicating a regional process of loss in forest cover. On the other hand, for the second period, two different trends can be identified. While the negative rate in Small_CR becomes even more negative (-11.5%), indicating an increase in the deforestation process, for th e GRs and Big_CR the loss of forest stopped and forest started to recover (0.71% in Old_G R, 2.3% in New_GR, and 4.6% in Big_CR). Differences in the rate of change of forest between period 1 and 2 for each zone were all significant at p<0.05. Trajectory Analysis No major differences were identified between th e two individual periods in the trajectory analysis and neither between those and the traj ectory summarizing the changes of the 21 years, so in order to maintain the simplicity, the later will be presented here. The trajectory analysis shows that the majority of the land cover did no t change across dates, though some differences 28

PAGE 29

can be identified between zones. In the Park 92.3% of the land cove r present in 1987 did not change during the period analyzed (Figure 3-3-A), indicating the st ability of vegetation covers in areas with strict conservation m anagement. Sm all_CR showed the next highest proportion of unchanged cover (80.2%), though here the lack of ch ange is due to the presence of agricultural areas, which represent 66.9% of the unchange d area. For Old_GR, New_GR and Big_CR the proportion of the area with unchanged land c overs was very similar (around 60%). The land cover composition of the unchanged areas was also analyzed. For the Park, 85.2% of the unchanged area between 1987 and 2008 wa s covered by forest, followed in extent by lagoons (11.6%). Then a gradient can be id entified between Old_G R, New_GR, Big_CR and Small_CR in terms of the forest and agricultural proportion of the unchanged areas. Forest cover decreases from Old_GR to Small_CR while the proportion of agriculture varies in the opposite direction (highest in Sm all_CR, Figure 3-3-B). The dynamic nature of the Caldenal system could be better described when the different types of changes were considered (Figure 3-3-C). In this analysis a distinct pattern was identified for the Park as compared to the other zones. In the Park, the overall cha nged area represents only 7.7%. Of that area, the proportion occupied by natura l changes, conversion to agriculture (in this case conversion did not represent increase in commercial agricultu re, but increase in the density and size of firebreaks) and regeneration were a lmost evenly distribute d. On the other hand, in GRs and CRs most of the changes occurring during the 21 year period were described as natural (on average 75% of the changed surface). Conversion to agriculture represents on average 15.0% of the changed area with no clear differences be tween GRs and CRs. Regeneration is in all cases smaller in extent than conversion to agriculture. 29

PAGE 30

In the Park most of the natural changes repr esen ted a process of cl osing of the vegetation as it matures towards forest cover. It is important to notice that the area in the park subject to natural changes was almost non exis tent (1.6%). More relevant ar e these natural changes in GRs and CRs, where they represent 12.3% of the area for Small_CR and around 30% for GRs and Big_CR. Both in GRs and CRs the increase in sh rub cover is the main change observed. This occurs via two different processes: 1) the process of opening from forest to shrublands, distributed all over the study area with no clear spatial pattern de scribing it (Figure 3-4-B); and 2) the process of closing from grasslands to shrublands, mainly concentrated in the western sector of the study area (Figure 3-4-C). Of the area converted to agricu lture, most of it was originally covered by forests, followed by areas covers by grasses (Figure 3-3-E). Ther e is also a process of regeneration from agriculture to natural land cover, though as alrea dy mentioned it is reduced in extent as compared to the conversion to agriculture (F igure 3-3-C). Differences are also evident in the type of natural cover being lost for agriculture and those being the product of th e regeneration. While conversion to agriculture occu rred mainly in forested ar eas (followed by grasslands), regeneration occurred from agriculture to grassla nds, and to a lesser extent forest and shrublands (Figure 3-3-F). This process of regeneration to gr asslands should be interp reted as establishment of planted pastures and not the regeneration of natural grasslands (as already explained this confusion is a product of the lim itations in the classification pr ocedure). Differences are also evident in the spatial distribution of these changes. While the areas of conversion from forest to agriculture are located to the east of the study area (Fi gure 3-5-C), conversion to pasture occurred in the central portion of the study area (Figure 3-5-D). 30

PAGE 31

CHAP TER 4 DISCUSSION Processes of Land Cover Change In the last century the Caldenal ecosystem has been subject to an intense process of conversion and degradation co mpromising its long term conservation (Mendez, 2007b). Different human interventions contributed to the present state of the sy stem. Extraction of wood, introduction of cattle, changes in fire freque ncy and intensity and finally conversion to agriculture have converted the Caldenal in one of the worlds crisis ecoregions (Hoekstra et al., 2005) and in a conservation priority at the nati onal level (APN, 2007). However, even when the need to take regional conservation measures to protect this system, its biodiversity and the services it provides to society has been acknow ledged, little has been done to increase the area protected or to develop sust ainable management practices (Mendez, 2007b). Game reserves appeared then as possible private land use practices with a conservation value for the maintenance of native habitats. Results indicate that game reserves were ini tially established in areas where forest cover was more abundant, and for the last 21 years this pattern has been maintained. Forest, shrubs and grasslands were more abundant in the areas where game reserves would later be established, as compared to the small cattle ranches, dominated by agriculture. Big cattle ranches presented in 1987 similar proportions of grasslan ds and shrublands as the game reserves, however forest cover was significantly lower. This preferential selection of forest by the game reserves might well be explained by the higher abunda nce of game species (especially Cervus elaphus ) in forested habitats of La Pampa. The fact that game reserves have been preferentially located in areas where the habitats of conser vation interest are located is va luable and should be considered in the process of regional conservation planning. 31

PAGE 32

Shifts between natural land covers (grassland, shrubland and forest) were identified as the more im portant changes in terms of extent, a nd again no differences we re identified between land uses. Savannas are highly dynamic systems in which fire and patterns of animal use influence the processes of change (Wisem an et al., 2004; Brook & Bowman, 2006). Depending on the intensity of the fire, consequences on the plant community can vary greatly. In the Caldenal region fires generate the senescence of the aerial structure of trees and shrubs, not killing them in most cases, followed by a vigor ous re-growth (Boo et al., 1997). How long it will take to reach the forest state depends on manage ment, especially cattle management (Mendez, 2007b). Cattle generates an increas e in the germination and establishment of Caldn seedlings (Villalobos et al., 2005). Studies in the area have shown that after cattle were introduced to an area (both forest and grassland) Caldn recruitment increased si gnificantly (Dussart et al., 1998). Both of these processes, fire changes and cattle introduction, have generate d in the last century an increase in the shrub cover in detriment of natural grasslan ds and open forests (Busso, 1997). Results from the trajectory analysis show the occurrence of both proce sses in the study area. Changes from grasslands to shrubl ands and from forest to shrubla nds are differentially located in the study region (the first concentrated main ly in the northwestern region and the second distributed all over the area), however no major di stinctions were identified between the land use types analyzed. Differences in the rates of conve rsion to agriculture were expected in the areas where game reserves were established as compared to cattl e ranches. However, no such differences were found. The rate of conversion to agriculture in the parcels wh ere old game reserves were established remained unchanged between the fi rst period (when game reserves were being established) and second (when game reserves were formally in operation). No differences were 32

PAGE 33

found either in the parcels where the new ga me reserves were established between the first period (these ranches were still working solely as cattle ranches) a nd the second (when game reserves were being established). If the rates of conversion to agriculture had increased over time in the rest of the Caldenal, the fact of mainta ining a steady rate of c onversion could have been interpreted as a positive outcome of the game reserves. However, no significant differences were found neither for big or small cattle ranches be tween both periods. These results do not support the initial hypothesis of differe ntial processes of land cover ch ange occurring in the game reserves as compared to the cattle ranches in the Caldenal region. On the other hand, forest rates of change did si gnificantly vary between periods in the four zones. Deforestation was a generalized pr ocess in the Caldenal between 1987 and 1999, occurring in the areas where game reserves wher e being established and in the cattle ranches (no reduction in forest cover was detect ed inside the protected area). A different pattern is identified for the second period, when deforestation stopped in the areas of old and new game reserves and in the big cattle ranches, but it increased in magn itude in the areas of small cattle ranches. The average annual rate of change in forest cover in the small cattle ranche s is very high (-11.5% for the second period); however it should considered th at forest cover was already scarce in this areas, and the loss of a small area of forest is computed as a large percentage of loss. Nevertheless, there is an increase in the rate of loss of forest in th e small cattle ranches that is not seen in the other zones analyzed. Apparently other factors differe nt from the land use type have influenced the processes of forest conversion in this region. The size of the property appears th en to have an effect on the rates of forest change more than the implementation or not of game reserves. In big properties (both game reserves and cattle ra nches) the process of forest lo ss stopped, while it became more 33

PAGE 34

intens e in the small ones. However, the size of the individual propertie s is a consequence of environmental and technological factors that determine the capability of the land to produce and generate income. The spatial distribution of the different zones analyzed (F igure 2-3) shows that old and new game reserves and big cattle ranche s are located primarily (85%) in areas unsuitable for agriculture, while the small cattle ranches are mainly located in the areas suitable for crop production (73%). These results reinforce the idea of the presence of an active agricultural frontier to the east of the Caldenal (Lell, 2004). Forest loss in La Pampa is at a maximum to the eastern, where environmental conditions are more appropriate for the development of agriculture. The pressure, and cons equently the conversion, is lowe r the farther west the property is located regardless of the establishment or not of game reserves. These results indicate that it is not the land use, but the presence of more appropriate environmental conditions (especially soil and water) that is influencing the processes of land cover change in the area. In present days, game reserves and big cattle ranches are not su ffering the processes of fo rest loss, however if new crops or more efficient farming techniques are developed for the drier environmental conditions of the west, the remainder of the Caldenal may follow the same conversion path occurred in the actual agricultural fields. Game Reserves and Conservation In a world where environmental degradation, habitat loss and species extinction is everyday more common (Fearnside, 2005; MEA, 2005) there is an urgent need for more and new effective conservation practices. Government ally owned and managed protected areas have been the cornerstone of conservation in the last century (Margules & Pressey, 2000), and its increase in number and extent is a clear proof of that. However, those protected areas have been recognized as not enough to assure the long term maintenance of global biodiversity and functioning ecosystems (Chape et al., 2005). More conservation is needed, but how? New 34

PAGE 35

conservation practices have starte d to develop in the last decade s try ing to bring together the provision of goods and services la ndowners need to su rvive with the comm on goal of long term biodiversity conservation (Frank et al., 2003; Sanchez-Azofeifa et al., 2007; Frost & Bond, 2008). Game reserves are one of these private enterp rises identified with high conservation value given their relatively low impact on local natural resources and th eir high profitabi lity (Barnes & deJager, 1996; Tomlinson et al., 2002; Lindsey et al., 2006a). In Africa na tive wildlife utilization as a land use is growing, given its proven eco logic and economic benefits to landowners (Bond et al., 2004). Fee hunting in North America is a growing industry also with positive effects both for the land owner and the ecosystem conser vation (Williams & Lathbury, 1996; Butler et al., 2005). However, in Argentina fee hunting in game reserves is a relativel y new activity; and its benefits for the conservation of the Caldenal are not completely clear. Game reserves in central Argentina were establ ished in the areas of highest forest cover, potentially very beneficial in a landscape desc ribed as suffering high rates of deforestation. However, results from this analysis showed that deforestation is happening in the more productive areas to the east of the study area, and not in the west, where most of the game reserves are located. On the other hand, the development of hun ting enterprises based primarily in the offer of exotic species greatly reduces the conservation value of this land use. Exotic species are one of the biggest threats to the cons ervation of biodiversity worldwide, and if game reserves in this region conti nue to actively introduce new sp ecies for hunting, they could generate negative impacts, precise ly in the areas where the habitat of conservation interest is located. The potential benefit of be ing located in areas of highest fo rest cover could be easily out weighted by the negative effects of the introduction of exotic species, that is why a deeper 35

PAGE 36

understanding on the effects of gam e reserves as sources of exotic species, and on the effect of those species on the rest of the system is needed before the re al value of game reserves as conservation tools could be determined. 36

PAGE 37

37 Figure 2-1. Relative location of th e study area in La Pampa Province, Argentina. Game reserves are located mainly in the central part of the Caldenal region. On the other hand, only one terrestrial protected area (Parque Luro) is present in the area. Figure 2-2. Natural gradients. Rainfall map co rresponds to the average annual rainfall for the period 1978-2008. Soil agricultu ral suitability map from Cano (1980). Economic units determined by local legislation.

PAGE 38

Figure 2-3. Spatial distribution of the parcels analyzed. Soil agricultural suitability map from Cano (1980). Old_GR: Old game reserves; New_GR: New game reserves; Big_CR: Big cattle ranches; and Small_CR: Small cattle ranches. Figure 3-1. Land cover maps of the cen tral Caldenal region of 1987, 1999 and 2008. 38

PAGE 39

Figure 3-2. Mean land cover composition of 1987, 1999 and 2008 (error bars represent the standard error of the mean). Old_GR: Old game reserves; New_GR: New game reserves; Big_CR: Big cattle ranches; and Small_CR: Small cattle ranches. 39

PAGE 40

Figure 3-3. Mean land-cover change traject ories between 1987, 1999 and 2008. A) % of area changed and unchanged by zone. B) Land-c over composition of the unchanged area. C) Land cover changes relative to the changed area. D) Trajectory of change relative to the natural changes area of the zone. E) Original la nd-cover of the areas converted to agriculture relative to the area converted to agriculture for the zone. F) Land-cover to which agricultural fields were regenera ted relative to the area regenerated for the zone. In parenthesis is the percentage of area relative to the total area of the zone. Old_GR: Old game reserves; New_GR: Ne w game reserves; Big_CR: Big cattle ranches; and Small_CR: Small cattle ranches. 40

PAGE 41

Figure 3-4. Spatial distribution of areas in which natural changes o ccurred between 1987, 1999 and 2008. A) O_grass: the natural vegetation opened, converting into grasslands. B) O_shrub: the natural vegetation opened, conve rting into shrublands. C) C_shrub: the natural vegetation closed, c onverting into shrublands. D) C_forest: the natural vegetation closed, converting into forest. 41

PAGE 42

Figure 3-5. Spatial distribution of agricultural changes be tween 1987 and 2008. A, B and C) Areas converted to agriculture between 1987, 1999 and 2008. C_grass: grasslands converted to agricultu re; C_shrub: shrublands convert ed to agriculture; C_forest: forest converted to agriculture. D,E and F) : Areas regenerated from agriculture to natural land covers. R_grass: regenerated in to grasslands; R_shrub: regenerated into shrublands; R_forest: re generated into forest. 42

PAGE 43

43 Table 2-1. Satellite images used in for the analysis. Sources: Global Land Cover Facility (GLCF) and Ecology Agency of La Pampa (EALP). Year Path Row Fecha Satelite Sensor Sources 1987 228 86 4/18/1987 Landsat 5 TM GLCF 229 85 2/20/1987 Landsat 5 TM GLCF 229 86 2/20/1987 Landsat 5 TM GLCF 1999 228 86 2/14/1999 Landsat 5 TM EALP 229 85 3/9/1999 Landsat 5 TM EALP 229 86 3/9/1999 Landsat 5 TM EALP 2008 228 86 3/10/2008 Landsat 5 TM EALP 229 85 3/1/2008 Landsat 5 TM EALP 229 86 3/1/2008 Landsat 5 TM EALP Table 2-2. Confusion matrix for the 2008 land-cover classification. Overall Kappa Statistics = 0.8525. OE: omission error in %; ProAc: produc ers accuracy in %; CoE: commission error in %; UssAc: users accuracy in %. Reference Data Classified Data 1 2 3 4 5 6 Total CoE UssAc 1 Agriculture 58 2 0 1 0 0 61 4.9 95.1 2 Grassland 2 17 4 2 0 0 25 32.0 68.0 3 Shrubland 1 2 28 5 0 0 36 22.2 77.8 4 Forest 1 2 6 70 0 0 79 11.4 88.6 5 Burned 0 0 0 0 7 0 7 0.0 100 6 Lagoon 0 0 0 0 0 36 36 0.0 100 Total 62 23 38 78 7 36 244 OE 6.5 26.1 26.3 10.3 0.0 0.0 ProAc 93.5 73.9 73.7 89.7 100 100 88.5

PAGE 44

Table 2-3. Trajectory analysis scheme. Each of the 216 possible trajectori es was classified in one of these 18 possible main trajectories at 4 different levels: LEVEL01: Changed or unchanged areas; LEVEL02 (for changed ar eas): Conversion to agriculture, regeneration from agriculture to natural land covers or natural changes; LEVEL03 (for natural changes): Opening of vegeta tion, closing of vegetation, or changes in extent of lagoons; LEVEL04 (has different m eanings for the different sublevels): 1-3: represent the original class be ing converted to agriculture; 4-6: represent the class to which agricultural fields were regenerated to ; 7-8: represent trajectories in which the natural vegetation is opening to grasslands or shrublands; 9-10: areas in which the natural vegetation is closing to shrublands or forest; 11: areas in which water has increase; 12: areas in which water has decr eased; 13-18: areas in which the land cover has not changed during the 21 years of analysis. LEVEL01 LEVEL02 LEVEL03 LEVEL04 Changed Conversion to agriculture Grassland 1 Shrubland 2 Forest 3 Regeneration Grassland 4 Shrubland 5 Forest 6 Natural changes Opening Grassland 7 Shrubland 8 Closing Shrubland 9 Forest 10 Water increase 11 Water decrease 12 Unchanged Agriculture 13 Grassland 14 Shrubland 15 Forest 16 Burned 17 Lagoon 18 Table 2-4. Statistical summary of the different zones analyzed. St andard errors are presented in parenthesis (no standard e rror for Park due to n=1). Zone Description N Mean (ha) Park Population 1 7,589 (-) Game Reserves Population 59 5,460 (597) New Sub population 21 7,144 (1,088) Old Sub population 38 4,530 (658) Cattle Ranches Population 2,389 1,087 (34) Big Sample 40 7,276 (570) Small Sample 40 880 (136) 44

PAGE 45

Table 3-1. Annual rate of convers ion to agriculture. Standard erro rs are presented in parenthesis (no standard error for Park due to N=1) Old_GR: Old gam e reserves; New_GR: New game reserves; Big_CR: Big cattle ranches; and Small_CR: Small cattle ranches. Period 1 Period 2 Zone Mean Mean Park 0.1% (-) 0.3% (-) Old_GR 0.6% (0.2%) 0.5% (0.2%) New_GR 0.9% (0.2%) 0.9% (0.1%) Big_CR 0.7% (0.2%) 0.7% (0.2%) Small_CR 8.4% (2.6%) 11.6% (3.6%) Table 3-2. Annual rate of change of natural land cover classes. St andard errors are presented in parenthesis (no standard error for Park due to n=1). Old_GR: Old game reserves; New_GR: New game reserves; Big_CR: Bi g cattle ranches; and Small_CR: Small cattle ranches. Grassland Shrubland Forest Period 1 Period2 Period 1 Period2 Period 1 Period2 Zone Mean Mean Mean Mean Mean Mean Park -3.9% (-) 1.8% (-) 0.0% (-) 0.0% (-) 0.3% (-) -0.2% (-) Old_GR -1.8% (5.3%) -1.8% (2.5%) 6.3% (9.7%) 0.1% (3.5%) -5.1% (1.5%) 0.7% (4.5%) New_GR 1.3% (1.5%) 2.7% (2.4%) 1.8% (3.8%) 5.9% (3.1%) -3.6% (1.2%) 2.3% (1.7%) Big_CR 2.7% (3.7%) 1.9% (1.7%) 9.7% (5.9%) 6.9% (3.0%) -4.7% (1.5%) 4.6% (2.0%) Small_CR -6.7% (4.7%) -11.8% (6.1%) 5.2% (4.5%) 2.7% (3.2%) -2.1% (4.0%) -11.5% (4.4%) 45

PAGE 46

46 REFERENCES Alcaraz-Segura, D., J. Cabello & J. Paruelo. (2009). Baseline ch aracterization of major Iberian vegetation types based on the NDVI dynamics. Plant Ecology, 202(1), 13. Altmoos, M. & K. Henle. (2007). Differences in characteristics of rese rve network selection using population data vers us habitat surrogates. Biodiversity and Conservation, 16(1), 113135. Amieva, E. O. (1993). El Parque Luro. Santa Rosa: Fondo Editorial Pampeano. Andrade, G. I. (2005). Science and so ciety at the World Parks Congress. Conservation Biology, 19(1), 4. APN. (2007). Las areas protegidas de la Argentina, herramienta superior para la conservacion de nuestro patrimonio natural y cultural Administracion de Par que Nacionales. Buenos Aires. pp. 87. Azevedo-Ramos, C., B. D. do Amaral, D. C. Neps tad, B. Soares & R. Nasi. (2006). Integrating ecosystem management, protected areas, and mammal conservation in the Brazilian Amazon. Ecology and Society, 11(2), 17. Barnes, J. I. & J. L. V. deJager. (1996). Ec onomic and financial incentives for wildlife use on private land in Namibia and the implications for policy. South African Journal of Wildlife Research, 26(2), 37. Bawa, K. S. & R. Seidler. (1998). Natural forest management and conservation of biodiversity in tropical forests. Conservation Biology, 12(1), 46. Bond, I., B. Child, D. de la Harpe, B. Jones, J. Barnes & H. Anderson (2004). Private land contribution to conservation is southern Africa. In: B Child (Eds) Parks in transition London: Earthscan, pp. 267. Boo, R. M., D. V. Pelaez, S. C. Bunting, M. D. Mayor & O. R. Elia. (1997). Effect of fire on woody species in central semi-arid Argentina. Journal of Arid Environments, 35(1), 87. Brockington, D. (2007). Forests, community c onservation, and local government performance: The village forest reserves of Tanzania. Society & Natural Resources, 20(9), 835. Brook, B. W. & D. Bowman. (2006). Postcards fr om the past: charting the landscape-scale conversion of tropical Australian savanna to closed forest during the 20th century. Landscape Ecology, 21(8), 1253. Brooks, T. M., R. A. Mittermeier, G. A. B. da Fonseca, J. Gerlach, M. Hoffmann, J. F. Lamoreux, C. G. Mittermeier, J. D. Pilgrim & A. S. L. Rodrigues. (2006). Global biodiversity conservation priorities. Science, 313(5783), 58.

PAGE 47

Bruner, A. G., R. E. Gullison, R. E. Rice & G. A. B. da Fonseca. (2001 ). Effectiveness of parks in protecting tropical biodiversity. Science, 291(5501), 125. Busso, C. A. (1997). Towards an increased and sustainable producti on in semi-arid rangelands of central Argentina: two decades of research. Journal of Arid Environments, 36, 197. Butler, M. J., A. P. Teaschner, W. B. Balla rd & B. K. McGee. (2005). Commentary: Wildlife ranching in North America argum ents, issues, and perspectives. Wildlife Society Bulletin, 33(1), 381. Cabrera, A. L. (1994). Enciclopedia Argentina de agricultura y jardinera, Tomo II, Fascculo 1: Regiones fitogeogrficas Argentinas Buenos Aires: Acme. Campbell, J. B. (2002). Introduction to Remote Sensing Ney York: Guilford Press. Cano, E., B. Fernndez & A. Montes. (1980). Inventario integrado de los recursos naturales de la provincia de La Pampa. Buenos Aires: Unlpam, Gobierno de La Pampa e INTA. Caro, T. M. (2001). Species richness and abundan ce of small mammals inside and outside an African national park. Biological Conservation, 98(3), 251. Caro, T. M. & G. O'Doherty. (1999). On the us e of surrogate species in conservation biology. Conservation Biology, 13(4), 805. Chape, S., J. Harrison, M. Spalding & I. Lysenko. (2005). Mesaruring the extent and effectiveness of protected areas as an indica tor for meeting global biodiversity targets. Philosophical Transactions of The Royal Society, 360, 443. Chapin, F. S., E. S. Zavaleta, V. T. Eviner, R. L. Naylor, P. M. Vitousek, H. L. Reynolds, D. U. Hooper, S. Lavorel, O. E. Sala, S. E. Hobbie, M. C. Mack & S. Diaz. (2000). Consequences of changing biodiversity. Nature, 405(6783), 234. Chen, S. & P. Rao. (2008). Land degradation mo nitoring using multi-temporal Landsat TM/ETM data in a transition zone between gra ssland and cropland of northeast China. International Journal of Remote Sensing, 29(7), 2055. Clerici, N., A. Bodini, H. Eva, J. M. Gre goire, D. Dulieu & C. Paolini. (2007). Increased isolation of two biosphere reserves and surrounding protected ar eas (WAP ecological complex, West Africa). Journal for Nature Conservation, 15(1), 26. Compumine. (2009). Available: http://www.compumine.com/web/public/home Crist, E. P. & R. J. Kauth. (1986). T he tasseled cap de-mystified. Photogrammetric Engineering and Remote Sensing, 52(1), 81. Daniels, A. E. (2006). Incorpor ating domain knowledge and spatia l relationships into land cover classifications: a rule-based approach. International Journal of Remote Sensing, 27(14), 2949. 47

PAGE 48

De La Maza, M., M. Lim a, P. L. Meserve, J. R. Gutierrez & F. M. Jaksic. (2009). Primary production dynamics and climate variability: eco logical consequences in semiarid Chile. Global Change Biology, 15(5), 1116. Dirzo, R. & M. C. Garcia. (1992). Rates of deforestation in Los Tuxlas, a neotropical are in southeast Mexico. Conservation Biology, 6(1), 84. Dussart, E., P. Lerner & R. Peinetti. (1998). Long term dynamics of 2 populations of Prosopis caldenia Burkart. Journal of Range Management, 51(6), 685. Engel, J. B. (2007). The development, status, an d viability of the cons ervation easement as a private land conservation tool in the Western United-States. Urban Lawyer, 39(1), 19. Engel, S., S. Pagiola & S. Wunder. (2008). Desi gning payments for environmental services in theory and practice: An overview of the issues. Ecological Economics, 65(4), 663. FAO. (2002). Global Forest Resources Assessment 2000. Main Report. Rome. pp. 512 Fearnside, P. M. (2005). Deforestation in Brazi lian Amazonia: History, rates, and consequences. Conservation Biology, 19(3), 680. Foley, J. A., R. DeFries, G. P. Asner, C. Barford, G. Bonan, S. R. Carpenter, F. S. Chapin, M. T. Coe, G. C. Daily, H. K. Gibbs, J. H. Helkowski, T. Holloway, E. A. Howard, C. J. Kucharik, C. Monfreda, J. A. Patz, I. C. Pren tice, N. Ramankutty & P. K. Snyder. (2005). Global consequences of land use. Science, 309(5734), 570. Frank, G., F. Muller & Kv. (2003). Voluntary appro aches in protection of forests in Austria. Environmental Science & Policy, 6(3), 261. Frost, P. G. H. & I. Bond. (2008). The CAMPFIRE programme in Zimbabwe: Payments for wildlife services. Ecological Economics, 65(4), 776. Gaston, K. J., S. E. Jackson, L. Cantu-Sa lazar & G. Cruz-Pinon. (2008). The Ecological Performance of Protected Areas. Annual Review of Ecology Evolution and Systematics, 39, 93. Gobbi, J. A. (1994). Coto de caza activity in La Pampa province, Argentina. Center for Latinamerican Studies, University of Fl orida. Master of Arts Thesis. pp. 93. Green, G. M., C. M. Schweik & J. C. Ra ndolph (2005a). Retrieving land-cover change information from Landsat satellite images by minimizing other sources of reflectance variability. In EF Moran and E Ostrom (Eds) Seeing the Forest and the Trees Cambridge: The MIT Press, pp. 442. Green, R. E., A. Balmford, P. R. Crane, G. M. Mace, J. D. Reynolds & R. K. Turner. (2005b). A framework for improved monitoring of biodive rsity: Responses to the World Summit on Sustainable Development. Conservation Biology, 19(1), 56. 48

PAGE 49

Hansen, M. C., R. S. De fries, J. R. G. Townshend & R. Sohlberg. (2000). Global land cover classification at 1km spatial resolution using a classification tree approach. International Journal of Remote Sensing, 21(6), 1331. Hoekstra, J. M., T. M. Boucher, T. H. Ricketts & C. Roberts. (2005). Confronting a biome crisis: global disparities of ha bitat loss and protection. Ecology Letters, 8(1), 23. Houghton, R. A. (1994). The worldwid e extent of land-use change. Bioscience, 44(5), 305. Huang, C. Q., S. Kim, A. Altstatt, J. R. G. Townshend, P. Davis, K. Song, C. J. Tucker, O. Rodas, A. Yanosky, R. Clay & J. Musinsky. (2007). Rapid loss of Paraguay's Atlantic forest and the status of protec ted areas A Landsat assessment. Remote Sensing of Environment, 106(4), 460. Hutton, J., W. M. Adams & J. C. Murombedzi. (2005). Back to the Barriers? Changing Narratives in Biodive rsity Conservation. NUPI Forum for Development Studies, 2, 341 370. IPCC. (2007). Intergovernmental Panel on Clim ate Change, Climate Change 2007: Synthesis Report, from http://www.ipcc.ch/ipccreports/ar4-syr.htm IUCN. (1994). Guidelines for Protected Areas Managem ent Categories Retrieved 09/19/2007, from http://www.unep-wcmc.org/protected_areas/categories/index.htm l. Jenkins, M., S. J. Scherr & M. I nbar. (2004). Markets for biodiversity services Potential roles and challeng es. Environment, 46(6), 32. Jensen, J. R. (2005). Introductory digital im age processing: A remote sensing perspective Upper Saddle River: Pearson Prentice Hall. Jin, S. M. & S. A. Sader. (2005). Comparison of time series tasseled cap wetness and the normalized difference moisture index in detecting forest disturbances. Remote Sensing of Environment, 94(3), 364. Joppa, L. N., S. R. Loarie & S. L. Pimm. (2008). On the protection of "protected areas". Proceedings of the National Academy of Sc iences of the United States of America, 105(18), 6673. Jupiter, S. D. & G. S. Marion. (2008). Changes in forest area along stream networks in an agricultural catchment of th e Great Barrier Reef Lagoon. Environmental Management, 42(1), 66. Kareiva, P. & M. Marvier. (2003). Conserving bi odiversity coldspots Recent calls to direct conservation funding to the world's biodiversit y hotspots may be bad investment advise. American Scientist, 91(4), 344. 49

PAGE 50

Karnieli, A., U. Gilad, M. Ponzet, T. Svora y, R. Mirzadinov & O. Fedorina. (2008). Assessing land-cover change and degradation in the Ce ntr al Asian deserts using satellite image processing and geostatistical methods. Journal of Arid Environments, 72(11), 2093. Kiage, L. M., K. B. Liu, N. D. Walker, N. Lam & O. K. Huh. (2007). Recent land-cover/use change associated with land degradation in the Lake Baringo catchment, Kenya, East Africa: evidence from Landsat TM and ETM. International Journal of Remote Sensing, 28(19), 4285. Klimek, S., A. R. Kemmermann, H. H. Steinmann, J. Freese & J. Isselstein. (2008). Rewarding farmers for delivering vascular plant diversity in managed grasslands : A transdisciplinary case-study approach. Biological Conservation, 141(11), 2888. Koutche, V. & J. Carmelich. (1936). Estudio forestal del caldn. Boletin del Ministerio de Agricultura de la Nacion, XXXVII(1), 1. Kruskal, W. H. & W. A. Wallis. (1952). Use of Ranks in One-Criterion Variance Analysis. Journal of the American Statistical Association, 47(260), 583. Lambin, E. F. & H. J. Geist. (1999). Global land use and land cover change: what have we learned so far? Land-Use and Land-Cover Change (LUCC): Implementation Strategy. IGBP Report #48 and IHDP Report #10. Langholz, J. A. & G. I. H. Kerley. (2006). Combining Conservation and Development on Private Lands: An Assessment of Ecotourism-based Pr ivate Game Reserves in the Eastern Cape Center for African Conservation Ecology. Port Elizabeth. pp. 29. Langholz, J. A. & J. Lassoie. (2001). Combin ing Conservation and Development on Private Lands: Lessons from Costa Rica. Environment, Development and Sustainability, 3, 309 322. Lell, J. D. (2004). El Caldenal: una visin pano rmica del mismo enfatiza ndo en su uso. In MF Arturi, JL Frangi and JF Goya (Eds) Ecologa y manejo de los bosques de Argentina pp. 18. Lindsay, M. J., H. M. Patterson & S. E. Swearer. (2008). Habitat as a surrogate measure of reef fish diversity in the zoning of the Lo rd Howe Island Marine Park, Australia. Marine Ecology-Progress Series, 353, 265. Lindsey, P. A., R. Alexander, L. G. Frank, A. Mathieson & S. S. Romanach. (2006a). Potential of trophy hunting to create incen tives for wildlife c onservation in Africa where alternative wildlife-based land uses may not be viable. Animal Conservation, 9(3), 283. Lindsey, P. A., P. A. Roulet & S. S. Romanach. (2006b). Economic and conservation significance of the trophy hunting in dustry in sub-Saharan Africa. Biological Conservation, 134(4), 455. 50

PAGE 51

Lom bard, A. T., R. M. Cowling, R. L. Pressey & A. G. Rebelo. (2003). Effectiveness of land classes as surrogates for species in conserva tion planning for the Ca pe Floristic Region. Biological Conservation, 112(1), 45. Loveland, T. R., B. C. Reed, J. F. Brown, D. O. Ohlen, Z. Zhu, L. Yang & J. W. Merchant. (2000). Development of a global land cover characteristics database and IGBP DISCover from 1 km AVHRR data. International Journal of Remote Sensing, 21(6), 1303. Margules, C. R. & R. L. Pressey. (2 000). Systematic c onservation planning. Nature, 405(6783), 243. MEA. (2005). Millennium Ecosystem Assessment: Ecos ystems and human well-being: Current state and trends: Island Press. Medina, A. A. (2007). Reconstruccin de los regmenes de fuego en un bosque de Prosopis caldenia provincia de La Pampa, Argentina. Bosque, 28(3), 234. Mendez, J. L. (2007a). Primer inventario nacional de bosques nativos. Segunda etapa, inventario de campo de la regin del Espinal: Distritos del caldn y andubay. Direccin de nacional de bosques. Buenos Aires. pp. 236. Mendez, J. L. (2007b). Primer inventario nacional de bosques nativos. Segunda etapa, inventario de campo de la regin del espi nal: distritos del caldn y andubay. Anexo 1: Estado de conservacin del Caldenal Direccin de nacional de bosques. Buenos Aires. pp. 86. Mittermeier, R. A., C. G. Mittermeier, T. M. Brooks J. D. Pilgrim, W. R. Konstant, G. A. B. da Fonseca & C. Kormos. (2003). Wilderne ss and biodiversity conservation. Proceedings of the National Academy of Sciences of the United States of America, 100(18), 10309. Mittermeier, R. A., N. Myers, J. B. Thomsen, G. A. B. da Fonseca & S. Olivieri. (1998). Biodiversity hotspots and majo r tropical wilderness areas: Approaches to setting conservation priorities. Conservation Biology, 12(3), 516. Nagendra, H. (2001). Using remote sensing to assess biodiversity. International Journal of Remote Sensing, 22(12), 2377. Narumalani, S., D. R. Mishra, R. G. Rothwe ll & Vf. (2004). Change detection and landscape metrics for inferring anthropogenic pro cesses in the greater EFMO area. Remote Sensing of Environment, 91(3), 478. NASA. (2009). Shuttle Radar Topography Mission. Available: http://www2.jpl.nasa.gov/srtm/ Ojim a, D. S., K. A. Galvin & B. L. Turner (1994). The Global Impact of Land-Use Change. Bioscience, 44(5), 300. 51

PAGE 52

Olson, D. M., E. Dinerstein, E. D. Wikram anayak e, N. D. Burgess, G. V. N. Powell, E. C. Underwood, J. A. D'Amico, I. Itoua, H. E. Strand, J. C. Morrison, C. J. Loucks, T. F. Allnutt, T. H. Ricketts, Y. Kura, J. F. Lamoreux, W. W. Wettengel, P. Hedao & K. R. Kassem. (2001). Terrestrial ecoregions of the worlds: A new map of life on Earth. Bioscience, 51(11), 933. Camara de Diputados de la Provincia de La Pampa. (1973). Fijando normas para el fracionamiento de predios rurales. Ley 468. Petit, C., T. Scudder & E. Lambin. (2001). Qu antifying processes of land-cover change by remote sensing: resettlement and rapid la nd-cover changes in south-eastern Zambia. International Journal of Remote Sensing, 22(17), 3435. Redford, K. H. (1992). The empty forest. Bioscience, 42(6), 412. Rissman, A. R., L. Lozier, T. Co mendant, P. Kareiva, J. M. Kiesecker, M. R. Shaw & A. M. Merenlender. (2007). Conservation easements: Biodiversity protection and private use. Conservation Biology, 21(3), 709. Rissman, A. R. & A. M. Merenlender. (2008). The Conservation Contributions of Conservation Easements: Analysis of the San Francisco Bay Area Protected Lands Spatial Database. Ecology and Society, 13(1), 25. Rodrigues, A. S. L., H. R. Akcakaya, S. J. Ande lman, M. I. Bakarr, L. Boitani, T. M. Brooks, J. S. Chanson, L. D. C. Fishpool, G. A. B. Da Fonseca, K. J. Gaston, M. Hoffmann, P. A. Marquet, J. D. Pilgrim, R. L. Pressey, J. Schipper, W. Sechrest, S. N. Stuart, L. G. Underhill, R. W. Waller, M. E. J. Watts & X. Yan. (2004). Global gap analysis: Priority regions for expanding the gl obal protected-area network. Bioscience, 54(12), 1092. Sanchez-Azofeifa, G. A., G. C. Daily, A. S. P. Pfaff, C. Busch & Nm. (2003). Integrity and isolation of Costa Rica's national parks and biological reserves: examining the dynamics of land-cover change. Biological Conservation, 109(1), 123. Sanchez-Azofeifa, G. A., A. Pf aff, J. A. Robalino & J. P. Boomhower. (2007). Costa Rica's payment for environmental services program: Intention, implementation, and impact. Conservation Biology, 21(5), 1165. Sanchez-Azofeifa, G. A., C. Quesada-Mateo, P. Gonzalez-Quesada, S. Dayanandan & K. S. Bawa. (1999). Protected areas and conservation of biodi versity in the tropics. Conservation Biology, 13(2), 407411. Sanchez-Azofeifa, G. A., B. Rivard, J. Cal vo & I. Moorthy. (2002). Dynamics of tropical deforestation around national parks: Remote sens ing of forest change on the Osa Peninsula of Costa Rica. Mountain Research and Development, 22(4), 352. Shaffer, J. P. (1995). Multiple Hypothesis Testing. Annual Review of Psychology, 46, 561. 52

PAGE 53

Siwe, R. N. & B. Koch. (2008). Change vector analysis to categorise land cover change processes using the tasselled cap as biophysical indicator. Environmental Monitoring and Assessment, 145(1), 227. Soul, M. E., J. Terborgh & Pd. (1999). Conservi ng nature at regional and continental scales a scientific program for North America. Bioscience, 49(10), 809. Southworth, J., D. Munroe & H. Nagendra. (2004a). Land cover change and landscape fragmentation comparing the utility of continuous and disc rete analyses for a western Honduras region. Agriculture Ecosystems & Environment, 101(2), 185. Southworth, J., H. Nagendra, L. A. Carlson & C. Tucker. (2004b). A ssessing the impact of Celaque National Park on forest fragmentation in western Honduras. Applied Geography, 24(4), 303. Stoner, C., T. Caro, S. Mduma, C. Mlingwa, G. Sabuni & M. Borner. (2007). Assessment of effectiveness of protection strategies in Ta nzania based on a decade of survey data for large herbivores. Conservation Biology 21 (3), 635. Subsecretara de Asuntos Agrari os, Gobierno de la Provincia de La Pampa. (2004). Cotos de caza. Disposicin 07/04. Tallis, H., P. Kareiva, M. Ma rvier & A. Chang. (2008). An ecosystem services framework to support both practical conservation and economic development. Proceedings of the National Academy of Sciences of the United States of America, 105(28), 9457. Terborgh, J. (2004). Reflections of a sc ientist on the Worl d Parks Congress. Conservation Biology, 18(3), 619. Terborgh, J., G. Nunez-Iturri, N. C. A. Pitman, F. H. C. Valverde, P. Alvarez, V. Swamy, E. G. Pringle & C. E. T. Paine. (2008). Tree recruitment in an empty forest. Ecology, 89(6), 1757. Tole, L. (2006). Choosing reserve sites pr obabilistically: A Colombian Amazon case study. Ecological Modelling, 194(4), 344. Tomlinson, K. W., J. W. Hearne & R. R. Alexander. (2002). An a pproach to evaluate the effect of property size on land-use options in semi -arid rangelands. Ecological Modelling, 149, 85. Turner, B. L., E. F. Lambin & A. Reenberg. (2007). The emergence of land change science for global environmental change and sustainability. Proceedings of the National Academy of Sciences of the United States of America, 104(52), 20666. Turner, W., S. Spector, N. Gardiner, M. Fladeland, E. Sterling & M. Steiniger. (2003). Remote sensing for biodiversity sc ience and conservation. TRENDS in Ecology and Evolution, 18(6), 306. 53

PAGE 54

Vanclay, J. X. (2001). The effectiveness of parks. Science, 293(5532), U1U2. Villalobos, A. E. d., D. V. Pelaez & O. R. Elia. (2005). Growth of Prosopis caldenia Burk : seedlings in central sem i-ari d rangelands of Argentina. Journal of Arid Environments, 61, 345. Vogt, N. D., A. Y. Banana, W. Gombya-Ssemb ajjwe & J. Bahati. (2006). Understanding the stability of forest reserve boundaries in the West Mengo region of Uganda. Ecology and Society, 11(1), 38. Weller, T. J. (2008). Using occupancy estimatio n to assess the effectiveness of a regional multiple-species conservation plan: Bats in the Pacific Northwest. Biological Conservation, 141(9), 2279. Williams, C. E. & M. E. Lathbury. (1996). Ec onomic incentives for habitat conservation on private land: Applications to the Inland Pacific Northwest. Wildlife Society Bulletin, 24(2), 187. Wilshusen, P. R., S. R. Brechin, C. L. Fort wangler & P. C. West (2002). Reinventing a SquareWheel: Critique of a Resurgent ``Protection Paradigm in International Biodiversity Conservation. Society and Natural Resources, 15, 17. Wiseman, R., B. R. Page & T. G. O'Connor. (2004). Woody vegetation change in response to browsing in Ithala Game Reserve, South Africa. South African Journal of Wildlife Research, 34(1), 25. Wright, S. J., G. A. Sanchez-Azofeifa, C. Po rtillo-Quintero & D. Da vies. (2007). Poverty and corruption compromise tropical forest reserves. Ecological Applications, 17(5), 1259 1266. 54

PAGE 55

55 BIOGRAPHICAL SKETCH Mariano Gonzalez Roglich was born in Allen, Rio Negro, Argentina in September of 1980. Since he was a kid, he enjoyed exploring the most diverse landscapes of Patagonia, and that is how his love for nature and his interest in c onservation of those pristin e landscapes emerged. In 2005, he graduated with a Bachelor of Science from La Pampa National University. In 2006, he earned a degree of Expert in Geographic Informa tion Systems from the International University of Andalucia, Spain. In 2009, he earned his Master of Science in interd isciplinary ecology from the University of Florida, US. He will pursue a do ctoral degree from Duke University, starting in fall 2009. He plans to build a career as researcher and conservationi st, going back to Argentina to work toward the conservation of that beautiful and pristine land that in spired him as a child.