Indigenousacculturation and Deforestation in Kavanayén, Gran Sabana, Venezuela Interciencia, Vol

Interciencia ISSN: 0378-1844 [email protected] Asociación Interciencia Venezuela

Kingsbury, Nancy D. Impacts of land use and cultural changein a fragile environment: indigenousacculturation and deforestation in Kavanayén, gran sabana, Venezuela Interciencia, vol. 26, núm. 8, agosto, 2001, pp. 327-336 Asociación Interciencia Caracas, Venezuela

Available in: http://www.redalyc.org/articulo.oa?id=33905903

How to cite Complete issue Scientific Information System More information about this article Network of Scientific Journals from Latin America, the Caribbean, Spain and Portugal Journal's homepage in redalyc.org Non-profit academic project, developed under the open access initiative IMPACTS OF LAND USE AND CULTURAL CHANGE IN A FRAGILE ENVIRONMENT: INDIGENOUS ACCULTURATION AND DEFORESTATION IN KAVANAYÉN, GRAN SABANA, VENEZUELA

NANCY D. KINGSBURY

o even the most casual eration is occurring primarily as a result where forest has been replaced by sa- observer, it appears that of declining soil fertility and chemical vanna. Rather, this theory has been pro- wide-scale deforestation and hydrological stress. This process has posed by connecting various observations: has recently occurred in the Gran Sabana. also been tied to shifting cultivation. It accounts and observations of deforesta- For the Pemón Amerindians, who have oc- has been suggested (Dezzeo, 1994) that tion and savanna extension with analyses cupied the Gran Sabana for at least sev- after a field is cut in primary forest, “an that indicate acidic soils, low soil fertility eral centuries, the forests are the primary authentic, although not very vigorous sec- and high levels of soil aluminum, and; natural resource. They are the location of ondary succession” takes place. However observations of intensified shifting culti- their conucos (shifting cultivation fields) after a second round of cultivation, the vation. and the source of game animals. Given the successional vegetation is either “mato- Previously no commu- extreme infertility of savanna soils (Kings- rral” (“scrub” or “brush”) or a secondary nity specific data on the rate of forest re- bury 1999) and their unsuitability for cul- forest in transition to matorral due to an establishment or forest deflection to sa- tivation, deforestation thus has potentially unspecified “degradation in the regenera- vanna after cultivation or on changes in profound impacts on Pemón subsistence, tion process.” vegetation cover and land use over time as well as the obvious ecological impacts It has also been pro- have been provided to support these as- such as loss of biodiversity. The larger posed (Dezzeo, 1994; Elcoro and Vera, sertions. This paper presents such data Venezuelan society is also affected as the 1986; Morales and Gorzula, 1986) that and examines the role of agriculturalists efficiency of other resource use in the re- forest displacement by savanna is a result in the changing ecology of the region and gion, such as the generation of hydro-elec- of the region’s “ecological fragility” the role of changes in indigenous culture tric power, requires stability of the hydro- which is principally related to high sus- –most notably settlement patterns– on the logical regime that results from the overall ceptibility to erosion and edaphic condi- indigenous system of agriculture. maintenance of forest cover. tions such as highly acidic soils, low con- As part of a larger study While some of the de- centrations of soil nutrients and low Ca/ on the influence of human settlement, forestation –e.g., numerous dead, black- Al ratios. Taken together, it is suggested population growth, shifting cultivation ened tree stumps in fields of otherwise that these factors result in marginal and forest ecology in the Gran Sabana pure savanna– is likely the result of cata- chemical and hydrological conditions (Kingsbury, 1999), changes in land use strophic fires (Kingsbury 1999), it has (Elcoro and Vera, 1986) resulting in a and vegetation cover over a ten-year pe- been asserted that other deforestation is slow rate of forest re-establishment after riod (1985 to 1995) were mapped and the result of a combination of the disturbance (Morales and Gorzula, 1986) analyzed in areas of intensive cultivation region’s particular ecological characteris- and frequent forest transformation to sa- near the central Gran Sabana community tics, identified by some authors as its vanna. of Kavanayén. “ecological fragility” and as a result of The evidence for attrib- intensified shifting cultivation by the in- uting the slow rate of forest re-establish- Location of the Study Area and digenous population of the region, the ment to marginal conditions is indirect. General Environmental Characteristics Pemón. There have been few surveys of disturbed Fölster (1986) has sug- sites and no comparisons of edaphic con- Part of the Guiana Shield, gested that spontaneous vegetation degen- ditions in undisturbed and disturbed sites the Gran Sabana is a plateau region in

KEYWORDS / Shifting Cultivation / Pemón Amerindians / Deforestation / Acculturation / Recibido:13/02/2001. Modificado: 06/07/2001. Aceptado: 19/07/2001

Nancy D. Kingsbury. Ph.D. in Geography, York University, Canada. Post-Doctoral Fellow, American Association for the Advancement of Science. Address: 5759 Avenue de l’Esplanade, H2T 2Z9 Montreal, Canada. e-mail: [email protected].

AUG 2001, VOL. 26 Nº 8 0378-1844/01/08/327-10 $ 3.00/0 327 the southeastern corner of Bolívar state of Venezuela (Figure 1). The southern edge is on the Brazilian border and the northern limit is formed by the Sierra de Lema mountain system. The northeastern edge is the border with Guyana and the western boundary is the Caroní” River. A large portion of the Gran Sabana is in the Canaima National Park. Almost the entire Gran Sabana is located within the Ca- roní” river basin which provides water for the Guri Dam, the largest hydro-electric project in the country and the second largest in the world. The dam provides 72 per cent of Venezuela’s domestic hydro- electric power (Miranda et al., 1998) and is managed by the national hydro-electric corporation, Electrificación del Caroní (EDELCA). Kavanayén is located at 5o35’N, 61o47’W (Figure 1). Average annual temperatures vary between 20 and 25oC; annual precipitation varies between 1800 and 2500mm (Kingsbury, 1999). El- evations are between 1000 and 1200m above sea level (Elcoro and Vera, 1986), with the “town” higher on the plateau and the major areas of cultivation in the valleys a few hundred meters below. The direction of the wind is generally from the East (Northeast and Southeast) to West and reaches velocities of up to 20km/h (Elcoro and Vera, 1986). Annual relative humidity is between 45 and 96%. Soil fertility in the region is strongly influenced by the under- lying geology. Located on a sandstone plateau, soils in the community of Kava- nayén are generally very sandy and infertile (Kingsbury and Kellman, 1997). As indicated in Table I, comparisons of surface soils below undisturbed old growth forests in Kavanayén with those in the more southern Gran Sabana community Figure 1. Gran Sabana, Venezuela. Location of study site and sites mentioned in the text. of Monte Bello (see Figure 1) indicate the relative infertility of soils in the former area. Kavanayén were adjacent to areas previ- Conucos are traditionally General characteristics ously cultivated and the sample may cut in the forest fragments, preferably of old and secondary growth forests in therefore be biased in favour of more fer- medium-high forests (Dezzeo, 1994) adja- the Gran Sabana are presented in the fol- tile sites). cent to the open savanna where settle- lowing Table II. ments are located. With the exception of High forests were de- Pemón Agriculture the centuries-old introduction of steel fined by Hernández (1994) as those with tools, the tool inventory and labor in- canopy heights >18m; medium-high for- Traditional agricultural practices volved in agriculture has remained re- ests are those with canopy heights be- markably similar for centuries. There tween 15 and 18m. These two forest Pemón subsistence is tra- have been, however, some notable types can be taken as synonymous with ditionally based on shifting cultivation. changes in agricultural site selection “old growth forest” as defined in this Like other horticultural societies of the characteristics in some Pemón communi- study. Low forest was defined by Hernán- Orinoco Basin (Harris, 1969), Pemón agri- ties. dez (1994) as forest with canopy heights culture is characterized by a reliance on Earlier studies of the <14m and can be taken as approximates fast-growing tubers: yuca, sweet potatoes Pemón stated that site selection for new to secondary growth forests in Kavana- and yams. Criollo foodstuffs are available conucos was “dictated exclusively on yén. Higher basal areas in old growth for sale in some communities and wage pragmatic grounds” (Urbina, 1979). Sites forests in Kavanayén as compared to workers there purchase more food than do had to be well-drained, preferably near those in the southern Gran Sabana may non-wage workers. However, cultivation water and free of large quantities of sur- reflect sampling bias (forests surveyed in remains the primary means of subsistence. face or shallow roots which make plant-

328 AUG 2001, VOL. 26 Nº 8 TABLE I the Gran Sabana by scientists (Elcoro and SURFACE SOIL CHEMICAL PROPERTIES (TOP 7cm). KAVANAYÉN AND Vera, 1986; Fölster, 1986). A possible ex- MONTE BELLO, GRAN SABANA planation for this discrepancy is that, for the Pemón, a sufficient fallow period is Kavanayén (n=41) Monte Bello (n=24) one where the forest has recovered to the mean (+- SD) mean (+- SD) point where it can be cut again and pro- duce an adequate crop, while for the sci- pH 4.19 (+-2.31) 5.06 (+-0.78) entists it typically means when a second- ECEC (cmol[+]/kg) 1.69 (+-0.59) 11.91 (+-15.69) ary forest similar in structure and species composition to the primary forest has be- Extractable cations (cmol[+] /kg) come established. As other researchers (for Ca 0.70 (+-0.06) 8.68 (+-12.09) example Azuaje, 1986) have found, ques- Mg 0.09 (+-0.05) 2.11 (+-3.46) tioning the Pemón about the length of an K 0.16 (+-0.08) 0.34 (+-0.25) “appropriate” fallow period is problem- Al 1.36 (+-0.54) 0.51 (+-0.70) atic due to difficulties in translation of P (ug/g)* 2.98 (+-2.23) 23.23 (+-26.73) both words and concepts. The Pemón answer is, typically, “depende” (“it de- *Resin-extractable pends”). When pressed for a quantifiable response, Pemón will answer that a forest is ready to be cut again when the trees ing difficult. Pemón also preferred to lo- months (Im Thurn, 1883). Three harvests are “como un hombre” (“like a man” or cate their settlements in flat areas of sa- from a single conuco would then mean a “as big as a man”), meaning the diam- vanna, near a year-round water source, period of active cultivation of 27 to 30 eter, as opposed to the height, of an adult and with easy access to forests (Urbina, months at the shortest, and 36 to 40 man. Except on very fertile tropical soils, 1979). Once a good settlement location months at the longest. it requires much more than the 15 to 26.5 was found, nearby forest fragments would The length of fallow years (the fallow period currently prac- be used intensively until adequate practiced by the Pemón varies between 5 ticed by the Pemón) in order for trees to amounts of preferred medium-high forest and 15 years in the southern part of the attain such diameters. no longer remained. Gran Sabana (Azuaje, 1986), to 15 years During the slow process The period of active cul- in the central part near Tuaukén (Urbina, of forest re-establishment after cultiva- tivation (planting, weeding and harvest- 1979), and 12 to 26.5 years in Kavanayén tion, light-demanding herbaceous species ing) for an individual conuco varies from in the northern part of the Gran Sabana (weeds and grasses), stump sprouts (cop- 2 to 3 years. Though a possible response (Elcoro and Vera, 1986). As Azuaje notes, pice) from trees that survived the initial by a shifting cultivator to declining forest actual fallow periods are dependent on burn and fast-growing pioneer tree spe- resources is to extend the cultivation the recuperative capacity of the forest cies first establish. The vegetation mix is cycle, the length of cultivation in the and on land use pressure. His informants slowly transformed as tree species form Gran Sabana appears to have remained reported that a fallow was ready to be young and then old secondary forests constant over more than 100 years. In the recultivated when there were no shrubs or (Nye and Greenland, 1960), asserting 1880s in the Roraima region, Im Thurn grass and one could see clearly - there dominance and excluding weedy herbs. In (1967, reprint of 1883 edition) observed was little “tangled” vegetation in the site, the absence of other disturbances, such as that “a field is deserted after three or in which case it was described as being fire, a regional forest flora gradually be- four crops have been taken from it; and a “poco enmarañado” (Azuaje, 1986). comes dominant again and soil nutrient new clearing is then made and planted.” These actual fallow peri- levels approach pre-disturbance levels. The length of time observed between ods are significantly shorter than the 50 In stable systems of planting and harvesting was nine to 10 to 100 years suggested as necessary in shifting cultivation forest is not replaced by savanna during the fallow period. The factors that exert a controlling influence TABLE II over forest succession after a disturbance FOREST CHARACTERISTICS, GRAN SABANA such as cultivation include the availability of regenerating germplasm (e.g. sprouting Forest type Height DBH Basal Area Number stumps, tree seed banks), of microhabitats m mean cm m2/hectare of sites for plant establishment, and of nutrients for plant growth (Uhl, 1987). Site selection character- Southern Gran Sabana Forests* istics and land use practices, as well as High 19.50 21.50 46.10 10 pre-existing resource endowments (i.e., Medium-high 17.10 19.70 29.30 8 forest characteristics and ecology; soil Low 12.60 13.50 9.50 6 fertility), all influence these factors. Pemón prefer to culti- Kavanayén Forests ** vate in the same areas that were previously cultivated by their family (parents Old Growth 15.41 20.05 64.38 3 and grandparents). This produces an in- Secondary Growth 11.90 12.15 24.99 4 formal type of hereditary land rights sources: * Hernández, 1994 within a communal property rights sys- ** Kingsbury, 1999 tem.

AUG 2001, VOL. 26 Nº 8 329 All land is held in com- TABLE III mon by the community. Current use (ac- SURROUNDING VEGETATION OF CONUCOS IN THE KAVANAYÉN VALLEY tive conucos) confer exclusive land use 1985 TO 1995 rights for the family that cuts and culti- vates the site. While it may be an indi- high forest 1 side 2 sides 3 sides all sides none vidual husband-wife pair that expends the vast majority of labor in site preparation Kavanayén C. 1985 12.1% 12.1% 9.1% 21.2% 45.4% and cultivation, use rights (to the prod- Kavanayén C. 1995 17.3% 28.8% 34.6% 7.7% 11.5% ucts of the conuco) also commonly extend to the couple’s parents, siblings and low forest 1 side 2 sides 3 sides all sides none their children (the primary couple’s nieces and nephews). When a couple de- Kavanayén C. 1985 18.2% 12.1% 9.1% 39.4% 21.2% cides to cut a new conuco, they usually Kavanayén C. 1995 23.1% 15.4% 7.7% 3.8% 50.0% do so near conucos of other family mem- bers and this fits in with the mixture of Kavanayén C. 1985 (n=33); Kavanayén C. 1995 (n=52); Kavanayén R. 1995 (n=46) rights accorded to individual couples and individual extended family groups. TABLE IV These individualized PEMÓN POPULATION IN VENEZUELA rights within the community property system also extend through the fallow period, as the collection of older crops and 1950 4,054 (Colson, 1971) wood for cooking fuel are also tied to the 1960 2,700 (Venezuelan National Census 1960 in Wilbert, 1971) family that previously cultivated the site. 1970 4,000 (Thomas, 1973) Once a site is fallowed, it reverts to the 7,000 (Monsonyi, 1972) community and “theoretically anyone 1982 11,462 (República de Venezuela, 1985) may cultivate it when it regenerates suffi- 1992 18,871 (República de Venezuela, 1993, Tomo I) ciently to sustain cultivation” (Thomas, 1982). The situation is more complex, due to the preference of relatives for As a Capuchin mission pressures and different economic opportu- maintaining fields close to each other and and educational centre, the growth of Ka- nities began to encourage communities of the practice of returning to recultivate vanayén is illustrative of the population several hundreds” (Colson, 1985). The one’s former fields. agglomeration that occurred around all largest Pemón community, Kavanayén, Capuchin centres. After the founding of had an estimated average population in Contemporary agricultural practices: the Mission in Kavanayén in 1942, the 1995 of approximately 925, densely con- The role of cultural change Capuchins encouraged Pemón in small centrated around the community centre communities around Kavanayén to send (the Capuchin Mission). In the national Pemón traditionally had their children to the Mission's school. Re- Indigenous Census of 1992, the com- a very diffusely-distributed population of ligious Orders in Venezuela were offi- munity’s settlement pattern is described low local density and small settlements. cially delegated the responsibility for the as “non-dispersed” (República de Venezu- In the 1970s, Thomas (1973) reported an formal education of indigenous people. In ela, 1993, Tomo II). overall density of 1 person per 4.1km2, 1945 (three years after the founding of Village fissioning no while an earlier source reported an aver- the Mission) there were 48 internados longer occurs also due to the attraction of age of one person per 3.8km2 (Beard, (children resident in the school whose services (schools, clinics, etc.) in the de- 1953 in Means, 1995). parents did not live in Kavanayén). In marcated indigenous communities, and the Compared with other 1995, however, their number (in this case restrictions on mobility imposed by the South American indigenous peoples, it all were girls) had fallen to 30 (Mission presence of non-indigenous settlements was noted that the Pemón “may be singu- Records, Sta. Teresita de Kavanayén). Af- and activities (hydro-electric projects, min- lar in their degree of dispersion” (Tho- ter the establishment of the Mission ing operations). Pressures against village mas, 1982). This can be seen as a his- school, many parents permanently moved fissioning exist at a bureaucratic level as toric adaptation to the non abundant re- to Kavanayén from other areas. To a well. Only demarcated, recognized indig- sources. The degree of dispersion began lesser extent, the migration to Kavanayén enous communities are entitled to govern- to change during the late nineteenth and continues. ment services such as school supplies and early twentieth centuries when the spread In the early part of the teachers, building materials and visits of the syncretic Hallelujah religion and XX century, Pemón communities were from medical personnel. other religious movements brought people substantially smaller. In the late 1920s, While communities and together in central settlement locations Tate visited one typical Pemón commu- residences are now permanent, the nature for religious ceremonies (Thomas, 1973). nity, the village of Tekwanno near Mount of shifting cultivation requires that fields Since the 1930s, the missionary activities Roraima (see Figure 1), which consisted are non-permanent. However, given that of the Capuchins and the Seventh Day of “about 6 houses inhabited by a fluctu- the forests around settlements are limited, Adventists have resulted in the “agglom- ating population, varying from twenty to the result is that land use has necessarily eration of populations surrounding mis- forty” (Tate, 1930). The traditional Pe- become more intensive. sion sites at Santa Elena, Kavanayén, món response to increases in localized Following the increasing Kamarata and Uonkén” (Thomas, 1973). population density was the establishment trend to sedentarism, increasing size of Similar agglomeration occurred around of new, smaller communities. The small communities and increasing influence of Adventist communities in the southern and mobile nature of communities began factors which favor cultivation of less part of the Gran Sabana. to change in the 1960s when “national distant sites, the distance from primary

330 AUG 2001, VOL. 26 Nº 8 TABLE V a 45min walk from the community cen- KAVANAYÉN POPULATION ter, this wage dependent site selection characteristic (distance from community 1982 1985 1992 1995 center), as well as the overall preference to locate fields so that services in the Kavanayén 3711 4752 6513 850-10004 community center are still accessible on a daily basis, has significant implications References: 1 República de Venezuela, 1985; 2 Elcoro and Vera, 1986; 3 República de Venezuela, 4 for the overall sustainability of the agri- 1993, Tomo I; pers. comm., V. Contasti, Padre Jesús, 1995 cultural system and the maintenance of the forest cover, or alternately, the loss of TABLE VI forest. PER ANNUM POPULATION INCREASES, KAVANAYÉN * Pemón Demographics. Overall and Kavanayén 1982 - 1992 1992 - 1995 Kavanayén 2.8% 6.63%** As seen in Tables IV, V and VI, population growth rates in Kava- * assuming constant linear growth for the periods in question nayén, like for the overall Pemón popula- ** based on a 1995 population estimate of 850 tion in Estado Bolívar, are high. While there are clearly errors in some of the data that appears in residences to fields has become shorter - true vegetation types– are not subject to Table IV (either the 1950 estimate was in 1995, conucos were located an average changes in the fallow period. Only data too high or the 1960 and lower 1970 esti- of 112.4min walk away while fallows of from conucos in the valleys in front on the mates were too low) the Pemón popula- 5 years or more were an average of downtown area of Kavanayén, the area tion is growing rapidly, both as a result 137.23min away, for a decrease of mapped by Elcoro and Vera in 1985, is in- of natural population increase and immi- 49.6min round-trip (Kingsbury, 1999). cluded in Table III. gration from neighbouring areas. The centre of the community (around the In 1985, more conucos Capuchin Mission) has remained un- were cut in the middle of high forest Vegetation and Land Use Changes changed since Kavanayén was founded (21.2% of fields with high forest on all in Kavanayén, 1985 to 1995 and primary residences have always been sides compared to 7.7% in 1995). The located in this general area. use of low forest has also declined dra- In 1985, Elcoro and There were only 30.5% matically. Vera (1986) mapped the vegetation and of rastrojos one hour or less away from While the practice of land use in the two major areas of culti- the center of town while 34.6% of cutting conucos adjacent to savanna is vation in Kavanayén - the east-west and conucos were located this close to town. still rare, their proximity to savanna has north-south valleys below the cliff in At two hours distance from town, the increased. Even assuming that some of front of the “downtown” area of the com- change in proximity of fields is even the savanna near rastrojos is secondary munity. Area 1 covers approximately more marked - 43.3% of rastrojos were grassland that established in the fallow 2496 hectares; Area 2 covers approxi- more than two hours walk from primary period, a smaller percentage of rastrojos mately 1833ha. Using a combination of residence, while only 34.5% of conucos have nearby savanna than do current extensive aerial photography and ground were located this far from the center of conucos (41.7% of rastrojos are ≤100m surveys, these researchers distinguished the community (Kingsbury, 1999). While from savanna compared to 62% of among a variety of vegetation types (in- this at first appears counter-intuitive - conucos). The average distance to sa- cluding several types of forests) and vari- there is more old growth forest further vanna has also decreased – savanna was ous land uses such as pastures, conucos from the center of the community - it re- an average of 531m from rastrojos and and casitas. In 1995, the same areas were flects a change in site selection prefer- only 369m from conucos (Kingsbury, re-mapped in order to determine what ences over time. This change in field lo- 1999). changes in forest cover and land use had cation then reflects a change in prefer- In Kavanayén, notable occurred in the 10-year period. Through ence for proximity to residence over type differences in site selection and agricul- reference to particular vegetation types in of forest (high or low) for cultivation. tural participation can also be observed the 1985 maps, the extent of some pre- It is clear from the data between Pemón who do and do not par- 1985 deforestation can also be deter- in Table III that there have been changes ticipate in wage labor (either on a perma- mined. in the surrounding vegetation types and nent or seasonal basis), with waged border characteristics of fields. Data drawn workers locating their conucos closer to Land use and vegetation type mapping from from Elcoro and Vera’s (1986) “Land the community center where formal em- methods Use and Vegetation Maps” which indicated ployment is located (Kingsbury, 1999). the vegetation type around the conucos Approximately 10% of the adult popula- Elcoro and Vera (1986) surveyed in 1985, is listed as “Kavanayén tion engages in full-time permanent or distinguished several vegetation types and C. 1985” and compared to similar data seasonal wage labor with a smaller per- land uses including: from the 1995 conuco surveys done in the centage involved in casual or part-time same areas in this study, labeled “Kavana- economic activities such as the sale of - high forest (canopy height ≥30m) yén C. 1995”. Of the latter-most data set, foodstuffs and the repair and sewing of - medium forest (canopy height ≥25m) only rastrojos with water or cliff borders clothes (Kingsbury, 1999). - low forest (canopy height <25m) were included. This smaller group was Given that no extensive - old secondary forest rastrojo chosen as these types of borders –unlike tracts of old growth forests remain within (>15 years)

AUG 2001, VOL. 26 Nº 8 331 the 1985 maps, new 1995 maps indicat- ing changes in vegetation and land use were then produced. Areas of new secondary vegetation (plain savanna, treed savanna and matorrales) were measured using a grid overlay of squares of 1mm2 and counting the number of squares covered by each area; following the map scale, measurements were then converted to m2 and then to ha. Areas of tree savanna indicated on the 1985 maps were similarly measured. Elcoro and Vera’s 1985 maps have been modified here to show the following, more general, vegetation types and land uses: Figure 2. Kavanayén Land Use and Vegetation, Area 1, 1985. - primary forest (includes gallery forest, high, medium and low forests) - secondary forest (includes old and - young secondary forest rastrojo trees in this type of savanna (scattered young secondary forests) (<15 years) living trees or dead, burnt stumps) is evi- - matorrales (all types) - high matorral (a vegetation mix of dence that the savanna is a secondary - savanna (with no trees) small trees, shrubs, ferns and grasses grassland. The few trees indicate that - tree savanna (sabana arbolada) with an average height of 1 to 3m) while some trees can establish in second- - conucos (active conucos and forest cut - low matorral (<1m in height) ary grasslands, or survive from the previ- for conucos) - edaphic matorral ous forest, a vigorous secondary forest - casitas. - pastures with a closed canopy cannot establish, - savanna particularly if the area is subject to fre- Comparisons of the - tree savanna (burnt tree stumps in the quent combustion. 1985 maps (Figures 2 and 4) and 1995 middle of savanna grass, or areas dom- After consultation with maps (Figures 3 and 5) show the extent inated by grass with the occasional one of the authors (N. Vera, pers. of deforestation and degradation of veg- short tree) comm.), the 1985 maps were redrawn to etation that has occurred in the 10-year - bosque talado (forest cut and cleared show changes in vegetation cover and period. The 1995 maps show only new for a conuco) land use. Dividing each of the original areas of plain savanna, treed savanna and - cultivated areas (conucos) maps into three sections, from various matorrales; vegetation types that were - casitas. points on the cliff top (1200 m.a.s.l.) easily identifiable. The distinction be- overlooking the valleys a few hundred tween the two types of savanna was Area 1 (East-West orien- meters below, and using large areas of made following the definitions used by tation) covers the valley directly in front savanna from the 1985 maps as initial Elcoro and Vera (1986): “tree savanna” is of the “downtown”, stretching from in reference points, changes were sketched grassland with burnt tree stumps or small front of the airport to the extreme west on top of the original maps. A visual trees and bushes in small or isolated (see Figure 2), to the end of the valley comparison of the map size of the refer- groups; “savanna” is pure grass and other (which ends in a cliff) to the east. In ence points to the actual size of the areas herbaceous vegetation. All the areas indi- 1985, there was a smaller percentage of was used as a guide for calculating the cated on these maps established in what high forest in this area than in Area 2. approximate size of areas of new vegeta- had previously been forest or a forest- This was interpreted as a consequence of tion. After these areas were drawn onto matorral mix (see Appendix I for a com- the past intensity of use and the concentration of conucos (Elcoro and Vera, 1986). Area 2 (North-South ori- entation) covers the valley west of the airport (see Figure 4). This is further from the downtown of Kavanayén and in 1985 vegetation was in the following order of descending proportion: high forest, late secondary forest, early secondary forest, high matorrales, medium-high forest, low forest, low matorrales, savanna (Elcoro and Vera, 1986). The 1985 maps can also be analyzed for data on pre-1985 deforestation. Vegetation identified in 1985 as “Sabana Arbolada” and matorrales are definitive evidence of earlier forest displacement. In the former, the presence of Figure 3. Kavanayén Land Use and Vegetation Changes, Area 1, 1995.

332 AUG 2001, VOL. 26 Nº 8 and 5) show significant increase in plain savanna and a consequent reduction of forest (both primary and secondary). There are also numerous new areas of tree savanna. In Section 1 (Southeast), savanna has extended into primary forest and has also extended into some 1985 rastrojos where secondary forest was establishing. There has also been a significant increase in matorrales in Section 1 from 1985 to 1995. In this Section there are numerous areas of matorrales and secondary forest within the primary forest fragments. There are also scattered dead trees within these fragments. Savanna has replaced Figure 4. Kavanayén Land Use and Vegetation, Area 2, 1985. high and low matorrales in Section 2. There has also been a significant extension of 1985 savannas. These are now plete list of degraded vegetation types in bordered by matorrales and scrub vegeta- patches of savanna with burnt tree stumps Areas 1 and 2). The 1995 maps show the tion. In 1985 by contrast, the forest-sa- as well as other numerous areas of treed extent of forest displacement by savanna vanna border was distinct. In Section 3, savanna in the middle of primary forest and matorrales and the extent of degrada- there is evidence of fire incursion into (Section 2, East). These are perhaps ras- tion from a forest-matorral mix to primary forest fragments (i.e. dead trees trojos where no forest re-establishment matorral-savanna mix. within forest fragments). has occurred, the burnt tree stumps hav- Conucos and casitas were In Area 1, savanna has ing been cut and burnt for the conuco. indicated on the map in their approximate established in high and low forest as well In 1985 in Area 2, there location. Due to both topography and the as in forest-matorrale mosaics. There are were five large areas of savanna Ar- height of high forest, the view of a num- 12 new areas of plain savanna covering bolada covering approximately 38ha (of a ber of casitas and conucos was likely ob- an area of approximately 60.9ha. In 6 of total map area of 1833ha). In 1995, there structed. As well, from the cliff top, dis- these areas, old and secondary forest was were two new areas of plain savanna cov- tant active conucos covered in mature displaced; in the 6 other areas, savanna ering approximately 15.23ha in areas that yuca and other crops can often be diffi- has now established in what was formerly in 1985 were forested with primary and cult to distinguish from young secondary a forest-matorrale mix. Two areas of new secondary forest; one area covers a 1985 forest. In these cases, the vegetation was vegetation are also associated with 1985 conuco. Additionally there were 17 new recorded as secondary forest. The num- conucos (see Appendix I). Two new areas areas of savanna Arbolada. These 19 new bers of casitas and conucos indicated in of matorrales cover an additional area of areas of both types of savanna cover an the 1995 maps is then likely lower than 2.7ha. area of approximately 83.8ha. the true number. The number of conucos In Area 2 the vast ma- Though numbers are has increased four-fold from 11 to 44 jority of new savannas of both types then approximate due to the re-mapping conucos in Area 1. The number of casitas (plain savanna and savanna Arbolada) methods used, particularly as a result of has also increased, from two to 16. have established in what was previously angles of observation, the data presented forest-only (as opposed to a forest- below, nonetheless, provides a picture of Area 2: matorrale mix). In 10 of the new areas of a degree of the extent of land use and Comparisons of Area 2 savanna, primary forest has been dis- land cover change in the areas of the in the same 10-year interval (Figures 4 placed; in 6 areas, savanna has displaced most intensive and longest-standing cultivation in Kavanayén.

Changes in agricultural intensity and forest loss

Area 1: There has been a significant increase in the amount of savanna in this valley. There is also evidence of fire intrusions into high forest (i.e. dead trees on the edges of forest fragments). On hillsides previously covered with high forest, there are now large areas of low forest, particularly in Section 3. In Sec- tion 2 there are large deforested tracts (areas of grassland) on forested hillsides. Many of the forest fragments, particularly in Section 2, are now Figure 5. Kavanayén Land Use and Vegetation Changes, Area 2, 1995.

AUG 2001, VOL. 26 Nº 8 333 a combination of primary and secondary TABLE VII forest. In the 4 other new areas of sa- CHANGES IN LAND USE AND VEGETATION. vanna, grassland has displaced a combi- KAVANAYÉN 1985 TO 1995 nation of forest and matorrales (see Ap- pendix I). Area 1 Area 2 Nine small areas of new Total map area (ha) 1,832.8 2,496 matorrales cover an additional 15.51ha. This vegetation has established in areas Vegetation Type (ha) 1985 1995 1985 1995 that were previously forested. The number of conucos Savanna Arbolada 128.31 0 38.07 68.59 has increased over two-fold, from 22 to Plain Savanna 18.12 60.9 237.51 15.23 53; the number of casitas has increased Matorrales 242.74 2.67 117.63 15.51 from zero to 11. Land Use Activities The total changes in land use and vegetation are presented in Conucos 11 44 22 53 Table VII. Casitas 2 16 0 11

Discussion and Interpretation Total Amount (ha) of Forest Displaced by Secondary Vegetation* (pre and post-1995) of Results 434.62 255.03 Comparison of the 1985 Notes: and 1995 maps show that: In Elcoro and Vera's original maps, adjacent areas of different types of matorrales were separately 1- Land use around Ka- identified; in the modifications of their maps appearing here (Figures 2 and 4), contiguous areas of all types of matorrales have been identified as single areas following the modifications of vegetation vanayén has become much more intensive types discussed previously. as the number and density of conucos * and casitas has increased substantially; Defined as pre-1985 sabana arbolada and matorrales; post-1985 plain savanna, savanna arbolada and matorral. In the 1985 maps, only sabana arbolada and matorrales were defined as secondary veg- therefore, there has been substantial for- etation. Plain savanna indicated in the 1985 maps may be primary grassland and therefore not an indi- est loss and displacement by savanna and cation of recent deforestation. matorral vegetation. Some of the reduction in high forest is a result of the increase in hood that savanna fires can penetrate the continues to be more intensive in the area the number of cultivators and conucos forest fragment. covered by Area 2 - in 1995, there were because, as mentioned, Pemón prefer to 3- Forest fragmentation 44 conucos in Area 1 as compared to 53 cut their conucos in high forest. In some has increased. There are numerous areas in Area 2 (compared to 1985 numbers of cases high forest has been replaced by where secondary forest appears to be 11 and 22 conucos respectively). As high low forest. While this represents a poten- well-established within primary forest forest is not establishing in Area 1 and tial reduction in species diversity as sec- fragments. In the absence of fire, these secondary forest has been replaced by sa- ondary tropical forests often have fewer areas are well-positioned to experience vanna and matorrales, conuqueros have tree species than do primary (high) for- continued forest succession as they have continued their more intensive use of more ests, it is not deforestation. However, nearby seed sources of forest tree species distant forests in Area 2. there has also been high forest displace- and edge microhabitats favourable to The density of conucos ment by savanna which entails both the seedling establishment. In many other ar- has also increased. When some of the costs of species loss as well as those of eas, by contrast (those areas identified in 1995 conucos become rastrojos (fallows), loss of forest animal habitat and the asso- 1985 as “rastrojo bosque secundario” or fire sources (clearing fires, cooking fires, ciated negative hydrological impacts. “secondary forest rastrojo”), the process communication fires) will be closer to Matorrales, like second- of forest re-establishment has been inter- flammable young secondary vegetation. ary forests, may be successional vegetation rupted and the successional forest has de- The potential for successful forest re-es- that, in five to ten years time may be re- graded to savanna. In 1985, Area 1 was tablishment may then be reduced if fire placed by young secondary forest. New dominated by secondary forest (as op- intrusion into the rastrojos occurs. savanna is also secondary vegetation, but posed to primary forest). Ten years later There has been signifi- is not likely transitional vegetation as it is much of the secondary forest has been cant deforestation and vegetation degra- unlikely that there will be forest re-estab- displaced by savanna, likely due either to dation in Kavanayén. What forest does lishment after grasses are well-established, recultivation after an inappropriately short remain is largely secondary (low) forest particularly when fire events are frequent. fallow period or intentionally lit fires that as opposed to primary (high) forest. In 2- There has been a de- inhibited continued succession in these 1985, much of the high forest was in ex- crease in the amount of high forest. secondary forests. tensive (forest-only) fragments. In 1995 Change in edge vegetation is another, al- 4- The process of forest many of the remaining fragments were beit minor, indication of forest loss. For- succession has been interrupted, or even vegetation mosaics of savanna, matorrales est fragments have become smaller and, truncated, in many secondary forests and forest. The substantial loss of sec- in many cases, previously distinct forest- (identified in the 1985 maps as “ras- ondary forest indicates that, in many savanna boundaries have degraded and trojos”) and there has been substantial for- cases, forest succession has been perma- now have matorrales as intermediate veg- est loss and displacement by savanna and nently truncated. Whereas, in the past, etation between the forest and savanna. matorrales. The increase in the number of the loss of primary forest was compen- The replacement of this forest tree edge conucos and casitas indicates agricultural sated for by the establishment of second- vegetation with more combustible edge intensification and change in the manner ary forest during the fallow period, forest vegetation may have increased the likeli- of forest use in that period. Cultivation loss now is often permanent. The role

334 AUG 2001, VOL. 26 Nº 8 that secondary forests play (as a future town, a cultivator has no option but to importantly in the Gran Sabana where agricultural resource, as animal habitat use younger secondary forests. The pres- fire is a significant factor in forest re-es- and stabilizing the hydrological regime) ence of extensive young secondary veg- tablishment, they also provide a type of has been permanently lost. etation around many of the conucos in fire break, protecting fallows from fire 5- There are no longer Kavanayén and the decrease in the incursion. any active pastures. amount of low forest from 1985 to 1995 Analysis of the signifi- 6- No forest has estab- are evidence that fallows have shortened cance of various site characteristics on lished in the 1985 savannas. Once sa- and that old secondary forests are not be- rates of forest re-establishment after cul- vanna vegetation is dominant in an area, ing allowed to establish. tivation (slope, soil fertility, root mat forest does not appear able to re-establish The so-called “corridor depth, distance to water and distance to on the infertile, non-diabase soils. This is of regeneration” (a strip of uncut high savanna), shows that the only factor true both for tree savanna (“sabana forest between fields) plays a major role which is statistically significant is dis- arbolada”) and the herbaceous savanna. in forest re-establishment after cultivation tance to savanna (Kingsbury, 1999). This (Uhl, 1990). This corridor not only pro- significant site characteristic can more Conclusion vides a nearby source of seeds for forest accurately be interpreted as the distance tree species and edge microhabitats suit- to sites of high combustibility and fre- The data presented con- able for tree seedling establishment, but quent combustion. As well, nearby sa- firm that there has recently been substantial deforestation around Kavanayén. De- forestation and replacement by savanna in Appendix I the Kavanayén valleys have occurred in the area in which agricultural use has VEGETATION TYPES DISPLACED BY SECONDARY SAVANNA AND MATORRALES, been increasingly intensified since 1942 KAVANAYÉN VALLEYS, 1985 TO 1995 when the community was founded. Judg- ing by the increase in the number of Figure 3, Area 1: conucos and casitas in these valleys, ag- New Areas of Plain Savanna ricultural use there has become even 1 -in former areas of young and old secondary forest more intensive in the period from 1985 to 2 -former high forest and one active conuco 1995. The deforestation in these valleys 3 -former old secondary forest is directly associated with intensive agri- 4 -former old secondary forest and high primary forest culture and indirectly with the increasing 5 -former old secondary forest and high primary forest use of fire that accompanies cultivation. 6 -former old and young secondary forest and matorrales While there is a complex 7 -former high primary forest, gallery forest and young secondary forest relationship between environmental and 8 -former old and young secondary forest, matorrales and three active conucos cultural changes, certain features of the 9 -former old secondary forest and matorrales cultural and physical environments are 10 -former old secondary forest and matorrales 11 -former young secondary forest and matorrales clearly more important in changing the 12 -former primary medium-high forest and matorrales ecology of the region. Among the cultural processes, clearly the most significant has New Areas of matorrales been the Pemón transition from being “mi- 1 -formerly a mix of high matorrales and old and young secondary forest gratory shifting cultivators” (Dezzeo, 2 -former young secondary forest and high matorrales 1994) to sedentary cultivators who con- tinue to shift fields but no longer shift Figure 5, Area 2: residences and community locations and the related trend towards larger settle- New Areas of Sabana Arbolada ments. There has been a concomitant 1 -former high primary forest change from site selection characteristics 2 -former old and young secondary forest 3 -former low primary forest and matorrales that traditionally emphasized primarily 4 -former matorrales ecological factors instead to those that 5 -former medium-high primary forest and matorrales now emphasize logistical factors. 6 -former medium-high primary forest, high primary forest, old and young When forest resources secondary forest (with an active conuco to the southeast edge) are reduced, the responses of shifting cul- 7 -former high primary forest and old secondary forest tivators are limited primarily to abbrevia- 8 -former high primary forest tion of the fallow period, locating fields 9 -former high primary forest more closely together, or abandoning cul- 10 -former high primary forest tivation. In the Gran Sabana, the second 11 -former high primary forest option in particular can result in rela- 12 -former high primary forest 13 -former high primary forest tively rapid environmental change and the 14 -former high primary forest third is generally not feasible in terms of 15 -former high primary forest and old secondary forest meeting subsistence needs. 16 -former high primary forest, young secondary forest and matorrales Abbreviated fallowing in 17 -former low and medium primary forest Kavanayén can be inferred from the combination of deforestation and population New Areas of Plain Savanna increases and concentration. When there 1 -former old and young secondary forest, high primary forest and once is pressure to stay close to town but there active conuco is little or no old growth forest close to 2 -former low primary forest forest

AUG 2001, VOL. 26 Nº 8 335 vanna is also a source of easily wind-dis- of ways – interviews on indigenous envi- Breakdown of Traditional Swidden Agricul- persed grass seeds. ronmental perception, forest surveys and ture. Scientia Guaianae No. 5: 65-78. Harris DR (1969) Agricultural systems, ecosys- It is also important to time series analysis of remotely sensed tems and the origins of agriculture. In Ucko note that once matorral or savanna veg- images. PJ, Dimbleby GW (Eds) The domestication etation is well-established, the likelihood While the process of de- and exploitation of plants and animals. of forest re-establishment is low. On non- forestation in the Gran Sabana may be Gerald Duckworth & Co. London. pp. 3-16. diabase soils in particular, soil impover- understood (even appearing obvious to Hernández L (1994) Bosques. In Ecología de la Altiplanicie de la Gran Sabana (Guayana ishment as a result of savanna displace- some, that is, that cultural and demo- Venezolana) I. Scientia Guaianae No. 4: 80- ment of forest (Kingsbury, 1999) and fire graphic change are resulting in deforesta- 94. frequency in non-forest vegetation likely tion) the pattern of this change and the Im Thurn EF (1967) Among the Indians of play a major role in precluding forest re- variations in this pattern at the regional, Guiana; Being Sketches Chiefly Anthropo- establishment in these areas. inter- and intra-community levels is not logic from the Interior of British Guiana. Dover Publications. New York. Reproduction Similar processes of en- so well understood and calls for further of the 1883 edition. vironmental and cultural change are study in order to prevent further defores- Kingsbury ND (1999) Increasing Use of De- likely occurring elsewhere in the Gran tation and its negative hydrologic and creasing Resources: A Case Study of Pemón Sabana in other Pemón communities, and cultural impacts. Amerindian Shifting Cultivation, Gran Saba- indeed among other indigenous shifting na, Venezuela. Ph.D. dissertation. York Uni- ACKNOWLEDGMENTS versity, Toronto, Canada. cultivators in South America who are also Kingsbury N, Kellman M (1997) Root Mat experiencing a process of rapid accultura- Depths and Surface Soil Chemistry in tion and incorporation into non-indig- Logistical support for Southeastern Venezuela. J. Tropical Ecology. enous societies and the accompanying this study was provided by the Departa- 475-479. land use changes. mento de Conservación de Cuencas of Means DB (1995) Fire Ecology of the Guyana Region, Northeastern South America. In Fire While the impacts of Electrificación del Caroní” (EDELCA). in Wetlands: A Management Perspective. Pro- this combination of resource reduction The author thanks Alexandre Barreto, ceedings of the Tall Timbers Fire Ecology and increases in resource use will vary Eduardo Gómez and Yoni Rivas at Conference No. 19. Tall Timbers Research depend on the characteristics of the par- EDELCA; Venancio Sucre, who provided Station. Tallahassee, Florida. pp. 61-77. ticular physical environment and the re- expert field assistance in Kavanayén; Miranda M with Blanco-Uribe A, Hernández L et al. (1998) All That Glitters is not Gold: gional fire ecology, in certain areas (e.g., Nelda Dezzeo and Judith Rosales, who Balancing Conservation and Development in Kavanayén) where soils are generally in- assisted during the research period. Field- Venezuela’s Frontier Forests. World Re- fertile, forests have a low resilience to work was undertaken with the financial sources Institute. Washington, D.C. disturbance and slow rates of re-growth, support of the International Development Monsonyi E (1972) Indian Groups in Venezuela. the environmental impacts of intensified Research Centre (Ottawa, Canada) and In The Situation of the Indian in South America. World Council of Churches. Gene- shifting cultivation (and the concomitant from a National Science and Engineering va. pp. 388-391. increases in cultivators, cultivated areas, Research Council of Canada grant to Pro- Morales L, Gorzula S (1986) The Interrelations in combustion frequencies and more com- fessor Martin Kellman of York University of the Caroní River Basin Ecosystems and bustible secondary vegetation) are both (Toronto, Canada). Hydroelectric Power Projects. Interciencia visible and negative. 11: 272-277. Nye PH, Greenland DJ (1960) The Soil Under While the social and mi- Shifting Cultivation. Commonwealth Agricul- cro-economic impacts of deforestation on REFERENCES tural Bureau. Bucks, England. the Pemón have not been documented, it República de Venezuela (1985) Censo Indígena de Venezuela, 1982 - Nomenclador de Co- is reasonable to conclude that these im- Azuaje A (1986) Caracterización de Conucos pacts will be (or in some communities, munidades y Colectividades. Oficina Central en la Cuenca Alta del Río Caroní. de Estadística e Informática Nº 180. Ca- are already) also negative, likely resulting Electrificación del Caroní. División de racas. Cuencas e Hidrología. Yuraní, Gran Sabana, in an increased consumption of and de- República de Venezuela (1993) Censo Indígena pendency on food imported into indig- Venezuela. de Venezuela, 1992. Oficina Central de enous communities, and dependency on Butt-Colson A (1971) Comparative Studies of the Estadística e Informática. Caracas. Tomo I wage labor to acquire resources with Social Structure of Guiana Indians and the & II. Problems of Acculturation. In Salano FM Tate GHH (1930) Notes on the Mount Roraima which to purchase foodstuffs. However, (Ed) The Ongoing Evolution of Latin Ameri- Region. Geographical Rev. 20: 53-68. infrastructural weaknesses in both the can Populations. Charles C. Thomas. Spring- Thomas DJ (1973) Pemon Demography, Kinship, transportation and employment sectors field, Ill. pp. 61-126. and Trade. Ph.D. dissertation. University of suggest that this subsequent cultural and Butt-Colson A (1985) Routes of knowledge: an Michigan. economic change may be problematic. aspect of regional integration in the circum- Thomas DJ (1982) Order without Government: This latter “data gap” Roraima area of the Guiana Highlands. Pemon Indian Society. University of Illinois Antropologia 63-64: 103-149. Press. Chicago. on the social and micro-economic im- Dezzeo N (1994) Ecology of the Gran Sabana Uhl C (1987) Factors Controlling Succession pacts of deforestation and the policy im- Uplands (Venezuelan Guayana). Scientia Following Slash-and-Burn Agriculture in plications of increased dependency on Guaianae No. 4: xv-xxxviii. Amazonia. J. Ecology. 75: 377-407. Uhl C (1990) Deforestation, Fire Susceptibility, non-agricultural labor for subsistence Elcoro S, Vera N (1986) Estudio de Rastrojos and Potential Tree Responses to Fire in the calls for further attention and study, and en Areas con Altas Concentraciones de Po- Eastern Amazon. Ecology 7: 437-449. blación en Kavanayén, La Gran Sabana, further participation of the Pemón in the Urbina L (1979) Adaptación Ecológico-Cultural management of their forest resources. Estado Bolívar. B.Sc. Thesis (Forestry Sci- ences). Universidad de los Andes, Mérida, de los Pemón-Arekuna: El Caso de Tuaukén. In terms of watershed Venezuela. M.Sc. dissertation. Instituto Venezolano de and forest management, the rate of defor- Investigaciones Científicas. Caracas. Fölster H (1986) Forest-Savana Dynamics and estation around other larger, growing and Wilbert J (1971) Cultural Variability in Venezu- Desertification Processes in the Gran elan Indian Tribes. In Salano FM (Ed.) The permanent communities is another impor- Sabana. Interciencia 11: 311-316. Ongoing Evolution of Latin American Popu- tant, and missing, set of data. Such data Fölster H (1995) Local Population Concentra- lations. Charles C. Thomas. Springfield, Ill. can, however, be produced in a number tions in the Venezuelan Guayana and the pp. 127-159.

336 AUG 2001, VOL. 26 Nº 8