Land-Use Change in the Cajamarca Catchment, Peru, 1975–1996

C.B. Bussink and R.J. Hijmans

We used land-use maps for different years and areas in the Cajamarca catchment in the northern Peruvian Andes to study changes in land use between 1975 and 1996. Despite population growth, agriculture has not been intensi- fied nor extensified. Neither did we find evidence for crop encroachment into the higher parts of the catchment. The agricultural area has decreased and the proportion of fallow land has increased. There has been a significant increase of shrubs and bare soil at the expense of natural pastures, perhaps indicating overgrazing. The area planted to specific crops has fluctuated, but there has been a clear decrease in the area planted to barley and an increase in wheat. The area with exotic tree species has also increased.

The global increase in population has led different patterns of land-use change are to important changes in agriculture, likely to occur simultaneously; there are through both intensification and reports of intensification as well as aban- extensification of production. donment from different parts of the Andes Extensification refers to the use of more (e.g., Mayer, 1979; Hervé and Ayangma, land for agriculture, while intensification 2000; Wiegers et al., 1999). Understanding refers to using agricultural land more the patterns and causes of land-use change intensively, e.g., with shorter fallow in the Andes could lead to the develop- periods. Both processes can have a ment of policies that would help avoid significant negative impact on the envi- some of the more negative consequences ronment. Agricultural extensification leads of these changes. Insight into the current to the loss of natural habitats, and intensi- processes of change could also help in the fication can lead to the loss of biodiversity design of development interventions. as well as to environmental pollution, as it often goes hand in hand with increased In the present study, we describe land-use pesticide and fertilizer use. However, change in the Cajamarca catchment intensification may also lead to increased between 1975–1996. We used a geo- economic land value and incentives to graphic information system (GIS) to conserve land resources through such compare a number of large-scale maps measures as soil conservation (Tiffen and (1:25,000) from different years and zones. Mortimore, 1994). The Cajamarca catchment is in the northern Peruvian Andes (Figure 1). It is It is important to note, however, that regarded as a region with a distinct patterns of land use change can be rather ecology, characterized by the jalca or different between regions and countries, grassland vegetation, in the highest areas, within countries, and even within village being a transition zone between the more territories. Particularly in areas of hetero- humid paramo Andes in the north and the geneous terrain, such as the Andes, drier puna Andes in the south (Troll, 1968).

CIP Program Report 1999 – 2000 421 Encanada

Banos del Inca Cajamarca Llanacanora Namora

Jesus Matara

Cajamarca San Marcos catchment

Lima

kilometers 0 5 10

Figure 1. Location of the Cajamarca catchment in Peru and the study areas: 1975/78 (the whole catchment), 1991 (A), 1992/1993 (B), 1996 (C), and the zone with three studies (D).

Near Cajamarca town, at 2650 m, average Tapia, 1996). Land-use systems in this area daily temperature is about 13° C and yearly are different from those found in the precipitation is 720 mm. The temperature central and southern Andes of Peru (see, is rather constant over the year, while e.g., Mayer, 1979; Morlon, 1992). For diurnal differences are about 17° C. There example, there is not much communally is a clear rainy season between October managed land in Cajamarca, and house- and April (De la Cruz et al., 1999). The hold access to different production zones altitude in the catchment is between 2000 is limited. m to 4200 m, with about 79% between 2600 m and 3800 m. The distribution of Between 1940 and 1993, the Cajamarca crops by altitude has been used to distin- Department had an annual population guish main agroecological zones: the growth rate of about 1.5%, and the smaller cultivated pasture zone (irrigated valley Cajamarca Province (which roughly bottoms), the maize zone (lower slopes), coincides with the Cajamarca catchment) the tuber zone (higher slopes), and the had a growth rate of 1.7%. Although this is natural pasture zone (jalca) (Kohler, 1986; below the national average (2.2% for the Kohler and Tillman, 1988; Seifert, 1990; same period), Cajamarca has a high

422 Enriching the Portfolio: CIP’s Global and Regional Partnerships growth rate for areas in the Peruvian Andes April). The maps were digitized using (INEI and UNFPA, 1996; INEI, 1995; ArcInfo (ERSI, Redlands, CA, USA) and Seifert, 1990; Wiegers et al., 1999). analyzed with ArcView (ESRI) and Microsoft Excel. Materials and methods Land-use change Land use To study changes in land use, the 1975/78 Our point of reference is a series of three data were compared with 1991, 1992/93, land-use maps for 1975, 1977, and 1978. and 1996. All comparisons were made on Together, these maps cover the whole the areas in common to the two maps. For Cajamarca catchment (211,854 ha), a small area of 4,369 ha (zone D in Figure except for a small part in the southwest 1), we had an overlap of three studies (25,932 ha). They also cover some areas (1975/78, 1991, and 1996). For all zones, outside the catchment, which we ex- we calculated the fallow ratio as the area cluded from our study. We compared these of land in fallow divided by the total maps with six land-use maps of smaller cultivated area (crops and fallow land). zones within the Cajamarca catchment, for different years between 1991 and 1996. Tabulating areas for one zone for two To facilitate the analysis and because of different dates allowed overall change to complementarity in location and similarity be assessed. As one land unit on a map of years, we merged the maps for 1975, can contain many land-use classes, we 1977, and 1978 (Gozalo et al., 1977; applied the following two rules to estimate Landa et al., 1978; Landa and Johansen, which land-use classes changed into 1978) to form what we will refer to as the which other classes. For all land units, the “1975/78 map”. We merged three maps for following applies: 1991 (zone A, 19,748 ha (UNC, 1991a, • If a land-use class is present in both 1991b, 1991c)) and two maps for 1992 and time periods, we assume that the 1993 (zone B, 6,916 ha (Chilón, 1993; minimum of the two areas occupied the Saldaña, 1994)). Zone C is based on one same location within that land unit in map for 1996 (18,732 ha (M. Jimenez, both periods (i.e., no change). UNC, unpublished)) (Figure 1). We also • The differences in area of one land-use made a single map for the “1990s,” class are proportionally subtracted from combining zones A, B, and C, using the (in the case of an increase) or added to most recent data when there was an (if there was a decrease) the areas of all overlap between zones. new land-use classes. The proportion is All the maps were made using the same based on the change in area. methodology. With air photos (most at a scale of 1:20,000) and fieldwork, map These rules are likely to result in conserva- units were delineated and the percentage tive estimates, meaning that some of different land-use classes within each changes may be underestimated, because unit was estimated. Land-use classes they do not take into account changes in include broad categories of vegetation as location of land-use classes within map well as different crops. For example, one units. map unit could contain 30% maize, 20% wheat, 40 % shrub land, and 10% natural Results pasture. We simplified the maps in some Land use in 1975/78 cases by merging classes: rye and oats; potatoes and other tubers; and pulses In 1975/78, the largest land-use class in (beans, lupines, lentils, and green peas). the Cajamarca catchment was natural Fieldwork was mostly carried out during pastures (40%); 38% of the land was the main growing season (November to cultivated (30% annual crops, 5% fallow,

CIP Program Report 1999 – 2000 423 and 3% cultivated pastures) (Table 1). The Changes in the other land-use classes were most important annual crops were barley generally smaller and inconsistent: in and wheat. Annual crops and fallow land some zones there was an increase and in were relatively over-represented in the other zones there was a decrease. three study zones in 1975/78, in compari- son with land use in the catchment as a Most of the individual land-use classes in whole (Table 1). 1975/78 had, for the most part, the same land use as in 1990 (Table 2). This was Land-use change especially true for land with shrubs, forests, cultivated pasture, and crops In all study areas, there was an increase in (around 70%). On average, over all zones, the area with shrubs (+10% for the 1990s), new forests have been planted on former forest (+3%), wheat (+3%), and fallow natural pasture (38% of the new forest), (+0.6%), whereas there was a decrease in cropland (21%), and on bare soil and shrub the areas with barley (-7%), pulses (-1%), land (19%). In the 1990s, 15% of the and natural pasture (-14%) (Figure 2). former area of bare soil was cultivated

Table 1. Land use in Cajamarca by different areas and years (%). Natural Cultivated Bare soil Area Year Shrubs Forest Crops Fallow pasture pasture & rocks Catchment 1975/78 40 13 0 3 30 5 9 Zone A 1975/78 27 8 1 0 45 9 10 1991 5 26 3 1 37 12 15 Zone B 1975/78 30 15 1 1 35 9 9 1992/93 20 23 3 1 34 9 11 Zone C 1975/78 32 6 1 7 37 7 10 1996 23 7 6 10 33 8 12 Zone 1975/78 30 8 1 4 39 9 9 A+B+C 1990s 16 19 5 5 35 9 12

Land use

Natural pasture

Shrubs

Forest

Cultivated pasture

Maize

Wheat Zone A (1975/78-91) Zone B (1975/78-92/93) Barley Zone C (1975/78-96) Rye & Oats Zone A+B+C

Pulses

Tubers

Fallow

Bare soil / rocks

-25 -20 -15 -10 -5 0 5 10 15 20

Area change (%)

Figure 2. Change in land use for three zones with different time series (1975/78–1991, 1975/78–1992/93, and 1975/78–1996).

424 Enriching the Portfolio: CIP’s Global and Regional Partnerships Table 2. Change in land use from one class to another between 1975/1978 and the 1990s (expressed in percent of area of each land-use class in 1975/78). 1990s Natural Cultivated Bare soil / 1975/78 Shrubs Forest Crops Fallow Total pasture pasture rocks Natural 39 20 6 3 15 5 11 100 pasture Shrubs 3 69 5 1 12 4 5 100

Forest 3 9 69 2 7 2 7 100 Cultivated 10 3 2 73 7 2 2 pasture Crops 6 11 2 2 66 8 6 100

Fallow 10 15 3 3 19 43 7 100 Bare soil / 5 12 5 1 15 5 57 100 rocks with crops again, 5% was fallow, and 1% and 1996) confirms the major trends: there was covered with cultivated pastures. New is an increase in the areas of bare soil, areas with bare soils or with shrubs were shrubs, forests, fallow, and wheat, and an mostly in areas that had natural pastures or important decrease in the areas with crops in 1978. barley and natural pastures. Although for most classes the changes between 1975/78 There was an average decrease of 4% of and 1991 and 1975/78 and 1996 point in the total arable land area (Table 1), which the same direction, the size of change can includes crops and fallow land, roughly be quite different (Figure 4). irrespective of altitude (Figure 3). The relative proportion of fallow land over crop Conclusions and Discussion land increased. In 1978, there was one ha of fallow for every 5.6 ha of cultivated We have not found any evidence for land (including fallow), while in the intensification of land use in the nineties there was one ha of fallow for Cajamarca catchment between 1975 and every 4.8 ha of cultivated land. 1996. The total area under cultivation had decreased slightly, and the fallow ratio Land-use change in zone D, for which we had increased. Neither did we find evi- had data for three years (1975/78, 1991, dence for an extension of the agricultural frontier into higher areas; this is often Relative area (%) suggested to be happening. 70 60 Our results are in line with the observa- 50 tions of Frias (1995) that, despite 40 1975/78 30 1990s population growth, the importance of 20 agriculture for employment is decreasing 10 in both absolute and relative terms. 0 2400-2600 2600-2800 2800-30003000-3200 3200-3400 3400-3600 3600-3800 However, the results seem to contradict Seifert (1990), who found that fallowing Altitude (m) had disappeared on farms of less than 5 ha. This could be because these small Figure 3. Agricultural area (crops and fallow) by farms, although they are the most common altitude class for 1975/78 and the 1990s.

CIP Program Report 1999 – 2000 425 Relative area

20 1975/78 15 1991 1996 10

0 Maize Wheat Barley Rye & Oats Pulses Tubers Fallow Cultivated pastureNatural pastureForest Shrubs Bare soil

Land use

Figure 4. Area of land use classes for three years (1975/78, 1991 and 1996) in one zone (Zone D; Figure 1). farm size, only cover 10% of the culti- of classification error overrides the effect vated area (Seifert, 1990). The decrease of of years. In zone A, the areas with shrubs, area planted to tubers and the increase of fallow land, and bare soil were probably area under cultivated pasture coincides overestimated at the expense of natural with Frias’s (1995) study for the whole pasture, because it is quite unlikely that Cajamarca Department. However, instead much of the soil that was bare in 1975/78 of the decrease in wheat described by would be cultivated again 20 years later, Frias (1995), we found an increase. as our data suggest. Rigorous definitions of these land-use classes are necessary to The largest change is an increase in shrub improve this type of study. Crops are easier land and bare soil, mainly at the expense to classify, and the results for zone D of natural pasture. Overgrazing may have indicate that there is considerable varia- led to soil erosion and loss of soil fertility, tion between years in areas with specific which in turn, may have led to an increase crops. in unpalatable shrubs. Alternatively, it might be a sign of vegetation recuperating For estimates of crop area, the timing of because of a decrease in grazing pressure. the survey is very important. In the off- The increase in forested areas matches the season, there would be much more fallow pattern of agricultural dis-intensification and the natural vegetation would be more and retraction. Also, over the past few likely to be classified as bare soil. The decades, there has been considerable spatial effect of rotation on land-use public investment in forestation in patterns should not be of much importance Cajamarca, particularly through a local for this study because it should average non-governmental organization. out within and between land units. One complication with our data is that we This study is descriptive in nature and does compared not only across different years not offer insights into the factors that drive but also across different zones. The only the changes we observed nor in other zone for which we had data for three important aspects of land use, such as periods was zone D. However, it appears productivity. Further research, including that for some land-use classes, the effect detailed fieldwork, would be needed for

426 Enriching the Portfolio: CIP’s Global and Regional Partnerships this. At the other end of the spectrum lies Frias, C. 1995. Pobreza campesina. Sólo the question about the representativeness un problema rural? Cajamarca: of our results for a larger area such as the economia, espacio y tecnologia. northern Peruvian Andes, which, for Intermediate Technology Development example, could be studied with satellite Group (ITDG), Lima, Peru. 153 p. images. Representativeness is an important Gozalo, V.J., C. Landa, and M. Garcia. issue because the Cajamarca catchment 1977. Uso Actual de la tierra de los may not be typical. Because the catch- PIAR’s Cajamarca y San Marcos. Vol. I. ment is located near an important Distritos de Jesus, Namora, Matara y departmental capital, land use may have parte de los de Cajamarca Llacanora y been relatively intensive for a long time; San Marcos. Ministry of Agriculture, then, over the past few decades, there Food, and Technical Cooperation, may have been more intensification in Belga, Cajamarca, Peru. other areas. Whereas in our study area, the Hervé, D. and S. Ayangma. 2000. area of exotic forest (eucalyptus, pine) Dynamique du l’occupation du sol dans increased, natural forests in nearby areas une communauté agropastorale de are believed to have decreased dramati- l’altiplano central bolivien. Revue de cally over the last 30 years (H. Willet, Géographic Alpine 2:69–84. personal communication; ONERN, 1975). INEI (Instituto Nacional de Estadística e Informática). 1995. Perú: Población The time period we studied is obviously total por área urbana y rural, según arbitrary and completely driven by the departamentos, provincias y distritos. presence of land-use maps. The results Instituto Nacional de Estadística e could be put into a wider context using Informática, Lima, Peru. census and household data in which an INEI (Instituto Nacional de Estadística e important milestone would be land reform Informática) and UNFPA (United of the early 1970s. Nations Population Fund). 1996. Perú: Proyecciones departamentales de la Acknowledgements población 1995–2015. Instituto We thank Jorge de la Cruz, Felipe de Nacional de Estadística e Informática Mendiburu, Edwin Rojas, and Percy and United Nations Population Fund, Zorogastúa for assistance; Jiefar Diaz, Lima, Peru. Wilfredo Poma (UNC), and Pablo Sanchez Kohler, A. 1986. Sistemas de uso de suelos (ASPADERUC) for providing access to old en laderas de Cajamarca - Perú. Boletin land-use maps; CONDESAN for partly de Lima 47:31–40. funding this research; and Joshua Posner Kohler, A. and H.J. Tillman. 1988. and Dominique Hervé for reviewing of this Campesinos y medio ambiente en paper. Cajamarca. Mosca Azul, Lima, Peru. Landa, E.C. and A. Johanson. 1978. Uso References actual de la tierra de los PIAR’s Cajamarca y San Marcos. Vol. III. Chilón C., M.I. 1993. Estudio del uso Estudio de parte de los distritos de actual de la tierra de la intercuenca del Ichocan, San Marcos, Encañada rio Sambar—Distrito Baños del Inca. Cajamarca y Chetilla. Ministry of Master’s thesis. Universidad Nacional Agriculture, Food, and Technical de Cajamarca, Peru. Cooperation, Belga, Cajamarca, Peru. De la Cruz, J., P. Zorogastúa, and R.J. Landa, E.C., A. Johanson, and M. Garcia. Hijmans. 1999. Atlas digital de los 1978. Uso actual de la tierra de los recursos naturales de Cajamarca. PIAR’s Cajamarca y San Marcos. Vol. II. Natural Resource Management Working Zona Norte, Distrito de Baños del Inca y Document No. 2. International Potato parte de la Encañada y Llacanora. Zona Center, Lima, Peru. 49 p.

CIP Program Report 1999 – 2000 427 Sur, Parte de los distritos de San Marcos capital and technology in growth and e Ichocan. Ministry of Agriculture, environment recovery in Kenya. World Food, and Technical Cooperation, Development 22(7):997–1010. Belga, Cajamarca, Peru. Troll, C. 1968. The cordilleras of the Mayer, E. 1979. Land use in the Andes: tropical Americas. In: Troll, C. (ed.). Ecology and agriculture in the Mantaro Geoecology of the mountainous regions Valley of Peru with special references to of the tropical Americas. Collequium potatoes. International Potato Center, Geographicum 9, Institut de Universitat Lima, Peru. 115 p. Bonn, Drummlers Verlag, Bonn, Morlon, P. 1992. Comprendre l’agriculture Germany. p. 15-56. paysanne dans les andes centrales. UNC (Universidad Nacional de Pérou-Bolivie. National Institute of Cajamarca). 1991a. Uso actual de la Agricultural Research, Paris, France. tierra. Map at 1:25.000 scale. Convenio HONREN (Oficina Nacional de CUSO – UNC. Proyecto de desarrollo Evaluación de Recursos Naturales). rural integral de Jesus. Universidad 1975. Inventario, evaluación y uso Nacional de Cajamarca, Cajamarca, racional de los recursos naturales de la Peru. (Unpublished.) zona sur del departamento de UNC (Universidad Nacional de Cajamarca. Volumen I. National Office Cajamarca). 1991b. Uso actual de la for the Evaluation of Natural Resources, tierra. Map at 1:25.000 scale. Servicio Lima, Peru. Silvo Agropecuaria (S.E.S.A). Saldaña A., M.D. 1994. Estudio del uso Universidad Nacional de Cajamarca, actual de la tierra de los caserios de Cajamarca, Peru. (Unpublished.) Chim-Chim, Carahuanga y la Collpa en UNC (Universidad Nacional de la cuenca del rio Quinuario – Distrito Cajamarca). 1991c. Uso actual de la Baños del Inca. Thesis, Universidad tierra. Map at 1:25.000 scale. Grupo Nacional de Cajamarca, Peru. Polivalente de Proyectos Sociales – Seifert, R. 1990. Cajamarca: Vía Chim Shaullo. Universidad Nacional de campesina y cuenca lechera. Privately Cajamarca, Cajamarca, Peru. (Unpub- published, Cajamarca, Peru. lished.) Tapia, M.E. 1996. Ecodesarrollo en los Wiegers, E.S., R.J. Hijmans, D. Hervé, and Andes altos. Friedrich Ebert Foundation, L.O. Fresco. 1999. Land use intensifica- Lima, Peru. tion and distintensification in the upper Tiffen, M. and M. Mortimore. 1994. Cañete Valley, Peru. Human Ecology 27 Malthus controverted: The role of (2): 319–339.

428 Enriching the Portfolio: CIP’s Global and Regional Partnerships