Introduction

In the Andean Cordillera, natural grasslands above 2500 m are commonly used for livestock grazing (Flores Ochoa 1977; Soto 1995). Extreme temperatures and arid climate result in poor soil development, unsuitable for agricultural practices (Tapia and Ruiz 1985). In these environments vegetation is steppic, distributed in patches among corridors of nude soil as in other arid or semiarid habitats at different altitudinal ranges (Montaña 1992; Bisigato and Bertiller 1996), and consists of rigid pastures, dwarf shrubs, and bushes or fleshy plants (Cabrera 1957; Ruthsatz 1974).

Studies on grazing effects on natural vegetation in the northern Argentinean Andes are scarce. Adler and Morales (1999) consider some structural features of natural grasslands to be related to grazing, and in agreement with Molinillo (1993), these authors find an association with some erosive processes with high temporary livestock concentrations.

In the study area, workshops with the shepherds were held to deal with the main problems of grazing related to availability and forage quality of the natural vegetation. As a result, 2 major vegetation types were distinguished according to physiognomy and use: scrubland, and grassland on high terraces. A grassland sector with a relatively small area but good forage quality was also identified, despite its high proportion of denuded soil. Peasants proposed different hypotheses (eg, “is it possible that animals fatten more there because the vegetation is more exposed to the sun in the morning?”) (Sánchez and Quiroga Mendiola 1999).

The study was initiated in this context, given the need to provide answers to intense hydric erosion recorded in the basin. The first part of the study is presented here. The present article aims to identify and describe floristic composition in the main vegetation units described in relation to grazing.

Study area

Santa Victoria Sierra is a mountain range to the east of the Andean Cordillera in northwestern Argentina (Figure 1), including the headsprings of the Iruya River. The main valleys have a north–south orientation, in response to geological structure (Turner and Mon 1979). Topography is irregular, with strong regional slope: from Santa Victoria, with a mean altitude of 4600 m, to the Bermejo River basin, at 600 m, there is a linear distance of 70 km, with a mean slope of 18% (Viera and Menéndez 1981).

The climate is highland semiarid, with different topo-climates caused by relief (Viera and Menéndez 1981). Mean annual rainfall of 300 mm occurs mostly in summer, and temperature shows pronounced daily and annual fluctuations (Bianchi 1981). Soils are rocky and shallow, with low organic content (Nadir and Chafatinos 1990). As a result of these features, the Iruya River contributes a great amount of sediment to the Bermejo River, in the high basin of the Río de la Plata (Konzewitsch 1958).

Local inhabitants are farmers and shepherds with a small livestock and subsistence economy, and sporadic employment as itinerant harvesters in different areas within the country (Quiroga Mendiola 2000). The most important population settlements are on alluvial terraces where soils are suitable for agriculture, irrigated by a wide net of channels. Grasslands and scrublands on mountain slopes are hard to access and non-irrigated. Land use is therefore limited to rainfed potato crops (Solanum tuberosum) and sheep and goat herding (Quiroga Mendiola 2000).

Methodology

The catchment area was divided into 3 vegetation units (grassland, sparse grassland and scrubland) based on differential physiognomic features. Pilot sampling to determine minimum plot size and minimum sample size (Milner and Hughes 1968) was conducted in each unit. Thirty 1-m² plots in grassland, 20 1-m² plots in sparse grassland, and 11 25-m² plots in scrubland were sampled for total cover and species cover. Plots were located left or right along 50-m line transects, and randomly distributed. Altitude, exposure, slope, and grazing pressure were measured on each plot (Krebs 1989).

Grazing pressure was estimated using the following index:

Vegetation units were defined a priori, and a multivariate analysis was performed on a matrix of 60 censuses and 70 species, with cover as the response variable, to explore the association among units and species, and correlate the ordering axes with environmental variables measured independently (grazing pressure, slope, exposition, altitude) (Gauch 1982).

Shrubs, gramineous/graminoids, and herbaceous species were classified according to life form and palatability. Information on species palatability was taken from Bertuche and Vorano (1977), Tapia and Ruiz (1985), Tovar (1988), Braun Wilke (1991), Tichit (1995), and Villca and Genin (1995). When no information could be found for a species, livestock behavior was observed for 1 year. These data were reviewed by interviewing experienced breeders. Species appetizing to sheep over the entire year were classified as “palatable;” those selected only by the end of the dry season when forage availability is lowest as “intermediate;” and species never chosen by livestock as “non-palatable.” Foraging quality among vegetation units (grassland, sparse grassland and scrubland) was analyzed by Variance Analysis, using cover of species grouped by life-form as a variable, and the proportion of different palatability levels with respect to total cover in each census. Tuckey tests (significance level 0.5) were applied a posteriori. The Levene test was used to confirm homocedasticity. Kruskall-Wallis non-parametric analysis was performed for data that did not meet these criteria (Zar 1984).

Results

Discussion

In the study area, differences between floristic composition and vegetation were related to grazing, altitude and slope. Although the sampling design does not reveal the relative importance of these factors, different authors found that grazing may affect the structure and floristic composition of mountain communities (Molinillo and Monasterio 1997; Adler and Morales 1999). The mechanisms involved in these changes are direct and indirect effects of grazing. Among the former, consumption of vegetative tissues can compromise sapling survival (Salihi and Norton 1987; Clements and Young 1996) and adult individuals (McLean and Wikeem 1985; O'Connor 1993). The latter operates through removal of reproductive tissues (O'Connor and Pickett 1992; O'Connor 1993; Bertiller and Coronato 1994) and generation of clearings for establishment of new individuals (Oesterheld and Sala 1990).

The effect of grazing on the floristic composition of communities depends not only on grazing pressure, but also on frequency and season (Adler and Morales 1999): pastures produce most of their biomass in spring, and their ulterior development is conditioned mainly by this initial production (Newman 1993). During the reproductive season, intensive grazing also impairs seed production (O'Connor and Pickett 1992; Bertiller 1994), favoring vegetative reproduction strategies (Belsky 1992).

In San Isidro, shrubby chamaephyte or suffruticose species dominate grasslands heavily grazed in spring, which sprout by the end of trimming season. The sites used in summer are mainly composed of grasses and species with rhizomes, bulbs or tuberous roots (Quiroga Mendiola 2000).

Grassland and sparse grassland share the species Pennisetum chilense, both with a disperse distribution and low abundance. However, it is the only palatable species that remains on sparse grassland, along with annual colonizing species. Sparse grassland probably belongs to the “grassland” vegetation unit, but in a transitional state of degradation, probably because of excessive use or lack of resting periods. P. chilense remains, despite its high palatability and thanks perhaps to its survival strategy through rhizomes.

Good cover recovery has been achieved on 7 artificial 1-year closures on both grassland types (Quiroga Mendiola 1996, 2000). However, grassland did not experience species shifts, but a rise in recruitment and plant height on sites where soil was removed for implanting the closure, suggesting the importance of the trampling effect in a selective process against therophyte species. On the other hand, sparse grassland exhibited different behavior after 1-year closure. Two new species were found: therophyte colonizers and an unusual growth of P. chilense. Reappearance of other herbaceous co-dominant species was not recorded, also suggesting the transitional degradation status of this grassland (Quiroga Mendiola 2000).

The effect of altitude, in turn, is a function of frequency and intensity of frost. In the highlands, largely exposed to winds and frost, species with meristems just below or above the soil surface will flourish, a strategy also effective as a defense against browsing (Westoby 1980; Coughenour 1985). This may explain why a heavily grazed unit such as grassland maintains a good cover of perennial palatable gramineous species.

In this study, altitude and slope are related to grazing. Intensive vegetation and steep slopes result in a nude site most susceptible to erosion—sparse grassland. However, environmental factors not measured in this work—such as organic matter content and soil moisture—can also help to explain plant distribution.

Grassland exhibited the greatest coverage with grasses and herbaceous dicotyledonous species. The other 2 communities exhibited lower foraging aptitude: mean coverage in sparse grassland was low, and species with low foraging values are dominant on scrubland. In the closure study, good primary areal annual productivity was found on grassland and scrubland, but it was very low on sparse grassland (Quiroga Mendiola 2000). This foraging aptitude ranking was related to actual land use intensity in the study area, the most productive areas moderately used in summer being those colonized by shrubs with no foraging value in spring, while those with intensive and constant use are practically denuded.

Conclusions

Physiognomic and floristic differences among the vegetation units correlate with altitude, slope, and grazing pressure. However, grazing seasonality, rather than grazing pressure, appears to play a major role in the conformation of floristic composition and vegetation structure, as revealed in analysis of the grazing pressure index in relation to vegetation data. On sites with uninterrupted grazing, annual colonizer species and non-palatable species become permanent. Only 1 palatable species survives thanks to its rhizome system. Grassland and sparse grassland species have good foraging aptitude; however, the second type is affected by permanent grazing, with a decline in species richness and mean cover, as well as in forage quality and soil protection.

The results of the present work make it possible to propose differential use of the 3 communities described. Grassland is a good site for pastoral use and does not appear intensely degraded in terms of cover and grass quality. Taking into account the potential effects of grazing on vegetation structure and composition, management practices to favor disseminule generation, such as a resting period during seed production and dispersion stages (rotational grazing use), should be implemented. Scrubland—because of the dominant vegetation type and steep slopes—could be used to provide only wood, as long as plants are not uprooted. Features of sparse grassland define it as a degraded unit, poorly fit for pastoral use, with high risk of erosion. Recovery measures such as permanent closure, planting of native species, and fertilizing should therefore be applied.

The poverty of the families in the study area, precarious land ownership, and changes in the domestic economy caused by transformations in regional production (Hocsman 2000; Quiroga Mendiola 2000) pose problems that will be difficult to solve when trying to optimize the system. The study team is presently working on implementation of management improvements in pastoral procedures and natural vegetation (soil systematization, emergency community aid with shifting funds, and irrigation systems for intensification of forage production, among other things).