Soil Erosion As Affected by Shrub Encroachment in North- Eastern Patagonia

Soil erosion as affected by shrub encroachment in northeastern Patagonia

Item Type text; Article

Authors Parizek, B.; Rostagno, C. M.; Sottini, R.

Citation Parizek, B., Rostagno, C. M., & Sottini, R. (2002). Soil erosion as affected by shrub encroachment in northeastern Patagonia. Journal of Range Management, 55(1), 43-48.

DOI 10.2307/4003261

Publisher Society for Range Management

Journal Journal of Range Management

Download date 02/10/2021 08:23:55

Item License http://rightsstatements.org/vocab/InC/1.0/

Version Final published version

Link to Item http://hdl.handle.net/10150/643623 J. Range Manage. 55: 43-48 January 2002 Soil erosion as affected by shrub encroachment in northeastern Patagonia.

BERNARDO PARIZEK, CESAR M. ROSTAGNO, AND ROBERTO SOTTINI

Authors are Range Ecologist, Vector Argentina S.A, Mendoza, Argentina, (at time of the research B. Parizek was a Research Assistant, Universidad Nacional de la Patagonia SJB, Puerto Madryn), Soil Scientist, Centro Nacional Patagonico, CONICET, Bvard. Brown s/n, 9120, Puerto Madryn, Chubut, Argentina, and Hydrologist, Instituto Nacional de Limnologia, CONICET, J. Macia 1933, 3016 Santo Tome, Santa Fe, Argentina.

Abstract Resumen

Soil erosion is the primary cause of irreversible loss of soil La erosion de los suelos es la causa principal de las perdi- productivity on most rangelands. In northeastern Patagonia, das irreversibles del potencial productivo de los suelos en la the increase in soil erosion has been closely associated with mayoria de los pastizales naturales. En el nordeste de the increase in shrub cover in the grass or shrub-grass Patagonia, el aumento de la erosion de los suelos ha estado steppes. We used rainfall simulation to compare infiltration estrechamente asociado al aumento de la cobertura de arbus- and sediment production from patches of grass, shrub-grass, tos en las estepas herbaceas o arbustivas-herbaceas. Nosotros and shrub steppes of the Punta Ninfas range site. Bare soil empleamos lluvia simulada para determinar la tasa de infil- and gravel covers were higher and litter cover was lower in tracion y la produccion de sedimentos en parches de estepas the shrub steppe than in the shrub-grass and the grass herbaceas, arbustivas-herbaceas y arbustivas del sitio steppes. In the shrub inter-spaces of the shrub steppe, bulk ecologico Punta Ninfas. Las coberturas de suelo desnudo y de density was greater and macroporosity and soil organic mat- gravas fueron mayores y la cobertura de mantillo menor en ter were lower (P 0.05) than in the mounds beneath shrubs la estepa arbustiva respecto a las estepas arbustiva- herbacea and in the grass and shrub-grass areas. Infiltration rate was y herbacea. En los espacios entre arbustos de la estepa arbus- 60 to 65% lower in the shrub steppe than in the grass and tiva, la densidad aparente fue mayor y la macroporosidad y shrub-grass steppes, respectively. On the contrary, total sedi- la materia organica fueron menores (P 0.05) que en los ment production and concentration were higher (P < 0.05) in monticulos debajo de los arbustos y en las estepas arbustiva- the shrub steppe as compared to the grass and the shrub- herbacea y herbacea. La tasa de infiltration fue un 60 y un grass areas. Gravel cover was the variable that best predicted 65% mas baja en la estepa arbustiva que en las estepas her- infiltration and sediment production. The organic matter bacea y arbustiva-herbacea, respectivamente. Por el con- content of the sediment, mostly litter, in the shrub and the trario, la produccion total y la concentracion de sedimentos shrub-grass steppes were similar and greater (P 0.05) than fueron mas altas (P s 0.05) en la estepa arbustiva comparado in the grass steppe. Runoff litter removal may represent one con las estepas herbacea y arbustiva-herbacea. La cobertura of the processes that drive the transition from shrub-grass to de gravas fue la variable que mejor predijo la tasa de infil- shrub steppes. High rates of sediment removal, mainly litter, tracion y la produccion de sedimentos. El contenido de mate- from the shrub interspaces of the shrub steppe may limit the ria organica de los sedimentos, mayormente mantillo, fue natural recovery of the soil physical and hydrological proper- similar en la estepa arbustiva y la arbustiva-herbacea y en ties. These degraded patches fail to capture incident rainfall ambas mayores (P 0.05) que en la estepa herbacea. La and restrict the possibilities for the recovery of perennial remocion de mantillo por el escurrimiento superficial posible- grasses favoring the dominance of shrubs. mente represente uno de los procesos que provocan la transi- cion de la estepa arbustiva-herbacea a la estepa arbustiva. Las altas tasas de remocion de sedimentos, principalmente Key Words: vegetation, litter, gravel cover; shrub, grass, mantillo, de los espacios entre arbustos de la estepa arbustiva shrub-grass steppes, organic matter removal, rainfall simula- pueden limitar la recuperacion natural de las propiedades tor, infiltration rate. fisicas a hidrologicas de los suelos. Estos parches degradados no pueden captar las lluvias incidentes, limitando asi las posi- bilidades de recuperacion de los pastos perennes y favore- Shrub encroachment and soil erosion have been identified as ciendo la dominancia de los arbustos. the main degradation processes in semi-arid grazing rangelands (Friedel 1991). Both processes can be closely related and in gen- eral erosion follows shrub invasion (Buffington and Herbel 1965). Schlesinger et al. (1990) suggest that overgrazing results in the redistribution of organic matter and nutrients and is the primary agent responsible for the current conversion of previously productive grasslands to unproductive mesquite (Prosopis glan- This work was funded by the CONICET. dulosa Ton.) shrublands. Manuscript accepted I 1 Mar. 01.

JOURNAL OF RANGE MANAGEMENT 55(1) January 2002 43 In most of the Patagonian rangelands, grazing appears to have modified the vegetation and accelerated soil erosion 24° processes (Soriano et al. 1983, Ares et al. 1990). Changes in vegetation include the increase of low forage quality shrub species such as neneo (Mulinum spinosum (Cav.) Pers.) and quilenbai (Chuquiraga avellanedae Cav.) (Bertiller 1993, 32° Beeskow et al. 1995) in some originally productive grassslands. In a range site of northeastern Patagonia, Beeskow et al. (1995) identified 3 stable states following the state and transition model proposed by Westoby et al. (1989). These states represent stable plant communities along a degradation gradient, with grass, shrub-grass, and 44° shrub steppes representing a good, fair, and poor condition from a forage production perspective. However, no information exists on how well these plant communities protect the site against accelerated erosion. Recently, the Task Group on Unity in Concepts and Terminology 52° (1995) proposed the concept of Desired Plant Community. The Desired Plant Community refers to any managed plant 7S° 72° 66° 6B° 54° 5®° community that has the capability to protect the site against accelerated erosion. In this study we used rainfall simulation to assess how different plant communities on a range site of northeastern Patagonia influenced infiltration and interrill erosion. 72° 64°

42° -I Materials and Methods

Study Area Study area The study was conducted in the upper part of a flank pediment that extends from the plateau border to a playa lake in the Punta Ninfas area, 70 km east of Puerto Madryn in the NE portion of the Chubut province (43:00'S, 64:30'W) (Fig. 1). It is an undulating plateau dissected by coastal 46° valleys and interrupted by numerous depressions with playa lakes. The climate Southern extreme of South America with Argentina and is and and temperate. Mean annual precip- the Chubut province. itation is 254 mm (1955-1992). Most of the rainfall occurs during the cold season Fig, 1. Location of the study area. from April to September. Mean annual temperature is 12.5°C. The mean annual The study area is located in the ecotone Dominant plant physiognomy is a wind velocity at 10 m above ground level between the Patagonian and Monte phyto- shrub-grass steppe although patches of is 4.6 m sec-1 (Barros 1983). Dominant geographic provinces. Principal species grass steppe and shrub steppe are present. soil in the flank pediment area is a Xeric representing the first province are quilen- These physiognomic patches have been Calciargids. This soil is shallow, with a bai and colapiche (Nassauvia fuegiana described as stable states (Beeskow et al. loamy sand A horizon 10-15 cm thick, (Speg.) Cabrera), and coiron (Stipa spe- 1995). Within the grass and grass-shrub and a sandy loam B2t horizon 15-20 cm ciosa Trin. and Rupr.). The cool season steppe patches a stable soil is present; in thick, underlain by a calcic horizon. A grasses flechilla (S. tenuis Phil.) and the shrub steppe, shrubs are associated gravelly sand to sandy clay alluvium 50 to flechilla negra (Piptochaetium napos- with mounds and are distributed in a 80 cm thick forms the soil substratum. taense (Speg. Hackel ap Stuckert.)) are the matrix of eroded soil with desert pavement This deposit of Holocene age rests on ter- main species of the Monte province. and a low vegetation cover. tiary sediments.

44 JOURNAL OF RANGE MANAGEMENT 55(1) January 2002 Sheep grazing for wool production was Table 1. Average canopy and ground cover (%) characteristics (standard deviation) for each plant introduced in the area at the beginning of community at the Punta Ninfas range site, Patagonia. the last century. Continuous grazing is practiced extensively in pastures exceed- Grass Shrub-Grass Shrub ing 2,500 ha. The mean stocking rate is Steppe Steppe Steppe ------0.25 sheep ha 1. (%) Shrubs 0 23.0 Perennial Grasses 24.4 (1.1) Experimental Procedures Perennial Dicot 7.4 (3.5) We selected 3 homogeneous areas, 10 x Annuals 12.8 (3.0) 10 m each representing, the grass, shrub- Bare Soil 18.8 (2.9) grass, and shrub steppes. Inside these Gravel 11.8 (1.9) areas, 10 plots measuring 0.63 x 1.60 m (1 Litter 24.0 (3.7) m 2) were located in each plant community. Slope was homogeneous across the 3 areas with an average of 7.2%. Sheet metal frames were dug into the perimeter cent to each plot. Roots were separated by Results and Discussion of the plots to channel the runoff generat- hand, washed, dried at 105° C and weighed; gravel was separated with a 2 ed by the simulated rainfall. Runoff leav- Surface characteristics ing the lower border of the plots was chan- yam sieve. Soil samples from this same Soil surface characteristics and plant neled through a 5 cm diameter pipe con- depth were collected and analyzed for tex- cover for each plant community are pre- nected to 5 liter containers. In December ture by the pipette method (Day 1965), sented in Table 1. Total vegetation cover of 1994 and 1995, simulated rainfall was and organic matter by the loss on ignition presented the highest value in the shrub- applied with a full cone, single nozzle method (Davies 1974). Total porosity was grass steppe and the lowest in the shrub rainfall simulator (Rostagno and Garayzar calculated as: (1-bulk density/particle den- steppe, respectively. In the shrub steppe, 1995) at an intensity of 100 mm hour-', sity) considering a particle density of 2.65 3 the main changes relative to the shrub- during 30 minutes. This rainfall simulator cm and macroporosity as the difference g grass community were decreased litter and produces a rainfall with a kinetic energy between water content at saturation and at herbaceous vegetation cover and increased equivalent to 66% of the kinetic energy of a field capacity (Helalia 1993). Field capaci- gravel cover. In the shrub steppe where natural rainfall at the same intensity. In the ty (moisture equivalent) was estimated by erosion has been more intense than in the study area, high intensity rainfall can occur centrifuging saturated samples (30 min; other plant communities, the A horizon from December to March. A rainfall event 2440 rpm). remaining beneath the shrubs as well as with the intensity and duration of the simu- the accumulation of wind blown and lated rainfall occurs about once per 100 Data Analysis splashed material give rise to small years (Vicenty et al. 1984). Runoff was col- Data were analyzed using correlation mounds associated with the shrubs, mainly lected at 1 and 2 minutes after rainfall initi- (Pearson) to assess the linear association quilenbai. However, the increase in bare ation, and then at 5 minutes intervals in of the variables and stepwise multiple soil cover in the shrub steppe was less separate containers and determined by vol- regression to identify the variables that than 2% compared to the grass steppe. ume. Time to runoff was recorded for each best predicted infiltration and sediment Soil characteristics of the grass and the plot. Infiltration rate was calculated for production. Analyses of variance were shrub-grass steppes were similar and dif- each interval as the difference between the performed on the terminal infiltration and fered greatly with those of the shrub applied rainfall and the runoff collected for sediment (mineral and organic) production steppe (Table 2). The 2 contrasting surface each interval. Total runoff was passed data using the SPSS package. Mean sepa- conditions present in the shrub steppe through a 44 pm sieve. The sediment coars- ration with the protected LSD was used to (mounds beneath the shrubs and desert er than 44 jim was recovered dried at 105° compare infiltration and sediment produc- pavement covered areas in the shrub inter- C for 24 hours and weighed. The weight of tion in each plant community. Significant spaces) were also different in term of soil the sediment < 44 pm was calculated using levels were determined at P 0.05. characteristics. Sand, organic matter, and 50 ml aliquot. After drying and weighing, root content were significantly lower; the total sediment <44 pm was calculated gravel and bulk density were higher in the considering total runoff volume, then it was eroded shrub interspaces of the shrub added to the > p4 mm sediment and con- verted to kg ha'. Previous to simulated rainfall applica- Table 2. Average soil characteristics (standard deviation), 0 to 5 cm soil depth, for each plant com- tion, ground cover (vegetation, litter, bare munity and 2 microsites in the shrub steppe at the Punta Ninfas range site, Patagonia. soil, and gravel) was determined using the point quadrat method (Goodall 1953). A Grass Shrub-Grass Shrub Steppe metal frame with holes every 5 cm was Steppe Steppe SI placed on the plot and 66 points were Bulk Density (Mg m 3) 1.06 (0.07) (0.11) (0.15) (0.05) measured in each plot. The slope of the Macroporosity (%) 31.9 (1.2) (1.8) plots was determined using the same Gravel Content (%) 17.4 (9.2) (14.5) (3.6) Sand (%) 73.0 (3.8) (5.1) (6.4) as frame for ground cover. Root Content (g m 2) 145.6 (31.1) (28.6) (9.3) (83.0) A 130 cm-3 soil core (0 to 5 cm) was Organic matter (%) 2.49 (0.29) (0.36) (0.19) (0.31) sampled for bulk density, root, and gravel 51= shrub interspaces content determinations from an area adja-

JOURNAL OF RANGE MANAGEMENT 55(1) January 2002 45 the soil infiltration (i.e., perennial grass and litter cover). To assess the differences in infiltration capacity of the degraded soils of the shrub interspaces and the mounds beneath shrubs in the shrub steppe, we determined their infiltration capacity separately by means of 15 cm diameter cylinders. Average (and standard deviation) infiltration rates for the first 10 minutes were 8.3 (3.5) and 175 (14.0) mm hour' for the shrub interspace and the mound, respectively. In the grass steppe the average infiltration rate was 103.7 (22.0) mm hour'. Although measurements of infiltration 0 5 10 15 20 25 30 35 capacity using cylinder infiltrometers tend to exceed the true (vertical) infiltration Time (min.) capacity (Tricker 1978), our results show Fig. 2. Average infiltration rate across the 30 minute simulated rainfall for the soils of the the large hydrological contrast between shrub-grass (S), grass (), and shrub (0) steppes at Punta Ninfas range site, Patagonia. R the crusted soil of the shrub interspaces is applied rainfall intensity. Plant community means with the same letter are not signifi- and the mounds. It is interesting to notice cantly different (P 0.05). that the infiltration rate of the soil beneath the shrub (mound) was also significantly steppe as compared to the soil of the other runoff (or infiltration rate) are contradicto- higher than in the soil of the grass steppe. plant communities. These differences were ry (Poesen 1992). Mehan (1986) showed Thus, although the quilenbai positively greater with respect to the soils beneath that coarse fragments on the soil surface of affected the infiltrability of the soil under the shrubs (mounds) of the shrub steppe. rangelands tended to increase infiltration its canopy, the patches where quilenbai is rates. It appears that as more soil surface dominant had the lowest infiltration. Infiltration Characteristics was exposed, more of the soil surface develops a crust thus reducing infiltration Infiltration curves for each plant commu- Sediment Production rates. nity are presented in Fig. 2. In the grass and Sediment production in the grass and the In our study, gravel cover integrated soil the shrub-grass communities, runoff started shrub-grass steppes were similar and sig- and soil surface characteristics which were at 5 and 7 min after rainfall initiation, nificantly lower than in the shrub steppe respectively. This difference can be a result considered to negatively affect the soil (Fig. 3). In the shrub steppe the sediment of the higher water interception/retention of infiltrability. Thus, in the shrub inter- concentration was also significantly higher the shrub canopy in the shrub-grass com- spaces of the shrub steppe, where the than in the other communities. Sediment munity. In these communities, the infiltra- higher gravel cover was found, gravel is concentration may provide a better mea- tion curves were similar after 15 min of the well embedded in a surface crust. Valentin sure of the erodibility of each soil surface rainfall initiation. In the shrub steppe, (1994) found a reduction in infiltration condition. However, sediment concentra- rate with increasing cover of embedded runoff started at less than 1 min after rain- tion was low, 2.4, 1.7, and 1.8 g liter"' for fall initiation. In this plant community, the rock fragments. Equally, the shrub inter- the shrub, shrub-grass, and grass steppes, infiltration rate decreased abruptly the first space soil presents a higher bulk density respectively. Sediment concentrations as (Table 2) as compared to the soils of the 10 min; after which it decreased more high as 40 g liter"' were recorded in a slowly until the end of the simulated rain- shrub-grass and the grass communities as shrub steppe with a Typic Torriorthent soil fall. well as the mound soils. The increase in with a natural rainfall event (Rostagno et Terminal infiltration rate in the grass bulk density of the shrub-interspace soil of al. 1999). The low bare soil cover in the and shrub-grass steppes were similar and the shrub steppe represents a proportional grass, shrub grass, and shrub steppes significantly higher than in the shrub decrease in macroporosity (Table 2). (<21%) and the low kinetic energy of the steppe. Stepwise regression analysis using Because of their texture (loamy sand and simulated rainfall (66% of a natural rain- the variables presented in Table 1 as inde- sandy loam), the soils of the study area are fall event of the same intensity) may have pendent variables, gave us the following susceptible to crust formation (Poesen limited sediment detachment by raindrop predictive model for infiltration rate (IR) 1988). However, a crust develops only in impact. at 30 min: the shrub interspaces of the shrub steppe, Stepwise regression analysis using the probably related to the low organic matter IR = 39.4 - 0.79 Gravel cover + variables presented in Table 1 as predic- litter and grass 1.33 Annuals; R2 = 0.69 (P = 0.001) (1) content and low cover of tors produced the following predictive this soil. Our results produced a negative rela- model for sediment production (SP): SP (g The cover of annual plants was selected tionship between gravel cover and infiltra- m 2) = 11.6 + 3.0 Gravel cover (%); R2= as the second variable of importance by tion rate (r -0.74). Similar results were 0.46 (P = 0.005), indicating that sediment = the regression analysis. Annual cover was found in a degraded shrub steppe with a production increased as gravel cover negatively related to gravel cover (r Typic Torriorthent soil (Rostagno 1989). = increased. This result is apparently contra- -0.56) and integrated other soil surface However, field-plot data on the relation- dictory, as gravel cover has been shown to ship between coarse fragments cover and variables considered to positively affect decrease interrill erosion in most cases

46 JOURNAL OF RANGE MANAGEMENT 55(1) January 2002 120 further increase the differences between N Organic the grass or grass-shrub and the shrub- dominated communities. This will also 100 a 0 Inorganic tend to increase the contrast between soils beneath shrubs and the soils of the shrub 0 80 N interspaces. Schlesinger et al. (1990) considered the increase in soil heterogeneity 0 60 in shrub invaded areas as an indicator of 0 a a. b rangeland degradation. 40 It is important to notice that differences 0 in total plant cover between the grass and 20 B B shrub steppes were small (i.e., < 5%). It a) to seems that more important than total plant cover as causative factor in interrill as well as rill erosion, is plant cover distribution SS SGS GS (Weltz et al. 1998). Thus, while in the Plant community grass steppe plant cover was homoge- neously distributed, in the shrub steppe it Fig. 3. Inorganic and organic sediment production for the shrub, shrub-grass, and grass was concentrated in small patches. Plant steppes at the Punta Ninfas range site, Patagonia. Plant community means with the same cover distribution determines the size of lowercase and uppercase letters are not significantly different (P 0.05) for the organic the patches of bare soil (or of gravel cov- and inorganic sediment production, respectively. ered and crusted soil). De Soyza et al. (1998) defined a bare patch index (bpi = (Poesen et al. 1994). In the study area, erodibility of these soils. In the grass com- mean size of bare patches x proportion of gravel cover was negatively associated to munity, litter cover accounted for 26% of bare soil) as a desertification indicator for herbaceous vegetation (r = -0.61) and lit- the ground cover and decreased to 6.4% in the Chihuahuan Desert. They found that ter cover (r = -0.68). The highest value of the shrub steppe. Decline in the production sites with high percentage of shrubs and gravel cover was present in the shrub of litter because of the low herbaceous sites near livestock watering points had interspaces of the shrub steppe where a cover in the shrub interspaces and the high the greatest bare patch index. soil surface condition of low infiltrability potential for litter removal by runoff from Equally important is the connection and high runoff production dominates. these areas may increase the hydrological among bare patches. In the shrub steppe, the Although gravel cover may decrease sedi- and physical differences between the com- degraded shrub interspaces connect each ment production, it seems to be less effec- munities. In the shrub steppe, the litter other allowing the runoff to leave the local tive than vegetal (Wilcox and Wood 1989) produced by shrubs (leaves, twigs, and plant community (i.e., the shrub steppe and litter cover (Benkobi et al. 1993) to inflorescence) accumulates beneath the patch) and eventually the range site. Under protect the soil from erosion. For example, shrubs and very little, if any, is spread present land use (i.e., continuous grazing) Benkobi et al. (1993) found that the litter beyond the shrub canopy. This decrease in an increase in the size and numbers of shrub and gravel cover necessary to reduce soil vegetation and litter cover in the shrub steppe patches (degraded patches) can be loss to 50% of bare ground was 60% and steppe is reflected in the lower organic expected, as shrubs are little affected by 85%, respectively. While both gravel and matter content of the shrub interspace soil grazing. The grass steppe can be considered litter decrease sediment detachment by (Table 2). A mechanism by which litter the Desired Plant Community (Task Group raindrop impact proportional to their may enter the shrub interspaces of the on Unity in Concept and Terminology is deposition of litter car- cover, litter more effectively absorbs the shrub steppe the 1995) for the site as it best meets forage by generated from the sur- erosive and transport energies of overland ried runoff production objectives. However, both, the flow, reducing flow velocities, and rounding grass or shrub-grass communi- grass and the shrub-grass steppes have the ties. In the shrub-grass steppe, the sedi- increasing deposition. capability to protect the site against acceler- ment organic matter content was also ated erosion. higher than in the grass steppe, and pro- In this range site where erosion dramati- Sediment Organic Matter Content portionally higher (percent of mineral sed- cally changes the hydrological properties The organic matter content of the sedi- iment) than in the shrub steppe. of the soil, the application of shrub control ment was significantly higher in shrub and techniques (i.e. prescribed fire) that effec- in the shrub-grass communities than in the Shrub Encroachment, Soil Erosion, tively help to restore the grass cover grass community (Fig. 3). Although a high should be applied in the first stages of proportion of organic matter in the sedi- and Sustainable Production shrub invasion (i.e., in the shrub-grass ment was plant residues, organic matter Although the main cause of the decline steppe state). In areas close to the study content in the sediment was inversely pro- in herbaceous vegetation producing the site accidental fires have temporarily portional to litter cover. In the shrub change from shrub-grass to shrub steppes reversed shrub-dominated communities (i.e., shrub compe- steppe where litter cover was much lower in shrub invaded areas back to grass steppes. It is also probable is clear, it than in the other plant communities, the tition, intense grazing, etc.) not that a grazing scheme that includes spring high amount of runoff generated in this is clear that past erosion in the shrub rest might favor perennial grasses in the community may explain the higher steppe has changed the characteristics of shrub-grass steppes, slowing or reversing amount of organic matter in the sediment. the soil and the distribution of soil shrub cover increase. Bork et al. (1998) The feedback between The decrease in litter cover following the resources. positive found in a sagebrush steppe that grazing decrease of grass cover may increase the past erosion and present erosion rate may

JOURNAL OF RANGE MANAGEMENT 55(1) January 2002 47 during the spring increased live shrubs and Beeskow, A.M., N.O. Elissalde, and C.M. Rostagno, C.M. 1989. Infiltration and sedi- decreased perennial grasses cover as com- Rostagno. 1995. Ecosystem changes associ- ment production as affected by soil surface pared to a fall-grazed treatment. In the ated with grazing intensity on the Punta conditions in a shrubland of Patagonia, shrub steppe where degradation has Ninfas rangelands of Patagonia, Argentina. J. Argentina. J. Range Manage. 42:382-385. Range Manage. 48: 517-522. Rostagno, C.M. and D. Garayzar. 1995. severely affected the soil characteristics, it Benkobi, L., M.J. Trlica, and J.L. Smith. Diseno de un simulador de Iluvia para estu- is less probable that herbaceous vegetation 1993. Soil loss as affected by different com- dios de infiltracion y erosion de suelos. cover may be restored and sustainable binations of surface litter and rock. J. Ciencia del Suelo 13:41-44. management be achieved, unless a tech- Environ. Qual. 22:657-661. Rostagno, C.M., H.F. del Valle, F. Coronato, nique of water conservation is applied. Bertiller, M.B. 1993. Estepas graminosas de and D. Puebla.1999. Runoff and erosion in However, because of drastic changes in Festuca pallescens en el SW del Chubut. five land units of a closed basin of northeast- soil properties, this plant community may p.14-22. En: Paruelo, J.M., M.B. Bertiller, ern Patagonia. Arid Soil Res. & persist and represent a different ecological T.M. Schlichter y F. Coronato (eds.), Rehabilitation. 13:281-292. site. Secuencias de deterioro en distintos ambi- Schlesinger, W.H., J.F Rynolds, G.L entes Patgonicos. Su caracterizacion medi- Cunningham, L.F. Huenneke, W.M. anteel modelo de Estados y Transiciones, Jarrel, R.A Virginia, and W.G. Whitford. Cooperacion INTA-GTZ, S.C. de Bariloche. 1990. Biological feedbacks in global deserti- Conclusions Bork, E.W., N.E. West, and J.W. Walker. fication. Sci. 247:1043-1048. 1998. Cover components on long-term sea- Soriano, A., W. Volkheimer, H. Walter, E.O. Soil surface characteristics as well as sonal sheep grazing treatments in three-tip Box, A.A. Marcolin, J.A. Valerini, C.P. soil infiltrability and sediment production sagebrush steppe. J. Range Manage. Movia, R.J. Leon, J.M. Gallardo, M. Rumboll, M. Canevari, P. Canevari, and differed significantly in shrub dominated 51:293-300. Buffington, L.C. and C. H. Herbel. 1965. W. Vasina. 1983. Deserts and semi-deserts and grass or grass-shrub dominated plant Vegetation changes on a semi-desert grass- of Patagonia. p. 423-460. In: N.E. West communities on the same range site. In the land range from 1858 to 1963. Ecol. Monogr. (ed.), Ecosystems of the world. Temperate Punta Ninfas range site, a continuous grass 35:139-164. deserts and semi-deserts. Elsevier, cover may coexist with a shrub cover of Davies, B.E. 1974. Loss-on-ignition as an esti- Amsterdam. approximately 25%. Although the shrub- mate of soil organic matter.Soil Sci. Soc. Task Group on Unity in Concept and Terms grass steppe is undesirable from a forage Amer. Proc. 38: 150-151. Committee Members. 1995. New concepts production perspective because the main Day, P.R. 1965. Particle fractionation and par- for Assessment of Rangeland Condition. J. shrubs have a very low forage value, soil ticle-size analysis, p. 545-567. In: C.A. Range Manage. 48:271-282. conservation is not different from the Black (ed.). Methods of soil analysis. Tricker, A.S. 1978. The infiltration cylinder: Monogr. Ser. Part I, No. 9, Amer. Soc. some comments on its use. J. Hydro. grass steppe. On the contrary, the shrub- Agron., Madison, Wis. 36:383-391. dominated community represents a poor De soyza, A.G., W.G. Whitford, J.E. Valentin, C. 1994. Surface sealing as affected condition from a forage production and a Herrick, J.W. Van Zee, and K.M. by various rock fragment covers in West soil and water conservation perspective. Havstad. 1998. Early warning indicators of Africa. CATENA 23:87-97. Although the development of a dense desertification: examples of tests in the Vicenty, O.A., J.J. Serra, and A.D. Gabetta. desert pavement in the eroded shrub inter- Chihuahuan Desert. J. Arid Environ. 39: 1984. Estudio de las precipitaciones en la spaces of the shrub steppe may limit sedi- 101-112. ciudad de Trelew, Chubut. 44 pp. MESOP, ment detachment, and thus, sediment pro- Friedel, M.H. 1991. Range condition assess- Rawson, Chubut, Argentina. duction, the high runoff production from ment and the concept of threshold: A view- Weltz, M.A., M. Kidwell, and H. Dale Fox. these areas keep the erosion rate of the point. J. Range Manage. 44:422-426. 1998. Influence of abiotic and biotic factors Goodall, D.W. 1953. Some considerations in in measuring and modeling soil erosion on shrub steppe above the erosion rates of the the use of point quadrats for the analysis of rangelands: State of knowledge. J. Range other communities. vegetation. Aust. J. Sci. Ser. B. 5:1-41. Manage. 51:482-495. The high rate of organic matter (mainly Helalia, A.M. 1993. The relation between soil Westoby, M., B. Walker, and I. Noy-Meir. litter) and water losses from the shrub infiltration and effective porosity in different 1989. Opportunistic management for range- steppe may limit the natural recovery of soils. Agr. Water Manage. 24:39-47. lands not at equilibrium. J. Range Manage. the soil physical and hydrological proper- Mehan, D. 1986. Effects of coarse fragments 42:266-274. ties and the possibilities for the re-estab- on infiltration rates and Green and Ampt Wilcox, B.P. and M.K. Wood. 1989. Factors lishement of the perennial grasses, and parameters. MS Thesis, Range Sci. Dept., influencing interrill erosion from semiarid favor the dominance of shrubs. Utah State Univ., Logan, Utah. slopes in New Mexico. J. Range Manage. Poesen, J.W. 1988. Surface sealing on sandy 42:66-70. and loamy soils: some aspects of seal formation and the influence of sealing on water Literature Cited erosion subprocesses. Quaderni de Scienza del Suolo l: 9-19. Ares, J.O., A.M. Beeskow, M.B. Bertiller, Poesen, J.W. 1992. Mechanisms of overland- C.M. Rostagno, M.P. Irisarri, J. flow generation and sediment production on Anchorena, G.E. Defosse, and C.A. loamy and sandy soils with and without rock Merino. 1990. Structural and dynamic char- fragments, p. 275-305. In: A.J. Parsons and acteristics of the overgrazed grasslands of A.D. Abrahams (eds.), Overland Flow northern Patagonia, Argentina. p. 149-175. Hydraulics and Erosion Mechanics. UCL In: A Breymeter (ed.), Managed Grasslands: Press, London. Regional Studies. Elsevier, Amterdam. Poesen, J.W., D. Torn, and K. Bunte. 1994. Barros, V. 1983. Atlas del potencial eolico de Effects of rock fragments on soil erosion by la Patagonia. Contribucion No. 69. Centro water at different spatial scales: a review. Nacional Patagonico. Puerto Madryn, Catena 23:141-166. Argentina.

48 JOURNAL OF RANGE MANAGEMENT 55(1) January 2002