Introduction

The cattle production industry in 75 % of continental Argentina, characterized by arid and semi-arid areas, is based on grazing of native vegetation (Busso & Fernández, 2018). Unfortunately, most of these areas have been exposed to overgrazing, and fire mismanagement, resulting in shrub encroachment and the local disappearance of acceptable, palat­able (desirable) perennial forage grasses for cattle grazing (Fernandez & Busso, 2018). As a result, rangelands of central Argentina have almost a unique warm­season, native perennial grass that is an acceptable forage to grazing livestock (Pappophorum vaginatum, Busso et al., 2004). It is imperative to increase desir­able perennial grass species in this region (Anderson, 1980) to address the need for additional forage and subsequent increase in cattle production. Bleak et al. (1966) emphasized that appropriate management of native vegetation would be the best approach to achieve a good plant cover of forage that is acceptable to grazing livestock in those rangelands. Severe water stress, especially during the warm season, is another constraint the vegetation might be exposed to in range­lands of central Argentina (Giorgetti et al., 1997).

In addition to P. vaginatum, other na­tive, warm-season, less abundant, pe­rennial grass species in this region are Aristida spegazzinii, A. subulata, and Sporobolus cryptandrus (Giorgetti et al., 2000a). This last species is a palatable perennial bunchgrass, but it is not only of lower abundance but also of low vig­or. The Aristida species are less palat­able and only eaten if a better forage is not available (Giorgetti et al., 1997). As a result, their basal area keeps expand­ing in these less grazed tussock peren­nial grasses. This results in more new shoot biomass production than that in the more palatable grazed P. vaginatum (Flemmer et al., 2003). Pappophorum vaginatum has resisted continuous graz­ing for decades, and its abundance in these rangelands is even greater than that of other less preferred perennial grasses (Aristida species, Giorgetti et al., 2000c). This suggests an important tolerance to defoliation or competitive ability in this species.

Several studies have been made on the morpho-physiology, demography, plant growth and responses to grazing in P. vaginatum (Torres et al., 2011, 2014). However, it is not yet clear which are the mechanisms responsible for its long-term persistence. Rapid reestablishment of a photosynthetic canopy is a distinc­tive feature of the perennial grasses toler­ant to defoliation (Caldwell et al., 1981). The number of leaves and tiller height are between the growth components that contribute most to photosynthetic cano­py reestablishment (Becker et al., 1997b). A high capacity for new tiller formation of rapid growth after forage removal will allow plants to reestablish the photo­synthetic tissue lost to herbivores, rap­idly recuperate the root-shoot balance, obtain a greater share of soil available resources, and maintain their competi­tive ability within the plant community (Busso & Richards, 1995). If defoliation has negative effects on one or more of these growth components, plant recu­peration to such event can result limited (Busso & Richards, 1995). In general, it has been proposed that the mechanisms of plant response to defoliation will also be affected by suboptimal environmen­tal conditions (Langer, 1963), including drought (Busso et al., 2003).

Tillering in communities of grasses ex­posed to defoliation is important for its survival. Even more, the exchange rates in tiller density and its weight are im­portant components in the production of dry matter (Wade, 1982). Daughter tiller production not only increases tiller density on plants but also contribute to increase the plant bud bank and imme­diate carbon gain because they are pho­tosynthetically active at the time parent tillers reach senescence (Olson & Rich­ards, 1988a,b). Species tolerant to de­foliation can produce a greater number of daughter tillers of fastest growth than those susceptible to defoliation (Cald­well et al. 1981; Busso & Richards, 1995).

Different studies have reported a va­riety of responses in the rate of leaf ap­pearance after a severe defoliation of perennial grasses, from increases to no changes (Sáenz & Deregibus, 2001) to even reductions (Chapman et al., 1983). Possible changes in the environmental conditions (e.g., light intensity and tem­perature), particularly to a level of the apical meristem, might modify the rate of leaf appearance on defoliated plants (Anslow, 1966). On the other hand, neg­ative responses have been associated with defoliations at late developmental morphology stages, in coincidence with high air temperatures and low soil mois­ture levels (Chapman et al., 1983).

The speed and magnitude of the pho­tosynthetic canopy reestablishment after defoliation depend on the number, na­ture and location of the removed mer­istems (Korner, 1991), which is closely associated to the developmental mor­phology stage of the plant. Greater total leaf lengths have been produced after active meristems remained on the plant after defoliation (Becker et al., 1997b). Because of this, the timing when the de­foliation events occur within the grow­ing season will have a direct effect on the subsequent forage production, survival, and recuperation of the plants (Olson & Richards, 1988a). Several studies ob­served increases in: (1) the stem + sheath lengths, and (2) total leaf lengths when defoliation occurred earlier during the growing season or when there was no defoliation of perennial grasses and her­baceous species (Paige, 1992). Knowl­edge of when a plant reaches certain stage of developmental morphology is very important for the synchronization of management guidelines to determine the most appropriate time for grazing, thus minimizing the negative effects of herbivory on the desirable components of vegetation.

Degradation of the plant population induced by grazing would eventual­ly reduce tiller number and total basal area, potentially reducing productivity and plant competitive ability within the community (Briske & Richards, 1995). However, Giorgetti et al. (2000b, 2006), working on exclosures, reported positive responses in the native perennial grasses P. vaginatum, A. subulata, A. spegazzinii y S. cryptandrus under different kinds of management and clipping frequencies. Sporobolus cryptandrus, a widely studied species in the United States of America, has shown increases, decreases or ab­sence of considerable changes in relation to its response to grazing (Bragg, 1978; Kleiner, 1983).

Given a competitive situation, a spe­cies increases its possibilities of captur­ing soil mobile and not mobile nutrients by increasing its root length (Ryser & Lambers, 1995). Under soil water stress conditions, which determine that water is absorbed by only a fraction of the root system, the amount of water that enters the plant might be closely related to the root length density (root length per unit soil volume; RLD) where water is avail­able. As a result, a greater RLD might favor water and nutrient uptake, thus contributing to increase competitive abilities. This is especially true in unpro­ductive environments where competi­tion for belowground resources is strong.

Root growth and respiration, and nu­trient uptake, depend upon the contin­uous supply of soluble carbohydrates produced by the plant photosynthetic tissues (White, 1973). Root activity may consume more than half of the available photoassimilates on mature plants (Fo­gel, 1985). Some authors have reported that carbon availability within the roots is immediately reduced after defoliation in response to a photosynthesis reduc­tion and to a preferential photosynthetic carbon allocation to meristems and de­velopment of new tissues (McNaughton, 1983). However, other authors have re­ported an increase in carbon allocation to roots, after hours from defoliation, on species adapted to grazing (Holland et al., 1996).

The role of grazing by domestic live­stock must be considered when evalu­ating strategies that contribute to com­petitive ability of rangeland vegetation. Factors that reduce photosynthetic ca­pacity (e.g., grazing) will also reduce soil resource acquisition mechanisms because these are dependent upon plant carbon fixation (Briske & Richards, 1995). Some grazing tolerant, competi­tive grass species (like Agropyron deser­torum (Briske & Richards, 1995) are able to preferentially allocate carbon to regrowing shoot sinks while curtailing root growth after defoliation which al­lows for rapid photosynthetic canopy reestablishment. In contrast, root growth has remained unabated after defoliation in other grazing sensitive, less competi­tive grasses (like Pseudoroegneria spica­ta). Differences in carbon partitioning to aboveground or belowground organs after defoliation could help explain spe­cies differential responses to defoliation (Caldwell et al., 1981; Chapin & Slack, 1979). Also, timing of defoliation at any particular developmental morphology stage can regulate the magnitude of car­bon partitioning between aboveground and belowground sinks in anyone spe­cies (Dunn & Frommelt, 1998).

Thereafter, there are various abo­veground (e.g., plant green basal area; total green tiller number; number of reproductive tillers; daughter tiller pro­duction; tiller height; number and length of total leaves [blades + sheaths, green + dry]; replacement of new leaves after defoliation) and belowground (RLD) morphophysiological traits that might contribute to herbivory tolerance, nu­trient acquisition, and plant biomass production and persistence in a com­munity (Briske & Richards, 1995; Dunn & Frommelt, 1998; Saint Pierre et al., 2004). Since P. vaginatum is practically the only perennial grass that has been exposed to grazing by livestock during decades in the warm season, how is it that this species currently persists and is the most abundant in the arid and semi­arid rangelands of central Argentina?

Our objectives were to examine the effect of plant defoliation twice a year, without removing most intercalary and apical meristems, on shoot dry weight production, and shoot and root traits on the native, warm season grasses P. vag­inatum, A. spegazzinii, A. subulata and S. cryptandrus.

Materials and methods

Study site

The study was performed in the Chacra Experimental Patagones, Buenos Aires, Argentina (40°39′S; 62°53′ W; 40 m a.s.l.), within the Phytogeographical Province of the Monte (Cabrera, 1976), during 2 years (2007-2008 and 2008-2009). Cli­mate is temperate arid to semi-arid, with higher precipitations during the summer and autumn. Long term (1981- 2010) an­nual rainfall is 416.7 mm, with a mean annual temperature of 14.1 °C, absolute minimum temperature of -7.6 °C (Au­gust), absolute maximum temperature of 43 °C (January), mean annual relative humidity of 60 %, and mean annual wind speed of 13 km h-1. Annual precipitations during 2007 and 2008 were 287.5 and 198.0 mm, respectively. Precipitations in 2007-2008 and 2008-2009 were 42 % and 30 % lower, respectively, than the long term annual mean. This provided us a unique opportunity to know the response of both species to water stress under field conditions. High variability in annual precipitation is typical of arid environments (Noy-Meir, 1993). Soil is a typical Haplocalcid. Average pH is 7, and the soil layer depth is not a constraint factor for root growth in the soil pro­file. The plant community at the Monte study region is characterized by an open, shrubby stratum which includes differ­ent quality, herbaceous species for cattle production (Giorgetti et al., 1997). Dom­inance of a particular grass or shrubby species in the study region is partially dependent on grazing history and fire frequency and intensity (Distel & Bóo, 1996).

Experimental design and defoliation managements

Plants of P. vaginatum, A. spegazzinii, A. subulata and S. cryptandrus were marked in the field, in a 1 ha exclosure. In this way, the presence of the Aristida species and S. cryptandrus were associ­ated to the same community as P. vagi­natum. A total of 48 plants were marked for sampling in 2007-2008 (4 species x 2 treatments [control versus defoliation] x 6 replicates/treatment). These plants had no neighbors within a radius of more than 0.2 m. This allowed most of RLD measurements to correspond to the marked plants. At the beginning of the second growing cycle (2008-2009), there were not enough plants of S. crypt­andrus so it had to be discarded during the second growing season studied. As a result, 36 new plants of P. vaginatum, A. spegazzinii, and A. subulata were marked for studying in 2008-2009, similarly to the previous year cycle. Only undefoli­ated plants were used in the determina­tions of plant developmental morpholo­gy stage. All plants grew under natural field conditions during the whole study; weed control was not conducted.

During the dormancy winter peri­od, in both growing cycles under study, all plants were clipped to 5 cm stubble height from the soil surface. In grazing management, it is possible to establish a rotational grazing system that includes this type of simulated clipping. Fall-win­ter-spring and spring-summer pastures, some deferred paddocks, use of natural grasslands, abandoned fields, hay, grains, etc. can be used to cover nutrient defi­ciencies at given times of the year. There is also the need of planning a forage chain with its corresponding foraging balance. As a result, the deferred paddock where P. vaginatum grew can be grazed only during winter time so during the rest of the year the animals graze the other paddocks included in the forage chain. Thus, when the paddock with plants of P. vaginatum is deferred, their seeds can be dispersed (by the wind) to the same or to the other paddocks. As a re­sult, only plant growth produced during each growing season was considered for measurements. During the springs of 2007 and 2008, half of the plants were defoliated twice (during the vegetative developmental morphology stage and immediately after the differentiation of the growth apex from vegetative to re­productive) to 5 cm stubble. The other half of the plants remained undefoliat­ed (control). Repeated clipping at 5 cm stubble can reduce the number, but not the hydration stage of buds under wa­ter stress conditions (Busso et al., 1989). Busso et al. (1989) demonstrated this by cutting buds in halves to determine their viability using TTC and Evans´s blue. After cutting the buds, they always were normally hydrated (Busso et al., 1989).

Measurements

Height of the apical meristem

In each growing cycle, tillers of each spe­cies were dissected every 2 weeks and observed under a binocular microscope to determine both the developmental stage and height of the apical meristem. This permitted to avoid its removal in the majority of plant tillers after applying the defoliation treatments. Defoliation was made by hand using scissors in all cases.

Tiller developmental morphology stages

Plant developmental morphology of undefoliated plants was determined monthly on plants of all the species during both growing cycles. With this purpose, one tiller was marked with a wire cable within each replicate plant. The studied developmental morphology stages were (1) vegetative, (2) boot stage, (3) beginning of exposed inflorescence, (4) exposed inflorescence, (5) anthe­sis (i.e., flowering stage), (6) immature grain, (7) mature grain, (8) seed disper­sal, (9) plant dormancy or death. De­terminations were made at a tiller scale (one tiller/plant) and on the same plants used for all the other measurements.

Tiller demography and growth

Plant circumference to determine green basal area was measured in each growing season. Expression of the study variables on a cm² basis was important for com­parative purposes because of the inher­ently differences in basal area among the study species.

On each of the study plants, one current season tiller was permanently marked with wire cables. Tiller number/ plant; tiller height; number of reproduc­tive tillers/cm²; number of daughter til­lers/cm²; number of total blades (green + dry)/cm² and total leaf length (blades + sheaths [green + dry]) /cm² were mea­sured periodically on each marked tiller during the growing seasons of 2007- 2008 and 2008-2009. Tiller height was measured from the soil surface area to the most distal portion of the longest leaf, maintaining the blade straight and in vertical position. These measurements followed those of Busso et al. (2003).

Root length density

Two destructive harvests were conduct­ed using a soil corer (volume: 181.5 cm³) at the time of defoliation and three weeks after the second defoliation, during 2007 and 2008. Six plants were harvested per species and defoliation treatment on each of the study years. At these dates, one soil core was taken diagonally from the periphery of each plant up to 0.15 m depth to assure that sampled roots corre­sponded to the sampled plant. One plant of each species was used per replicate at each sampling date. Roots from each soil core were used for determinations of root length.

Roots were removed from soil cores by manual washing using a 35-mesh sieve. Thereafter, roots from each plant were spread on a 20 x 30 cm glass, and after spreading, roots were covered with an­other glass of the same size. Then, each glass set was scanned. Obtained images were processed using the software Root­edge 2.3b (Kaspar & Ewing, 1997) to ob­tain root length. Since soil core volume was known, RLD (cm roots/cm³ soil) could be calculated.

Aboveground biomass production

Shoot dry weight production was mea­sured at the time of the defoliation treat­ments (C: clipping). At the end of the growing cycle, the amount of dry weight produced by undefoliated control plants during the whole study period, and that produced by defoliated plants from the time of treatment to the end of the grow­ing cycle (R: regrowth) were measured. Total dry weight produced by defoliated plants was calculated as C+R. All plant tissues obtained above clipping height were oven-dried at 60 °C and then weighed.

Statistical analyses

Data were analyzed using the statistical software INFOSTAT (Di Rienzo et al., 2009). Variables that were periodically evaluated were analyzed using repeated measures ANOVA where factors were the species, the defoliation treatments and sampling dates. The multivariate approximation was utilized using Wilks’ statistics (Wilks, 1932). When the in­teraction between any factor and time was no significant (p>0.05), data from all dates were averaged. When the in­teraction was significant (p<0.05) each sampling date was analyzed separately. RLD data were analyzed with a two-way ANOVA, where factors were the species and the defoliation treatments. When the interaction was significant, both fac­tors were analyzed separately. Mean sep­aration was conducted using the protect­ed LSD of Fisher, with a significance level of 0.05. Numbers of daughter tillers/cm² were transformed to √(x+0.5); RLD to √(x); and total dry-matter production/ cm² data to ln (x + 1) to comply with the normality and homoscedasticity as­sumptions of variance (Sokal & Rohlf, 1984). Untransformed values are shown in illustrations. Combined analysis across the study years was not conducted because they showed similar (p> 0.05) monthly precipitation and temperature values during the warm growing seasons in our study.

Results

Height of the apical meristem

Height of the apical meristem from the stem basal area was at the most 7.20 ± 1.41 mm (mean ± 1 E.E.) in 2007 and 7.24 ± 0.84 mm in 2008 on undefoliat­ed tillers (n = 12) of all the species at the time of defoliation.

Tiller developmental morphology stages

Plants of the four species were in the veg­etative developmental morphology stage in September 2007 (86.5 mm precipita­tion, Figure 1).

In the following month (7.5 mm), all species were found at the vegetative stage except for P. vaginatum which showed 66.7 % of the sampled til­lers at the reproductive developmental morphology state. In November 2007 (29 mm precipitation), the percentage of tillers in that species was greater than 80 % at the reproductive stage. At this time, only 16.7 % of tillers of A. spegazz­inii were at the boot stage. Tillers of the two remaining species continued at the vegetative stage. During December 2007, 16.7 % of the marked tillers of P. vagina­tum were at the stage of seed dispersal.

This developmental morphology stage was reached by A. spegazzinni in January 2008. The reproductive cycle of A. subu­lata was initiated in December 2007 and no seed dispersal was observed during the study. Even more, in December 2007 (2.5 mm), 16.7 % of their sampled tillers were dead. Fifty percent of marked til­lers of A. spegazzinii reached this stage a month later, in January 2008 (15 mm). In February 2008, P. vaginatum was the only species to reach such dead stage on 66.7 % of its marked tillers (16.5 mm). At this date, it was the only species dis­persing seeds. Sampled tillers of S. crypt­andrus remained at the vegetative stage during most part of the study, until Jan­uary 2008, when dead tillers started to show.

During the drier second study year (2008-2009), plant regrowth started a month later, in October 2008 (Septem­ber: 4.5 mm; October: 19 mm, Figure 1). The few plants of S. cryptandrus found were at a dormancy stage or dead. Because of this, S. cryptandrus was not included in the second study year. In November 2008 (6 mm), 83.3 % of P. vaginatum tillers were at the reproduc­tive developmental morphology stage. Seed dispersal started in December 2008 (32.5 mm) and continued until Febru­ary 2009 (January: 26 mm, February: 33.5 mm). Aristida spegazzinii presented a greater percentage of tillers at the re­productive stage in November (66.7 %) in relation to the same month during the previous year (16.7 %). This spe­cies reached the stage of seed dispersal in January 2009. Once again, tillers of A. subulata initiated later its reproduc­tive stage, in January 2009, with 66.7 % of their tillers at the stages of boot and beginning of exposed inflorescences. All marked tillers in all three species died by February 2009, with a greater percentage in A. spegazzinii (66.7 % versus 33.3 % in the other two species).

Tiller demography and growth

Green basal area

In 2007-2008, there were no differences (p>0.05) on green basal area between species or treatments (Figure 2).

During 2008-2009, however, A. spegazzinii showed the greatest (p<0.05) basal area of all the other species, in both treat­ments (Figure 2).

Tiller number/plant

This variable showed variation during the studied periods. Because of this, re­sults are presented for each sampling date (Table 1).

There were no differences between defoliation treatments (p>0.05) in both studied growing seasons (Ta­ble 1). During 2007-2008 A. spegazzinii showed the greatest (p<0.05) number of tillers-plant. The remaining species showed a similar (p>0.05) tiller number per plant during September and Oc­tober (Table 1). From November, and until the end of the study, A. spegazzinii produced the greatest (p<0.05) number of tillers/plant among all species (Table 1). Plants of A. subulata produced an in­termediate (p<0.05) tiller number, and those of P. vaginatum and S. cryptandrus produced the lower (p<0.05) number of tillers/plant.

During 2008-2009, the greatest (p<0.05) tiller number was shown on plants of A. spegazzinii (Table 1), and P. vaginatum and A. subulata produced a similar (p>0.05) tiller number.

Tiller height

In September 2007, plants of P. vagina­tum and both species of Aristida showed a greater (p<0.05) height than those of S. cryptandrus (Table 2).

From October to the end of the 2007-2008 growing cy­cle, plants of P. vaginatum showed the greatest height among all study species. In that month, the other species did not show significant differences (p>0.05) among themselves. From November onwards (after the first defoliation), and until the end of the growing season, the species showed a different response on each of the treatments (interaction spe­cies*treatment: p<0.05). Within the con­trol treatment, Aristida species showed greater values (p<0.05) than S. crypt­andrus in November and December. In January and February, only A. subulata, between the Aristida species, showed a greater (p<0.05) height than S. cryptan­drus. Within the defoliated treatment, P. vaginatum had once again the great­est (p<0.05) height, altought A. subula­ta showed greater (p<0.05) values than A. spegazzinii in November and Decem­ber 2007 and February 2008. Defoliation reduced (p<0.05) plant height on all four species between December 2007 and February 2008, except on S. cryptandrus in November, and on all four species in September and October (p>0.05). From November 2007 to February 2008, con­trol plants of A. subulata maintained a similar (p>0.05) height to A. spegazzinii.

Among all study species, S. cryptandrus was the species that most often had a lower (p<0.05) height during the grow­ing season 2007-2008.

At the beginning of the second grow­ing season, all four species showed a similar (p>0.05) height (Table 2). Con­trol plants of all four species showed a greater (p<0.05) height than the defoli­ated plants from the time of the first de­foliation, differences which were main­tained until the end of the study. From November onwards, when P. vaginatum initiated its reproductive cycle, a greater (p<0.05) height was shown in this one than in the other species. This differ­ence was maintained during December. In January, A. subulata, which initiated the production of reproductive struc­tures, showed a similar (p>0.05) height to P. vaginatum. Height of A. subulata, however, was lower (p<0.05) than on A. spegazzinii. In February, height of A. subulata was greater (p<0.05) than in the other two species (Table 2).

Number of reproductive tillers/cm²

In October 2007-2008, the number of reproductive tillers/cm² was at least 96 % greater (p<0.05) on P. vaginatum than in the other species on control plants (0.25 ± 0.11, mean ± SE), and at least 71.4 % greater (p<0.05) in this (0.07 ± 0.03) than in the other species on defoliated plants. Control plants (0.17±0.09) had at least 35.3 % more (p<0.05) reproductive tillers than defoliated plants (0.11±0.05) in Aristida spegazzinii in comparison to the other species from 24 November on­wards in the 2007-2008 growing season.

In the second year (2008-2009), the number of reproductive tillers was at least 34.6 % greater (p<0.05) (0.17± 0.05= P. vaginatum; 0.26±0.08= A. subulata; 26 February 2009) on control plants, and at least 12.5% greater (0.16± 0.05= P. vagi­natum; 0.14± 0.04=A. subulata; p<0.05) on defoliated plants in comparison to those in the other species from 18 No­vember 2008 to 22 January 2009. On 26 February, control and defoliated plants of P. vaginatum and A. subulata showed a similar (p>0.05) number of reproductive tillers on control and defoliated plants.

Daughter tiller production/cm²

During 2007-2008, parent tillers of P. vaginatum produced a greater (p<0.05) amount of daughter tillers than the oth­er species (Figure 3).

In 2007-2008 and 2008-2009, there were no differences in daughter tiller production between de­foliated and undefoliated plants (p>0.05; Figure 3).

Number of total leaves (green + dry)/cm²

Leaf production was not affected by defo­liation (p>0.05) in any of the study dates (Figure 4).

During the first year, S. crypt­andrus had a greater (p<0.05) number of total leaves (green + dry)/cm2 than A. spegazzinii (Figure 4). In 2008-2009, A. subulata showed a greater (p<0.05) num­ber of total leaves/cm² than A. spegazz­inii (Figure 4), but a similar (p>0.05) total leaf number as P. vaginatum.

Total leaf (blade + sheath [green + dry]) length/cm²

During 2007-2008, there were no differ­ences (p>0.05) neither among species nor among treatments (Figure 5).

In the following growing season (2008-2009), there were higher (p<0.05) values on control than on defoliated plants in P. vaginatum, A. spegazzinii and A. subulata (Figure 5). Also, defoliated and undefo­liated plants of P. vaginatum and A. sub­ulata showed greater (p < 0.05) total leaf (blade + sheath [green + dry]) length/ cm2 than A. spegazzinii (Figure 5).

Root length density

In 2007-2008, there was interaction (p<0.05) between time and the other factors. This is why results are report­ed separately for each sampling date. There were no significant differences (p>0.05) among treatments during this period (Table 3).

However, P. vaginatum showed greater values (p<0.05) than the other species at the time of the first de­foliation.

During the second year, there were no differences (p>0.05) between time and the other factors. As a result, val­ues obtained in the two sampling dates were averaged. During this period, no differences (p>0.05) were found neither among species nor among treatments (Table 3).

Aboveground biomass production

At the end of the first year of study, A. subulata showed the greatest (p<0.05) aboveground biomass production in comparison to the other species (Figure 6).

There were no significant differenc­es (p>0.05) between P. vaginatum and A. spegazzinii. Sporobolus cryptandrus was the lowest (p<0.05) productive spe­cies. At this time, aboveground plant biomass was greater (p<0.05) on defoli­ated than on undefoliated plants of each species (Figure 6).

In the second year, once again defoliat­ed and undefoliated plants of A. subulata showed the greatest (p<0.05) production of dry weight in comparison to the oth­er species (Figure 6), and there were no significant differences (p>0.05) between defoliated and undefoliated plants on each species (Figure 6). However, while there was a tendency for an increased dry weight production on defoliated than on undefoliated plants on P. vaginatum, an inverse tendency was shown on the other Aristida species (Figure 6).

Discussion

Stages of developmental morphology

In arid and semiarid ecosystems, plant growth and developmental morphology stages are greatly controlled by precipi­tation and plant water stress levels (Ber­tiller et al., 1991; Yuan et al., 2007). Some studies have reported that water stress causes an advance of the plant develop­mental morphology stages in various herbaceous species (Fresnillo Fedorenko et al., 1996; Giorgetti et al., 2000b). In this study, the long-term average annu­al precipitation (416.7 mm) was greater than the precipitations during the grow­ing seasons in 2007-2008 (September 2007 to April 2008: 178 mm) and 2008- 2009 (123 mm). This might have caused an advance on the plant developmental morphology stages in the drier grow­ing season (2008-2009), especially in P. vaginatum. For example, regrowth of all species, but S. cryptandrus, was delayed about a month in 2008-2009 in compar­ison with the previous year. However, P. vaginatum and A. spegazzinii initiated and completed their reproductive cycles despite the water stress conditions.

Knowledge of the advancement of the plant developmental morphology stages during dry periods in the native species is important to adjust manage­ment guidelines in a way that favors nat­ural reseeding. Giorgetti et al. (2000b) showed that a drier period advanced stem elongation and fructification on P. vaginatum, and seed dispersal in this species and on S. cryptandrus. These results disagree with those of Jones (1992), who reported that water stress delays blooming on perennial species. In our study, tillers of S. cryptandus did not reach the reproductive plant devel­opmental morphology stage in 2007- 2008; during this period, plant mortality was earlier on S. cryptandrus than on P. vaginatum. In addition, there was no regrowth in this species during 2008- 2009. As a result, it was excluded from the study. Cano (1988) reported that al­though the study species are considered as warm-season grasses, they can stay in a vegetative developmental morphol­ogy stage when conditions are not ade­quate for their growth. Plant dormancy appears to be a mechanism that allows plants to face seasonal droughts and extensive water deficits in semiarid re­gions. Growth of meristems during ad­verse conditions is detrimental to meri­stem survival, and thus perennials often have developed mechanisms for induc­ing dormancy during or prior to the adverse conditions in order to prevent meristem growth (Sarath et al., 2014). This allows the plant to resume growth with prior stored reserves when condi­tions are favorable. Seasonal dormancy can occur in the summer months, gen­erally induced by lack of water or high temperature (Sarath et al., 2014), thus reducing leaf exposure to water loss and heat gain during water stress peri­ods (Brown, 1995). Other authors have reported that S. cryptandrus has a high dependency on water availability for plant development (Kemp, 1983) and that plant dormancy is induced during drought periods. Dormant organs (rhi­zomes, crown buds) can survive during long periods (Van Andel & Ernst, 1985)

The beginning of the developmen­tal morphology stages shown on the Aristida species agrees with reports by Cano (1988). Aristida spegazzinii ad­vanced the reproductive developmen­tal morphology stage in response to a drier spring in 2008. At the same time, A. subulata was the only species that delayed their developmental morphol­ogy stages in that year. Giorgetti et al. (2000b) reported an advance in seed dispersal during dry years on Aristida pallens at the same study site. This re­sult was not observed in our study on Aristida subulata, which was unable to disperse seeds.

Pappophorum vaginatum showed the most extensive seed dispersal develop­mental morphology stage in both grow­ing seasons among all study species. This species can produce up to 1600 light weight, hairy spykelets per plant (Entío, 2019). It is possible that this response contributes to a possible greater seed germination and seedling establishment in this than in the other species, at a time (late spring-early summer) when precip­itations represent more than 25.7 % of the long-term (1981-2010: 416.7 mm) total annual precipitation. Thereafter, future research should investigate if ei­ther sexual or asexual reproduction is more important to determine the per­sistence of this species in the rangelands of central Argentina. Anyhow, and even though the Aristida species might differ in their survival strategies, their tempo­rary plasticity in their developmental morphology stages suggests that they are well adapted to the intra- and inter-an­nual precipitation variability that is com­mon at the Monte.

The advancement in the reproductive developmental morphology stages, a high production of reproductive tillers (even under grazing conditions), and a more extensive period during the seed dispersal developmental morphology stage in P. vaginatum might contribute to explain its greater abundance in compar­ison to the other species in the Monte, grazed rangelands.

Tiller demography and growth

Green basal area

Plants of A. spegazzinii showed a green basal area at least 40 % greater than in P. vaginatum and A. subulata at the end of the 2008-2009 growing cycle, despite the fact that precipitation from September 2008 to May 2009 was more than 24 % lower than in the first study growing sea­son. Undefoliated and defoliated plants of A. spegazzinii had at least a 51 % great­er tiller number than on defoliated and undefoliated plants of the other species by 26 February 2009 (Table 1). At the initiation of the study on 30 Septem­ber 2007, undefoliated (129.3 ± 25.0) and defoliated (138 ± 30.8) plants of A. spegazzinii had at least a 55 % greater til­ler number than the other species. Simi­larly, at the initiation of the second, dri­er growing season on 17 October 2008, tiller number of undefoliated (43.2 ± 15.9) and defoliated (42.3 ± 13.4) plants of A. spegazzinii was at least 24 % greater than in the other species. This species is only grazed when a better forage is not available (Cano, 1988). As a result, some plants of A. spegazzinii with a greater basal area than the other species might have been included in the exclosure by the time it was built. This is because the desirable species within the exclosure (i.e., Pappophorum vaginatum) might have been exposed to long-term live­stock grazing before the exclosure was made.

Tiller number/plant

The cumulative effects of two successive years of severe defoliations reduced tiller number on plants of all species. These adverse effects of defoliation on tillering of perennial grasses have been reported by various authors (Briske & Richards, 1995). Defoliation has caused (i) inhib­itory effects on activation and viability of axillary buds (Busso et al., 1989); (ii) a reduction in carbon reserve availabili­ty which reduce initial regrowth when a photosynthetic surface area is not avail­able after a disturbance (Busso et al., 1990); and (iii) a lower survival of grow­ing tillers (Busso et al., 1989).

The serious reduction in tiller num­bers, especially daughter tiller numbers, on defoliated and undefoliated plants across all species from the first to the second year might be in part the result of the dry growing season in 2008-2009. Tiller number is an indication of bud outgrowth (Busso et al., 1989). These au­thors emphasized that defoliated tillers on drought-treated plants of the bunch­grass Pseudoroegneria spicata (Syn: Ag­ropyron spicatum) showed a significantly lower number of physiologically active buds than tillers on undefoliated controls in that treatment. Flemmer et al. (2002a) also determined in various perennial grasses that the proportion of stem bas­es producing tillers per plant was much lower under water stress than under higher levels of soil moisture availability. Additionally, crown and total (crown + roots) pools of total nonstructural car­bohydrates were positively associated with early spring tiller regrowth on two perennial bunchgrass species (Busso et al., 1990). Therefore, we suggest that the great reduction in tiller numbers from the first to the second growing season in our study could be a sign of a depleted plant in both axillary bud bank and en­ergy reserves. Additionally, Hodgkinson (2010) reported that the density of live tillers generally declined on plants of the perennial grasses Austrodanthonia au­riculata, Bothriochloa macra, Eragrostis curvula, Phalaris aquatica and Themeda triandra as they remained under drought conditions. Furthermore, he showed that the death of plants in the drought treat­ment began for all species very soon after the foliage of plants had died. Hodgkin­son (2010) determined a critical thresh­old for days in drought beyond which all plants of any of his study species would die, which was species dependent: about 100 days for A. auriculata, P. aquatica and T. triandra, 200 days for B. macra and 300 days for E. curvula.

Tiller height

During 2007-2008 and 2008-2009, tillers were taller on defoliated and undefoli­ated P. vaginatum plants than on those of the other species on most sampling dates. Since all study plants of all species were defoliated to 5 cm stubble before the start of each growing season, a great­er tiller height on P. vaginatum was an indication of a fastest growth rate.

The second study year was drier than the first one. This conditioned, at least in part, plant responses to defoliation and interpretation of results. Daughter tiller production and plant growth are very sensitive to water stress (Brown, 1995). For example, water stress can prevent bud outgrowth (Busso, et al. 1989) and reduce photosynthetic surface area (Bitt­man & Simpson, 1987). This can be asso­ciated, at least in part, with: (1) reduction in plant height (Turner et al., 1986), (2) smaller number and size of individual leaves (Busso & Richards, 1995), and (3) reduction on tiller longevity (Caldwell et al., 1981). Reestablishment of plant surface area might be severely limited when defoliation occurs under water stress conditions, as it is usually the case on semiarid rangelands (Ludlow, 1986). In this case, P. vaginatum was the species that showed the greatest performance under conditions of defoliation under water stress.

After the second defoliation, by mid-spring 2007-2008 and early summer 2008-2009, overall mean relative growth rates (on a tiller height basis; cm cm-1 day-1) were greater on P. vaginatum than on the other species in 2007-2008 (De­foliated plants: P. vaginatum= 0.089, A. spegazzinii= 0.034, A. subulata= 0.057, S. cryptandrus= 0.041; Unde­foliated plants: P. vaginatum= 0.141, A. spegazzinii= 0.096, A. subulata= 0.098, S. cryptandrus= 0.040) and 2008- 2009 (Defoliated plants: P. vaginatum= 0.025, A. spegazzinii= 0.017, A. subula­ta= 0.021; Undefoliated plants: P. vagi­natum= 0.039, A. spegazzinii= 0.030, A. subulata=0.036). Rapid leaf tissue re­placement after defoliation is a critical component of defoliation tolerance in various perennial grass species (Briske & Richards, 1995). The number of active meristems and the amount of residual photosynthetic surface area have been reported to be more important than car­bon reserves in limiting regrowth rates on defoliated plants (Richards and Cald­well, 1985). Also, changes in environ­mental conditions, especially at the level of apical meristems, have fostered leaf growth on defoliated plants (Briske & Richards, 1995). Rapid replacement af­ter defoliation in other grasses has been attributed to compensatory photosyn­thesis on leaf and stem tissues, increase in tissue longevity or increase in the wa­ter status of defoliated plants (Briske & Richards, 1995).

Number of reproductive tillers/cm²

Each reproductive tiller in P. vaginatum can produce an 11-cm panicle with a mean of up to 70 spikelets of 1 cm each (Entío, 2019). This author also found that plants of P. vaginatum could pro­duce a mean of up to 1100 spikelets. These spikelets have a 2-mm antecio and awns of approximately 8 mm. The small size, ligth weight of these spikelets allow them to be blown away from the mother plant by wind. Results obtained by Entío (2019), taking into account that they are potential reseeding values, would allow to suppose that this mechanism of sexu­al reproduction might largely contribute to the long term persistence of this spe­cies in rangelands of central Argentina. In this way, studies should be conducted to determine which mechanism (sexual, asexual) is relatively more important in contributing to the persistence of P. vag­inatum to disturbances (grazing with or without drought conditions) in these rangelands.

During the first study year, there was a reduction in the production of repro­ductive structures because of defoliation. This result agrees with that of N’Guessan (2007) on perennial grasses. This could be the result of a greater photoassimila­tion demand by the vegetative develop­mental morphology stage, which must be reestablished after defoliation, with the subsequent reduction in carbon and nutrient partitioning to the reproductive tissues (Culvenor, 1993). It has been re­ported that the vegetative growth is the major way of reproduction in semiar­id rangelands (Belsky, 1992). However, even though new plant establishment as a result of sexual reproduction occurs in a sporadic way (Briske & Richards, 1995), it is necessary for maintaining the genet­ic diversity of populations. This diversi­ty allows species to adapt to particular environments and overcome changes to greater scales (Briske & Richards, 1995). During the 2008-2009 growing season, production of reproductive structures was highly reduced even on undefoliated plants. This period was characterized by an unusual drought, which caused a gen­eral reduction on the leaf area production traits on all study species.

Daughter tiller production/cm²

In both years, daughter-tiller production was similar on defoliated and undefoliat­ed plants in all species. This means that defoliation did not hinder daughter-til­ler production, which is known to con­tribute to defoliation tolerance (Briske & Richards, 1995), even under dry con­ditions. After clipping to 5 cm stubble under drought conditions, P. vaginatum showed in both years not only a high lev­el of daughter tiller production but also plant shoots. These results could explain in great part the persistence of P. vagina­tum under grazing.

Number of total leaves (green + dry)/cm²

The total number of leaves (green + dry) was not reduced by defoliation in 2007- 2008 and 2008-2009, even after plants were exposed to at least 57 % lower pre­cipitation conditions than the long-term mean (416.7 mm). This suggests that de­foliated plants, in general, could recover from disturbance. Even though they were unable to overcome the control plants, they were able to reach a similar leaf production, even after being defoliated twice every year. When aboveground tis­sues are removed, their reestablishment is faster from intercalary meristems, in­termediate from the apical meristem, and lower from the activation (and subsequent growth) of axillary crown buds (Hyder, 1972; Briske & Richards, 1995). Our results indicate that permanency of a high proportion of intercalary and api­cal meristems after defoliations at the be­ginning and middle of the growing sea­son allowed plants to recuperate the lost tissues. Briske & Richards (1995) report­ed that the number of active meristems and the residual amount of photosyn­thetic surface area are more important than the availability of carbon reserves in limiting the regrowth production rate on defoliated plants. In addition, changes in environmental conditions, especially to the level of apical meristems, can favor leaf growth on defoliated plants (Anslow, 1966). Detling et al. (1980) also deter­mined a greater photoassimilate alloca­tion to new leaf production than support structures on plants tolerant to defolia­tion. A rapid leaf replacement is a critical trait to explain grazing tolerance in var­ious perennial grass species (Caldwell et al., 1981; Briske & Richards, 1995).

Total leaf (blade + sheath [green + dry]) length/cm²

Leaf lengths per unit basal area be­tween both defoliation managements in all species were similar immediately after defoliations in 2008-2009. Rain­fall during November and December, when defoliation occurred, was 6 and 32.5 mm, respectively, and potential evapotranspiration was 139.8 and 155.4 mm, respectively. This very likely con­strained defoliated plants from growing faster than the undefoliated controls. The effect of water stress in constrain­ing growth of defoliated plants has been reported in various perennial grasses (Busso et al., 2003).

Height of the growth apex was below 1 cm from the soil surface in the study species during spring. This finding is im­portant because livestock could graze the studied species without removing most active intercalary and apical meristems. This indicates that growth will continue to be rapid after spring grazing, because growth rate is faster from intercalary meristems, intermediate from the leaf primordia in the growth apex, and low­er from the activation and subsequent outgrowth of the axillary buds (Briske & Richards, 1995). As active meristems remained on the plants after the defolia­tion events in 2007 and 2008, they were able to recuperate their tissues lost to de­foliation.

At the end of the 2008-2009 growing season, total leaf length produced be­tween the second defoliation in 2008, and the end of the experiment, was simi­lar on defoliated and undefoliated plants in all study species. This indicates that leaves grew faster on defoliated than on undefoliated plants after defoliation to reach equal total leaf lengths at the end of the growing season.

However, leaf replacement per unit surface area was similar (p>0.05) on de­foliated and undefoliated plants in 2007- 2008 and 2008-2009. In 2007-2008, such replacement was enough to reach simi­lar (p>0.05) total leaf lengths on defoli­ated and undefoliated plants of all spe­cies. However, it was not great enough in 2008-2009, when defoliated plants showed a lower total leaf length than un­defoliated plants on all study species.

Overall, and in relation to the number of total leaves, there were no significant differences neither between treatments nor among species during the first study year (287.5 mm annual precipitation). However, during the second, drier year (198 mm annual precipitation), control plants of P. vaginatum, A. subulata and A. spegazzinii showed a higher total leaf number than defoliated plants of these species. Also, total leaf length was higher on plants of P. vaginatum and A. subulata than on those of A. spegazzinii. The fact that P. vaginatum showed a similar re­sponse to other persistant, less palatable species (i.e., A. subulata and A. spegazz­inii) also demonstrates it persistence on grazed environments. This is, P. vagina­tum was the species which showed the higher yield after defoliation and during the water stress period (2008).

Root length density

Roots of various perennial grasses grew at soil depth deeper than 0-20 cm at the study site (Saint Pierre et al., 2002, 2004; Torres et al., 2014). However, lack of manpower made us to work within the 0-20 cm depth of the soil profile. Distel & Fernández (1986) reported that 67 % of the root biomass of the perennial grasses S. tenuis and Piptochaetium napostaense occurred in the upper 20 cm from the soil surface. Also, there are works that took soil samples from the first 0-15 cm of the soil profile to determine root length density on S. clarazii, S. tenuis and S. gynerioides (Busso & Bolletta, 2007).

Root length density in the first 20 cm of the soil profile was not reduced by defoliation on any species during 2007- 2008 and 2008-2009, even though the last growing season was 31 % drier than the first one. Other works on perennial grasses native to the Monte demonstrat­ed that root growth was not reduced af­ter defoliation, neither at different devel­opmental morphology stages (Becker et al., 1997c; Saint Pierre, 2002) nor under water stress conditions (Busso & Bolleta, 2007). This response could be interpret­ed as a strategy that allows the species a continuous exploration for resources in soils often exposed to water stress (Lud­low, 1986). In Nassella longiglumis and Jarava ichu, native perennial grasses in the Monte, Flemmer et al. (2002b) found that root growth was not reduced by wa­ter stress at the shallower soil horizons (0- 20 cm soil depth). In the same way, oth­er species also maintained shallow root systems in response to water stress (Brar & Palazzo, 1995). At our study site, this could be a strategy to make a better uti­lization of small rainfall events (< 5mm: Sala & Lauenroth, 1982). An analysis of 18 years (1983-2000) of rainfall records at our study site showed that 61 % of the rainfall events were <5 mm. Rainfall events of this magnitude were showed to rapidly stimulate ecological processes in rangeland grasses (Sala & Lauenroth, 1982). A continuous root growth during the year, even under conditions of water stress and defoliation, would contribute to a greater survival of the native species at the Monte rangelands.

Our results disagree with those of Richards (1984), who reported that a root growth reduction after defoliation is an effective mechanism that facilitates photosynthetic surface area reestablish­ment, and the balance between roots and shoots. As such, it contributes to both defoliation tolerance and maintenance of competitive ability.

In S. cryptandrus it has been reported that its shallow root system (Canfield, 1948) would limit its capacity of explor­ing deeper in the soil for resources (Wan et al., 1993). When plants of S. cryptan­drus were exposed to water stress, their photosynthetic capacity was progres­sively reduced at the same time that soil water stress increased at the shallow soil horizons (Wan et al., 1993). Despite of being considered a species adapted to water stress conditions, its survival in the semiarid rangelands would be more related to its capacity of avoiding long-term drought events (through an en­trance into a stage of induced dorman­cy: Canfield, 1948) than to its tolerance capacity.

Belowground biomass presents a greater inertia than that aboveground. This allows that the native species achieve a greater equilibrium with the environment where they develop (Gó­mez Gutiérrez et al., 1989), offering a sustained production through time. These characteristics highlight the value of P. vaginatum as a forage component of the rangelands at the Monte.

Aboveground biomass production

Defoliated and undefoliated plants of A. subulata had the greatest dry weight production among all four species at the end of both years of study (Fig­ure 6). Plant traits that contribute to explain that response were as follows: (1) Tiller numbers per cm² on defoliat­ed and undefoliated plants of A. subula­ta tended to be greater than those on P. vaginatum and S. cryptandrus, but not on A. spegazzinii, in 2007-2008 (overall means [tillers/cm2]: P. vaginatum= 0.9, A. spegazzinii= 4.9, A. subulata= 2.0, S. cryptandrus= 1.1). Also, tiller numbers per cm2 tended to be greater on A. subu­lata than on P. vaginatum and A. spegazz­inii in 2008-2009 (overall means [tillers/ cm²]: P. vaginatum= 0.5, A. spegazzinii= 1.1, A. subulata= 1.6); (2) in 2007-2008, tiller height of defoliated and undefoliat­ed plants of A. subulata was greater than on A. spegazzinii and S. cryptandus, but not on P. vaginatum (Table 2). Aristida subulata also had a greater tiller height than A. spegazzinii and P. vaginatum in 2008-2009 (Table 2); (3) Total leaf length of defoliated and undefoliated plants was similar on all species in 2007-2008 (Figure 5). However, total leaf length was greater on defoliated and undefo­liated plants of A. subulata than on A. spegazzinii, but not on P. vaginatum, in 2008-2009 (Figure 5). We have to rec­ognize that standing crop declines have been reported in other perennial grasses towards the end of the growing season, and have been attributed to transloca­tion, leaching, oxidation, leaf dropping and seed dissemination (Pieper, 1978).

We have to highlight that 2008-2009, being drier than 2007-2008, conditioned plant responses to defoliation and in­creased difficulty in the interpretation of results. In 2008-2009, precipitation was 131.5 mm from September 2008 to May 2009; 29.5 mm from September to No­vember (the most active growing period) 2008; 92 mm from December 2008 to February 2009, and 10 mm from March to May 2009. During 2007-2008, precip­itation was 223.1 mm from September 2007 to May 2008; 123 mm from Septem­ber to November 2007; 34 mm from De­cember 2007 to February 2008, and 65.5 mm from March to May 2008. It is well known that plant growth and tiller for­mation are very sensitive to water stress (Busso et al., 1989; Briske & Richards, 1995). Mohammad et al. (1982) found that the amount of regrowth in Agropy­ron desertorum and Elymus junceus was inversely related to the increased water stress to which plants had been previ­ously exposed. The low rainfall during the most active growth period (early to mid-spring: September to November) in 2008-2009 and the cumulative effects of two successive years of two severe defoli­ations annually (which determined sub­stantial reductions in [1] tiller numbers per plant, [2] production of daughter tillers per cm², and [3] total leaf length per cm2) contributed to reductions in plant dry weight per cm² between the first and second years. The negative ef­fects on plant growth of the combined influence of water stress periods and at least one severe defoliation during two or more consecutive years have been demonstrated in various perennial grass species from temperate regions (Busso et al., 1989, 2003). Like in our study, these research works indicated that the differ­ence in plant responses between consec­utive years can have a great biological significance.

In 2007-2008, dry weight production per cm² was greater on defoliated than on undefoliated plants of all species. This indicates that defoliated plants overcom­pensated the tissue lost to defoliation and it is an indication of defoliation tol­erance. This response agrees with results in several perennial grasses when the active meristems remained on the plants after defoliation (Becker et al., 1997b). Also, and although it was not significant, defoliated plants of P. vaginatum showed three times increase in aboveground dry weight per cm² in comparison to unde­foliated plants in 2008-2009. At the same time, however, aboveground dry weight per cm² of defoliated plants did not tend to be greater than on undefoliated plants on A. spegazzinii and A. subulata. Cli­mate on semi-arid and arid areas has a great influence on plant populations not only because precipitation is low but also because its distribution during the grow­ing season is unpredictable (Huxman et al.,2004). Plant mortality is correlated with extended low precipitation periods (Turner, 1990).

We observed overcompensation on forage production in 2007-2008, and ex­act compensation in 2008-2009 on defo­liated plants of all species. These results indicate that defoliated plants grew faster than those undefoliated, which allowed defoliated plants to overcompensate or exactly compensate forage production. These responses on defoliated plants have been reported for a great variety of perennial grasses (Noy-Meir, 1993) and other herbaceous species (Paige, 1992). Plants defoliated early or at the middle of the growing season had more time to re­cuperate their lost tissues to defoliation and most active meristems remained on the plants after defoliation (Becker et al., 1997a). In addition, values of leaf pro­duction remained similar on all plants in 2007-2008 and 2008-2009, even after two defoliation events during each grow­ing season.

Cavagnaro & Dalmasso (1983) also studied the response of Pappophorum caespitosum to clipping at different heights and frequencies at the Monte in Argentina. They found that clippings at 15 cm stubble height produced the highest values of productivity. Clippings at 5 cm stubble once a year and those made every two months greatly reduced productivity, reducing the alive crown to the 5 %. Drought also reduced dry matter production and plant survival. Even though it is a different species of the same genus, P. caespitosum always showed a rapid regrowth response to precipitations, even lower than 10 mm under field exclosures (without grazing conditions) (Dalmasso, 1994).

We have to acknowledge a limitation in this study: the native species are attempt­ed for grazing, but they were assessed under clipping. It is well known that clipping at a given height does not ade­quately mimic grazing. The primary rea­sons for this difference are that (i) graz­ing does not remove uniform amounts of forage from all tillers, hence tiller remov­al from plants is unrealistically severe, and (ii) grazing animals have substantial indirect effects such as soil compaction and recycling of nutrients via dung and urine (Asner et al., 2004). The greater production or growth on grazed plants than on clipped ones has been partially attributed to the non-uniform nature of herbivory (Parsons et al., 1984).

Conclusions

In the native species, and because of de­foliation, we observed a reduction in the production of reproductive structures during the first study year, in plant height in both growing seasons, and in blade + sheath length during the second growing season. Aristida subulata was character­ized for having a similar or greater basal area than the other species, with a great­er tiller production per plant. Sporobolus cryptandrus showed a greater leaf pro­duction per unit basal area. However, its plants showed a reduced size and vigor. Besides, this species almost did not pro­duce reproductive structures and had no activity during 2008-2009, possibly be­cause of the scarce precipitations which did not appear to be sufficient to simu­late regrowth in S. cryptandrus. During this year, the production of inflorescenc­es was reduced in P. vaginatum and in both Aristida species. A reduction in the reproductive effort because of the long-term effects of defoliation with or with­out drought conditions could eventually affect the persistence of these species in the plant community. Defoliated plants were able to have a similar leaf length during the first growing season, and a similar number of parent and daughter tillers and leaf numbers in both growing seasons, as control plants. Hilbert et al. (1981) reported that low growing spe­cies may require only a slight increase in their growth rates to replace tissues lost to defoliation. This suggests that the studied grazed species would be adapted to the grazing pressures to which they are exposed continuously. Species with a long herbivory history appeared to have developed certain tolerance to defolia­tion through evolution of morphological and physiological adaptations (Heady & Child, 1994). Pappophorum vaginatum showed a similar or greater daughter til­ler production, number of reproductive structures, plant height and total leaf length as compared to the other species. Relative growth rates for tiller height were greater on defoliated and undefoli­ated plants of P. vaginatum than on the other species by mid spring 2007-2008, and early summer 2008-2009. Because greater growth rates are positively relat­ed to competitive ability (Grime, 1979), our results suggest that P. vaginatum is a superior competitor when compared to the other study species. Rapid growth rates after defoliation also contribute to herbivory tolerance (Briske & Richards, 1995). In addition, at the same study site, P. vaginatum showed a greater root proliferation (Torres et al., 2015) and colonization of its root system by fungi forming arbuscular mycorrhiza (Torres, 2011). These above- and belowground traits in P. vaginatum might contribute to the persistence of this species under the effects of grazing and drought in range­lands of central Argentina.

Finally, studies of Entío (2019) on various populations of P. vaginatum showed that there was a great variabili­ty in the number of spykelets per plant on reproductive tillers between differ­ent genotypes within a same population (range=800-1600 spykelets per plant). This highlights the potential that this species has to be improved by genetic programs on the number of spykelets produced per plant. The production of such a high number of light, hairy spike­lets per plant during a growing season raises the question of whether sexual reproduction is more or less important than asexual reproduction to determine the long-term persistence of this species on grazed rangelands of central Argen­tina.

References

1. ANDERSON, D.L. 1980. La recuperación y mejoramiento de los pastizales naturales. Ecología Argentina 4: 9-12.

2. ANSLOW, R.C. 1966.The rate of appearance of leaves on tillers of the Gramineae. Herb. Abstr, 36: 149-155.

3. ASNER, G.P; ELMORE, A.J.; OLANDER, L.P; MARTIN, R.E. & THOMAS HARRIS, A.A. 2004. Grazing systems, ecosystem responses, and global change. Ann. Rev. Environ. Resour. 29: 261-299.

4. BECKER, G.F.; BUSSO, C.A; MONTANI, T. 1997a. Effects of defoliating Stipa tenuis and Piptochaetium napostaense at dif­ferent phenological stages. I. Axillar bud viability and growth. Journal of Arid En­vironments 35: 233-250.

5. BECKER, G.F; BUSSO, C.A; MONTANI, T.; BREVEDAN, R.E.; ORCHANSKY, A.; BURGOS, M.A. & FLEMMER, A.C. 1997b. Effects of defoliating Stipa tenuis and Piptochaetium napostaense at differ­ent phenological stages. II. Tiller demog­raphy and growth. Journal of Arid Envi­ronments, 35: 251-268

6. BECKER, G.F.; BUSSO, C.A.; MONTANI, T.; BURGOS, M.A.; FLEMMER, A. & TO­RIBIO, M. 1997c. Effects of defoliating Stipa tenuis and Piptochaetium napos­taenseat different phenological stages. III. Root growth. Journal of Arid Environ­ments 35: 269-283.

7. BELSKY, A.J.1992. Effects of grazing, com­petition, disturbance and fire on species competition and diversity in grass lands communities. Journal of Vegetation Sci­ence 3: 187-200.

8. BERTILLER, M. A.; BEESKOW, A. M; COR­ONATO, F. 1991. Seasonal environmental variation and plant phenology in arid Pa­tagonia (Argentina). Journal of Arid En­vironments 21: 1-11.

9. BITTMAN, S.; SIMPSON, G.M. 1987. Soil water deficit effect on yield, leaf area, and net assimilation rate of three forage grass­es: crested wheatgrass, smooth brome­grass, and altai wildrye. Agronomy Jour­nal 79: 768-774.

10. BLEAK, A.T.; FRISCHKNECHT, N.C.; PLUMMER, A.P. & ECKERT, R.E. 1966. Problems in artificial land natural reveg­etation of the arid shad scale vegetation zone of Utah and Nevada. Journal of Range 18: 59-65.

11. BRAGG, T.B.1978. Effects of burning, cattle grazing, and topography on vegetation of the choppy sands range site in the Ne­braska Sandhills Prairie. En: Proceedings 1st International Rangeland Congress. Hyder, D. N. (ed.). Society for Range Management, Denver, Colorado, USA, p. 248-253.

12. BRAR, G.S.; PALAZZO, A.J.1995.Tall and hard fescue responses to periodic soil water deficits. Journal of Agronomy and Crop Science 175: 221-229.

13. BRISKE, D.D.; RICHARDS, J.H. 1995. Plant responses to defoliation: a physiological, morphological and demographic evalua­tion. En: Wildland Plants: Physiological Ecology and Developmental Morpholo­gy. Bedunah, D. J.; Sosebee, R. E. (eds.). Society for Range Management, Denver, Colorado, USA, p.635-710.

14. BROWN, R.W. 1995.The water relations of range plants: Adaptations to water defi­cits. In: Bedunah, J.; Sosebee, R. E. (eds.), Wild and Plants Physiological Ecology and Developmental Morphology. Society for Range Management, Denver, Colora­do,USA, p.291-413.

15. BUSSO, C.A. & BOLLETA, A. I. 2007. Peren­nial grasses of different successional stag­es under various soil water inputs: do they differin root length density ?. Interciencia 32: 206-212.

16. BUSSO, C.A.; BREVEDAN, R.E.; FLEMMER, A.C. & BOLLETTA, A. 2003. Morpho­physiological and demographic responses of perennial grasses to defoliation under water stress.In: Hemantaranjan, A. (ed.), Plant Physiology and Plant Molecular Bi­ology in the New Millenium. Advances in Plant Physiology, Vol. V. Scientific Pub­lishers, Jodhpur, India, p. 341-395.

17. BUSSO, C.A.; FERNÁNDEZ, O.A. 2018. Arid and Semiarid Rangelands of Argen­tina. In: Climate Variability Impact on Land Use and Livelihoods in Drylands, Mahesh; K. Gaur y Victor R. Squires (Eds.). p. 261-292. SPRINGER, New York, U.S.A.348 p. eBook ISBN 978-3-319-56681-8, Hardcover

18. BUSSO, C A; MUELLER, R J; RICHARDS, J H. 1989. Effects of drought and defoliation on bud viability in two caespitose grasses. Annals of Botany 63: 477-485.

19. BUSSO C.A.; GIORGETTI, H.D.; MON­TENEGRO, O.A & RODRIGUEZ, G.D. 2004. Perennial grass species rich­ness and diversity on Argentine range­lands recovering from disturbance. Phy­ton International Journal of Experimental Botany, 73: 9-27.

20. BUSSO, C.A; RICHARDS, J.H. 1995. Drought and clipping effects on tiller demography and growth of two bunch grasses in Utah. Journal of Arid Environment 29: 239-251.

21. BUSSO, C.A.; RICHARDS, J.H.; CHATTER­TON, N.J. 1990. Nonstructural carbo­hydrates and spring re-growth of two cool-season grasses. Journal of Range Management 43: 336-343.

22. CABRERA, A.L. 1976. Regiones fitogeográfi­cas Argentinas. En: Enciclopedia Argen­tina de Agricultura y Jardinería. Ferreira Sobral, E.F. (ed.). ACME, Buenos Aires, Argentina, p.1-85.

23. CALDWELL, M. M.; RICHARDS, J.H.; JOHNSON, D.A.; NOWAK, R.S. & DZU­REC, R.S. et al. 1981. Coping with herbiv­ory: photosynthetic capacity and resource allocation in two semiarid Agropyron bunch grasses. Oecologia 50: 14-24.

24. CANFIELD, R.H. 1948. Perennial grass com­position as an indicator of condition of Southwestern mixed grass ranges. Ecolo­gy 29: 190-204.

25. CANO, E. 1988. Pastizales naturales de La Pampa. Descripción de las especies más importantes. Convenio AACREA-Pro­vincia de La Pampa, Argentina, 438pp.

26. CAVAGNARO, B. & DALMASSO, A. D. 1983. Respuesta a la intensidad y frecuencia de cortes en gramíneas nativas de Mendoza. I Pappophorum caespitosum y Trichloris crinita. Deserta, 7: 203-206. IADIZA.

27. CHAPIN, F.S. III & SLACK, M. 1979. Effect of defoliation upon root growth, phos­phate absorption and respiration in nu­trient-limited tundra graminoids. Oeco­logia 42: 67-79.

28. CHAPMAN, D. F.; CLARK, D.A.; LAND, C. A.; AND DYMOCK, N.1983. Leaf and til­ler growth of Lolium perenne and Agrostis spp. and leaf appearance rates of Trifolium repens in a set-stocked and rotationally grazed hill pastures. New Zealand Journal of Agricultural Research, 26: 159-168.

29. CULVENOR, R.A. 1993. Effect of cutting during reproductive development on the regrowth and the regenerative capacity of the perennial grass, Phalaris aquatica L. in a controlled environment. Annals of Botany, 72: 559-568.

30. DETLING, J.K.; DYER, M.I; PROCTER-GREGG, C & WINN, D.T.1980. Plant herbivore interactions: examination of potential effects of bison saliva on re­growth of Bouteloua gracilis (H.B.K.) Lag. Oecologia 45: 26-31.

31. DI RIENZO, J.A.; CASANOVES, F.; BALZA­RINI, M.G.; GONZALEZ, L.; TABLADA, M.; ROBLEDO, C.W. INFOSTAT versi.n (2009). Grupo INFOSTAT, FCA, Univer­sidad Nacional de Córdoba, Argentina.

32. DISTEL, R.A. & BÓO, R.M. 1996. Vegeta­tion states and transitions in temperate semiarid rangelands of Argentina. En: Proceedings of the VthInternational Rangeland Congress. Rangelands in a Sustainable Biosphere West E.N. (ed.), Society for Range Management, Salt Lake City, USA, p.117-118.

33. DISTEL, R.A. & FERNÁNDEZ, O.A. 1986. Productivity of Stipa tenuis Phil and Pip­tochaetium napostaense (Speg.) Hack in semiarid Argentina. Journal of Arid En­vironments, 11: 93-96.

34. DUNN, J.P. & FROMMELT, K. 1998. Effects of belowground herbivory by Diabrotica virgofera (Coleoptera) on biomass alloca­tion and carbohydrates storage of mayze. Applied Soil Ecology, 7: 213-218.

35. ENTÍO, L. 2019. Análisis de variabilidad en caracteres asociados a la implantación y la producción de forraje y semillas en poblaciones naturales de Pappophorum vaginatum. Doctoral Thesis, Universidad Nacional de la Plata, La Plata, Provincia de Buenos Aires, Argentina.197 pp.

36. FERNÁNDEZ, O. A.; BUSSO, C. A. 1999. Arid and semi-arid rangelands: two thirds of Argentina. En: Case Studies of Range­land Desertification. Arnalds O, Archer, S. (eds.). Agricultural Research Institute. Report Nº 200, Reykjavik, Island, p.41-60.

37. FLEMMER, A.C.; BUSSO, C.A.; FERNÁN­DEZ, O.A. 2002a. Bud viability in peren­nial grasses: Water stress and defoliation effects. Journal of Range Management 55: 150-163.

38. FLEMMER, A.C.; BUSSO, C.A.; FERNÁN­DEZ, O.A. & MONTANI, T. Root growth, appearance and disappearance in peren­nial grasses: Effects of the timing of water stress with or without defoliation. Cana­dian Journal of Plant Science, 82: 539-547.

39. FLEMMER, A.C.; BUSSO, C.A.; FERNÁN­DEZ, O.A. & MONTANI, T. 2003. Effects of defoliation under varying soil water regimes on aboveground biomass of pe­rennial grasses. Arid Soil Research and Management, 17: 139-152.

40. FOGEL, R. 1985. Roots as primary producers in below-ground ecosystems. En: Ecolog­ical Interactions in Soil: Plants Microbes and Animals. Fitter, A. H.; Atkinson, D.; Read, D. J.; Usher, M. B. (eds). Special publication Nº4 of the British Ecological Society, Blackwell Scientific Publications, Oxford, England, 451 pp.

41. FRESNILLO FEDORENKO, D.E; FERNÁN­DEZ, O.A; BUSSO, C.A; ELIA, O. E. 1996. Phenology of Medicago minima and Erodium cicutarium in semi-arid Argen­tina. Journal of Arid Environments 33: 409-416.

42. GIORGETTI, H. D.; BONTTI, E.E; BÓO, R.M; RODRÍGUEZ, G.D.; MONTENE­GRO, O.A.; ELÍA, O.R. & KUGLER, N. 2000a. Composición botánica de dietas de vacunos en la región del Monte de la Provincia de Buenos Aires. Revista Ar­gentina de Producción Animal 20:139- 140.

43. GIORGETTI, H.D.; MANUEL, Z.; MON­TENEGRO, O.A; RODRÍGUEZ, G.D. & BUSSO, C.A. 2000b. Phenology of some herbaceous and woody species in central, semiarid Argentina. Phyton International Journal of Experimental Botany, 69: 91- 108.

44. GIORGETTI, H.D.; MONTENEGRO, O.A.; RODRÍGUEZ, G.D. 2000C. Frecuencia de especies herbáceas y leñosas en pas­tizales naturales del centro de Argentina recobrándose de disturbios. 23º Congreso Argentino de Producción Animal, Corri­entes, Argentina, p. 138-139.

45. GIORGETTI, H.; MONTENEGRO, O.A.; RODRÍGUEZ, G.D.; BUSSO, C.A.; MONTANI, T.; BURGOS, M.A.; FLEM­MER, A.C.; TORIBIO, M.B. & HOR­VITZ, S.S.1997.The comparative influ­ence of past management and rainfall on range herbaceous standing crop in east-central Argentina: 14 years of obser­vations. Journal of Arid Environments 36: 623-637.

46. GIORGETTI, H.D.; MONTENEGRO, O.A.; RODRÍGUEZ, G.D. & BUSSO, C.A. 2006. Tasas de crecimiento relativo en gramíneas perennes de diferente palatab­ilidad en los pastizales del centro de Ar­gentina. Revista Argentina de Producción Animal, 26: 199-200.

47. GÓMEZ GUTIÉRREZ, J. M.; BARRERA MELLADO, I.; FERNÁNDEZ SANTOS, B.1989. Fitomasa subterránea en pastiza­les semiáridos de dehesa. Estudio com­parativo de cuatro transecciones. Pastos (18-19): 95-107.

48. GRIME, J.P. 1979. Plant strategies and vege­tation processes. Wiley Publishing Co., New York, USA, 222 pp.

49. HEADY, H.F.; CHILD, D.R. 1994. Rangeland Ecology and Management. Westview Press. Boulder, Colorado, USA, 519 pp.

50. HILBERT, D.W.; SWIFT, D.M.; DETLING, J.K. & DYER, M.I.1981. Relative growth rates and the grazing optimization hy­pothesis. Oecologia 51: 14-18.

51. HODGKINSON, K.C. 2010. Perennial grass survival in drought. CSIRO Sustainable Ecosystems, Canberra.

52. HOLLAND, J.N.; CHENG, W.; CROSSLEY JR., D.A. 1996. Herbivore-induced chang­es in plant carbon allocation: assessment of below-ground C fluxes using car­bon-14. Oecologia 107: 87-94.

53. HUXMAN, T.E.; KEIRITH, A.; SNYDER, A.; TISSUE, D.; LEFFLER, A.J.; OGLE, K.; POCKMAN, W.T.; SANDQUIS, D.R.; TOTTS, D. L. & SCHWINNING, S. 2004. Precipitation pulses and carbon fluxes in semiarid and arid ecosystems. Oecologia 141: 254–268.

54. HYDER, D.N. 1972. Defoliation in relation to vegetative growth. En: Biology and Utili­zation of Grasses. Younger, V. B.; McKell, C. M. (eds.). Academic Press, New York, USA, p. 302-317.

55. JACKSON, R.B.; CALDWELL, M.M. 1989. The timing and degree of root prolifer­ation in fertile-soil microsites for three cold-desert perennials. Oecologia 81: 149–153.

56. JONES, H.G. 1992. Plants and microclimate. A quantitative approach to environmen­tal plant physiology. Cambridge Univer­sity Press, Cambridge, England, 456 pp.

57. KASPAR, T.C.; EWING, R.P. 1997. ROOT­EDGE: Software for measuring root length from desktop scanner images. Agronomy Journal, 89: 932-940.

58. KEMP, P.R. 1983.Phenological patterns of Chihuahuan desert plants in relationship of the timing of water availability. Journal of Ecology 71: 427-436.

59. KLEINE, E.F. 1983. Successional trends in an ungrazed, arid grassland over a decade. Journal of Range Management 36: 114- 118.

60. KORNER, CH. 1991. Some often-overlooked plant characteristics as determinants of plant growth: a reconsideration. Func­tional Ecology, 5: 162-173.

61. LANGER, R.H.M.1963.Tillering in herbage grasses. Herbage Abstracts, 33: 141-148.

62. LUDLOW, M. M.1986. Simultaneous pressure of water stress and defoliation in range­land plants. En: Proceedings Symposium Rangelands: A Resource Under Siege. 2nd International Rangeland Congress. Joss, P. J.; Lynch, P. W.; Williams, O.B. (eds.). Cambridge University Press, Cambridge, England, p. 433-436.

63. MCNAUGHTON, S. J. 1983. Compensatory plant growth has a response of herbivory. Oikos, 40: 329-336.

64. MOHAMMAD, N.; DWYER, D.D. & BUS­BY, F.E.1982. Responses of crested wheat grass and Russian wildrye to water stress and defoliation. Journal of Range Man­agement, 35: 227-230.

65. N’GUESSAN, M. 2007. Effects of grazing on growth and morphology of rhizomatous and caespitose grasses in tallgrass prai­rie, M. S. Thesis, Kansas State University, Kansas, USA, 65pp.

66. NOY-MEIR, I. 1993. Compensating growth of grazed plants and its relevance to the use of rangelands. Ecological Applications 3: 32-34.

67. OFFICIAL METHODS OF ANALYSIS OF THE ASSOCIATION OF OFFICIAL AG­RICULTURAL CHEMISTS. 1965. Hor­witz W. (ed.). 10th Edition, Association of Official Agricultural Chemists, Washing­ton D C, USA, 957 pp.

68. OLSON, B E; RICHARDS, J H. 1988a. Annu­al replacement of the tillers of Agropyron desertorum following grazing. Oecologia 76:1-6.

69. OLSON, B.E.; RICHARDS, J.H. 1988b.Tus­sock regrowth after grazing: intercalary meristem and axillary bud activity of til­lers of Agropyron desertorum. Oikos, 51: 374-382.

70. PAIGE, K.N.1992.Over compensation in re­sponse to mammalian herbivory from mutualistic to antagonistic interactions. Ecology, 73: 2076-2085.

71. PARSONS, A.J; COLLET, B.; LEWIS, J. 1984. Changes in the structure and physiolo­gy of a perennial ryegrass sward when released from a continuous stocking management: implication for the use of exclusion cages in continuously stocked swards. Grass and Forage Sci. 39: 1-9.

72. PIEPER, R.D. 1978. Measurement tech­niques for herbaceous and shrubby veg­etation. New Mexico State University. Las Cruces, New Mexico, USA. 148 pp.

73. RICHARDS, J.H. 1984. Root growth re­sponse to defoliation in two Agropyron bunch grasses: field observations with an improved root periscope. Oecologia: 64, 21-25.

74. RICHARDS, J.H. 1993. Physiology of plants recovering from defoliation. Proceedings of the XVII International Grassland Con­gress, Palmerston, New Zealand, p. 85-94.

75. RICHARDS, J.H.; CALDWELL, M.M. 1985. Soluble carbohydrates, concurrent pho­tosynthesis and defficiency in regrowth following defoliation: A field study with Agropyronspecies. Journal of Applied Ecology 22: 907-920.

76. RÚGOLO DE AGRASAR, Z.E.; STEIBEL, P.E.; TROIANI, H.O. 2005. Manual ilus­trado de las gramíneas de la provincia de La Pampa. Universidad Nacional de La Pampa. Santa Rosa, La Pampa, Argenti­na. 359 p.

77. RYSER, P. & LAMBERS, H. 1995. Root and leaf attributes accounting for the perfor­mance of fast-and slow-growing grasses at different nutrient supply. Plant and Soil, 170: 251-265.

78. SÁENZ, A. M.; DEREGIBUS, V.A. 2001. Efecto de eventos de defoliación y perío­dos de sequía sobre el estado hídrico y la morfogénesis foliar de Poa ligularis. Re­vista Argentina de Producción Animal, 21: 172-173.

79. SAINT PIERRE, C.; BUSSO, C.A.; MON­TENEGRO, O.A.; RODRÍGUEZ, G.D.; GIORGETTI, H.D.; MONTANI, T.; BRAVO, O.A. 2002. Root proliferation in perennial grasses of low and high palat­ability. Plant Ecology, 165: 161-169

80. SAINT PIERRE, C.; BUSSO, C.A.; MON­TENEGRO, O.A; RODRÍGUEZ, G.D.; GIORGETTI, H.D.; MONTANI, T. & BRAVO, O.A. 2003. Root proliferation in perennial grasses of low and high palat­ability. Plant Ecology 165: 161-169.

81. SAINT PIERRE, C.; BUSSO, C.A.; MONTE­NEGRO, O.A. 2004. Soil resource acqui­sition mechanisms, nutrient concentra­tions and growth in perennial grasses. Interciencia, 29: 303-311.

82. SALA, O.E.; LAUENROTH, W.K. 1982. Small rainfall events, an ecological role in semiarid regions. Oecologia 53: 301–304.

83. SARATH, G.; BAIRD, L.; MITCHELL, R.B. 2014. Senescence, dormancy and til­lering in perennial C4 grasses. Plant Sci­ences, 217-218: 140-151.

84. SOKAL, R.R. & ROHLF, F.J. 1984. Introduc­ción a la Bioestadística. Editorial Reverté, S.A. Barcelona, España, 376 pp.

85. TORRES, Y.A; BUSSO, C.A.; MONTENE­GRO, O.; ITHURRART, L.; GIORGETTI, H.; RODRÍGUEZ, G.; BENTIVEGNA, D.; BREVEDAN, R.; FERNÁNDEZ, O.; MUJICA, M.M.; BAIONI, S.; ENTÍO, J.; FIORETTI, M. & TUCAT. G. 2011. De­foliation effects on the arbuscular mycor­rhizas of ten perennial grass genotypes in arid Patagonia, Argentina. Applied Soil Ecology 49: 208-214.

86. TORRES, Y.A.; BUSSO, C.A.; MONTENE­GRO, O.A.; ITHURRART, L.; GIORGET­TI, H.; RODRÍGUEZ, G.; BENTIVEG­NA, D.; BREVEDAN, R.; FERNÁNDEZ, O.; MUJICA, M.M.; BAIONI, S.; ENTÍO, J.; FIORETTI, M. & TUCAT. 2014. Root proliferation in perennial grasses in arid Patagonia, Argentina. Journal of Arid Land 6: 195-204.

87. TORRES, Y.A.; BUSSO, C.; ITHURRART, L.; MONTENEGRO, O.A.; CARDIL­LO, D.; AMBROSINO, M.; LEDESMA, M.; ORAZI, M.; GIORGETTI, H. Y RODRÍUEZ, G. 2015. Producción sub­terránea en gramíneas perennes de pas­tizales semiáridos. Naturalia Patagónica 7.P.30.

88. TURNER, G.T. 1990. Long-term vegetation change at a fully protected Sonoran desert site. Ecology 7: 464-477.

89. TURNER, N.C.; O’TOOLE, J. C.; CRUZ, R. T.; NAMUCO, O.S. & AHMAD, S. 1986. Re­sponses of seven diverse rice cultivars to water deficits. I. Stress development, can­opy temperature, leaf rolling and growth. Field Crops Research 13: 257-271.

90. VAN ANDEL, J.; ERNST, W.H.O. 1985. Ecophysiological adaptation, plastic re­sponses, and genetic variation of annuals, biennials, and perennials in woodland clearings. En: Structure and Functioning of Plant Populations. Haeck, J.; Wold­endorp, J. W. (eds.). Elsieve Science and Technology, Amsterdam, Holland, 408pp.

91. WADE, M.H. 1982. Influence of tiller density as a yield component of cultivated pas­tures. Revista Argentina de Producción Animal 9: 181-188.

92. WAN, C.; SOSEBEE, R.E.; MCMICHAEL, B. L. 1993. Soil water extraction and pho­tosynthesis in Gutierrezia sarothrae and Sporobolus cryptandrus. Journal of Range Management 46: 425-430.

93. WHITE, L. M. 1973. Carbohydrate reserves of grasses: A review. Journal of Range Man­agement 26:13-18.

94. WILKS, S.S. 1932. Certain generalizations in the analysis of the variance. Biometrika 24:471-494.

95. YUAN, W.; ZHOU, G.; WANG, Y.; HAN, X. & WANG, Y. 2007. Simulating phenolog­ical characteristics of two dominant grass species in a semi-arid steppe ecosystem. Ecological Research 22:784-791.

Enlace alternativo

https://www.mendoza.conicet.gov.ar/portal/multequina/indice/pdf/29/03-ART-Torres-2020.pdf (pdf)