Does Buffelgrass Have a Long Permanence in an Established Pasture? An Analysis of the Population Dynamics of This Exotic Grass in Central Sonora, Mexico

Abstract

The introduction of exotic forage species to new environments for livestock purposes is a common practice to increase productivity. Unfortunately, the population dynamics of introduced species as well as that of native species that persist in grasslands has been poorly studied. In Sonora, the introduction of exotic buffelgrass pasture has caused substantial modifications in the structure of desert scrublands. In this study, an evaluation of the population dynamics of buffelgrass pasture in two grasslands with different times (10 and 50 years) was carried out using classification by size category according to the total number of stems per plant. For each size category of stems, the probabilities of permanence, transition, and regression, and for estimating seed establishment and fecundity were evaluated. The results obtained indicate that in both grasslands, the population growth values (λ) were slightly greater than 1 (λ > 1), which indicates that the populations are stable. The results of this study suggest that the permanence of individual buffelgrass plants in established grasslands is the determining factor in λ. Likewise, our results suggest that in both grasslands, pasture management plays an important role in the permanence or deterioration of buffelgrass pastures.

1. Introduction

In arid regions, the conversion of natural ecosystems by pasture for livestock purposes has caused serious changes that alter the environment and reduce biological diversity [1,2]. This conversion promotes a series of severe changes, such as the loss of crucial soil nutrients, changes in soil water infiltration, increased erosion, and overgrazing, which is considered the most impactful factor in grassland degradation [3,4]. In addition, in places where conversion of natural vegetation to grasslands occurs, the tree elements of the landscape are considerably reduced, decreasing structural heterogeneity and the herbaceous stratum becomes dominant [3]. All these changes can affect the population dynamics of native plant species that remain and those that are established in the grasslands because demographic processes such as growth, survival, and reproduction can be modified. Alterations in vegetation cover directly impact the abundance and population structure of plant species, as well as their regeneration and reproduction [3,5,6]. Furthermore, these new environmental conditions in grasslands can modify the recruitment and regeneration of new individuals [7]. Unfortunately, the population dynamics of plant species that remain in established grasslands have been poorly studied. In this context, demographic studies are a useful tool to explore population dynamics in detail [8,9,10]. Matrix models allow us to understand population structure and behavior through stable age distribution, reproductive values, and population growth rate (λ), while elasticity analyses allow us to identify those life cycle stages that show a greater effect on population growth [10,11,12]. The evaluation of these demographic aspects is crucial for any conservation and recovery effort [12].

In the arid areas of northern Mexico, deforestation is mainly due to clearing for the establishment of buffelgrass (Cenchrus ciliaris (syn. Pennisetum ciliare) [L] Link) grasslands for cattle feed [13]. However, once established, buffelgrass can invade adjacent habitats and spread rapidly, altering fire regimens and biogeochemical cycles and displacing native plants [13]. Generally, buffelgrass is apomictic, though rare sexual individuals exist, and its seeds are easily spread in disturbed environments. Additionally, buffelgrass can reproduce vegetatively via rhizomes and stolons, leading to a variety of plant forms, from dense monotypic stands to small clumps or lone tussocks [13,14]. In established pastures, the scarcity of seedlings despite the dominance of adult plants suggests recruitment depending on specific environmental conditions [13], which remain unquantified in long-term demographic studies. In the central region of Sonora, animal husbandry records from 2006 estimated that 140,000 ha had been converted to buffelgrass grasslands [15]. Nowadays, it is estimated that buffelgrass has invaded 8 to 10 million ha of the Sonoran Desert ecosystem in the United States and Mexico [16]. Currently, the increase in the establishment of grasslands and land invasions by buffelgrass in Sonora is exponential, and it is estimated that in the short term, the area of grasslands occupied will be easily duplicated [17]. Unfortunately, despite its extensive distribution, we know very little about the population behavior of buffelgrass pasture. Actually, it is unknown for the life cycle stages with the most sensitive effect on the buffelgrass population. Under this scenario, matrix models are optimal for studying buffelgrass population dynamics. It has been documented that stage-structured matrix models are widely used to analyze perennial plant dynamics, as they identify critical life stage transitions under varying environmental conditions [10]. For buffelgrass, this approach is relevant due to its reproductive plasticity and recruitment variability. During land conversion to establish buffelgrass pastures, some trees and columnar cacti are left as shade for livestock. Demographic studies in native plants that persist in pastures show that buffelgrass establishment negatively impacts native seedling regeneration, leading to a population structure composed solely of adult native individuals [18,19]. However, at present, there are no studies that have evaluated the population dynamics of buffelgrass and its regeneration mechanisms in pasture sites. The information available on demographic studies in pastures indicates that in other regions of the world, the recruitment of new individuals to the population plays an important role in population abundance [20]. However, in the case of buffelgrass pastures, its population performance is unknown.

In this study, the population dynamics of buffelgrass in two grasslands with different times since conversion to buffelgrass are explored: older (>50 years) versus newer (<10 years). Matrix models and elasticity analysis over a two-year period were used to determine the population dynamics and to identify the most sensitive life cycle stages for impacting population growth (λ). Here, we test two hypotheses: (H1) population growth rate (λ) will be higher in newer pastures due to reduced intraspecific competition; (H2) the seedling stage will exhibit the highest elasticity, indicating λ’s critical sensitivity to recruitment. Matrix simulations were performed to explore the importance of the seedling establishment process in λ under three establishment probabilities. This study aims to explore the variation in population dynamics over two years and determine the most sensitive category of demographic process in the buffelgrass population.

2. Materials and Methods

2.1. Study Area

The study areas are located in “El Diamante” (28°41.782′ N, 110°15.804′ W) and “La Noria” (28°41.352′ N, 110°17.707′ W), both study sites placed in the central-eastern region of the state of Sonora (Figure 1). The principal land use in this area is extensive livestock for cattle ranching. Prairies of buffelgrass have been established in these sites since the 1960s [19]. During the study period (2006–2008), the pasture located at “El Diamante” had a conversion time of 10 years, and the pasture at “La Noria” had a conversion time of 50 years. In the area, for both sites, the average annual temperature is 23 °C and the average annual precipitation is 498 mm (25 years of data); 80% of the annual precipitation falls between the months of July and September (CONAGUA, Hermosillo, Sonora). The dominant soils are regosols and lithosols with luvic xerosol and haplic phaeozole [21]. The dominant vegetation in the region is thornscrub [22], dominated by columnar cacti (i.e., Pachycereus pecten-aboriginum, Carnegiea gigantea, Stenocereus thurberi) and tree species (i.e., Parkinsonia praecox, P. microphylla, Prosopis glandulosa, Acacia coulteri, A. cochliacantha, Jatropha cordata, Guaiacum coulteri).

2.2. Field Experiment

Research permits were obtained from the ranch owners before data collection. During the study period (2006–2008), field work was carried out during the months of October and November, after the dry season. At each site, permanent observation plots were set up and randomly placed within homogeneous buffelgrass-dominated areas, avoiding edges. Three plots were set up in the buffelgrass pasture of “El Diamante”, and three in the buffelgrass pasture of “La Noria”; both sites utilize an extensive grazing management system. Three permanent 5 × 5 m observation plots were set up at each site in 2006. All individual buffelgrass plants within each plot were counted. To ensure sufficient sample size, at each site, a sample of plants that exceeded 100 individuals was labeled for later identification. The number of stems of each plant in the sample was determined in the initial year (2006). Growth for each plant was evaluated by counting the stems for each individual during subsequent years (2007 and 2008). During the reproductive season, for each evaluation year, the number of inflorescences (panicles) of each plant was recorded within the study plots in order to define the average number of panicles per size category. A sample of 50 panicles was also collected in order to estimate the number of seeds per panicle and the average number of seeds produced by each size category. The comparison of the total number of stems per plant was analyzed using the repeated-measures ANOVA model through JMP software v16 (SAS Institute, Cary, NC, USA).

2.3. Demographic Model

To evaluate population dynamics, classification by size categories was performed according to the total number of stems per plant. The marked plants were classified into 6 size categories to perform the matrix models. The categories expressed in number of stems per plant were (1) 1–10, (2) 11–20, (3) 21–40, (4) 41–60, (5) 61–80, and (6) >81. For each size category, calculations were made regarding the probabilities of permanence, transition, retrogression, and fecundity. The probability of permanence was calculated as the proportion of individuals in a size category that remain in the same category after a given time interval. The probability of transition was calculated as the proportion of individuals in a size category that move to the next higher category. The probability of retrogression was calculated as the proportion of individuals in a size category that return to the previous category. Fecundity was estimated by multiplying the number of panicles by the average number of seeds per panicle, by the germination percentage, and by the survival percentage. Although an attempt was made to obtain a true value of fecundity, the data obtained were not considered reliable for the matrix design, because constant observations of seed establishment in situ could not be made during the time in which the field work was carried out. Nevertheless, fecundity for each matrix was simulated with three values of probability of plant establishment (0.001, 0.01, and 0.1) as described in Vega and Montaña [23]. These probabilities were chosen based on observed seedling recruitment ranges in semi-arid grasses [23] and prior sensitivity analyses [24]. Seed category was not included in the demographic model due to the difficulty of obtaining data to define the buffelgrass seed bank. However, field observations suggest that a large number of seeds that fall to the ground are quickly removed. A Lefkovitch matrix model was constructed to calculate the finite rate of population growth (λ) [10]. The equation used for the analysis of population dynamics was the following:

nt + 1= A n(t)

where n indicates the column vector whose n1 elements are the number of individuals in each category at time t or t + 1, and A represents the square matrix with the transition probabilities between size categories from one period to the next. An elasticity analysis was also performed to explore the relative contribution of demographic processes (permanence, growth, and fecundity) and size categories to the population growth rate [11]. The analyses were performed using the Démographe software v2.1 [25], with code available in Supplementary Materials. Confidence intervals (95%) for the value of λ were derived from 10,000 bootstrap iterations of transition matrices. The relative contribution of the different demographic processes (permanence, transition, retrogression, and fertility) was evaluated through elasticity analyses [11].

3. Results

3.1. Comparison of the Total Number of Stems per Plant

The initial number of stems did not vary between sites (El Diamante vs. La Noria) (Figure 2). However, during the study period, despite observing a slight decrease in stems compared to the initial period, no significant differences were detected between the study populations (2006; F = 0.0434, df = 1, P = 0.8392, 2007; F = 0.1562, df = 1, P = 0.7010, 2008; F = 0.0062, df = 1, P = 0.9389) (Figure 2). Conversely, in the El Diamante site, the comparison between years of study showed significant differences in the number of stems produced per plant (F = 4.2713, df = 2, P < 0.0340). In contrast, no significant differences were detected at the La Noria site (F = 1.8629, df = 2, P = 0.1894).

3.2. Population Growth Rates

The matrix model showed that finite rates of population growth (λ) varied slightly between sites and years (

Table 1. Finite rate increase and confidence interval values for buffelgrass pasture populations in two consecutive years. Seedling establishment probabilities of 0.001, 0.01, and 0.1 were used to model the projection matrices (Supplementary Materials S1: matrix structures and simulations parameters).

Period	Probability	El Diamante (10 yrs After Conversion)	La Noria (50 yrs After Conversion)
2006–2007		λ (95% confidence interval)	λ (95% confidence interval)
	0.001	1.0018 (1.0016–1.0019)	1.0015 (1.0014–1.0016)
	0.01	1.0166 (1.0164–1.0168)	1.0376 (1.0373–1.0378)
	0.1	1.1050 (1.1048–1.1052)	1.0414 (1.0412–1.0416)
2007–2008
	0.001	1.0002 (1.0001–1.0004)	1.0013 (1.0011–1.0015)
	0.01	1.0033 (1.0030–1.0034)	1.0317 (1.0315–1.0319)
	0.1	1.0233 (1.0232–1.0235)	1.0355 (1.0352–1.0357)

) (Supplementary Materials S1: matrix models simulations). In the initial year (2006–2007), lambda values from El Diamante were slightly higher than those obtained from La Noria, although the values were greater than 1 in all cases (Table 1, Supplementary Materials S1). In the subsequent models, the use of grassland by livestock in both sites corresponded with a significant decrease in stems, which was reflected in the reduction in the finite rate increase (λ), although in neither case was it less than 1 (Table 1, Supplementary Materials S1).

Despite slight decreases in plant establishment simulations (0.001, 0.01, and 0.1), λ remained > 1, indicating stable/increasing populations (Table 1). However, marginal declines in El Diamante (2007–2008) could reflect grazing pressure in the site (Table 1).

In both populations, the elasticity of the size categories was similar for both study periods (Figure 3). In all cases, permanence was the demographic process with the highest relative contribution followed by growth and fecundity (Figure 3).

4. Discussion

Landowners are primarily concerned about the deterioration of buffelgrass pastures, which manifests as reduced plant density, coverage, and forage production for livestock over time [26]. In addition, in many cases, this is accompanied by invasion of shrubs and woody species of little or no forage value for livestock [26]. In our research, it was observed that in the youngest site (El Diamante: 10 years conversion), there were a large number of shrub species such as Encelia farinosa, Acacia cochliacantha, Opuntia sp., and Prosopis glandulosa. These shrub encroachments in El Diamante may result from reduced buffelgrass competition (due to grazing) or direct facilitation by cattle (e.g., seed dispersal). Further studies should disentangle these drivers. Despite a λ > 1 typically indicating population stability, physical damage to buffelgrass plants at the El Diamante site suggests localized stress, likely from cattle grazing. This damage, coupled with a lack of growth during the study period, indicates that vegetation growth at the site is influenced by livestock presence. The observed population stability, despite this damage, may reflect compensatory growth or clonal resilience, which could be buffering the plants against the impacts of grazing. On the contrary, at the oldest site (La Noria: 50 years conversion) a low abundance of shrub species was observed. In a previous work at Sonora, Franklin and Molina-Freaner [27] found that the age of a buffelgrass pasture was unrelated to species richness within the pasture. This suggests that passive recovery of species richness to pre-conversion levels is unlikely and demonstrates that land conversion can result in large losses of plant species richness at local and regional scales.

During the study period, it was observed that management in El Diamante was more intense as opposed to management in La Noria. The available evidence indicates that site management is a determining factor in the permanence of buffelgrass pastures. In our case, the fact that the buffelgrass plants with the greatest deterioration were detected at El Diamante as opposed to those found at La Noria indicates that management at the site plays a considerable role, despite not detecting significant differences in the population growth rate. According to the study by Ibarra et al. [28], analyzing 200 sites with buffelgrass in Mexico and Texas, they found that previous overgrazing was one of the factors that determined the null or poor establishment of buffelgrass plants in pastures. In our case, the results obtained showed that during the field work in El Diamante, distinct signs of overgrazing were detected through very damaged buffelgrass plants, in contrast to buffelgrass plants found at La Noria, which show less damage. The speed at which buffelgrass plants deteriorate after establishment is very variable between sites, and seems to depend on different factors such as type of soil, amount of vegetation that remains standing, degree of erosion, and rainfall [26]. Consequently, it is very important to use the buffelgrass pasture appropriately, that is, pay attention to management and stocking rate so that in the short term the pasture can self-regenerate and prevent its productivity from decreasing, turning into a degraded area. We consider that the management of buffelgrass pastures is a key factor if we want to ensure the permanence of the established buffelgrass pastures for cattle production.

Elasticity Analysis

The elasticity analysis allows us to identify the stages of the life cycle of the study species that show a greater effect in relation to population growth [10,11,12]. In the present study, while fecundity had low elasticity, simulations revealed λ’s dependence on seedling establishment a critical factor during pasture recovery or invasion. This underscores the need to monitor seedbanks in management. Likewise, this low elasticity detected in fecundity suggests clonal propagation, not seedling recruitment, sustains these populations. This contrasts with obligate seeders like Pachycereus, highlighting buffelgrass’s invasion resilience. However, the most relevant result was to find that permanence is the one that determines the population status in both pastures, despite the difference in years in their conversion between the two pastures. Ecological theory related to life-history strategies suggests that adult stages are more important for the persistence of long-lived species [29]. Our study in the buffelgrass population supports this theory, since adult categories (4 and 5) were found to be the most important. In a study carried out in northwestern Sonora, where population growth rate was also evaluated and an elasticity analysis was performed with the species Pachycereus pecten-aboriginum, it was determined that permanence was the demographic process with the highest relative contribution [16]. These results and those obtained in our research are similar to those reported by other authors for long-lived plants in which the survival of adult plants is considered the stage of the life cycle with the greatest contribution to the growth rate [11,30,31,32]. For this reason, we consider that the result obtained in the elasticity analysis is of great relevance since these data help us to estimate which stage or process has the greatest effect on population growth and the permanence of the species in the place.

5. Conclusions

In this study, an analysis of the population dynamics of buffelgrass was performed in two pastures with different conversion times (El Diamante: 10 years and La Noria: 50 years) using classification by size categories according to the total number of stems per plant. For each size category, calculations were made regarding the probabilities of permanence, transition, retrogression, and fertility.

For both pastures in both study periods, population growth values slightly greater than 1 (λ > 1) were obtained, which indicated that the population is stable. The results obtained in this research suggest that the permanence of the buffelgrass plants in pastures is the determining factor in the population growth rate. In turn, the simulated fertility scenarios highlight the role of recruitment in post-disturbance recovery, warranting further study. Likewise, our results obtained suggest that in both sites, site management plays a fundamental role in the permanence or deterioration of buffelgrass.

Based on our results, it can be concluded that buffelgrass pastures show potential for persistence (λ > 1) under moderate grazing, though this does not equate to ecological sustainability or forage productivity. However, under conditions of excessive grazing, the pastures cease to be productive and quickly show conditions of invasion by shrubs not desired for livestock production purposes or extreme degradation.

This study informs ecological restoration strategies, such as targeting vulnerable life stages (i.e., adult plant removal) in buffelgrass control programs.

It is necessary to take into account that our work only considered two sites with defined conversion times and with certain characteristics, so it is not possible to generalize our results to other sites. Long-term monitoring across diverse sites, coupled with seed bank dynamics, is needed to generalize findings. In addition, targeting adult removal and adopting rotational grazing to reduce pressure on vegetative stages could enhance control efforts. However, we consider that the results obtained will be useful to better understand the behavior of buffelgrass when it is established in Sonoran ecosystems. It is expected that this information will be useful to propose better management strategies for this species and thus avoid the degradation of arid lands.

Finally, this study provides the first demographic analysis of buffelgrass in pastures of differing ages, offering a baseline for invasive species management in semi-arid ecosystems.