Effects of Scale, Temporal Variation and Grazing on Diversity in an Endemic Pasture in Sierra de Zapaliname, Coahuila, Mexico

Abstract

Grasslands and pastures are extensively studied due to their geographic variation, species richness, ecological functioning, and economic importance. They are vital components of land use in many parts of the world. The impact of grassland management on species diversity and species composition has also been widely discussed, but results have been contradictory. It is well known that the relationship between species richness and the sampled area is perhaps one of the most consistent rules in plant ecology. This relationship is particularly important in biodiversity studies as it helps to predict richness at larger scales. Additionally, species richness is also influenced by absolute plant abundance, spatial patterns, and the degree of species mixing. However, species richness also depends on absolute plant abundance, spatial patterns, and the degree of mixing species. To assess this relationship, we analyzed the impact of cattle grazing on species richness at a sampling scale in the Sierra of Zapaliname, a protected area in northern Mexico. Our results revealed that the increase in plant species concerning the sampling area significantly differed in the plots excluded from grazing from the control (grazed) plots, and these relationships are differently detected in the function of the scale. Despite the lack of differences in previous studies on species richness without considering the scale, once the scale is incorporated, differences arise among both treatments. As indicated in previous studies, grazing exclusion can lead to a decrease in species richness, but we suggest that some areas of the pasture could be excluded from grazing for longer periods, as long as it is compatible with the economic needs of the local inhabitants, to investigate changes and promote diversity, especially for plant species associated with areas excluded from grazing.

1. Introduction

The primary goal of conservation managers around the world is to maintain biodiversity [1,2]. In the case of plant communities subjected to long-term effects of herbivore grazing, it has been suggested that maintaining low to medium grazing intensities could enhance plant diversity [3,4]. However, research results are conflicting and depend on grazing management [5], environmental conditions [6], climate gradient [7], sampling scale [8], site productivity [9], and grazing intensity [10], among others.

Grasslands and pastures have been extensively studied due to their geographic variation, species richness, ecological functioning, and economic importance [11,12]. They are vital components of land use in many parts of the world [13]. The impact of grassland management has also been widely discussed, but the results have been contradictory. For instance, ungulate grazing has been reported to increase [14,15] or decrease [16] species richness. Depending on the successional state of the grassland and pasture, environmental conditions will have a higher impact on species composition than on species relationships. However, there are several mechanisms explaining species richness and coexistence in pastures and grasslands such as biomass, weather conditions, and ecological competition abilities between species [17].

Pastures in North America have a combination of tall and short grasses from southern Canada to central Mexico [18]. Semiarid pastures in Mexico are classified as shortgrass prairies, forming a part of a broader plant community that extends its distribution across a vast expanse, ranging from the northern reaches of Alberta down to the southern reaches of Arizona, New Mexico, Texas, and northern Mexico. This expansive geographical range underscores the adaptability of this plant community to a diverse array of climatic and ecological conditions, while the classification as shortgrass prairies signifies a distinctive ecological profile characterized by hardy, drought-tolerant vegetation. The presence of this community across such a range highlights its ecological resilience and its ability to thrive in regions with limited water availability and fluctuating environmental conditions. This ecological significance, spanning multiple regions and encompassing a spectrum of ecosystems, accentuates the importance of understanding its dynamics and management strategies in order to ensure its conservation and sustainable utilization. [19]. The genus Bouteloua comprising 29 species and 13 subspecies, is a dominant component of these ecosystems in Mexico [20]. Overgrazing by livestock is a widespread issue in northern Mexico [21,22]. It is recommended to establish recovery periods to enhance biomass production and promote the growth of highly palatable plants for livestock while managing negative impacts on shrub species. Similar outcomes to those observed in Mexico due to overgrazing have been reported in South Africa [23]. The consequences of overgrazing include a decline in rangeland condition, a reduction in palatable forages, and changes in plant species composition [24]. Rangeland management should consider the perceptual evidence of changes in soil and vegetation patterns, as well as socioeconomic factors such as land tenure and forms of organization [25].

The relationship between species richness and sampled area is a consistent rule in plant ecology [26] and has particular importance in biodiversity studies as it enables predictions of richness at larger scales [27]. However, species richness is also influenced by absolute plant abundance, spatial patterns, and the degree of species mixing [28].

Due to this, it is crucial to consider the scale at which biodiversity should be analyzed and assessed [29], as long as there are variations across different scales, and studying it at multiple scales would provide a more accurate representation of the overall patterns and drivers of biodiversity in pastures [30].

Scale is important because biodiversity is influenced by various factors operating at different spatial and temporal scales [31]. Local-scale factors, such as habitat structure and management practices, can have a significant impact on the composition and abundance of species within a small area [32]. On the other hand, landscape-scale factors, such as land use patterns and connectivity, can influence the movement and dispersal of plant species across larger areas [33]. Moreover, regional and global-scale processes, such as climate change and species invasions, can have far-reaching consequences for pasture biodiversity [34].

By considering multiple scales, researchers can gain insights into the interactions between these different factors and their effects on biodiversity [35]. This approach allows for a more holistic understanding of the mechanisms shaping biodiversity patterns in pastures. Furthermore, considering scale helps to identify the appropriate spatial and temporal scales at which conservation and management interventions should be implemented to maximize their effectiveness [36].

In this study, we hypothesized that grazing enhances diversity through microenvironmental disturbance caused by livestock. An additional hypothesis was that the sampling scale (from 0.01 m² to 100 m²) alters plant diversity on control and grazing-excluded plots. By supplying pertinent information based on empirical findings, these results will equip range managers with the knowledge necessary to make informed decisions, fostering a proactive and adaptive approach to the sustainable management of the ecosystem.

2. Materials and Methods

2.1. Study Area

The study site is located in southeastern Coahuila State, which serves as a transition zone between the Chihuahuan Desert and the Sierra Madre Oriental physiographic province (25°13′57.48″–25°14′ 57.25″ N and 100°56′ 44.62″–101°01′5.17″ W). The study area lies within the Sierra de Zapaliname natural protected area (Figure 1).

The climate is arid to semiarid and falls under the BSKw classification of semiarid temperate weather (a cold steppe climate with dry winters, characterized by limited rainfall, relatively moderate temperatures, and a vegetation pattern dominated by grasslands and pastures), with precipitation occurring mainly in summer [37]. The study plots were established at elevations ranging from 2102 to 2268 masl. The site features calcareous rocks and deep, well-drained soils. The average annual temperature is 16.9 °C, and the average annual precipitation is 498 mm. The plant community in the area is dominated by Bouteloua curtipendula, B. dactyloides, B. gracilis, B. uniflora, Aristida havardii, A. pansa, and Muhlenbergia phleoides [38]. The woody species scattered in the area are Alloberberis trifoliolata, Buddleja scordioides, Gymnosperma glutinosum, Mimosa biuncifera, and Prosopis glandulosa.

Figure 1. Study site showing the sampling plots (black points for pair of plots) positioned through the pastures studied and the location in the protected area Sierra de Zapaliname, Coahuila State, Mexico [39].

Figure 1. Study site showing the sampling plots (black points for pair of plots) positioned through the pastures studied and the location in the protected area Sierra de Zapaliname, Coahuila State, Mexico [39].

Agricultural activities began in the late 19th century [40]. Cropland is largely devoted to wheat, corn, beans, and barley. Additionally, fruit trees were cultivated in the pasture areas and alluvial valleys. Currently, approximately 400 ha of pasture in the study site are grazed by cattle and horses, with a relatively constant population of 63 cattle heads and 37 equines. This total number of animals has remained constant in the last decade with a regular use of the pastures (personal communication).

Samplings were conducted between 2017 and 2021 during the humid period of the year (August). In 2016, the annual rainfall exceeded 500 mm. However, over the subsequent years, rainfall decreased to approximately 200 mm (much drier than the average). The average annual temperature remained relatively constant, with minimal fluctuations of less than 0.5 °C (Figure 2).

2.2. Sampling Design and Sample Collection

On March 2017, a systematic survey of the main pasture community in the natural protected area of Sierra Zapaliname was conducted. Along a transect, we established eight pairs of square plots (20 × 20 m²) approximately 1000 m apart from each other. One of the plots in each pair was excluded from livestock grazing, using barbed wire, while the other was used as a control. Within each plot, 10 × 10 m² permanent plots were established, from which plant samples were collected. The control and grazing-excluded plots were separated by a minimum of 10 m. We used a global positioning system (GPS; Etrex, Garmin Ltd., Olathe, KS, USA) to register plot position and elevation.

To examine the effects of scale, we assessed vegetation at various sampling scales. We followed the procedure of Peet et al. [41,42] and recorded all vascular plant species in the nested square quadrats at each corner of the plots. The quadrats’ areas were 0.01, 0.1, 1, and 10 m². We also observed new plant species in the remaining 100 m² plot. Cover percentage of each plant species in the 100 m² plot was estimated using a 10-point scale (1 = trace, 2 = <1% cover, 3 = 1–2%, 4 = 2–5%, 5 = 5–10%, 6 = 10–25%, 7 = 25–50%, 8 = 50–75%, 9 = >75%, 10 = 100%). Latitude, longitude, altitude, and slope, were also recorded (

Table 1. Plot characteristics (“E” for exclusion plots and “C” for control plots). Aspect measured in degrees using as reference the north and slope in sexagesimal degrees. (*) Meters above sea level.

Plot	Alt (m.a.s.l.) *	Aspect	Slope
E1	2243	140	15
C1	2246	140	15
E2	2231	140	23
C2	2235	140	23
E3	2213	140	20
C3	2220	140	20
E4	2194	135	25
C4	2198	135	25
E5	2171	140	18
C5	2180	140	18
E6	2131	140	12
C6	2138	140	12
E7	2129	140	10
C7	2125	140	10
E8	2112	140	10
C8	2115	140	10

).

Plant specimens were collected, and their taxonomic identities were determined. Vouchers were deposited at the ANSM herbarium (Autonomous Agraria University Antonio Narro’s herbarium), and species names were confirmed using the checklist of vascular plants of the Sierra of Zapaliname [38].

2.3. Statistical Analysis

The use of power models in ecological research to describe species–area relationships has been established in various studies [26,43,44]. The equation S = cAz (where “S” is the number of species, “A” is the sampled area that can be related to the total amount of resources or primary productivity, “c” is a constant that represents the number of species that can be supported in a minimum area, and “z” is a scaling exponent that characterizes the relationship between area and species richness) is commonly used to understand the effects of scale on biodiversity and to compare biodiversity among different areas. In the present study, we tested the adjustment of the power function to our log–log data using the Pearson correlation coefficient (p < 0.05). The species–area relationships were examined using the number of species observed at each plot for each scale. For each plot, we obtained four data points for each scale except for the largest scale (100 m²), where only one data point existed. To estimate the parameters “c” and “z”, we employed logarithms to linearize the data. These parameters were estimated for each plot across different years.

In this study, we aimed to assess the impact of two factors, control vs. excluded and sampling year, on the “c” and “z” values of the plots. These effects were analyzed using the Generalized Linear Model (GLM) procedure in SAS (SAS Institute Inc., Cary, NC, USA). The control vs. excluded factor refers to whether the plots were subjected to grazing control or excluded from grazing. The sampling year factor indicates the specific year in which data were collected.

We treated the main effects (control vs. excluded and sampling year) as fixed effects, meaning that they were factors intentionally manipulated or recorded. Additionally, we included the pair of plots as random effects, accounting for the potential variability between specific plot pairs.

To ensure the validity of this analysis, the Breusch–Pagan test was conducted, which checks for homogeneity of variances. This test helps to determine if the assumption of equal variances across the different groups or treatments is met. We considered a significance level of p < 0.05 to determine the presence of any significant deviations from homogeneity.

Rarefaction analysis is a valuable tool in ecology as it accounts for variations in sample sizes and allows for fair comparisons of species richness between different sampling efforts, providing a more comprehensive understanding of biodiversity patterns and comparisons. We implemented rarefaction analyses and calculated the expected number of species as a function of sampling effort using the species presence in subplots (except for 100 m² plots, where the entire plot was used). We calculated rarefaction curves for grazed and non-grazed plots at different scales (160 plots for 0.01, 0.1, 1, and 10 m² scales; and 40 plots for the 100 m² scale). The accumulation curve incorporated the 95% confidence interval and was represented together for each different scale. Basic statistical methods followed that of Legendre and Legendre [45] and were carried out in the Vegan R software (version 4.1.3) [46].

3. Results

The study area had consistent environmental features under similar management conditions throughout the transect, with minimal altitudinal changes of less than 150 m, uniform aspect, and slope ranging from 10 to 20 degrees as shown in Table 1.

A comprehensive survey of the study area yielded 161 plant species (Appendix A). Out of these species, only three were identified as introduced, namely Asphodelus fistulosus, Malva parviflora, and Tribulus terrestris. However, these introduced species did not exhibit dominance over the control and grazing-excluded plots. Based on the classification of functional groups, the prevailing group was forbs, with 106 species, followed by grasses with 30 species, and then shrubs and cacti with 13 and 10 species, respectively, dominating these pastures. Out of the total species count, only 60 were deemed palatable. One fern was also found during the studied period (Ophioglossum engelmannii). The richness of species in these pastures not only supports local biodiversity, but also sustains important ecological processes, including pollination, nutrient cycling, and wildlife movement.

The estimated power function (

Table 2. Power function constants for each plot at different years (tre = treatment; co: control; ex: grazing excluded).

		2017		2018		2019		2020		2021
Plot	tre	z	c	z	c	z	c	z	c	z	c
1	co	0.23	0.69	0.32	0.55	0.31	0.55	0.30	0.61	0.25	0.75
2	co	0.31	0.60	0.36	0.51	0.37	0.43	0.40	0.42	0.34	0.53
3	co	0.28	0.62	0.31	0.52	0.33	0.47	0.32	0.56	0.27	0.66
4	co	0.29	0.65	0.35	0.50	0.27	0.69	0.33	0.62	0.25	0.76
5	co	0.34	0.37	0.34	0.38	0.33	0.42	0.40	0.26	0.28	0.57
6	co	0.28	0.62	0.27	0.68	0.24	0.78	0.26	0.72	0.25	0.72
7	co	0.23	0.54	0.24	0.56	0.23	0.62	0.20	0.77	0.24	0.64
8	co	0.27	0.59	0.28	0.57	0.26	0.73	0.23	0.79	0.23	0.75
1	ex	0.28	0.62	0.33	0.52	0.28	0.66	0.31	0.54	0.27	0.63
2	ex	0.32	0.55	0.32	0.55	0.37	0.43	0.38	0.45	0.34	0.56
3	ex	0.35	0.55	0.37	0.40	0.38	0.30	0.38	0.43	0.32	0.62
4	ex	0.28	0.71	0.30	0.63	0.32	0.46	0.33	0.57	0.28	0.70
5	ex	0.35	0.32	0.33	0.39	0.39	0.18	0.36	0.32	0.34	0.37
6	ex	0.28	0.58	0.30	0.53	0.33	0.51	0.27	0.68	0.29	0.60
7	ex	0.34	0.30	0.29	0.48	0.33	0.40	0.37	0.34	0.34	0.38
8	ex	0.31	0.41	0.32	0.43	0.37	0.40	0.28	0.63	0.27	0.63

) for all plots along the five years of sampling was significant in all cases (p < 0.001, the adjustment to the log–log), meaning that the z and c constants estimated can be reliable. When comparing these constants with the fixed factors of treatment and year, and the random factor pair of plots, for the constant c, we found significant differences for management, but not for year, with higher values in the case of the grazing-excluded plots (F_1,80 = 13.11, p < 0.001), while these results were not significant for the year effect (F_4,80 = 2.11, p > 0.05; Figure 3a). In the case of the constant z, similar results were observed and we found significant differences in the case of management, with higher values in the excluded plots (F_1,80 = 13.82, p < 0.001), while these results were not affected by year (F_4,80 = 2.20, p > 0.05; Figure 3b).

The rarefaction estimated curves offered different results for the studied scales (Figure 4). Control plots revealed a rarefaction curve over the curve of the ungrazed plot at the scales of 1 dm² and 10,000 dm², being only significant in the first case over more than 100 accumulated plots. In 10 and 100 dm², the estimated curves for excluded plots were over the accumulated curve of control plots, but only significant over 100 accumulated. In the case of the accumulation curve of 1000 dm² plots, there were no differences (p > 0.05).

4. Discussion

The study examined the effects of five years of cattle grazing exclusion in a native pasture in northern Mexico. Previous studies on this site [39] revealed no significant differences in species richness, evenness, and soil nutrients between the grazed and ungrazed areas. However, some species showed a higher prevalence in the ungrazed compared to the grazed plots. Moreover, grazing exclusion of the rangeland led to the expansion of grasses, while forbs increased in the grazed areas, but only for a few species. These findings suggest that medium-term grazing exclusion did not significantly affect soil nutrient content, but promoted grass growth.

The analysis of the coefficient power functions c and z revealed that grazing management affected the accumulation of species along with the increase in sampling area as well as the overall species richness in this ecosystem. However, year did not affect the vegetation, which means that management variability is the most important factor driving diversity. Based on this power function, grazing will affect these diversity parameters. However, variation over the years did not have an effect, being considered important in other studies [7,47]. This hypothesis is supported by the relatively constant weather conditions of the study period.

Higher values of c (Figure 2a) for the exclusion plots indicated a higher species richness as the intercept with the logS–logA space. In practical terms, the higher intercept of c in the logS–logA graph for exclusion plots signifies a stronger sensitivity to area, indicating that relatively small changes in sampling area can lead to more pronounced shifts in the species richness. In the case of the z values, we also found higher values for the slope, which means a higher increase in richness concerning the area [26]. Consequently, higher z values emphasize the significance of landscape heterogeneity in fostering species coexistence and accentuate the need to consider both area and habitat diversity in conservation and management strategies. In previous studies on this site, general richness analyses revealed non-significant differences [39], so the consideration of the scale should be incorporated in the analyses to detect the impact of management on the plots [41].

When comparing the rarefaction curves of control and excluded subplots at different scales, we observed significant differences at different levels. Specifically, at small scales (1 dm²), the control plots had a higher species richness accumulation over 100 subplots. However, at 100 dm², the richness was higher in the grazed plots, with significant differences over 100 plots. These differences in the curves can indicate different levels of richness, as previously suggested [48], and reveals that grazing has a stronger effect at smaller scales. Grazing by goats can create microhabitats by consuming dominant plants, which promotes diversity and coexistence among species by reducing competition for light and nutrients [49,50]. However, this effect disappears at larger scales (10,000 dm²) where environmental conditions become the main factor affecting species richness [7], suggesting that grazing exclusion effects are significant only at short study scales. It is important to note that while the impacts of scale might be more evident in smaller areas, larger areas are not immune to scale-related effects. Understanding the scale at which ecological processes operate is crucial for effective management and conservation, regardless of the size of the area under consideration.

The study of plant biodiversity in pastures is crucial for understanding the ecological dynamics of these ecosystems. However, it is important to recognize the importance of scale in such studies [51,52]. The scale at which biodiversity is measured can greatly influence our understanding of the relationships between plant communities, environmental factors, and management practices [53]. Therefore, researchers need to carefully consider the appropriate spatial and temporal scales when studying plant biodiversity in pastures. In our case, we have a particular pasture with a specific animal density for different species, but we obtained consistent results with regard to the impact of the scale. It has been demonstrated that varying degrees of herbivory across different spatial scales can lead to a heterogeneous distribution of plant species, enhancing overall biodiversity within grazing landscapes [6]. Moreover, the study of Bakker [54] underscores how the variable consumption patterns of herbivores contribute to niche differentiation among plant species, thereby influencing the assembly and composition of plant communities. The intricate relationship between animal variability and its ecological ramifications underscores the need for a comprehensive understanding of these interactions to effectively manage and conserve pastures’ biodiversity and species composition.

The scale refers to the spatial extent and resolution at which biodiversity data are collected and analyzed, while temporal variability refers to changes in biodiversity over time. Considering scale is important because plant diversity can exhibit variations across different spatial scales, such as local, landscape, and regional scales, which are influenced by factors like habitat structure, land use patterns, and climate change [55,56]. Additionally, temporal variability in biodiversity is influenced by factors like seasonal changes, natural disturbances, and human activities, and understanding these fluctuations is vital for long-term monitoring and adaptive management strategies [57,58]. By considering both scale and temporal variability, researchers and land managers can gain a comprehensive understanding of the factors shaping plant biodiversity in pastures and develop informed conservation strategies that account for spatial and temporal dynamics. Managing recommendations can arise from these results.

5. Conclusions

These findings suggest that after five years of cattle and equine grazing exclusion, there were no significant differences in species richness concerning the scale over five years, but management revealed significant differences with a higher number of plant species and a higher increase in plant species accumulation along the scale in the case of the grazing-excluded plots. Therefore, our results indicate that medium-term grazing as the one occurring in the study area will have an impact on species richness as long as the scale is considered. As indicated in previous studies, grazing exclusion can lead to a decrease in species richness, but these data suggest that some areas of the pasture could be excluded from grazing for longer periods, as long as it is compatible with the economic needs of the local habitants, to investigate changes and promote diversity, especially for species associated with exclusion areas. However, it is imperative to strike a balance between conservation goals and the livelihoods of local communities. An effective implementation of grazing exclusion requires the careful consideration of traditional land use practices and the economic realities of those dependent on pasture resources. Collaborative efforts involving ecologists, community members, and policymakers can facilitate the establishment of well-defined exclusion zones that align with both ecological restoration objectives and the socioeconomic needs of the region.