Livestock Grazing Impact on Species Composition and Richness Understory of the Pinus cembroides Zucc. Forest in Northeastern Mexico

Abstract

In the pine forests of Mexico, disturbances are primarily due to cattle, horses, goat, and sheep grazing, particularly in communal grazing lands. The most evident disturbances are low tree recruitment, invasive shrubs establishment, species composition changes, and invasion of weeds dispersed mainly by livestock. The Sierra de Zapalinamé is a mountain range and natural protected area of northeast Mexico. We conducted the current study in this area in a forest stand of Pinus cembroides excluded from grazing in the last 25 years (1200 ha with pine forest vegetation and mountain chaparral) and another area nearby subjected to livestock grazing. Forest structure (basal area and density), tree species richness, total understory species richness, and understory species composition were analyzed at the control and grazed sites. Our results revealed that grazing has modified the understory species composition and reduced the evenness in the control plots. Therefore, to maintain species diversity and forest structure, we concluded that extensive grazing should be restricted for some areas or the number of animals reduced in zones of high ecological value.

1. Introduction

Livestock grazing is one of the most important activities worldwide [1], requiring correct techniques to maintain species composition and soil conservation in grasslands [2], as these cases can cause remarkable and significant variation in plant species composition [3,4,5]. In the pine forests of northern Mexico, grasslands disturbance is due to overgrazing by cattle, horses, goats, and sheep, particularly in communal areas [6]. The most evident disturbances are low tree recruitment and the establishment of invasive shrubs and weeds dispersed mainly by livestock.

Mexico is second in the diversity of pinyon pine woodlands after Eurasia [7]. The distribution of Mexican pinyons includes mountainous zones of Arizona, New Mexico, and Texas in the southwest United States of America and from northern and central Mexico down to the state of Puebla [8]. The pinyon pine forests, dominated by Pinus cembroides Zucc., occupy extensive areas in the eastern and western Sierra Madre mountains [9]. These drought-tolerant pines develop at altitudes of 1800 to 2800 m, on dry soils and rocky slopes of mountains and hills with poorly developed soils, in temperate dry to temperate sub-humid zones. They are common in transition zones between xeric and relatively mesic forest communities at higher elevations [10].

The Sierra de Zapalinamé is a mountainous protected area enacted by the state of Coahuila, Mexico [11]. Here, the pinyon pine forest grows on low slopes with slight slants and inter-mountain valleys with deep soils in temperate sites at altitudes from 2150 to 2650 m [12]. Little is known about the effects of heavy grazing in this type of vegetation on forest stands and associated vegetation, even though forest grazing is a common management practice applied worldwide [13]. Thus, it was considered pertinent to shed light on vegetation diversity in pinyon pine forest to elaborate guides for pinyon pine forest richness conservation in northeastern Mexico. Also, it would be valuable to improve the soil quality, restore vegetation and recover wildlife in an overgrazed pinyon pine forest.

The main objective of this study was to analyze the impact of livestock grazing exclusion for 25 years on species composition and soil nutrients in a pinyon pine forest. We hypothesized that Pinus cembroides forests excluded from grazing for a prolonged period have a positive impact on species richness and species composition and that the effect on nutrient composition will also be significant. Therefore, determining species composition when combining pines and cattle can be valuable for managers in the decision-making procedure for Pinus cembroides forests.

2. Materials and Methods

2.1. Study Site

The Sierra of Zapalinamé is located in northeastern Mexico, and it has an area of 45,000 ha. It is located south of the city of Saltillo, between 25°15′00″ N and 25°25′58″ N and between 100°47′14″ W and 101°05′03″ W (Figure 1) and belongs to the Gran Sierra Plegada physiographic subprovince. The elevation ranges from 1590 m in the foothills to 3140 m in the highest mountain, with inter-mountain valleys averaging 2200 m a.s.l. Rocks of the area are sedimentary, belonging to the Jurassic and Cretaceous periods; limestone covers 43% of the area, while 17% is sandstone and conglomerates [14]. Alluvial soils occupy 30% of the area with variable depth and are mainly found in the plains with alluvial fans at the mountain base. Soils in the valleys are deep. There are also smaller areas of calcium and phaeozem calcaric xerosols.

The prevailing climate of the study area is the dry type (BSkw), while the upper parts of the mountain have a temperate type (C(w0)) following Köppen’s classification. The average annual temperature is 16.9 °C, and the mean annual rainfall is 498 mm [14]. Rains are convective and occur mainly in the warmest months of the year. Different plant communities have been recorded for this area, including rosetophyllous scrub, pine forest, fir forest, oak forest, and montane chaparral. In the protected area, pinyon pine forest occupies 12.54% of the area surrounded by a xeric scrubland (9.55% of the protected area) [15]. Pinyon pine forests are distributed mainly in the Cuauhtémoc and Sierra Hermosa canyons; Pinus cembroides grow scattered among Juniperus flaccida Schltdl. and J. deppeana Steud. On the branches of pine trees, the epiphytic Tillandsia recurvata (L.) L. is common [16]. In more conserved areas, the herb layer is dominated by the grasses Piptochaetium fimbriatum (Kunth) Hitchc. and Bouteloua dactyloides (Nutt.) Columbus. In areas with intense perturbation by cattle grazing, Asphodelus fistulosus L. and Gymnosperma glutinosum Less. are abundant [15]. In this forest, the greatest perturbation is caused by human intervention, through land-use change to establish agricultural areas, causing fragmentation and surface area reduction.

This study was conducted in a forest stand excluded from grazing for 25 years on the San José del Anhelo private property (1200 ha with pinyon pine forest and mountain chaparral). This area was used to establish the grazing-excluded plots. This exclusion management technique was intended to improve soil quality, restore vegetation and recover wildlife. Outside the private property, in the same potential vegetation stand, we located the grazed plots (Ejido Cuauhtémoc, 270 ha), where extensive grazing takes place with 104 cows, 18 donkeys, 28 horses, 683 goats, and 84 sheep (this information has remained relatively constant in the studied area for the last 20 years according to personal communication from the protected area managers). Owners of this site selectively log pines for house and fences construction, Christmas trees, and harvest of pinyon nuts. These activities have become the main economic activity of the surrounding settlements. Therefore, we avoided the areas with intensive management or use in our study.

2.2. Sampling Design

In August 2017, we systematically located 16 (30 × 30 m) rectangular plots in a stand of Pinus cembroides in the Sierra of Zapalinamé natural protected area. We established eight plots in the grazed area (grazed plots) and eight in the grazing-excluded area (control plots). Plots were located along a transect in the center of the stand at a distance of 100 m (avoiding trails or human disturbances). In each plot, we measured land altitude and slope and estimated the canopy cover of the stand using a convex spherical crown densitometer [17]. We visually estimated the percentage cover of rock, bare soil, and litter cover in each plot. Grass cover (only grasses excluding forbs) and understory woody species cover also were visually estimated. We considered trees to be individuals whose stems were ≥2.5 cm DBH (diameter at breast height). We measure DBH for all trees alive in the 30 × 30 m plots to estimate basal area and density to ha. Previous studies recommended these classifications following the physiognomy of these species [18]. This category will be considered as part of the canopy for its description.

We identified all herbs and shrubs in a concentric plot of 10 × 10 m. Cover for all the species on plot surfaces was estimated and recorded on a scale of 1 to 9 (cover classes: 1: traces; 2: >1% of cover in the plot; 3: 1%–2%; 4: 2%–5%; 5: 5%–10%; 6: 10%–25%; 7: 25%–50%; 8: 50%–75%; 9: >75%). Taxonomic identities of collected plant specimens were determined and vouchers were deposited at the ANSM herbarium. For species names, we followed the checklist of vascular plants of the Sierra of Zapalinamé [12]. Plot position and elevation were measured using a global positioning system (GPS; Etrex, Garmin Ltd., Olathe, KS, USA).

We took four soil samples (0 to 10 cm in depth in each corner of the plots). These were mixed, dried, and sifted through a 2-mm sieve; debris and stones were eliminated. Organic matter content was determined by the Walkley and Black method [19], and pH was measured in a soil-to-water ratio of 1:5 extract. Soil total nitrogen (the Kjeldahl method), phosphorus Olsen (0.5 M NaHCO₃ with a soil/solution ratio of 1:20 [20]), K, Na, Mg, Ca, and electrical conductivity were determined. After adding lanthanum to a solution, Ca and Mg contents were determined by the ASS Manual (Wollongbar Agricultural Institute, Australia). Soil K and Na concentrations were determined after extraction with the AL-method (open vessel extraction with ammonium lactate and acetic acid). We also calculated Cation Exchange Capacity. The qualitative levels baseline for nutrients obtained, such as N, P, K, and organic matter, were according to SEMARNAT (2000) [21] and Fernandez-Linares et al. (2006) [22].

2.3. Statistical Analysis

A one-way distance-based permutational t-statistic [23] was performed for comparison between grazed and control plots (as factors) for species richness, Smith and Wilson evenness [24], basal area (m²/ha), and tree density (individuals/ha). The analyses were based on the Bray–Curtis distance of the raw data, with p-values < 0.05 obtained with 9999 permutations and a Monte Carlo correction where necessary. Primer 6 and Permanova+ (PRIMER-E Ltd., Plymouth, UK) were used to perform all PERMANOVA statistical procedures. We used the same comparison analysis for the following soil nutrient and environmental variables: pH, EC (exchangeable cations, dS/m), Polsen (Phosphorus Olsen extraction in ppm), organic matter (% OM), available cations in meq/100 (Na, K, Ca, Mg), cation exchange capacity (CEC) and total nitrogen (% TN), soil cover, grass cover, woody species cover, litter cover estimated in percentages and canopy cover.

As a technique of direct gradient analysis, we used partial Canonical Correspondence Analysis (CCA; [25]) in CANOCO 5.1. [26] to examine how species composition changed over the different plots as a function of the environmental characteristics included in the analysis. In the environmental matrix, we used the following variables: pH, EC (exchangeable cations, dS/m), Polsen (Phosphorus Olsen extraction in ppm), % OM, available cations in meq/100 (Na, K, Ca, Mg), CEC, a and %TN. Additionally, we included soil cover, grass cover, woody species cover, litter cover estimated in percentages, and canopy cover. As a biotic matrix, we used the total species composition based on the cover of the 10 × 10 m plots. We selected the three most informative environmental variables, applying a forward selection procedure to remove the variables that did not explain a significant portion of the variability reported by the analysis when performing the axes (Monte Carlo permutation test with 9999 interactions for p < 0.05). Axes I and II are graphically displayed with the selected environmental variables and plots enclosed in a different polygon for grazed vs. control. Species are presented separately in the same bio-dimensional space of CCA axes I and II.

An MRPP (Multi-response Permutation Procedure) was used to determine changes in species composition between grazed and control plots with a matrix base in cover. The Bray-Curtis distance was used for this analysis [27]. For the same data matrix, a Species Indicator Analysis was used to determine the significant representative species in each group [28]. The analyses were carried out in the vegan R package [29].

3. Results

The altitude of the plots ranged between 2350–2500 and as for slope and canopy cover, differences were not significant for control vs. grazed plots. Grass cover was higher (PseudoF_1,14 = 19.87, p < 0.01) in the control plots, while woody species cover, soil, and litter were higher in the grazed plots (PseudoF_1,14 = 8.20, PseudoF_1,14 = 5.45 and PseudoF_1,14 = 11.18 respectively, with a p < 0.05 for the first two and p < 0.01 for the last one). The rest of the variables, including the nutrient content, did not reveal significant differences (

Table 1. General abiotic information of experimental plots and soil nutrients and characteristics of experimental plots. The values′ average (Avg) and standard deviations (Std) for grazed and control plots are included.

								% Cover
Plots	Treatment (1)		Alt (m)	Slope	Grass *	Woody *	Rock	Soil *	Litter *	Canopy
PG1	Grazed		2351	10	25	25	3	1	60	53.6
PG2	Grazed		2346	11	60	5	2	2	20	47.2
PG3	Grazed		2342	39	20	10	3	10	10	56.6
PG4	Grazed		2356	28	30	15	15	20	10	52.8
PG5	Grazed		2372	10	40	10	2	30	15	55.6
PG6	Grazed		2372	10	55	15	1	30	20	49.8
PG7	Grazed		2379	9	50	10	1	30	20	52
PG8	Grazed		2394	10	45	10	1	45	15	55.2
Average			2364.0	15.9	40.6	12.5	3.5	21.0	21.3	52.9
SD			18.1	11.3	14.5	6.0	4.7	15.6	16.2	3.2
PC2	Control		2419	10	80	5	1	2	2	55.4
PC3	Control		2436	10	70	10	2	5	3	57.2
PC4	Control		2450	12	70	10	1	2	7	43.6
PC5	Control		2468	10	85	5	1	2	10	52.2
PC6	Control		2466	11	85	7	1	1	2	36.6
PC7	Control		2498	11	65	5	3	15	7	53.6
PC8	Control		2501	11	70	5	1	1	10	42
Average			2458.4	10.8	75.6	6.5	1.4	4.0	5.8	49.8
SD			30.5	0.7	7.8	2.3	0.7	4.7	3.3	7.9
		%	%	ppm	µS/cm			meq/100 g
Plots	pH	OM	TN	P Ols	EC	Ca	Mg	Na	K	CEC
PG1	7.5	15.5	0.63	21.65	1048	4.4	12.4	10,463.2	0.2	39
PG2	7.2	13.8	0.64	20.13	1223	12	16	12,202	0.2	12
PG3	5.2	14.5	0.81	25.36	1387	12	6	13,852	0.28	29
PG4	7.4	17.5	0.75	11.36	1471	12	8	14,690	0.18	30
PG5	4.0	14.5	0.47	8.35	1580	10	10	15,780	0.31	25
PG6	7.6	6.9	0.4	17.37	1262	20	4.8	12,595.2	0.22	21
PG7	6.8	7.8	0.27	3.01	916	8.4	11.6	9140	0.26	20
PG8	2.3	6.2	0.26	3.28	374	12	20	3708	0.22	30
Avg	6.0	12.1	0.5	13.8	1157.6	11.4	11.1	11,553.8	0.2	25.8
Std	2.0	4.4	0.2	8.5	383.4	4.4	5.1	3837.0	0.0	8.2
PC1	7.3	17.8	0.58	11.36	1238	8.4	12	12,359.6	0.22	30
PC2	4.0	7.4	0.27	11.64	971	10.8	9.2	9690	0.2	24
PC3	7.3	18.2	0.5	6.71	1282	14	10	12,796	0.41	32
PC4	7.7	7.8	0.31	3.56	1022	8.8	12	10,199.2	0.17	30
PC5	7.5	9.1	0.46	10.54	970	8	4	9688	0.27	25
PC6	7.6	24.5	0.77	17.65	1366	14	6	13,640	0.17	14
PC7	7.3	14.5	0.77	17.65	1027	10	10.4	10,249.6	0.23	25
PC8	7.0	17.5	0.71	18.89	1365	15.2	16.8	13,618	0.27	12
Avg	6.9	14.6	0.5	12.3	1155.1	11.2	10.1	11,530.1	0.2	24.0
Std	1.2	6.1	0.2	5.5	174.8	2.9	3.9	1744.0	0.1	7.4

(*) Indicate significant differences between grazed vs. control plots for a p < 0.05. (1) “Grazed” for plots where grazing has been active for the last 25 years, and “control” for non-grazed plots for all this period. OM: Organic matter; TN: Total nitrogen; P Ols: Phosphorus Olsen; EC: Electric Conductivity; CEC: Cation exchange capacity.

).

We found 12 tree species in the study site, 6 in the grazed plots, and 11 in the control plots. Fraxinus greggii A. Gray was not present in the control plots, while Arbutus xalapensis Kunth, Arctostaphylos pungens Kunth, Pinus arizonica Engelm., Rhus virens Lindh. ex A. Gray and Yucca carnerosana (Trel.) McKelvey were not present in the grazed plots. Pinus cembroides, Juniperus deppeana, J. flaccida, J. coahuilensis (Martínez) Gaussen ex R.P. Adams, Quercus saltillensis Trel. and Q. microphylla Née were present in both study sites.

Pinus cembroides was the dominant species with higher values of basal area in grazed plots (85.6%) than in control plots (72.0%). Tree species richness differed significantly between sites with 3.6 ± 1.6 (mean ± standard deviation) for the eight plots in grazed plots and 6.0 ± 1.3 (mean ± standard deviation) in control plots (PseudoF_1,14 = 10.51, p < 0.05). There were no significant differences for density and basal area (PseudoF_1,14 = 2.83 and PseudoF_1,14 = 1.50, p = n.s. respectively) between sites. More detailed information about the canopy can be found in Arévalo et al. [16].

A total richness of 163 understory species was found in the pinyon pine forest; the families with the most plant species were Asteraceae with 45 species, Poaceae 18, and Fabaceae 8. Herbs were the most common biological form, with 119 species. In addition, we recorded 55 species in the control plots, which did not occur in the grazed plots, while 34 species were present only in grazed plots (Appendix A).

Concerning species richness, non-significant differences existed between experimental sites. At the same time, Smith and Wilson Evenness revealed significant differences, with higher values in the grazed than the control plots (PseudoF_1,14 = 1.47, p = n.s. and PseudoF_1,14 = 5.20, p < 0.05 respectively; Figure 2).

The CCA analysis revealed that of all the physical parameters analyzed, only three were significant for the distribution of the species composition: grass cover, canopy cover, and soil total nitrogen. First, the CCA axis discriminated grazed vs. control plots, with grass cover more representative in the control plots and canopy cover more representative in the grazed plots. However, axis 2 was related to %TN with higher values in the grazed plots (Figure 3). In the case of the species, a gradient of variation in species composition was revealed from control to grazed plots.

Differences between treatments based on species cover were significant (MRPP) with a T = −7.102 and group probability correction of A = 0.087 for a p < 0.01.

Finally, the ISA base in 1000 permutations revealed that the indicator species for the grazed plots were Bouteloua dactyloides, Crusea diversifolia, and Dichondra brachypoda, while Dalea radicans, Muhlenbergia rigida (Kunth) Kunth, Arbutus xalapensis, Hedeoma costata A. Gray, Juniperus deppeana, and Malaxis brachystachys (Lindl.) Rchb. f. were indicators for the control plots (p < 0.01 for all species). These indicator species were represented in the CCA biplot in bold letters (Figure 4).

4. Discussion

Some environmental characteristics markedly differed between sites, such as grass cover, soil cover, litter and woody species cover, and characteristics related to grazing intensity [30]. However, differences in nutrient content were not significant. The intensity of herbivores′ effect on plant communities varies along environmental gradients or vegetation stands [31], even being insignificant or null at low grazing intensities [32,33,34].

For forest structure (basal area and density), differences were not significant, although the tree species richness was higher in control plots than in grazed plots, indicating an adverse grazing impact on some species such as Arbutus xalapensis or Yucca carnerosana [16]. These species are highly palatable for goats and cattle [35,36]. Some studies have demonstrated that grazing impacts species richness [37,38].

Although it is dependent on a spatial scale and highly related to climate variability [39,40] and resource availability [38,41]. In the present study, this impact was significant for tree species, although basal area and density of trees did not differ, suggesting that grazing exclusion did not promote changes in forest structure for that period.

Regarding total species number, we found non-significant differences. However, evenness was higher in grazed plots. This can be seen with the reduction through grazing of dominant species and the larger values of grass cover in the control plots. Several studies have found higher evenness in grazed than ungrazed treatments [40,42,43]. This was also reflected in the current study. However, some studies measuring grazing effects on plant species composition and species richness have traditionally been inconsistent and conflicting in their results, lacking a general model that predicts the response of grazing intensity or abandonment [37,44], and the lack of consistent results has been attributed to high factors variability such as the evolutionary history of grazing, productivity gradients or grazing intensity [33].

The ordination analyses revealed that grazed plots vs. control plots are discriminated based on the total species cover, and only three environmental variables significantly explained the species composition. They were grass cover, canopy cover, and total nitrogen. Nitrogen was not an important discriminant variable among sites and was more related to particular conditions of the soil. However, grass cover and canopy cover were important variables explaining species composition in control and grazed plots. The reduction of grass cover is a typical result [45], but grazing does not typically impact canopy cover. Canopy cover and grass cover are inversely related [46], but some studies have revealed a low relationship between canopy cover and grazing [47].

In control plots, two terrestrial orchids grow: Malaxis brachystachys and Goodyera oblongifolia Raf. Moreover, species like Dahlia tubulata P.D. Sørensen, Geranium semannii, Gibasis geniculata (Jacq.) Rohweder and Salvia regla Cav. Are frequent but did not occur in the grazed plots. Both species groups prefer to grow in mesic forest conditions and undisturbed sites. These species are common in oak forests in this region [18], growing in humid canyons.

The grasses with high coverage in the herbaceous stratum in the control plots were Muhlenbergia rigida and Piptochaetium fimbriatum, both tufted grasses that grow up to 1.0 m tall. On the other hand, in the grazed plots, Bouteloua dactyloides was the dominant species. This is a cespitose, stoloniferous short grass that forms a short carpet up to 10 cm high [48]. According to Encina-Domínguez et al. (2019) [15], it is a common grass in the Zapalinamé mountain range, growing in remnant semi-desert grasslands located in valleys with deep soils that support intense cattle and horse grazing.

In grazed plots, annual weeds [49], such as Bidens odorata, Dyssodia papposa (Vent.) Hitchc., D. pinnata (Cav.) B.L. Rob., Euphorbia dentata Michx., Lepidium virginicum L. and Viguiera dentata (Cav.) Spreng also grow. These are dispersed by grazing livestock from croplands adjacent to the forest. It is noteworthy that Prunus cercocarpifolia Villarreal was found only in grazed forests, a shrubby rhizomatous species endemic to this mountain range [12]. In these areas, the cacti Echinocereus knippelianus Liebner, in conservation status by the Mexican government, also grows [50].

We only found one exotic species on these analyzed pastures. In grazed plots, a perennial weed Asphodelus fistulosus L. an exotic species from Eurasia grows with scattered individuals. In the mountain range, it is common along the roads, abandoned agricultural fields, and overgrazed areas. In this area, with a long dry season and a very cold winter, both are two strong environmental filters that may limit the establishment of many native ruderal and exotic species [51], which can explain its low number.

Species composition was well discriminated, with some species as the control or grazed plots indicators. Shrubs and trees were indicators on control plots, together with some other herbs, while herbs only are indicators of grazed plots. Livestock grazing reduced highly palatable shrubs, particularly by goats. In general, there was a species turn-over from the grazed to control plots, being more similar to the climax vegetation of these forests. Future research directions may also be highlighted.

5. Conclusions

Uncontrolled livestock grazing has modified the species composition in the Pinus cembroides forest. The number of trees has been affected negatively, reducing the number of species. Other studies have also revealed an adverse effect of cattle grazing on species richness and plants [52,53,54].

We conclude that extensive grazing carried out for decades in the Pinus cembroides stand of the Sierra de Zapalinamé should be restricted or the number of animals reduced in zones of high ecological value, to maintain diversity and forest structure. Livestock grazing is a necessary activity for the economy of farmers in communal lands because of the meat products obtained from cattle, goats, and sheep in the local area. We suggest the application of controlled grazing pressure for some areas of particular conservation interest to restore mature, persistent P. cembroides forest to a more historical condition.