Seasonal Effects of Non-selective Grazing on Soil Infiltration, Compaction, and Vegetation Structure in a Semi-arid Grassland of Northeast Mexico

2.1. Study Area

The study was conducted in a 32-hectare experimental area dominated by A. clandestina in north-eastern Mexico, located at the “Los Ángeles” Experimental Ranch. The coordinates are 25° 06' 30”N and 100° 59' 18” W, at an altitude of 2,162 meters (Fig. 1). According to the Köppen classification, modified by Peel, [13], the climate is BS_okw(e'), dry, with a precipitation/temperature ratio of 22.9 mm/16 °C, warm summers, and an average temperature of 7 to 14 °C. The average annual rainfall is 350 mm, divided into two periods: May to October (87%) and November to April (13%) (Fig. 2). Historical records (1971-2010) reflect the high interannual variability typical of semi-arid climates in northeastern Mexico. Annual precipitation ranges from 200 to 550 mm (x̄ = 345 mm), with extreme temperatures ranging from -12.5 to 40 °C. In addition, a regional warming trend of +3.4 °C has been observed over the last three decades, and the basic soil properties include a fine clayey-sandy texture and a shallow soil type [14].

2.2. Methodology

A randomized complete block design was employed, with the four seasons of the year serving as blocks and three replicates per season, providing sufficient statistical power for seasonal block comparisons. The experimental units consisted of 600 m² plots (10 x 60 m), spatially distributed in a randomized block array within the 32-hectare area. Grazing was carried out at the beginning of each season using non-selective grazing [15] (operationally defined as short-duration, high-density grazing designed to non-selectively consume all available biomass), with an instantaneous stocking rate of 333 livestock units (LU ha^-1) (where 1 LU is defined as 450 kg live weight) at full forage availability [16] (quantified as >3,000 kg DM ha^-1). This stocking rate represents an instantaneous, high-intensity grazing event applied over a single day per season (winter: January 29, spring: March 31, summer: June 23, autumn: September 23, 2023), following the non-selective grazing protocol established by Acocks [15]. Measurements were conducted immediately before (0-5 hours prior) and immediately after (0-5 hours post) each grazing event.

2.3. Variables Evaluated

2.4. Statistical Analysis

The instantaneous and cumulative infiltration rates were modelled using the Kostiakov-Lewis equation, and model fit was evaluated based on the coefficient of determination (R² > 0.90), indicating adequate predictive capacity for seasonal variation in infiltration. Prior to inferential analyses, the normality of residuals was assessed using the Shapiro-Wilk test, and homogeneity of variances was verified with Levene's test. When both assumptions were met, ANOVA was applied to evaluate differences in infiltration, compaction, and soil moisture between sites and seasons [21]. The separation of means was performed using Tukey's HSD test for seasonal comparisons [22] and Student's t-test for grazing condition contrasts (p < 0.05). To estimate the magnitude and precision of the observed differences, effect sizes and 95% confidence intervals were calculated and reported alongside all p-values. For variables that did not meet the normality assumption, including basal cover and soil moisture percentage, a nonparametric approach was adopted. Associations between physical and productive variables were therefore analysed using Spearman's rank correlation, which does not assume normality or linearity and is robust to the presence of outliers [23]. All analyses were performed in R (R Core Team, 2025) using RStudio (version 2025.05.1+513).

3.1. Instantaneous Infiltration Rate and Accumulated Infiltration

The seasonal values ​​of cumulative and instantaneous infiltration in response to grazing management are summarized in Figs. (3 and 4), demonstrating a hydrological dynamic with defined temporal variations. Under pre-grazing conditions, cumulative infiltration ranged from 9.0 cm (autumn) to 13.5 cm (winter), while the instantaneous infiltration rate ranged from 32.4 cm h^-1 (autumn) to 58.5 cm h^-1 (winter). After grazing, the highest cumulative infiltration was recorded in spring (14.8 cm) and the lowest in autumn (7.1 cm); the instantaneous infiltration rate followed a similar pattern, reaching its maximum in spring (47.5 cm h^-1) and decreasing significantly in autumn (15.2 cm h^-1).

The Kostiakov-Lewis model provided a robust fit across all seasons and treatments (R² > 0.90; Table 1), confirming its high predictive capacity despite seasonal variability in infiltration. This behaviour indicates that the dynamics of water infiltration in the soil can be adequately described by power functions, which accurately model the progressive decrease in the infiltration rate over time (Figs. 3 and 4).

3.2. Infiltration Rate

Figure 5 shows that, when comparing infiltration rates across seasons (inter-seasonal comparison), water infiltration rates did not show significant differences ( p ≥ 0.05) between pre- and post-grazing conditions, except for the summer season. In contrast, when comparing pre- and post-grazing infiltration within each season (intra-seasonal comparison), only the summer season showed a significant difference ( p < 0.05). This distinction is important: seasonality, not grazing pressure, was the primary driver of inter-seasonal variation in infiltration rates.

3.3. Soil Moisture

In all seasons evaluated, statistically significant differences (p < 0.05) were observed in soil moisture before and after non-selective grazing (Fig. 6). The autumn season recorded the highest percentage of post-grazing moisture (59%), while the lowest moisture corresponded to the winter season with 21% in grazed areas. On the other hand, before grazing, statistically significant differences (p < 0.05) were detected in moisture content. Under these initial conditions, the autumn season again had the highest moisture content (50%), followed by summer (30%), spring (23%), and winter (10%). In contrast, when comparing soil moisture before and after grazing by season, there was a significant difference (p < 0.05) only in winter between the two events.

3.4. Soil Resistance to Penetration

Figure 7a shows that the winter season showed a statistically significant difference (p < 0.05) at all four depths (2.5, 5.0, 7.5 and 10 cm); the highest compaction was recorded at 10.0 cm pre-grazing, with a penetration resistance exceeding 3096 kPa, while post-grazing highest values were 2360 and 2270 kPa at 10.0 and 7.5 cm, respectively. During spring (Fig. 7b), post-grazing compaction showed significant differences (p < 0.05) at all depths, with the highest post-grazing compaction at 10.0 and 7.5 cm, reaching 2,495 and 1,926 kPa, respectively. The lowest pre-grazing penetration resistance was recorded at 2.5 cm, with 179 kPa.

In the summer season (Fig. 7c), statistical differences (p < 0.05) were recorded in soil compaction at all four depths. However, before the impact of grazing, the highest compaction was recorded at 10.0 and 7.5 cm, with 1429 and 1240 kPa, respectively, while the highest post-grazing compaction was also recorded at 10.0 and 7.5 cm, with 2183 and 2027 kPa, respectively. Finally, in autumn (Fig. 7d), compaction after grazing showed significant differences (p < 0.05) between depths. Post-grazing, there was a statistical difference, with the highest compaction at 10 cm (penetration resistance of 1844 kPa) and the lowest at 2.5 cm pre-grazing (243 kPa). Regarding soil compaction between pre- and post-grazing conditions at the same depth, a significant difference was detected in autumn at 2.5 and 10 cm. In contrast, no significant differences (p ≥ 0.05) were observed in winter, spring, and summer.

3.5. Vegetation Structure

Figure 8a shows the kg DM ha^-1 present in each season before each grazing event, where the highest dry matter production was recorded in spring with 3347 kg DM ha^-1, followed by winter, summer, and autumn with 3307, 3010, and 1907 kg DM ha^-1, respectively. However, significant differences (p < 0.05) were recorded between pre- and post-grazing within each season. Post-grazing dry matter production showed a decrease, with the highest amount of dry matter recorded in summer, followed by spring, winter, and autumn, with values of 661, 437, 398, and 218 kg DM ha^-1, respectively. In contrast, when comparing kg DM ha^-1 between pre- and post-grazing conditions within the same season, significant differences (p < 0.05) were found in all four seasons, with pre-grazing dry matter being higher than post-grazing dry matter in all cases.

Plant height is shown in Fig. (8b), which recorded a statistically significant difference (p < 0.05) between pre- and post-grazing. However, the pre-grazing winter season showed the highest grass height at 53 cm, followed by spring, summer, and autumn, with values of 50, 46, and 37 cm, respectively. Post-grazing, height showed significant differences (p < 0.05) between seasons, with the highest post-grazing grass height occurring in spring at 16 cm, followed by winter, autumn, and summer at 14, 13, and 11 cm, respectively. In contrast, the comparison of height between pre- and post-grazing within a season showed significant differences (p < 0.05) in all seasons, with higher values in pre-grazing.

On the other hand, basal cover (Fig. 8c) in each season before and after grazing showed no significant difference (p ≥ 0.05). However, the highest basal cover was recorded before grazing in summer at 4%, followed by spring and autumn at 3.1% and 2.9%, respectively. The lowest value was recorded in winter with 1.7% basal cover. Post-grazing basal cover did not show a statistically significant difference (p ≥ 0.05); however, winter had the highest cover at 3.7%, followed by autumn, spring, and summer at 2.2, 1.8, and 1.7%, respectively. The comparison of basal cover between pre- and post-grazing within a season showed significant differences (p < 0.05) in all seasons, with higher values in pre-grazing. Aerial coverage (Fig. 8d) pre- and post-grazing showed a statistical difference (p < 0.05) between seasons, with spring showing the highest aerial coverage at 26%, followed by summer (25%), winter (23%), and autumn (18%). On the other hand, post-grazing aerial coverage in the winter season recorded the highest percentage at 11%, followed by autumn, summer, and winter at 6, 5, and 3%, respectively. The comparison of aerial coverage between pre- and post-grazing in the same season showed significant differences (p < 0.05) in all seasons, with higher values in pre-grazing.

3.6. Correlation

Correlation analysis (Table 2) revealed positive relationships between dry matter production (DMP) and the following variables. Strong associations with height and above-ground cover (0.86 and 0.85, respectively) stood out, while the correlations with basal cover (0.50) and infiltration (0.38) were low to moderate. On the other hand, height (HEI) showed a positive association with aerial cover (AC) (0.80), whereas it showed negative associations with basal cover (BC) and infiltration (INF) (0.36 and 0.23, respectively). Basal cover correlated moderately with both aerial cover and infiltration (0.44 in both cases) and weakly with soil moisture percentage (SMP) (0.22).

Finally, the relationship between aerial cover and infiltration had a coefficient of 0.36. Positive correlations in soil compaction were observed across all depths. The 2.5 cm stratum was strongly associated with the 5.0 cm stratum (0.92), with the relationship decreasing in the 7.5 and 10 cm layers (0.64 and 0.49, respectively). In turn, the 5 cm depth correlated with the 7.5 and 10 cm depths (0.79 and 0.61), highlighting the strong association between the deep 7.5 and 10 cm strata (0.91).

4.1. Instantaneous and Cumulative Infiltration

The pattern of water infiltration into the soil depends closely on factors such as porosity and external factors, including vegetation cover, seasonal rainfall, and management practices [24]. The highest cumulative infiltration was in spring, followed by winter, which can be attributed to lower soil compaction in the top 2.5 cm (Fig. 7) after grazing, as well as to initial soil moisture conditions, where lower water content was observed in the winter and spring seasons (Fig. 6). In addition, greater aerial cover was recorded in these seasons (Fig. 8). In contrast, lower values were recorded in summer and autumn, due to lower structural resistance and higher soil moisture associated with high precipitation (Fig. 2), which reduces infiltration [25]. In this context, the instantaneous infiltration rate exhibits typical behaviour across all seasons, starting with high values during the first 20 minutes followed by a progressive decline towards an asymptote. This pattern is consistent with Kostiakov's law [26] and other potential models that describe the decrease in infiltration as the soil becomes saturated [27]. In particular, the higher initial rate present in spring and winter could be associated with a looser soil structure and macropores formed by biological activity and wetting-drying cycles [28]. Additionally, the potential adjustment for cumulative infiltration (Table 1, Fig. 4) yields high R² values across all seasons, indicating a robust representation of the phenomenon [29].

4.2. Infiltration Rate

When comparing instantaneous infiltration rates within each season, no statistically significant differences due to grazing were found, except in the summer season. However, seasonal variation was observed, suggesting that factors such as initial soil moisture, vegetation cover, and precipitation have a greater influence than grazing pressure [17]. Nevertheless, the lower water infiltration rate (Fig. 5) indicates that cattle trampling compacts the soil, increasing its apparent density and resistance to the penetrometer, thereby limiting infiltration [30]. This compaction reduces infiltration by up to 81% due to lower soil porosity and higher density [8]. In semi-arid climates, spring infiltration may be increased by vegetation cover, which protects the soil surface from direct compaction [31].

In summer, heavy rainfall favours initial infiltration, except when extreme drought forms crusts that limit water absorption [32]. In addition, the interaction between temperature and humidity alters the soil structure and influences its microporosity [33]. In winter, low evaporation and reduced biological activity favour structural conservation and, with it, infiltration [34]. Therefore, the results show that seasonality is the main determinant of infiltration, even though grazing did not produce significant changes. It should therefore be considered that the interaction between livestock and the ecosystem is complex and can manifest itself cumulatively and in the long term [17]. In this context, the claim that non-selective grazing can improve infiltration [9] should be interpreted with caution. This effect has been documented primarily with low to moderate stocking rates or on soils with very pronounced crusts, where hoof action alters surface sealing and increases microporosity. However, field studies in Mediterranean silvopastoral systems of the Iberian Peninsula [10], tallgrass prairies of the United States [11], and global meta-analyses [9] consistently report that non-selective grazing reduces infiltration due to surface sealing and subsurface compaction, particularly when applied under high soil moisture conditions, findings that are consistent with those of the present study. Internationally, comparable reductions in infiltration have been documented with high livestock loads in semi-arid systems in Australia, Sub-Saharan Africa, and South America [9], which reinforces the conclusion that the grazing-infiltration relationship is not linear and is mediated by the intensity of the livestock load, soil texture, and previous moisture conditions.

4.3. Soil Moisture

The results obtained show variations in soil moisture before and after grazing, influenced by vegetation type, grazing intensity, and climatic factors [35]. The higher moisture content in the autumn season after grazing may be associated with more frequent rainfall and moderate temperatures, conditions that favour water retention in the soil [36]. In this context, compaction and water infiltration are influenced by soil moisture, as shown in Fig. (6). The autumn season had the highest moisture content; however, there was no significant relationship with compaction. It should be noted that in the spring season, torrential rains were recorded where the cattle were grazing, and therefore compaction was greater than in the other seasons. In this sense, an increase in water content can lead to higher bulk density and soil resistance, especially in clayey and silty soils [37].

4.4. Soil Resistance to Penetration

Wet-dry cycles can reduce soil compaction by increasing microporosity, especially in soils with lower grazing intensity [38]. These cycles facilitate soil structure recovery by allowing soil expansion and compaction, which helps reduce hardness [39]. Freeze-thaw cycles promote soil recovery by generating cracks and volumetric expansion, reducing compaction. Their effectiveness depends on soil type and climatic conditions [40]. In addition to physical processes, the presence of roots and macrofauna activity contribute significantly to soil recovery. Roots break down compacted layers, and macrofauna generate micropores that promote infiltration [39]. These mechanisms in arid ecosystems are significantly influenced by precipitation, and the processes can persist for decades. However, there is a contrasting perspective that suggests that non-selective grazing results in vegetation recovery, as the action of hooves associated with short-term, high-intensity grazing becomes an alternative mechanism for increasing and maintaining the productivity of arid grasslands by incorporating seeds into the soil [41]. On the other hand, vegetation cover such as grasslands and shrubs significantly affects infiltration and runoff during heavy rainfall events (4-6 mm min^-1) [42]. These compaction patterns are consistent with results reported in comparable ecosystems internationally. In the montane Mediterranean grasslands of Portugal, [10] a study recorded penetration resistance values exceeding 2000 kPa at depths of 5-10 cm following intensive livestock trampling under wet conditions, which closely resembles the fall results of the present study. In tallgrass prairies of the southern United States, [11] another study documented persistent subsurface compaction under continuous intensive grazing, with structural recovery observed only after prolonged rest periods. In contrast, [9] a global meta-analysis demonstrated that light to moderate grazing can maintain or reduce compaction relative to ungrazed controls, highlighting grazing intensity as the critical variable across diverse grassland ecosystems.

4.5. Vegetation Structure

During this study, dry matter production prior to grazing was highest in spring, followed by summer, winter, and autumn, coinciding with higher levels of solar radiation and precipitation [43]. Post-grazing, the reduction in forage was clearly noticeable and attributed to the high intensity of grazing by animals on the biomass present [44]. At the same time, pre-grazing plant height showed differences between seasons, with winter standing out, as seasonal phenology either favours or hinders plant development [45]. After grazing, the greatest height was recorded in spring, coinciding with studies that indicate rapid plant recovery in seasons with better temperature conditions and water availability [46]. No differences (p ≥ 0.05) were observed in basal cover between seasons, either before or after grazing. However, it was higher in winter, due to the strategy of certain species to concentrate their basal growth in response to defoliation or stress [47].

Nevertheless, after grazing, the highest aerial cover was observed in winter, followed by autumn, summer and spring, suggesting greater structural resilience in the season with less extreme climatic conditions [48]. The results show the influence of seasonality and grazing on grassland dynamics, highlighting the importance of regulating grazing intensity and frequency to maintain species productivity and persistence throughout the year [43].

4.6. Correlation

Several studies show that plant height and density are positively correlated with total biomass; taller and denser species have higher aerial and forage productivity [49]. The results show that soil compaction is negatively correlated with productivity and water infiltration, leading to lower biomass production. Mora and Lázaro [50] describe how a high apparent density negatively affects plant growth by reducing root growth and water uptake. This is why compaction promotes small species or those with shallow roots, thus reducing dry matter. Research also indicates that soil organic matter improves water infiltration and retention [51]. Although these studies focus on crops, the principle applies to grasslands: well-nourished, aerated soils under natural vegetation infiltrate better than bare or eroded soils [52].

4.7. Limitations, Practical Implications, and Future Perspectives

It is important to acknowledge several limitations of this study. The low number of replicates, the single annual cycle, and the spatial heterogeneity inherent in semi-arid grasslands limit the generalizability of the results and prevent the assessment of long-term cumulative effects. Despite these limitations, the seasonal patterns documented here have direct practical implications: grazing during spring, when soil moisture is low and vegetation cover provides surface protection, had the least impact on infiltration capacity, whereas grazing in autumn under conditions of maximum soil moisture intensified subsurface compaction more markedly [9]. These findings suggest that reducing stocking rates and postponing grazing events during periods of high moisture, combined with minimum recovery intervals of a full growing season, could mitigate structural soil degradation [49]. Furthermore, rotational, adaptive multi-pasture systems and non-selective grazing distribute grazing pressure more evenly across space and time and have proven effective in reducing compaction and maintaining infiltration capacity in semi-arid grasslands [11], representing a viable alternative for grasslands in northeastern Mexico.

Future research should prioritize the integration of precision livestock management technologies, such as GPS tracking, wireless soil sensors, and drone-based remote sensing, to facilitate real-time adjustments in grazing pressure [53]. This integration is essential to prevent soil compaction from exceeding irreversible thresholds, thereby strengthening adaptive management responses to climate variability in the semi-arid grasslands of northeastern Mexico. Likewise, the development of multi-year monitoring programs that incorporate these digital tools will help elucidate the relative contributions of stocking rates, grazing chronology, and seasonal climate fluctuations to long-term soil structural recovery and primary productivity [54].