Effects of Environmental Factors on Plant Productivity in the Mountain Grassland of the Mountain Zebra National Park, Eastern Cape, South Africa

Vegetation biomass distribution is a vital concept in plant life history, providing the basis for our understanding of plants’ responses or adaptive strategies [1,2]. Plant biomass determines the ability of an ecosystem to acquire energy, thus playing a vital role in shaping the community structure and the function of the ecosystem [3,4,5,6]. This is because vegetation biomass is a key contributor to soil organic matter, which can influence greenhouse gas emissions in terrestrial ecosystems [7]. Thus, vegetation biomass is considered to have a particular function in the global carbon (C) cycle [8]. The general prediction is that terrestrial ecosystems globally will be affected by climate change. This is because some of the anthropogenic impacts of climate change include changes in precipitation, atmospheric C, and plant biomass production [9,10]. Plant biomass contributes to food webs and the functioning of the overall ecosystem [11]. With an increase in atmospheric C concentrations over the past several decades, studying plant biomass is essential for understanding vegetation dynamics, terrestrial ecosystem C stocks, and their responses to environmental changes [12]. Plant biomass represents the amount of total energy stored in a given unit of surface area during an observation period. It is influenced by the availability of limited resources such as water, carbon dioxide, and nitrogen (N) [13]; however, the interactive effects of these environmental factors on plant biomass remain unclear [14]. Environmental factors play an important role in C accumulation and contribute to C transformation [15]. Environmental factors influence C absorption, thereby influencing plant growth and biomass [16,17].

Grassland ecosystems cover approximately 40% of the Earth’s land surface, and they account for 1/10 of global C storage [18]. Consequently, it is critical to investigate the plant productivity of these ecosystems to understand their vegetation dynamics, terrestrial ecosystem C stocks, and response to environmental changes [18]. A shift in environmental conditions would affect vegetation biomass [19]. Soil nutrients influence the C absorption of plants, thereby influencing plant growth and causing increases in biomass [16,17].

Several studies have demonstrated that plant biomass influences the structure of a community and the function of ecosystems and is influenced by environmental factors [12,13,14]. However, little is known about the interactive effects among these environmental factors on plant biomass [20,21]. The responses of plant biomass to different environmental factors are complex, and research on the interactive effects of potential environmental driving factors (such as the climate, soil properties, and topographic properties) on plant biomass distribution in the Mountain Zebra National Park (MZNP) is scarce. Therefore, in the present study, we assessed plant productivity across the MZNP and examined the influence of environmental factors on plant productivity. We focused on aboveground biomass (AGB), since it can be measured without causing any disturbance to the topsoil, unlike belowground biomass (BGB) measurements, which require the entire plant to be uprooted.

The objectives of the present study were to determine whether biomass is evenly distributed across the park, whether environmental factors play a role in plant biomass production, and whether different plant genera account for different biomass values. We hypothesized that a higher availability of N would enhance high plant productivity because N is one of the basic nutrients that limits plant growth and, subsequently, plant productivity/biomass production [22].

First, the biomass pattern across all 2594 (50 × 50 + 47 × 2 plots) locations was examined. A graphical illustration of this is shown in Figure 2, and a quantitative summary is presented in Table 1. The distribution was positively skewed, with only a few of the values greater than 15 g/m². Although the maximum biomass recorded was 36 g/m², 75% of the records did not exceed 10 g/m². Therefore, it is not surprising that the data contained significant evidence (p-value of approximately 0.0000) against the hypothesis that biomass was normally distributed across the MZNP. The p-values provided in red indicate the existence of statistically significant evidence that goes against the hypothesis, i.e., that shows the corresponding data to be normally distributed.

Second, the distribution of average biomass as a function of the observed genus was investigated. See Figure 2 and Figure 3 for a visual illustration and a quantitative summary of this, respectively. The distribution was approximately normally distributed, given the concept of the central limit theorem [40]. There was insufficient evidence in the observed data (p-value = ~0.4011) to reject the claim that mean biomass, as a function of genera, was normally distributed. It can, however, be inferred that the distribution showed a mild left tail caused by the low average biomass of the Felicia genus.

The average biomass distribution pattern was further investigated in terms of plots and with respect to species (Figure 2 and Figure 3). In both instances, the inferences were similar to those obtained, with respect to the approximate normal distribution of genera and high p-values that indicated insignificant evidence against the normal distribution claim (Table 2). As seen for the genera, the other pairs of mean biomass distributions had tails. The tail with respect to the plots was mild and caused by the relatively high average biomasses of Plot 3 and Plot 35. However, the tail in the case of the species was heavy on both ends of the distribution. This was because of the particularly high mean biomass recorded for Asparagus striatus and Diospyros austro-africana, as well as the relatively low average biomass recorded for Felicia filifolia. The two species with large mean biomasses had to be temporarily excluded from the data so that the normal distribution hypothesis would not be rejected. The average biomass computed plot-wise contained the least evidence against a normal distribution hypothesis. Therefore, this justified the utilization of the GLS method to subsequently analyze the relationship between average biomass values and different environmental factors (Table 3).

Vegetation unit and water pH (“p-value” approximately 0.0385) were the two variables that were each identified as influencing the average biomass (Table 3). An increase in water pH of 0.10 tended to be associated with a decrease in average plot biomass of approximately 0.069 g/m². The average plot biomass seemed to be similar across Vus, except for VU2–3 (“p-value” ≈ 0.0333). Compared with that at VU1, the average biomass tended to be approximately 3.1 g/m² higher at VU2–3. The relationship between average biomass and each of the average yearly rainfall and soil organic C values merits further investigation as the significance of these associations only became weakened after the p-value adjustment. Similar post-adjustment evidence weakening was observed for VU9–11. Therefore, despite the identification of a pair of variables that were significantly associated with average biomass, more extensive studying is required with data collected from more plots.

We further sought to understand the optimal linear combination of the observed environmental factors, including the standard deviation index, which explained the recorded variation in average biomass (Table 4). A three-variable model that included aspect, VU, and water pH was identified as the optimal model. It was estimated that the model explained approximately 43.65% of the observed variation in average biomass. Given that VU and water pH have been previously identified to have a significant pairwise relationship with average biomass, their inclusion in the parsimonious model was not surprising. Based on a similar analogy, the presence of aspect in the model was unexpected. A possible explanation for this is that the aspect variable had a confounding effect. This hypothesis was investigated further.

Figure 4 and Figure 5 were used to investigate whether the importance of the aspect variable in the optimal biomass model was because of its potential role as a confounder. For the purpose of representation, the aspect variable used to generate the graphs in the figures was adjusted. North-east and north-west were re-classified as north. Similarly, south-east and south-west were re-classified as south.

Figure 4 shows that VU9–11 had the highest average in the eastern and southern aspects, whereas this did not appear in the west. Moreover, VU2–3 was among those with the highest mean across all aspects, except in the east, where it did not appear. Therefore, the aspect played a confounding role in the association between mean biomass and VU.

More stimulating features were evident in Figure 5, where the confounding effects of the aspect on the association between water pH and average biomass were explored. Except for the south, all the aspect categories had extreme water pH values that skewed the distributions leftward. Notably, the extreme values were influential. The directions of the slopes of the regression lines changed from negative to positive for the eastern and northern aspects. For the western aspect, excluding only the extreme values caused the regression slope to increase. Therefore, it is plausible to claim that there was a negative association between water pH and average biomass for the western aspect. Conversely, for the southern aspect, the association between water pH and average biomass appeared to be strictly positive. This contrast validates the suspected confounding attribute of the aspect variable on the pairwise relationship between the water pH and the average biomass.

The MZNP is the interface between three biomes, namely, grassland, Nama Karoo, and Albany thicket. The biomass across the park was not normally distributed, owing to the high biomass of the genera Asparagus and Diospyros as well as the relatively low average biomass recorded for F. filifolia. Moreover, D. austro-africana and Asparagus are dominant within the Nama Karoo biome, while F. filifolia dominates the grassland biome within the park. The high-biomass species were visibly dominant in the study area while F. filifolia was not dominant, which may explain their higher biomass contribution. Soil pH was the sole factor investigated in this study that markedly affected plant biomass production. Other soil factors that were examined did not demonstrate significant effects on plant biomass production. This is inconsistent with the findings of the studies by Baraloto et al., Sun et al., and Yang et al. [44,45,46], which report significant relationships between biomass and soil properties in grasslands. Furthermore, Sun et al. [45] reported a positive relationship between soil C density and AGB.

Soil N content did not influence plant biomass production in the present study. This is in line with the findings of several studies stating that soil N content has little to no influence on plant biomass production in most terrestrial ecosystems [47,48]. Bhandari and Zhang [49] observed a negative relationship between soil N content and plant biomass production, in which a relatively high N content resulted in a relatively low biomass production, while a relatively low N content translated to a relatively high biomass production. A study conducted in an alpine meadow also revealed that N content is a major factor affecting AGB, influencing the species richness [50].

Vegetation biomass is also an important factor [51] as biomass is a major contributor to soil organic matter, which can influence greenhouse gas emissions in the terrestrial ecosystem. Biomass also plays a major role in energy production, which is one of the greatest essential quantitative characteristics for understanding community structure and ecosystem function.

The positive association observed between plant biomass and soil pH in the present study is consistent with the findings of a study conducted by Bhandari and Zhang [49], which states that soil pH and plant biomass production have a positive relationship at the southern aspect [52]. Although most of the environmental factors did not influence biomass distribution in the present study, studies that examine environmental factors and plant biomass remain invaluable for an overall understanding of plant biomass distribution in an area.

Regionally, AGB is primarily determined according to altitude, climate, and soil fertility [53,54]. However, in the present study, there was no significant relationship between AGB and mean annual precipitation.