Anthropogenic Pressure on Tree Species Diversity, Composition, and Growth of Balanites aegyptiaca in Dinder Biosphere Reserve, Sudan

Anthropogenic disturbances, such as illegal harvesting and livestock browsing, often affect natural forests. However, the resulting tree species diversity, composition, and population structure have rarely been quantified. We assessed tree species diversity and importance value indices and, in particular, Balanites aegyptiaca (L.) Del. population structure, across 100 sample plots of 25 m × 40 m in disturbed and non-disturbed sites at the Dinder Biosphere Reserve, Sudan, from April 2019 to April 2020. We found that the tree species diversity in non-disturbed sites was more than double that of disturbed sites (p < 0.001, T = 32.6), and seedlings and saplings comprised more than 72% of the entire tree population (F2,48 = 116.4, p = 0.034; F2,48 = 163.2, p = 0.021, respectively). The tree density of B. aegyptiaca in the disturbed site was less than half that of the non-disturbed site (p = 0.018, T = 2.6). Balanites aegyptiaca was seven times more aggregated in disturbed sites compared to more regularly spaced trees in non-disturbed sites (T = 39.3 and p < 0.001). The poor B. aegyptiaca population status of the disturbed site shows that the conservation of this vulnerable species is essential for a sustainable management and utilization scheme.


Introduction
Natural forests, forest plantations, biosphere reserves, and other protected areas have always consolidated human lives ecologically and economically [1][2][3]. The ecological support of the forests can be observed through its hydrological functions, carbon sequestration, sites of recreation, biodiversity protection and conservation [1,[4][5][6]. Their economic contribution includes the provision of food, timber and non-timber products, fodder for livestock, medicinal extracts, and shelter for humans and animals [2,7,8]. Biosphere reserves constitute a massive stock of biodiversity and perform a principal role towards the satisfaction of various needs for the local communities worldwide [9,10]. Particularly in sub-Saharan countries and the Sahel region, the recent and rapidly increasing over-exploitation and other human disturbances may strongly degrade the dynamics of these resources [11,12].
In eastern Africa, trees and shrubs, often growing inside or neighboring protected areas, provide multiple products and services to local people [13,14]. This high anthropogenic pressure can disturb the population dynamics of trees by eliminating their natural  Moreover, our findings showed that the proportion of the seven most common tree species were significantly different for adult trees (F 6,404 = 22.3; p = 0.012), saplings (F 6,404 = 31.1; p = 0.01), and seedlings (F 6,404 = 27.6; p = 0.013) between the non-disturbed and disturbed sites, with high percentages of seedlings and saplings in the non-disturbed site Moreover, our findings showed that the proportion of the seven most common tree species were significantly different for adult trees (F6,404 = 22.3; p = 0.012), saplings (F6,404 = 31.1; p = 0.01), and seedlings (F6,404 = 27.6; p = 0.013) between the non-disturbed and disturbed sites, with high percentages of seedlings and saplings in the non-disturbed site and adult trees dominating the disturbed site ( Figure 2). The percentage of the adult trees for B. aegyptiaca, C. hartmannianum, L. fruticosa, and Ziziphus spina-christi (L.) Desf., in the disturbed site were double that in non-disturbed site, while their seedlings were more than three times as high as those in the disturbed site ( Figure 2). However, Anogeissus leiocarpus illustrated an inverse pattern, with higher proportions of seedlings and saplings in the disturbed sites and more adult trees in the non-disturbed sites.  Asterisks above the bars show significant differences across the sites for each species according to Tukey's Post-Hoc tests (* = p < 0.05; ** = p < 0.01; *** = p < 0.001; n = 406).

High Importance Value Index (IVI) and High Dominance in Disturbed Sites
Both the disturbed and non-disturbed sites were dominated by B. aegyptiaca, C. hartmannianum and Acacia seyal var seyal, all of which had a species importance value index  (Tables A1 and A2). Balanites aegyptiaca was the most dominant tree species in the non-disturbed site, but less so in the disturbed site (F 1,43 = 138.8, p < 0.001). The importance value index (IVI) for the frequently encountered tree species was significantly higher in the disturbed vs. non-disturbed sites (F 6,406 = 88.3, p = 0.0123).
We also found that relative abundance, dominance, and frequency for most tree species in the DBR were lower in the disturbed sites, except for Acacia seyal and Acacia senegal (L.) Willd., demonstrating the inverse trend (F 1,43 = 205.4, p < 0.001). Acacia seyal was four times as dominant in the disturbed than in the non-disturbed site, while its abundance was almost double that of the non-disturbed site (F 1,43 = 164.3, p < 0.01). The same trend was observed for the importance value index (Tables A1-A3). Further, the total number of mature trees for B. aegyptiaca, Anogeissus leiocarpus and Lannea fruticosa in the non-disturbed site was almost double and three times that of the disturbed site, respectively (Table A3).
Our principal component analysis on the IVI values showed that both the highest and lowest IVI values were found in disturbed sites. The first group (tree species with high IVI) was dominated by Acacia seyal, Combretum hartmannianum, Balanites aegyptiaca, Ziziphus spina-christi, and Acacia senegal), while the second and third groups (medium and low IVI values, respectively) were discriminated by Anogeissus leiocarpus, Acacia polyacantha, Terminalia brownii Fresen, Lannea fruticosa, Dichrostachys cinerea (L.) Wight and Arn., Tamarindus indica L., and Sterculia setigera Del ( Figure 3). In addition, the analysis effectively expressed and explained 78% of the variation among the disturbed and non-disturbed sites.

High Importance Value Index (IVI) and High Dominance in Disturbed Sites
Both the disturbed and non-disturbed sites were dominated by B. aegyptiaca, C. hartmannianum and Acacia seyal var seyal, all of which had a species importance value index (IVI) of >28 (Tables A1 and A2). Balanites aegyptiaca was the most dominant tree species in the non-disturbed site, but less so in the disturbed site (F1,43 = 138.8, p < 0.001). The importance value index (IVI) for the frequently encountered tree species was significantly higher in the disturbed vs. non-disturbed sites (F6,406 = 88.3, p = 0.0123).
We also found that relative abundance, dominance, and frequency for most tree species in the DBR were lower in the disturbed sites, except for Acacia seyal and Acacia senegal (L.) Willd., demonstrating the inverse trend (F1,43 = 205.4, p < 0.001). Acacia seyal was four times as dominant in the disturbed than in the non-disturbed site, while its abundance was almost double that of the non-disturbed site (F1,43 = 164.3, p < 0.01). The same trend was observed for the importance value index (Tables A1-A3). Further, the total number of mature trees for B. aegyptiaca, Anogeissus leiocarpus and Lannea fruticosa in the non-disturbed site was almost double and three times that of the disturbed site, respectively (Table A3).
Our principal component analysis on the IVI values showed that both the highest and lowest IVI values were found in disturbed sites. The first group (tree species with high IVI) was dominated by Acacia seyal, Combretum hartmannianum, Balanites aegyptiaca, Ziziphus spina-christi, and Acacia senegal), while the second and third groups (medium and low IVI values, respectively) were discriminated by Anogeissus leiocarpus, Acacia polyacantha, Terminalia brownii Fresen, Lannea fruticosa, Dichrostachys cinerea (L.) Wight and Arn., Tamarindus indica L., and Sterculia setigera Del ( Figure 3). In addition, the analysis effectively expressed and explained 78% of the variation among the disturbed and non-disturbed sites.

Natural Regeneration and Growth form Distribution Are Highly Limited in Disturbed Sites
We found a vigorous natural regeneration at the non-disturbed site, where saplings and seedlings together represented more than 72% of the site population, while the disturbed site was dominated by mature trees with less than 15% saplings and seedlings ( Figure 4). The growth form distributions (mature tree, sapling, and seedling) were significantly different within and between the disturbed and non-disturbed sites, with an inverse distribution pattern (F 2,48 = 116.4, p = 0.034 and F 2,48 = 163.2, p = 0.021, respectively; Figure 4).

Natural Regeneration and Growth form Distribution Are Highly Limited in Disturbed Sites
We found a vigorous natural regeneration at the non-disturbed site, where saplings and seedlings together represented more than 72% of the site population, while the disturbed site was dominated by mature trees with less than 15% saplings and seedlings ( Figure 4). The growth form distributions (mature tree, sapling, and seedling) were significantly different within and between the disturbed and non-disturbed sites, with an inverse distribution pattern (F2,48 = 116.4, p = 0.034 and F2,48 = 163.2, p = 0.021, respectively; Figure 4). . Average ±SE number of stems of mature trees, saplings, and seedlings per ha across all tree species plotted along the disturbed and non-disturbed sites. Asterisks above bars signpost significant differences within the growth stages across the disturbed and non-disturbed sites according to Tukey's Post-Hoc tests (* = p < 0.05; ** = p < 0.01; *** = p < 0.001).

High Density and Crown Width of Balanites aegyptiaca in the Non-Disturbed Site
The density of B. aegyptiaca in the disturbed site was less than half that of the non-disturbed site (Table 1) while its tree crowns were twice as wide in the non-disturbed than in the disturbed sites. The basal area contribution (%) for B. aegyptiaca was significantly more aggregated in the disturbed site and more regularly spaced in the non-disturbed site (Blackman index, Table 1).

Figure 4.
Average ±SE number of stems of mature trees, saplings, and seedlings per ha across all tree species plotted along the disturbed and non-disturbed sites. Asterisks above bars signpost significant differences within the growth stages across the disturbed and non-disturbed sites according to Tukey's Post-Hoc tests (* = p < 0.05; ** = p < 0.01; *** = p < 0.001).

High Density and Crown Width of Balanites aegyptiaca in the Non-Disturbed Site
The density of B. aegyptiaca in the disturbed site was less than half that of the nondisturbed site (Table 1) while its tree crowns were twice as wide in the non-disturbed than in the disturbed sites. The basal area contribution (%) for B. aegyptiaca was significantly more aggregated in the disturbed site and more regularly spaced in the non-disturbed site (Blackman index, Table 1). The diameter at breast height (DBH) of B. aegyptiaca in the non-disturbed site exhibited a bell-shaped normal distribution while it resembled an inverse J-shape, positive asymmetrical Weibull distribution in the disturbed site (W = 0.853; Figure 5). The distribution of height classes displayed a different trend, with juvenile trees dominating in the nondisturbed site and mature trees in the disturbed one (W = 0.845; Figure 6). Further, the correlation between crown width and DBH showed a strong positive relationship with R 2 = 0.6 for both the disturbed and non-disturbed sites (Figure 7). At the disturbed sites, the slope of crown width vs. DBH was lower, indicating that the site was dominated by the medium-sized trees.  The diameter at breast height (DBH) of B. aegyptiaca in the non-disturbed site exhibited a bell-shaped normal distribution while it resembled an inverse J-shape, positive asymmetrical Weibull distribution in the disturbed site (W = 0.853; Figure 5). The distribution of height classes displayed a different trend, with juvenile trees dominating in the non-disturbed site and mature trees in the disturbed one (W = 0.845; Figure 6). Further, the correlation between crown width and DBH showed a strong positive relationship with R 2 = 0.6 for both the disturbed and non-disturbed sites (Figure 7). At the disturbed sites, the slope of crown width vs. DBH was lower, indicating that the site was dominated by the medium-sized trees.

Low Tree Species Diversity and Density in Disturbed Sites
Our findings revealed that the disturbed site has lower species richness, diversity, and evenness compared to the non-disturbed site, which is consistent with [1,12]. Only a few tree species dominated the disturbed sites, which might be more tolerant to livestock browsing. However, as tree diversity is an essential component for a healthy forest system [45,46], this limited number of dominant tree species raises a critical concern. Researchers [47] concluded that 76% of perennial plants in southern Jeddah, Saudi Arabia, disappeared due to ecological and anthropogenic disturbances, while [48] reported that tree species diversity and richness of Slovenian forests were reduced from 272 to 243 species because of deer browsing and livestock grazing. In contrast, [49] documented that

Low Tree Species Diversity and Density in Disturbed Sites
Our findings revealed that the disturbed site has lower species richness, diversity, and evenness compared to the non-disturbed site, which is consistent with [1,12]. Only a few tree species dominated the disturbed sites, which might be more tolerant to livestock browsing. However, as tree diversity is an essential component for a healthy forest system [45,46], this limited number of dominant tree species raises a critical concern. Researchers [47] concluded that 76% of perennial plants in southern Jeddah, Saudi Arabia, disappeared due to ecological and anthropogenic disturbances, while [48] reported that tree species diversity and richness of Slovenian forests were reduced from 272 to 243 species because of deer browsing and livestock grazing. In contrast, [49] documented that

Low Tree Species Diversity and Density in Disturbed Sites
Our findings revealed that the disturbed site has lower species richness, diversity, and evenness compared to the non-disturbed site, which is consistent with [1,12]. Only a few tree species dominated the disturbed sites, which might be more tolerant to livestock browsing. However, as tree diversity is an essential component for a healthy forest system [45,46], this limited number of dominant tree species raises a critical concern. Researchers [47] concluded that 76% of perennial plants in southern Jeddah, Saudi Arabia, disappeared due to ecological and anthropogenic disturbances, while [48] reported that tree species diversity and richness of Slovenian forests were reduced from 272 to 243 species because of deer browsing and livestock grazing. In contrast, [49] documented that low and intermediate disturbances increased the tree species richness, diversity, and functional richness of the boreal forests in Ontario. Our low species richness and diversity in the disturbed site may be attributed to the intensive livestock browsing and illegal harvesting in the area.
Accordingly, controlling livestock grazing and human trespassing in the Dinder Biosphere Reserve is crucial for minimizing this pressure and paving the way for tree recovery. We recommend that the introduction of community forestry in the degraded areas around the villages can contribute efficiently to controlling harvesting activities in reserves. However, we claim that awareness-raising programs are also urgently needed to reduce the current and future damage to natural regeneration in the reserve.
The inverse pattern of Anogeissus leiocarpus showing more seedlings and saplings in the disturbed sites and more adults in the non-disturbed site might be due to the intensive illegal harvesting of adult trees by the local community, as this species seems in high demand across eastern Africa [50][51][52][53]. Further, the high tannin content in its seedlings and saplings may reduce its palatability for livestock, and, hence, might have minimized the browsing impact [54].

High Importance Value Index (IVI) and Dominance in Disturbed Sites
Our results elucidate that Acacia leata, A. mellifera (Vahl) Benth, A. polyacantha, A. senegal, and A. seyal var fistula were frequent at the disturbed site. These tree species are characterized by a high growth rate and the ability to quickly encroach degraded and open spaced rangeland [55,56]. Hence, our disturbed areas might already show signs of bush encroachment through these species due to the intensive anthropogenic use. Accordingly, high densities of these Acacia species, combined with the ongoing illegal felling and livestock browsing, may compromise the recruitment of B. aegyptiaca seedlings and saplings in the disturbed site and eliminate its population, as was found for fruit and nut tree species [26,57,58]. The overutilization by livestock, bush encroachment, and overharvesting have also led to the degradation of useful tree species and a biodiversity decline in Ethiopia, Tanzania, and Botswana, respectively [56][57][58]. Our high value of IVI for Acacia seyal trees in disturbed sites, with a huge gap to the nearest tree species, is an indicator of the unbalanced ecosystem and needs urgent action to protect the vulnerable tree species in the area.
An unbalanced ecosystem generally appears as a result of biotic and abiotic disturbances or both, within or across different species [29,47,59]. For the disturbed sites in the DBR, both types were observed; the aggregated distribution patterns of B. aegyptiaca illustrated that seedlings and saplings were located in patches. In addition, the acrossspecies trend was shown by the low tree species richness in the disturbed compared to the non-disturbed sites.
Furthermore, the tree stand density is linearly proportional to the basal area, a key parameter of relative occurrence (relative dominance) [60]. Our results highlight that tree species such as B. aegyptiaca, T. brownii, and T. indica have been over-utilized and their populations severely disturbed. These results are in line with [34] and [60,61], who concluded that illegal harvesting and livestock browsing affected the stand density and the IVI in the natural forests in Sudan.

Natural Regeneration and Growth form Distribution Limited in Disturbed Sites
We found that more than 85% of the tree species in the disturbed site have a poor regeneration status, and some species such as T. indica were only found in their mature stage. In contrast, seedlings and saplings of B. aegyptiaca, Combretum hartmannianum, A. seyal, Anogeissus leiocarpus, and Lannea fruticosa represented 72% of the tree population at the non-disturbed site. With the assumption that all tree species in DBR are native [14,62], disturbances resulting from environmental factors should affect them equally. Wild animals are present in the non-disturbed site and the buffer zone of the DBR [17,63,64], where we did not find any reduced tree recruitment. Therefore, the reduction in natural regeneration and crown width we found in our disturbed site likely results from intensive browsing by livestock and a reduction of seed production as branches are cut and edible fruits are collected by the local communities. Our findings are in line with previous studies in Kenya and Sudan [2,60,61], where felling mainly affected mature trees while new recruitment was hampered by over-browsing by livestock.
Intensive browsing by livestock influences the recruitment of the tree seedlings and saplings and disturbs the population dynamics [2,19,65]. The density of seedlings of Quercus ithaburensis Decne and Quercus agrifolia Ne'e in Israel and north California, respectively, declined to <50% of the initial population due to intensive browsing and overgrazing of livestock [66,67] and a 50% tree species decline was observed in the Galician oak forest of Northwest Spain [68]. In our study, the low percentage of seedlings and saplings in the disturbed sites (15%) indicates that the recruitment of seedlings and saplings to adult trees was severely interrupted. Likewise, illegal harvesting through the cutting of branches, debarking of the stem, and the removal of entire trees eliminate the vulnerable tree species and reduce the species diversity in the highly disturbed stands [1,12,60], which we could also see for B. aegyptiaca tree species. Therefore, we highlight that all areas with new regeneration of B. aegyptiaca and other vulnerable tree species in the transition zone of the DBR must be strongly protected to allow natural recovery. In our study, we did not include the effect of forest fire and drought, which might also reduce the success of natural regeneration and damage tree seedlings and saplings [20,69,70] as there was only limited information on these factors. However, we can show that already, human and livestock activities alone leave a significant negative effect on recruitment and persistence of most tree species in the DBR, highlighting the fragility of this ecosystem.

High Density and Crown Width of Balanites aegyptiaca in the Non-Disturbed Sites
We found the high density and crown width but the low basal area contribution of B. aegyptiaca in the non-disturbed sites. This situation might result from seasonal foraging patterns as, during the dry season, pastoralists and livestock keepers debranch mature and taller trees for livestock to browse [1,71,72]. During the wet season, livestock utilize and depend on perennial and herbaceous plants that are generally available in and around the villages [41,47,73]. A healthy population structure is usually characterized by higher seedling and sapling (young generation) proportions compared to mature trees [2,19,60]. Our results of higher seedling and sapling densities in the non-disturbed sites agree with these conclusions. Moreover, the lower slope of the correlation between the crown width and DBH of B. aegyptiaca in the disturbed site demonstrates that even healthy trees might be under high competition in disturbed sites and cannot fully expand for effective photosynthesis.
The variation in class frequency distribution across disturbance sites may result from the massive medicinal uses of B. aegyptiaca by local people as they prune the foliage, debark the stem and dig for tree roots [31,34,74]. Intensive pruning and bark removal will reduce the quantity of foliage and interrupts tree growth [1]. Although minor disturbance can enhance the biological diversity in natural forests [75], extensive use may accelerate forest degradation [71,76]. While our study is only a snapshot in time, i.e., covering only one year of assessment, it provides information that can be used to initiate comprehensive monitoring and awareness-raising programs about the current degraded situation and help managing the reserve on a sustainable basis, particularly B. aegyptiaca stands.
Other factors that might cause a fluctuation in natural regeneration are the type of browsers that dominate the site [21,77]. Goats are characterized by an extensive browsing habit with the preference for woody plants and broadleaved tree species [77][78][79], while cattle and camels prefer grazing on herbaceous plants and browsing the crown of mature trees, respectively [19,21,80]. Although grasses might protect small tree seedlings from browsing livestock [81,82], in the DBR, only little grass cover remained during the dry season, and grasses were severely overgrazed by livestock [14,73,83]. Researchers [78] have documented that more than 85% of Pinus sylvestris L saplings in Mediterranean mountains were browsed more than once per season by goats, and more than 30% of the apical shoot had been consumed after establishment. Such browsing activity can eliminate the species in its critical stage of development and hinder the tree population dynamic, particularly for vulnerable species such as B. aegyptiaca.

Study Area
The study was conducted in the Dinder Biosphere Reserve (DBR), Sudan, located at 12 • 26 N, 12 • 42 N, 34 • 48 E and 35 • 02 E (Figure 8), with a total area of 10,291 km 2 [73]. The average monthly minimum and maximum temperatures were 18 • C and 30 • C, respectively, while the average annual rainfall was 775 mm [64,84]. The DBR consists of three zones, which are the transition, buffer, and core zone [17,18] (Figure 8). The transition zone is under high human pressure, encompassing more than 20 villages (disturbed), while the core zone is fully protected (non-disturbed) [73]. The disturbed site is under high anthropogenic pressure by frequent illegal harvesting, livestock grazing and agriculture [18,73] while at the non-disturbed site, timber harvesting, pastoral and agricultural as well as charcoal production are prohibited [14,73].
browsing livestock [81,82], in the DBR, only little grass cover remained during the dry season, and grasses were severely overgrazed by livestock [14,73,83]. Researchers [78] have documented that more than 85% of Pinus sylvestris L saplings in Mediterranean mountains were browsed more than once per season by goats, and more than 30% of the apical shoot had been consumed after establishment. Such browsing activity can eliminate the species in its critical stage of development and hinder the tree population dynamic, particularly for vulnerable species such as B. aegyptiaca.

Study Area
The study was conducted in the Dinder Biosphere Reserve (DBR), Sudan, located at 12°26′ N, 12°42′ N, 34°48′ E and 35°02′ E (Figure 8), with a total area of 10,291 km 2 [73]. The average monthly minimum and maximum temperatures were 18 °C and 30 °C , respectively, while the average annual rainfall was 775 mm [64,84]. The DBR consists of three zones, which are the transition, buffer, and core zone [17,18] (Figure 8). The transition zone is under high human pressure, encompassing more than 20 villages (disturbed), while the core zone is fully protected (non-disturbed) [73]. The disturbed site is under high anthropogenic pressure by frequent illegal harvesting, livestock grazing and agriculture [18,73] while at the non-disturbed site, timber harvesting, pastoral and agricultural as well as charcoal production are prohibited [14,73].

Data Collection
We randomly laid out 100 rectangular sample plots of 25 m × 40 m (1000 m 2 ) across the study area, 50 each in the disturbed and non-disturbed site, respectively, from April 2019 to April 2020. Within each sample plot, we identified all tree species and grouped them into adult trees with a stem diameter at breast height (diameter at 1.3 m above the ground) of DBH > 7 cm [60], saplings (core diameter < 7 cm and >3 cm; [10]) and seedlings (core diameter < 3 cm) and a height of H < 1 m [85,86]. For all adult trees, we measured the DBH, total tree height (H), crown width (CW) and crown height (CH). Calipers (65 cm) were used for measuring the diameter at breast height (DBH) for the small trees and diameter tapes (5 m) for the larger ones. We used a Suunto clinometer for measuring H and a Spiegel Relaskope for the measurements of CH as recommended by [86][87][88]. The CW was measured by using a tape measure (50 m) in eight directions from the main tree stem (every 45 degree) to the vertically projected edge of the crown [88]. All assessed tree species in this study were identified by a botanist and the literature [28,89].

Data Analysis
For each sample plot, we calculated the means of the tree size parameters (diameter at breast height, height, and crown width) as recommended by [1,12]. We also computed tree basal area (m 2 ), tree density (number of trees ha −1 ), and importance value index (IVI) [1,12,60,61], ( Table 2). This index results from the integration of relative frequency, relative abundance, and relative dominance, which are all expressed in percentages [1]. The values of IVI vary from 0 to 300, with 0 indicating no importance while values close to 300 show high importance of the species within the tree species community [12,90]. Table 2. Utilized equations for the calculation of dendrometric and structural parameters of the tree species collected in the Dinder Biosphere Reserve, Sudan, over a period of one year in sites of different anthropogenic pressure. DBH i is the diameter at breast height of the ith tree (cm), n represents the total number of trees in the sample, and s is the area sampled. σ 2 N and µ N represent the variance and mean, respectively, of the B. aegyptiaca tree density of the site. Density is the number of stems per sampled area. G B is the basal area of B. aegyptiaca trees while G is the basal area for all tree species encountered per site (stand basal area).

Parameter Equation Reference
Tree basal area (g) g = (π ÷ 4) × DBH i Species abundance (A) A = Total number of trees for a species ÷ Area sampled [61] Relative abundance (RA) RA = (Abundance of a species ÷ Total abundance for all) × 100 [12] Species dominance (D) D = Total basal area of a species ÷ area sampled [61] Relative dominance (RD) RD = (Coverage of a species ÷ Total coverage for all) × 100 [12] Species frequency (F) F = Occurrence or absence of species in a sample plot [60] Relative frequency (RF) RF = (Frequency of a species ÷ Total frequency for all) × 100 [61] Importance Value Index IVI = RA + RD + RF [12] Blackman Index (IB) The species richness (S) was calculated as the number of species recorded in each site [1]. Moreover, to measure the stand diversity, Simpson's reciprocal Index of Diversity (I D ) and Simpson's evenness (Eq) were calculated by applying the following equations [1,12]: Simpson s evenness (Eq) = n i is the number of trees of species i, n is the overall number of trees cruised in the plot, and sp is the number of species found in the considered plot. I max represents the maximum value of the Simpson's diversity index and S is the species richness. The mean and coefficient of variation for all calculated parameters were computed for all sample plots and species by using Minitab 17. One-way ANOVA in JAMOVI (version 1.1.7) was performed to compare mean numbers of trees, saplings, and seedlings in the non-disturbed and disturbed sites. A paired sampled T-test was accomplished using JAMOVI to compare the studied dendrometric parameters between the sites. The same tests also applied for the comparison between diversity indices in the two studied sites as recommended by [1,12]. Further, we ran a hierarchical cluster analysis (HCA) in Past (version 3.6) to group the values of IVI, and the three resulted groups for both the disturbed and non-disturbed sites were projected in the principal component analysis (PCA).
Weibull distribution was used to distinguish between size class distributions of B. aegyptiaca in the disturbed and non-disturbed sites [1,12] in Minitab 17.

Conclusions
Our study findings elucidated that the disturbed sites in the Dinder Biosphere Reserve have low species richness, diversity and natural regeneration compared to the nondisturbed sites. Moreover, vulnerable species such as B. aegyptiaca are severely impacted by anthropogenic disturbances, and their natural regeneration might be at risk. The results also revealed that A. seyal was the most dominant tree species in the disturbed site, which may show its bush encroachment properties [55,56]. We also demonstrated that more than 85% of the tree species in the disturbed sites have poor natural regeneration, which may mainly result from the intensive livestock grazing and browsing. We showed that species such as Tamarindus indica, Adansonia digitata, Terminalia brownii, and Piliostigma reticulatum have the lowest importance value index, and even, some of them are found only in the mature stage. Such species need urgent protection and conservation measures to recover and restore naturally before they collapse and disappear.
Furthermore, we found a low tree diversity and natural regeneration, particularly in areas that were exposed to intensive goats browsing (see also Mohammed et al. 2021, in review). Goats generally browse a diversified range of tree species at different heights from close to ground level up to the height of 2.10 m [21], which makes goats the most harmful livestock to natural regeneration.
We claim that the current policy of managing the transition zone (disturbed site) of the Dinder Biosphere Reserve must be revised, and more restrictions should be endorsed to eliminate this intensive illegal harvesting. We further recommended a comprehensive monitoring program in the transition zone (disturbed site) to control the browsing of livestock within the reserve and promote the natural regeneration of B. aegyptiaca and other vulnerable tree species.