Potential of Soil Seed Bank and Ungulate-mediated Endozoochory in Old Field Restoration

Soil seed bank plays a key role in old field restoration because it provides information about plant species that may regenerate (Hopfensperger 2007). Most studies on old field restoration have concentrated on floristic similarities between soil seed bank and standing vegetation (Török et al. 2018). However, investigating the effects of proximity to natural/old field edge (habitat fragmentation border) on soil seed bank could yield better recovery predictions. Proximity to natural/old field edge may favour recruitment of a different plant community than that found in the old field interior. This due to favorable abiotic (e.g., soil moisture) and biotic (e.g., seed dispersal) microenvironments created near the edge. On the other hand, ungulate-mediated endozoochory has been shown to potentially facilitate ecological restoration in degraded systems, given that ungulates disperse diaspores across different landscapes (Baltzinger et al. 2019). However, ungulate-mediated endozoochory dispersal in degraded old fields is affected by several factors that include availability of nearby natural perches that attract dispersers, structural vegetation complexities, presence of food especially fruits, and feeding regimes (Baltzinger et al. 2019).

This study examined the role of soil seed bank and ungulate-mediated endozoochory on seedling emergency, diversity, and composition in old fields targeted for ecological restoration. The study predicts that: i) proximity to natural/old field edge, which has favourable biotic and abiotic microenvironments, facilitates increased soil seed bank emergence and diversity compared to areas far from the edge into the old field, ii) basic functional traits (feeding regime) facilitate old field seed dispersal by ungulates, meaning seedlings of trees and shrubs will dominate dung of browsers as compared to seedlings of grasses and forbs which are edible by most ungulates. The study is motivated by the need to find an effective old field restoration initiative at Lapalala Wilderness to restore bushveld ecosystem structure and function.

The soil seed bank and ungulate dung were collected in old fields located at Lapalala Wilderness in Limpopo province, South Africa. Vegetation in the reserve is classified as Waterberg Mountain Bushveld, and soils are predominantly sandy (Mucina and Rutherford 2006). Most rain falls in summer (September to April), and temperatures are mild in winter and hot in summer. The old fields where the study was conducted were previously used for tobacco farming 35 years ago but are currently used for wild animal grazing. Three old fields, that share a farm road boundary with adjacent natural areas, were identified in the reserve. The farm road that separated the areas acted as the edge (fragmentation border) for both the natural area and old field. At each of the above-mentioned three sites, five line transects which extended 50 m from the edge into both the natural area (referred to as negative direction [—NA]) and old field area (referred to as positive direction [+OF]), with the edge being zero, were set up. The total number of transects were 30 (five transects × six sites, i.e. three in natural and three in old field area). At each transect, plots measuring 10 × 10 m were set up at distances of 10, 30, and 50 m from the edge into the old fields and at 10 and 50 m from the edge into the natural areas. The expectation was that there would be no marked soil seed bank variation in natural areas to warrant plots at 30 m distance. Within each plot, four soil cores (10-cm diameter × 10-cm depth) were collected from each corner. The four soils were mixed together to form a sample per plot, before being sieved using a 2-mm diameter sieve to remove debris and placed into plastic germination trays measuring 20-cm wide × 20-cm long × 10-cm deep. In total, 75 soil samples were collected and transported to a passively ventilated greenhouse located at the University of Venda in Limpopo province, South Africa.

Dung of five native ungulates (Ceratotherium simum— white rhinoceros, Aepyceros melampus—impala, Equus quagga—zebra, Connochaetes taurinus—wildebeest, and Tragelaphus strepsiceros—kudu) that frequently visit both the old field and natural areas were collected at each site. Ungulate dung was collected through randomly walking in each old field site. No dung was collected in the natural areas because the animals move freely between the two areas. Ungulate dung was collected in buckets every afternoon for five days, mixed together (per species), and left to air dry for two weeks. After drying, the dung was gently crushed, weighed to 2 kg per species, and spread into germinations trays with similar above-mentioned dimensions (no soils were added). A total of 75 trays were filled and placed in the greenhouse.

In the greenhouse, trays with soil and ungulate dung samples were arranged based on the three sites and further based on field type (natural and old field areas) for soil seed bank trays, and animal species for dung trays. All trays were rotated monthly to account for temperature and light variations. All trays were watered for nine months from April to December 2017. Emerging seedlings were counted and identified monthly. Identified species were assigned to four growth forms (trees, shrubs, forbs, and grasses) based on morphology. For the soil seed bank experiment, the effects of distance from the edge on seedlings density (the number of individual seedlings per square meter), species richness and Shannon-Wiener diversity index (H') were compared using ANOVA. For the ungulate-mediated endozoochory experiment the effects of different ungulates on seedlings density (the number of individual seedlings per kg), species richness and diversity index (H') were compared using ANOVA. Proof of normality and homogeneity of variance were tested using the Shapiro-Wilk's and Levene's test, respectively and data were normally distributed. All ANOVA analysis were done using STATISTICA version 13.1. Correspondence Analysis (CA), using Canoco for Windows 5, was used to investigate how distance from edge affects species assemblage.

A total of 233 seedlings from 34 plant species emerged from the soil seed bank experiment, of which 14 were trees and shrubs, 10 were forbs and 10 were grasses (Supplementary Material). Both seedling density and diversity (H') were significantly (p < 0.001) higher at 10 and 50 m from the edge into the natural area compared to all measured distances from the edge into the old field (Table 1). Seedling richness was significantly (p < 0.001) higher 50 m into the old field area compared to 30 and 10 m into the old field from the edge (Table 1). Seedling richness for trees and shrubs significantly (p < 0.001) increased with distance from old field to natural area (Table 1). In contrast, seedling richness for grasses significantly (p < 0.001) decreased with distance from old field to natural area. Seedling richness for forbs showed no significant (p > 0.05) differences at different distances from the edge (Table 1).

Correspondence analysis (CA) showed that most of the 34 plant species which germinated from the soil samples were found in both the natural field area and 10 m into the old field area. A less diverse suite of species germinated from the 30 and 50 m into the old field area (Figure 1). The first two CA axes showed high eigenvalues of 0.33 and 0.11, respectively and accounted for 66% of the variance. Clear distinctions can be seen regarding assemblages of some individual species in relation to distance from the edge, e.g., tree species of Senegalia caffra, S. senegal, Vachellia tortilis, and Grewia spp. assembled more in the natural area than grass species of Eragrostis lehmanniana and Enneapogon cenchroides which assembled more in old field far from the edge (Figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Correspondence analysis (CA) ordination of thirty-four plant (■) species that emerged from soil collected at various distance from the natural/old field edge. OF means Old Fields (positive transect direction from edge) and NA means Natural Areas (negative transect direction from edge). Full plant names are in Supplementary material.

Only 14 plant species emerged from ungulate dung collected from the five different species, of these, three were tree and shrub, five were forbs and six were grasses (Supplementary material). The three trees and shrubs of V nilotica, V. karroo, and L. javanica were present only in the dung of A. melampus, C. taurinus, and T. strepsiceros, whilst grass species of Heteropogon contortus, Aristida congesta, Cynodon dactylon, Panicum maximum, and E. lehmanniana were present in dung from all ungulates. Seedling density showed no significant (p > 0.05) differences among the ungulates (Table 2). Seedling richness and diversity (H') were significantly (p < 0.01) higher in A. melampus dung compared to E. quagga and C. taurinus dung, but there were no significant (p > 0.05) difference among A. melampus, C. simun, and T. strepsiceros dung (Table 2). Higher richness for trees and shrubs were recorded in A. melampus and T. strepsiceros dung compared to C. taurinus, C. simum, and E. quagga (Table 2). Seedling richness for forbs showed no significant (p > 0.05) differences among the ungulates (Table 2). Higher grass diversity was recorded in C. simum dung compared to C. taurinus and T. strepsiceros dung, but there were no significant (p > 0.05) differences between A. melampus and E. quagga dung (Table 2).

Results show that old field areas that are close to natural area border (edge proximity) have high soil seed banks, though increased seedling diversity near the edge was more visible for trees and shrubs than for forbs and grasses. This result concurs with the prediction that proximity to natural/old field edge facilitates increased soil seed bank emergence and diversity and can contribute to vegetation recovery in areas close to the edge. The abundance of soil seed bank close to the edge could result from favourable microenvironment conditions, especially increased seed dispersal rates near the edge (Cubina and Aide 2001). Both seed dispersal and pollination are higher in old field areas close to natural vegetation edge than in areas far from the edge. Also, the presence of more standing vegetation close to the edge could explain the recorded increased soil seed bank near the edge. The dominance of forbs and grasses in areas far from the edge into the old fields could point to a limited role that soil seed bank is likely to play on old field vegetation recovery in such areas. The persistence of forb and grass species in old fields has been shown to slow down restoration initiatives given that forbs and grasses outcompete woody species for resources, preventing their germination and growth (Blumenthal et al. 2003).

Results on ungulate-mediated endozoochory experiment showed the predominance of grasses and forbs in the ungulate dung of the five species, whereas trees and shrubs were present in dung of most browsers, (A. melampus, C. taurinus, and T. strepsiceros). This result concurs with some previous studies that reported the prevalence of graminoids and annuals in the dung of most herbivores (Shiponeni and Milton 2006). This trend also agrees with the prediction that basic functional traits (feeding regime) favour ungulate-mediated dispersal of different growth form. Seed dispersal via animal ingestion is known to be effective for seeds that survive in the digestive tracts of ungulates (Baltzinger et al. 2019).

Although results for this study are for one year (inter-annual studies may yield varying results), the study concludes that: i) soil seed bank can potentially aid old field vegetation recovery in areas that are close to the edge, ii) soil seed bank is likely to offer limited contribution towards old field vegetation recovery in areas that are far from the edge, and iii) ungulate dung plays a limited seed dispersal role of woody species in old fields. Restoration implications from this study include, i) fencing old field areas close to the edge to facilitate vegetation recovery through grazing exclusion, ii) improving soil seed bank in areas far from the edge through seeding, and iii) promoting seed dispersal in old fields through artificial perching.