Effect of tree species on understory vegetation, herbaceous biomass and soil nutrients in a semi-arid savanna of Ethiopia

Journal of Arid Environments 139 (2017) 76e84

Contents lists available at ScienceDirect

Journal of Arid Environments journal homepage: www.elsevier.com/locate/jaridenv

Effect of tree species on understory vegetation, herbaceous biomass and soil nutrients in a semi-arid savanna of Ethiopia Zewdu K. Tessema a, *, Ejigu F. Belay b a

Rangeland Ecology and Biodiversity Program, School of Animal and Range Sciences, College of Agriculture and Environmental Sciences, Haramaya University, PO Box 138, Dire Dawa, Ethiopia b Department of Animal Sciences, Faculty of Agriculture, Wollega University, PO Box 38, Shambu Campus, Ethiopia

a r t i c l e i n f o

a b s t r a c t

Article history: Received 23 January 2016 Received in revised form 29 November 2016 Accepted 21 December 2016

The effect of tree species on understory vegetation, herbaceous biomass and soil nutrients were studied in a semi-arid savanna of Ethiopia. Twenty large trees, from each of the species: Acacia robusta, Ziziphus spina-christi, and Balanites aegyptiaca, a total of 60 trees, and 480 samples were used for measuring understory vegetation, herbaceous biomass, and soil nutrients during the study. The inside tree canopies had a higher speciesdiversityand plant abundance than the outside tree canopies. Acacia robusta had a higher number of species and plant abundance in the understory vegetation compared to other tree species. The biomass yield of herbaceous vegetation under the inside canopies of A. robustawas higher than the canopies of other tree species. Similarly, most soil nutrient contents were higher under A. robusta than other tree species, and the inside canopies had a higher soil nutrient contents compared to outside tree canopies. Hence, the presence of larger trees in semi-arid African savannas confirmed to maintain more species composition and diversity of understory vegetation, higher herbaceous biomass and improved soil nutrients. Therefore, conservation of larger tree species is crucial for proper utilization and ecological stability of semi-arid African savannas under the changing climate and global warming. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Canopy cover Functional groups Plant abundance Species composition Species diversity Soil fertility

1. Introduction Savannas are found in tropical or subtropical ecosystems (Scholes and Archer, 1997), occupying about 50% of the southern continents (Sankaran et al., 2004). It is highly dynamic over temporal and spatial scales, and varies with changes in rainfall, soil nutrients, fire and herbivory (Rietkerk and van de Koppel, 1997), characterized by the coexistence of scattered trees and continuous grass layers (Scholes and Archer, 1997; Sankaran et al., 2004). These scattered trees alter the composition and spatial distribution of herbaceous vegetation in a semi-arid savanna. The spatial pattern and abundance of grass species in semi-arid savannas are dictated by a complex and dynamic interactions between trees and grasses (Scholes and Archer, 1997). Various plant communities consisting of mixtures of grasses, annual forbs and seedlings and saplings grow together under the canopy of larger trees in semi-arid savannas. These plant communities play an important role in primary production, and as habitats

* Corresponding author. E-mail address: [email protected] (Z.K. Tessema). http://dx.doi.org/10.1016/j.jaridenv.2016.12.007 0140-1963/© 2017 Elsevier Ltd. All rights reserved.

for insects, birds, mammals and other species (Legare et al., 2001). Previous studies in semi-arid savannas showed that grass composition is higher under trees compared to open areas (Weltzin and Coughenour, 1990), as a result of increased soil nutrients and shade under tree canopies (Belsky, 1994; Ludwig et al., 2001). Increased soil moisture availability due to hydraulic lift (Ludwig et al., 2001) could also potentially increase grass composition and productivity under tree canopies. Moreover, the accumulation of soil nutrients under tree canopies is often higher under leguminous trees than non-leguminous trees (Belsky et al., 1993). In semi-arid African savannas, where there is an extreme variation in water and nutrients, larger trees usually modify the micro-climate and soil properties, leading to complex local interactions between vegetation and soils under their canopies (Mitchell et al., 2012). These trees also create micro-sites, which exert influences on plant communities grown under their canopies, differently from open areas (Jeltsch et al., 1996). However, larger trees in semi-arid eastern and southern African savannas are being cleared for charcoal, firewood and timber production (Caro et al., 2005; Tessema et al., 2011), despite their importance in maintaining biological diversity and ecological

Z.K. Tessema, E.F. Belay / Journal of Arid Environments 139 (2017) 76e84

stability of the system. Consequently, this causes a reduction in understory herbaceous vegetation composition, herbaceous biomass, and depletion of soil quality (Mekuria et al., 2007; Tessema et al., 2011). However, the effects of larger trees on understory vegetation vary with the environment (Burrows, 1993), as different herbaceous species respond differently to different tree species (Jeltsch et al., 1996; Kahi et al., 2009). In addition, the influence of larger trees on plant communities underneath is reported to be site and plant species-specific in semi-arid savannas (Kahi et al., 2009; Mitchell et al., 2012). However, information is lacking about the impact of larger trees on the composition of herbaceous vegetation growing under tree canopies compared to the outside canopies in semi-arid Ethiopian savannas. Moreover, results from one area with specific tree species cannot be extrapolated to other areas with different tree and herbaceous plant species (Kahi et al., 2009). Indeed, information about the composition of understory vegetation, herbaceous biomass and soil nutrients in relation to larger tree species under their canopies and open areas are lacking in semi-arid savannas of Ethiopia. Therefore, we studied the effect of leguminous and non-leguminous tree species on the composition of understory vegetation, herbaceous biomass and soil nutrients in an experimental setup, and tested the following hypotheses: (i) inside tree canopies amplify increased understory vegetation composition, herbaceous biomass and soil nutrients than the surrounding open areas, and (ii) the composition of understory vegetation, herbaceous biomass and soil nutrients are higher under leguminous tree species than non-leguminous species in a semi-arid savanna of Ethiopia. 2. Materials and methods 2.1. Description of the study area The study was conducted at Babile Elephant Sanctuary (BES: 08
220 3000 - 09
000 3000 N; 42
0101000 - 43
050 5000 E; 850e1785 m asl), located in the eastern lowlands of Ethiopia (Fig. 1). The BES was selected because there are larger tree species found in isolation and it was possible to contrast tree canopies between the inside canopies versus the outside open areas for understory composition, herbaceous biomass and soil nutrients. The BES was established in 1970, covers 6982 km2 and is located 560 km southeast of Addis Ababa, in a semi-arid trans-boundary between Oromia and Somali regions of Ethiopia (Demeke, 2008). The mean annual rainfall was 702.9 mm, ranging between 452 and 1116.9 mm, and was highly variable among years. Its main rainy season is from JulyeSeptember, with a short rainy season from MarcheApril. The mean daily minimum and maximum temperatures are 11.9
C and 27.2
C, respectively, with a mean daily temperature of 19.6
C (Demeke, 2008). The BES was established to protect the only surviving African elephant (Loxodonta africana orleansi) population in the Horn of Africa (Barnes et al., 1999). The area is also known for its diverse groups of wild animals, which include Crested porcupine (Hystrix cristata), Abyssinian hare (Lepus habessinicus), Grivet monkey (Cercopithecus aethiops), Lesser galago (Galago senegalensis), Blackbacked jackal (Canis mesomelas), White-tailed mongoose (Ichneumia albicauda), Dwarf mongoose (Helogale parvula), Spotted hyena (Crocuta crocuta), Large-spotted genet (Genetta macullata), Caracal (Lynx caracal), Rock hyrax (Procavia capensis), Warthogs (Phacochoerus africanus and P. aethiopicus), Lesser kudus (Tragelaphus imberbis) and Greater kudus (T. strepsiceros), Bush duiker (Sylvica pragrimmia), Phillip's dik-dik (Madaqua saltiana) and Guenther's dik-dik (Rhyncho tragus guentheri) (Demeke, 2008). The vegetation of BES was represented by Acacia e Commiphora woodland, semidesert scrubland and evergreen scrub ecosystems, dominated

77

with A. robusta Burch., Tamirandus indica L., Oncoba spinosa Forsk, A. tortilis, Balanites aegyptiaca, and Ziziphus spina-christi. Lantana camara, Grewia shweinfurtii and Glycine spp. are the dominant shrub species in addition to other herbaceous vegetation available in BES (Belayneh et al., 2011). 2.2. Selection of sampling sites Based on visual field observation and previous vegetation studies (Belayneh et al., 2011), three dominant tree species, representing one leguminous (Acacia robusta Burch) and two nonleguminous tree species (Ziziphus spina-christi and Balanites aegyptiaca (L.) Del), found in isolation, were selected for this study. The species used in this study are representative of the dominant trees in the region (Belayneh et al., 2011; Biru and Bekele, 2012). Based on their abundance (distribution), canopy sizes, basal areas and tree heights, compared to other shrubby woody species, they represent suitable species for a systematic study of the effects of tree canopy cover on understory vegetation and soil nutrient dynamics. Accordingly, 20 matured trees, from each species, were systematically selected based on their similar canopy size (z25 m2 diameter) and tree height (z8 m) according to previous studies (Ludwig et al., 2004; Kahi et al., 2009), without shrubs or termite mounds under or close to their canopies. Moreover, exclosures were erected around all the experimental trees and adjacent open areas before the start of the main rain, beginning of June 2012, to keep off wild and domestic herbivores. In total, 60 trees (3 tree species  20 trees/species) were selected for this study. The height of each individual tree from each tree species was estimated by walking away from the tree bending forward and look through two legs back to the tree by stopping as reached at a point where it is possible to see the top of the tree (at 45
) and measuring the distance along the ground to the tree. It is assumed that this distance is equivalent to the height of the tree (Savadogo and Elfving, 2007). The canopy diameter (CD) of trees was measured by using measuring tape on ground level throughout the canopy length in two dimensions, at right angle to each other. According to Savadogo and Elfving (2007), the vertical projected canopy area of each tree species was calculated using the formula: CA ¼ (CD1 x CD2) x p/4, where CD1 and CD2 are the two canopy diameter in two dimensions at right angle to each other. 2.3. Sampling of understory vegetation Under each selected individual tree, the species composition of understory vegetation, including, herbaceous species, as well as seedlings and saplings of woody species, were assessed, recorded and identified using 1-m2 quadrat under inside and outside canopies of individual trees (Fig. 2) in September 2012, during the flowering stage of most herbaceous species. Four quadrats in four directions (north, south, east and west) were used under the inside and outside canopy of each individual tree, totaling 480 samples (3 tree species x 20 trees/species x 2 canopy cover x 4 directions as sample quadrats). For those species that were difficult to identify in the field, their local names recorded, herbarium specimens collected, pressed and dried properly and transported to Haramaya University Herbarium, for further identification. The species were further classified in to grasses (annuals and perennials), herbaceous legumes, forbs and woody species to determine the contribution of each functional group. Individual plants were counted in each quadrat to determine the relative abundance of each species. The herbaceous biomass (in DM basis) was determined by harvesting the whole fresh biomass within each quadrat, and oven-drying at 70
C for 48 h and weighing. Plant nomenclature follows Cufodontis (1953e1972), Fromann and Persson (1974), and Philips (1995).

78

Z.K. Tessema, E.F. Belay / Journal of Arid Environments 139 (2017) 76e84

Fig. 1. Location of the study area, Babile Elephant Sanctuary (BES: 08
220 3000 - 09
220 000 3000 N; 42
220 0101000 - 43
220 050 5000 E), eastern Ethiopia.

Fig. 2. Sampling design for data collection of the composition of understory vegetation, herbaceous biomass and soil nutrients under the inside and outside tree canopy at Babile Elephant Sanctuary, eastern Ethiopia. Understory vegetation, herbaceous biomass and soil nutrient contents were recorded on 1  1 m2 quadrats, and sampling plots were laid out in four directions (north, south, east and west) within the canopy and outside the canopy of each tree.

Z.K. Tessema, E.F. Belay / Journal of Arid Environments 139 (2017) 76e84

2.4. Soil sampling and analyses Four soil samples in 1-m2 were collected in four directions (north, south, east and west) at depth of 0e10 cm, using a soil auger, under the inside and outside canopies in each tree (Fig. 2), yielding a total of 480 samples (3 tree species x 20 trees/species x 2 canopy cover x 4 directions as sample quadrats) for the soil nutrient analyses. The soil samples from the same tree species under each canopy cover were pooled and mixed together to form a composite soil sample. Finally, each of the 24 composite soil samples (3 tree species x 2 canopy covers x 4 directions as quadrat samples) was divided into three equal parts, out of which one was randomly chosen and stored in plastic bags, labelled, sealed and transported to the soil laboratory of Haramaya University in Ethiopia for chemical analyses. The soil samples were oven-dried at 70
C for 24 h for analysis using the following techniques: pH was determined in a 1:2.5 soil water ratio suspensions using the Bouyoucos hydrometer method (Bouyoucos, 1962), while electrical conductivity (EC) was determined using the sodium saturation ratio (Van Reeuwijk, 1992). The percentage organic carbon (OC) was determined according to the Walkley and Black (1934) method, and total N using the Kjeldahl procedure (Jackson, 1970). Available phosphorus (P), exchangeable potassium (K), calcium (Ca), magnesium (Mg) and sodium (Na) were analyzed according to Olsen et al. (1954). Cation exchangeable capacity (CEC) was analyzed using the method of NRC (1996).

79

canopy under each tree species in terms of species composition of the understory vegetation (Fig. 3). The first and second ordination axis explained cumulatively 87% of the total variance extracted by the PCA. The inside tree canopies were dominated by annual forbs compared with the outside tree canopies in the present study. The proportion of annual grasses, total grasses and annual forbs were higher under Z. spina-christi tree species than other tree species (Table 1). However, the interaction of tree species by canopy cover had no significant (P > 0.05) effects on all functional groups of the understory vegetation in the present study. The inside canopy had a higher number of species compared with the outside canopy, with a mean of 8.4 and 7.7 species/m2, respectively. Moreover, a higher Shannon-Wiener species diversity index (F1, 474 ¼ 6.77, P < 0.001) was recorded under the inside tree canopies than outside tree canopies (Table 2). Acacia robusta had significantly a higher number of species (F2, 474 ¼ 15.27, P < 0.0001) compared with B. aegyptiaca and Z. spina-christi. Similarly, a higher Shannon-Wiener species diversity (F2, 474 ¼ 20.03, P < 0.0001) was recorded under A. robusta than other tree species (Table 2). The inside canopy had a higher (F1, 474 ¼ 67.95, P < 0.0001) herbaceous plant abundance than the outside canopy, with a mean of 102.4 and 72.2 plants/m2, respectively. The leguminous tree species, A. robusta had higher plant abundance (F2, 474 ¼ 13.32, P < 0.0001) than the non-leguminous tree species (Table 2). However, the interaction effect of tree species by canopy cover had no significant (P > 0.05) effect on the number of species, plant abundance and Shannon-Wiener species diversity index in the present study.

2.5. Data analyses 3.2. Herbaceous biomass The number of species (species richness), species composition and functional groups (grasses, herbaceous legumes, forbs and woody species) and plant abundance (number of individual plants in each species) were recorded in the understory vegetation. Species diversity was calculated using Shannon-Wiener diversity index (H’), and the similarity in species composition of the understory vegetation under the inside and outside canopies of ach tree species was determined by Jaccard Coefficient of Similarity (Magurran, 2004). Moreover, ordination of sampling sites under the inside and outside canopies in each tree species was also carried out by multivariate technique (Canoco 4.5, Ter Braak, 1997) to compare the similarity in species composition between understory vegetation, using Principal Component Analysis (PCA). First, we confirmed the length of gradient on the 1st ordination axis is whether linear (<3) or unimodal (>4) by Detrended Correspondence Analysis (DCA) using the abundance data of herbaceous vegetation under the inside and outside canopies of each tree species before running a PCA analysis. To test for differences in all data recorded, a General Linear Model (GLM) was applied, with tree species, canopy cover and their interactions, as independent factors. Data were analyzed using SAS Software (SAS Inc., 2009), and results are presented as mean ± 95% C.I. Tukey - HSD test was employed to investigate significant differences between means at P  0.05. Dependent proportional data were arcsine transformed to meet the assumptions of normality and homogeneity of variance. 3. Results 3.1. Understory vegetation Out of the 87 species identified in the understory vegetation, the number of grass species, annual forbs and woody species (including seedlings and saplings) were 32, 46 and 9, respectively. In total, 20 annual and 12 perennial grass species were recorded (Appendix A). The ordination result showed a clear separation of sampling sites, as the inside canopy are separately clustered from the outside

There were canopy specific differences on biomass yield of herbaceous vegetation (F1, 474 ¼ 37.13; P < 0.0001, Table 3), as the inside tree canopy had a higher biomass yield than the outside tree canopies. Tree species also had a highly significant (F2, 474 ¼ 14.73, P < 0.0001, Table 3) effect on biomass yield of herbaceous vegetation in the present study. Accordingly, A. robusta had the highest biomass yield of herbaceous vegetation compared to Z. spina-christi and B. aegyptiaca (Fig. 4). Moreover, the inside canopy of A. robusta was higher in biomass yield of herbaceous vegetation than the canopies of other tree species (interaction term of tree species x canopy cover). 3.3. Soil nutrients The inside tree canopies of A. robusta and Z. spina-christi had a higher K content of the soil than B. Aegyptiaca (Table 4). In a similar fashion, the inside canopy of A. robustahad a significantly higher CEC contents compared with other tree species. The EC content of the soil was higher under both the inside and outside canopies of A. robusta than Z. spina-christi and B. aegyptiaca. Moreover, the OC contents were higher under the inside tree canopies of A. robusta followed by the inside canopies of Z. spinaechristi tree species, with a mean values of 4.9% and 3.6%, respectively. The P contents were higher under the canopies of A. robusta compared with other tree species. However, the pH of the soil was not affected by the canopies of tree species, as it was slightly, but not significantly, lower under the outside tree canopies than the inside tree canopies across all tree species. In addition, tree canopies did not affect the Na content of the soil. Tree species significantly affected the pH of the soil, as the highest soil pH was obtained under B. aegyptiaca (>8), indicating that the soils under B. Aegyptiaca were more alkaline. Moreover, A. robusta had significantly higher Na and P contents than those of Z. spina-christi and B. Aegyptiaca. However, neither tree species nor their canopy cover had significant effects on most nutrients of the soil in the present study (Table 4).

80

Z.K. Tessema, E.F. Belay / Journal of Arid Environments 139 (2017) 76e84

Fig. 3. Ordination diagram of the inside tree canopies and their corresponding outside tree canopies of three selected tree species using Principal Component Analysis (PCA). ARIC ¼ Acacia robusta Inside Canopy; AROC ¼ Acacia robusta Outside Canopy; BAIC ¼ Balanites aegyptiaca Inside Canopy; BAOC ¼ Balanites aegyptiaca Outside Canopy; ZSIC ¼ Ziziphusspina-Christi Inside Canopy; ZSOC ¼ Ziziphus spina-Christi Outside Canopy.

Table 1 Effect of tree species, canopy cover and their interactions on functional groups of understory vegetation, with statistical results of the GLM (F, P, R2adjusted). Functional groups (Means ± 95% C.I)

Acacia robusta Inside canopy Outside canopy Balanites aegyptiaca Inside canopy Outside canopy Ziziphus spina-Christi Inside canopy Outside canopy Tree species (TS) F (df ¼ 2, 474) P Canopy Cover (CC) F (df ¼ 1, 474) P TS  CC F (df ¼ 2, 474) P Adjusted R2

Annual grass

Perennial grass

Total grass

Herbaceous Forbs

Woody species

1.95 ± 0.20 1.95 ± 0.15

1.58 ± 0.12 1.35 ± 0.11

3.25 ± 0.24 3.25 ± 0.2

3.53 ± 0.24 2.65 ± 0.22

0.75 ± 0.11 0.70 ± 0.14

2.03 ± 0.22 2.45 ± 0.23

1.45 ± 0.14 1.45 ± 0.12

3.48 ± 0.22 3.90 ± 0.24

3.20 ± 0.3 3.23 ± 0.3

1.58 ± 0.1 1.68 ± 0.2

2.73 ± 0.20 2.63 ± 0.20

1.30 ± 0.13 1.30 ± 0.14

4.30 ± 0.24 3.98 ± 0.2

3.93 ± 0.3 3.33 ± 0.3

0.78 ± 0.1 0.40 ± 0.1

6.97 0.001

1.01 0.366

8.30 0.000

2.39 0.094

37.86 0.000

0.46 0.498

0.52 0.471

0.04 0.8515 0.852

5.29 0.022

1.05 0.307

1.02 0.364 0.656

0.52 0.594 0.511

1.49 0.227 0.774

1.61 0.203 0.537

1.76 0.175 0.525

4. Discussion 4.1. Effect of tree species on understory vegetation We found more species composition and a higher species diversity and plant abundance under the inside tree canopies than outside canopies in the present study, which might be due to the contribution of increasing soil nutrients, reducing solar radiation and soil temperature under the inside canopies than the outside tree canopies. According to Abule et al. (2005) different species composition under tree canopies was recorded compared with open areas in the rift valley of Ethiopia due to low soil fertility as a result of pronounced grazing impacts under the outside tree canopies. Ludwig et al. (2004) also indicated that older and larger trees found to enhance soil nutrients under tree canopies than young trees in semi-arid African savannas. Scholes (1990) mentioned that it seems economical to clear single e standing large trees, but it is doubtful whether the shrubs that persist under overgrazed semiarid African savanna conditions would facilitate the same positive

influences as larger trees on the composition of understory vegetation (Tobler et al., 2003). On the contrary, Ludwig et al. (2004) suggested that competition for moisture and soil nutrients between trees and herbaceous species, particularly, grasses in semiarid African savannas, may lead to lower herbaceous swards under tree canopies as opposed to the outside canopies. Our study, however, confirmed that the facilitative effects of trees on the composition of understory vegetation out-weighed the competition effects of trees and herbaceous species, which might be the presence of scattered trees according to field observations. In addition, the availability of more species composition, species diversity and plant abundance in our study under the inside tree canopies than the outside canopies might be also explained due to litter fall and their root system, as well as due to the deposition of faeces by birds and browsing animals under the tree canopies that improves soil nutrient conditions (Barbier et al., 2008; Augusto et al., 2013). According to Ludwig et al. (2001, 2003) the hydraulic lift e effects together with shading can also lead to increased moisture availability in semi-arid African savannas, which might be

Z.K. Tessema, E.F. Belay / Journal of Arid Environments 139 (2017) 76e84

81

Table 2 Effect of tree species, canopy cover and their interactions on species richness, species diversity and relative plant abundance (n/m2) of understory vegetation, with statistical results of the GLM (F, P, R2adjusted). Means ± 95% CI

Acacia robusta Inside canopy Outside canopy Balanites aegyptiaca Inside canopy Outside canopy Ziziphus spina-Christi Inside canopy Outside canopy Tree Species (TS) F (df ¼ 2, 474) P Canopy Cover (CC) F (df ¼ 1, 474) P TS  CC F (df ¼ 2, 474) P Adjusted R2

Species richness (n/m2)

Shannon Wieners species diversity

Plant abundance (n/m2)

9.43 ± 0.32 7.70 ± 0.30

1.70 ± 0.05 1.59 ± 0.04

125.53 ± 7.4 63.08 ± 2.8

8.28 ± 0.3 8.83 ± 0.3

1.70 ± 0.05 1.58 ± 0.06

63.18 ± 3.0 84.70 ± 3.61

7.53 ± 0.23 6.65 ± 0.30

1.42 ± 0.03 1.30 ± 0.07

118.33 ± 5.5 68.85 ± 2.3

15.27 0.0001

20.03 0.0001

13.32 0.0001

5.47 0.0202

6.77 0.010

67.95 0.0001

5.78 0.004 0.690

0.04 0.096 0.670

50.97 0.0001 0.547

Table 3 Analysis of variance (ANOVA) for the effect of tree species, canopy cover and their interactions on biomass yield of herbaceous vegetation in Babile Elephant Sanctuary, eastern Ethiopia. Source

Degree of freedom

Type III sum of squares

Mean square

F value

Probability

Tree species (TS) Canopy cover (CC) TS x CC Error

2 1 2 474

231,968.07 296,054.54 113,285.11 3,731,663.17

115,984.23 296,054.54 56,642.55 7872.71

14.73 37.13 7.19

0.0001 0.0001 0.0008

Fig. 4. Biomass yield of herbaceous vegetation under the inside canopy and outside canopy of selected tree species at Babile Elephant Sanctuary, eastern Ethiopia.

essential for the growth of understory herbaceous vegetation. Thus, our result confirmed that the facilitative effects of trees would lead to more number of species and plant abundance under tree canopies. Similarly, older trees widely spaced in a medium density (Wiegand et al., 2006) showed an optimal effect on herbaceous layer in semi-arid savannas compared to more densely wooded savannas, where grasses are inferior competitors for light, moisture, and soil nutrients (Dohan et al., 2013; Moustakas et al., 2013).

Because herbaceous species grown outside tree canopies face higher solar radiation and temperatures that could lower moisture and nutrient availabilities in the soil than herbaceous species grown under the canopies of large trees (Augusto et al., 2013; Moustakas et al., 2013). As expected in our hypothesis, the composition of herbaceous vegetation is higher under leguminous tree species than non e leguminous tree species in the present study, since A. robusta had

82

Z.K. Tessema, E.F. Belay / Journal of Arid Environments 139 (2017) 76e84

Table 4 Effect of tree species, canopy cover and their interaction on soil nutrient contents with statistical results of the GLM (F, P, R2adjusted).

pH EC CEC TN P N:P Ca K Mg Na OM

CC

R2 adjusted

Acacia robusta

Ziziphus spina-christi

Balanites aegyptiaca

TS

TS x CC

IC

OC

IC

OC

IC

OC

F(df ¼ 2,18)

P

F(df ¼ 2,18)

P

F(df ¼ 2,18)

P

7.9 ± 0.20 0.4 ± 0.09 30.5 ± 1.30 2.1 ± 0.14 13.1 ± 1.05 0.16 ± 0.01 16.9 ± 3.80 2.8 ± 0.23 6.7 ± 0.68 0.5 ± 0.05 4.9 ± 0.28

7.5 ± 0.20 0.8 ± 0.25 24.3 ± 3.29 1.7 ± 0.02 10.5 ± 2.12 0.19 ± 0.05 15.8 ± 2.54 1.7 ± 0.28 7.9 ± 0.89 0.5 ± 0.03 3.1 ± 0.27

8.1 ± 0.05 0.2 ± 0.02 25.6 ± 1.60 2.1 ± 0.24 11.9 ± 1.34 0.18 ± 0.02 13.4 ± 1.76 3.02 ± 0.07 6.4 ± 0.72 0.3 ± 0.02 3.6 ± 0.38

7.8 ± 0.14 0.19 ± 0.03 21.2 ± 2.40 0.9 ± 1.46 6.4 ± 1.08 0.17 ± 0.05 17.1 ± 1.56 2.0 ± 0.12 7.5 ± 0.29 0.2 ± 0.04 2.5 ± 0.30

8.2 ± 0.07 0.13 ± 0.01 20.1 ± 1.95 1.3 ± 0.13 5.7 ± 1.05 0.23 ± 0.02 13.5 ± 1.58 1.8 ± 0.25 5.6 ± 0.60 0.3 ± 0.04 2.5 ± 0.37

8.1 ± 0.06 0.16 ± 0.01 17.1 ± 3.19 1.5 ± 0.30 5.5 ± 0.87 0.30 ± 0.08 13.4 ± 0.99 1.7 ± 0.25 5.8 ± 0.26 0.3 ± 0.02 2.8 ± 0.79

5.04 8.37 6.77 3.90 14.67 2.69 0.85 6.25 3.50 33.67 5.04

0.018 0.003 0.006 0.039 0.000 0.095 0.442 0.009 0.052 0.000 0.018

4.59 2.49 5.36 7.17 9.75 0.60 0.21 15.42 2.79 0.09 5.49

0.046 0.132 0.033 0.015 0.006 0.448 0.653 0.001 0.112 0.770 0.031

0.90 2.15 0.23 7.72 2.28 0.38 0.66 3.71 0.52 0.02 3.07

0.424 0.146 0.793 0.004 0.131 0.687 0.531 0.045 0.605 0.985 0.071

0.48 0.56 0.52 0.63 0.52 0.27 0.15 0.66 0.38 0.79 0.55

EC ¼ electrical conductivity (mmhos/Cm); CEC ¼ cation exchange capacity (Meq/100 gm soil; TN ¼ total nitrogen (%); Av. P ¼ Available phosphorous (mg/kg soil); N:P ¼ nitrogen phosphorous ratio; Ca ¼ calcium (Cmol (þ)/kg soil); Av. K ¼ Available potassium (Cmol (þ)/kg soil); Na ¼ sodium (Cmol (þ)/kg soil); Mg ¼ magnesium (Cmol (þ)/kg soil); OM ¼ organic matter (%).

had a higher number of species and plant abundance compared with Z. spina-christi and B. aegyptiaca. According to previous studies (e.g., Belsky et al., 1989; Power et al., 2003; Hart and Chen, 2006; Barbier et al., 2008) trees with N2 - fixing abilities improve soil N levels, and facilitate higher plant growth than non e N2 e fixing trees. Atmospheric N2-fixation by leguminous Acacia spp. enhance soil nutrient levels (Danso et al., 1992; Riegel and Miller, 1992) that would lead to the presence of more species diversity and higher growth of herbaceous vegetation. Similarly, Ludwig et al. (2004) indicated that the presence of scattered trees in semi-arid African savannas are referred to as ‘islands of fertility’ because elevated soil nutrients are found under inside tree canopies, together with decreased solar radiation, reduced evapotranspiration and soil temperatures, and these conditions might favor the growth of herbaceous vegetation beneath tree crowns compared to the open areas. In semi-arid African savannas, where soil fertility is very low, large trees found in isolation, can create patches of grasses and other herbaceous species under their canopies (Treydte et al., 2007). Hence, these understory vegetation can serve as feed resources for both wild and domestic undulates in and around the BES in Ethiopia. The contribution of leguminous trees to the understory herbaceous vegetation, however, could be affected compared to non e leguminous trees in semi-arid African savannas. For instance, leguminous trees may not fix N - efficiently during dry seasons under condition of high temperature, low available P and high browsing intensity (Hartwig, 1998), which are also typical features in semi-arid Ethiopian savanna. Under these circumstances, the leguminous trees may not have beneficial effect to the surrounding vegetation since N e fixing of Acacia spp. have a higher N and P requirement than non e leguminous trees in arid and semiarid savannas (Treydte et al., 2007). In addition, anthropogenic pressures and extraction of biomass have been causing the degradation of habitat (Biru and Bekele, 2012), leading to the deforestation of few tree species that are found in isolation in Babile Elephant Sanctuary in Ethiopia.

(Dohan et al., 2013; Moustakas et al., 2013). The higher productivity of herbaceous vegetation under the canopy of tree species can be further explained with a reduction in temperature and evapotranspiration (Gilliam, 2007; Augusto et al., 2013), as the productivity of herbaceous vegetation under the shades of trees is reported to increase in arid and semi-arid tropical rangelands (Hart and Chen, 2006; Dohan et al., 2013). Tree species had a significant effect on the biomass yield of herbaceous vegetation, as the highest biomass yield of herbaceous vegetation was obtained under A. robusta than Z. spina-christi and B. aegyptiaca in the present study. This might be related to a higher canopy size and basal areas of A. robusta (Belayneh et al., 2011) than Z. spina-christi and B. aegyptiaca as observed in the field. Furthermore, a higher concentration of total N, available P and OM contents observed under leguminous tree species in our study might show an improved fertility status of the soil, leading to a higher biomass production of herbaceous vegetation under leguminous tree species than non-leguminous tree species. Similarly, Cech et al. (2008) reported that biomass production of herbaceous vegetation is influenced due to higher soil N and P accumulation under leguminous trees. According to Kahi et al. (2009) the mean biomass yield of herbaceous vegetation under A. tortilis was twice the biomass yield of herbaceous vegetation under non e leguminous tree species, indicating that the N2 e fixation of leguminous trees improves the biomass production of understory vegetation in Babile Elephant Sanctuary of Ethiopia in particular and in semi-arid African savannas at large. This might be related to the addition of nutrients to the top soil layers by the N-fixing of leguminous trees, which can easily be accessible by shallow rooted understory vegetation, leading to promote their growth. Accordingly, trees in our study sites have confirmed to increase the biomass production of understory herbaceous vegetation under the inside canopies. Thus, proper management and conservation of these remnant mature trees found in isolation is very crucial in Babile Elephant Sanctuary in Ethiopia under the changing climate and global warming.

4.2. Effect of tree species on herbaceous biomass 4.3. Effect of tree species on soil nutrients The biomass yield of herbaceous vegetation was significantly affected by the canopy cover of tree species in our study, indicating that a higher amount of herbaceous biomass was recorded under tree canopies compared to the outside tree canopies. This might be related to improved soil fertility due to higher litter fall and animal excreta (faeces and urine) deposition under the inside tree canopies, as animals usually gathered and spent the time during hot and dry seasons under shade of larger trees in semi-arid savannas

The pH of soil was significantly influenced by the canopy cover of trees in the present study, as a higher pH value was observed under the inside canopies than the outside tree canopies, which might be might be due to accumulation of litter fall and the roots of understory vegetation. Similarly, Karim et al. (2009) and Kahi et al. (2009) reported a higher pH values of soils under the inside tree canopies than outside tree canopies in a semi-arid areas of Kenya.

Z.K. Tessema, E.F. Belay / Journal of Arid Environments 139 (2017) 76e84

This might be explained due to the lower OM contents and a higher CEC values of the soils away from the canopies of larger trees compared to the higher OM contents under the inside tree canopies due to accumulation of faeces, fallen leaves and litter falls. However, Dunham (1991) reported higher soil pH outside tree canopies than inside tree canopies, which was inconsistence to our findings. Higher percentage of total N was obtained under the inside canopies of trees compared to outside tree canopies in our study, which might be due to the availability of higher herbaceous vegetation under the inside tree canopies that can add more N and OM to the soils through soil microbial decay and decomposition (Augusto et al., 2013; Prescott and Grayston, 2013). The higher content of total N under tree canopies might be due to the seasonal litter fall and reduction of leaching compared to outside tree canopies (Kahi et al., 2009). Because the canopy cover of large trees had a significant effect on soil OM contents, as a higher OM values were recorded under the inside tree canopies compared to the outside tree canopies in the present study, which might be due to the availability of high moisture retention under the shade of trees that could facilitate the fast decomposition of animal dung, nests of birds and litter fall under the tree canopies. Similarly, Treydte et al. (2007) reported a higher OM contents of soils under the inside tree canopies in areas where there is low amount of rainfall, which might be due to enrichment of soils by leguminous woody species (Abule et al., 2007). The availability of a higher P and K contents of soils under the inside tree canopies than outside tree canopies might indicate the favorability of the micro-environment under the inside tree canopies compared with outside tree canopies in Babile Elephant Sanctuary of Ethiopia, a typical feature of semi-arid African savannas. Total N content of soil was affected by tree species, as a higher total N was observed under A. robusta compared to Z. spina-christi and B. aegyptiaca. This result was in line with Abule et al. (2007) who reported that leguminous trees contribute to a higher N content of soils in the rift valley of Ethiopia, as a result of N2 - fixing rhizobium bacteria association to the root nodules of leguminous trees compared to non e leguminous trees. In the present study, the higher soil OM contents under A. robusta might be due to the accumulation of higher litter fall and roosting birds due to the larger canopy size of A. robusta compared to B. aegyptiaca as observed in the field. Moreover, Ca and available K of the soils were higher under the leguminous tree species than non-leguminous tree species in our study. However, the N: P ratio of the soils was not affected by different tree species, indicating that N fixation by leguminous tree species in the present study did not lead to P limitation of the soils, since soils under leguminous tree species might be co-dominated with both N and P in tropical savannas (Cech et al., 2008). Therefore, the presence of remnant matured tree species found in isolation have contributed significantly to the species composition and biomass production of understory vegetation serving as major forages for both wild and domestic ungulates in semi-arid Ethiopian savannas. 5. Conclusions In the present study, a higher number of species and plant abundance were observed under the inside tree canopies than the outside canopies. Similarly, the biomass yield of understory vegetation was higher under the inside canopies of A. robusta than the other tree species. Most of the soil nutrient contents were higher under the inside canopies of A. robusta compared to outside tree canopies in the present study. In addition, A. robusta had more species diversity and plant abundance in the understory vegetation compared to other tree species. In semi-arid Ethiopian savannas, where rainfall is low with inter e seasonal variability, the presence

83

of larger trees are highly important in maintaining the composition and diversity, and increasing the herbaceous biomass of understory vegetation for both domestic and wild herbivores, as well as improving the soil conditions, under the changing climate and global warming. Hence, any planning on how many trees would be cleared in Babile Elephant Sanctuary in particular and in semi-arid African savannas in general should take into account the importance of under canopy areas for both grazing and resting of wild and domestic animals. Therefore, careful utilization and proper conservation of larger trees found in isolation are significantly needed to maintain understory species diversity and ecological stability of semi-arid African savannas. Acknowledgements The authors would like to acknowledge Haramaya University, Ethiopia, for providing transport during the field work and for carrying out the soil analyses. We are grateful to the management of Babile Elephant Sanctuary in Ethiopia for allowing us to conduct our research. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jaridenv.2016.12.007. References Abule, E., bro, Smit, G.N., Snyman, H.A., 2005. The influence of woody plants and livestock grazing on grass composition, yield and soil nutrients in the Middle Awash Valley of Ethiopia. J. Arid. Environ. 60, 343e358. Abule, E., Snyman, H.A., Smit, G.N., 2007. Rangeland evaluation in the middle Awash valley of Ethiopia: II. Woody vegetation. J. Arid. Environ 70, 272e292. Augusto, L., Dupouey, J.-L., Ranger, J., 2013. Effects of tree species on understory vegetation and environmental condition in temperate forest. Ann. For. Sci. 60 (8), 823e831. Barbier, S., Gosselin, F., Balandier, P., 2008. Influence of tree species on understory vegetation diversity and mechanisms involved-A critical review for temperate and boreal forests. For. Ecol. Manag. 254, 1e15. Barnes, R.F.W., Craig, G.C., Dublin, H.T., Overton, G., Simons, W., Thouless, C.R., 1999. African Elephant Database 1998. IUCN/SSC African Elephant Specialist Group, Gland, Switzerland/Cambridge, UK. Belayneh, A., Bekele, T., Demissew, S., 2011. The natural vegetation of Babile elephant sanctuary, eastern Ethiopia: implications for biodiversity conservation. Ethiop. J. Biol. Sci. 10 (2), 137e152. Belsky, A.J., 1994. Influences of trees on savanna productivity - tests of shade, nutrients and tree - grass competition. Ecol 75, 922e932. Belsky, A.J., Amundson, R.G., Duxburg, J.M., Riha, S.J., Ali, A.R., Mwonga, S.M., 1989. The effect of trees on their physical, chemical and biological environments in a semi-arid savanna in Kenya. J. Appl. Ecol. 26, 1005e1024. Belsky, A.J., Amundson, R.G., Duxbury, R.M., Riha, S.J., Ali, A.R., Mwonge, S.M., 1993. Comparative effects of isolated trees on their under canopy environments in high and low rainfall savannas. J. Appl. Ecol. 30, 143e155. Biru, Y., Bekele, D., 2012. Food habits of African elephant (Loxodonta africana) in Babile elephant sanctuary. Ethiop. Trop. Ecol. 58 (1), 43e52. Bouyoucos, G.J., 1962. Hydrometer method improved for making particle size analysis of soils. Am. Soc. Agro. J. 54, 4644. Burrows, W.H., 1993. Deforested in Savannah context: problems and benefits for pastoralism. Proc. XVII Int. Grassl. Congr. 2223e2230. Caro, T.M., Sungula, M., Schwartz, M.W., Bella, E.M., 2005. Recruitment of Pterocarpus angolensis in the wild. Forest.Ecol. Manag. 219, 169e175. Cech, P.G., Kuster, T., Edwards, P.J., Venterink, H.O., 2008. Effects of herbivory, fire and N2-fixation on nutrient limitation in a humid african savanna. Ecosyst 3, 991e1004. Cufodontis, G., 1953-1972. EnumeratioplantarumAethiopiac spermatophyte (sequentia). Bull. Jard. 1 Etat Brux. 30, 653e708. Danso, S.K.A., Bowen, G.D., Sanginga, N., 1992. Biological nitrogen fixation in trees in agro-ecosystems. Plant Soil 141, 177e196. Demeke, Y., 2008. The Ecology and Conservation of the Relic Elephant Population in the Horn of Africa. Australia, PhD. Thesis University of Melbourne, pp. 32e41. Dohan, J., Demb cl c, F., Karembe, M., Moustakas, A., Am cvor, K.A., Hana, N.P., 2013. Tree effects on grass growth in savannas: completion, facilitation and the stress-gradient hypothesis. J. Ecol. 101, 202e209. Dunham, K.M., 1991. Comparative effects of Acacia albida and Kigelia African trees on soil characteristics in Zambezi riverine woodlands. J. Trop. Ecol. 7, 215e220. Fromman, B., Pearson, S., 1974. An Illustrated Guide to the Grass of Ethiopia, Chilalo

84

Z.K. Tessema, E.F. Belay / Journal of Arid Environments 139 (2017) 76e84

Agricultural Development Unit, Asella, Ethiopia. Gilliam, F.S., 2007. The ecological significance of the herbaceous layers in temperate forest ecosystems. BioScience 57 (10), 845e858. Hart, S.A., Chen, H.Y.H., 2006. Understory vegetation dynamics of North American boreal forests. Crit. Rev. Plant Sci. 25, 381e397. Hartwig, U.A., 1998. The regulation of symbiotic N2 fixation: a conceptual model of N feedback from the ecosystem to the gene expression level. Persp. Plant Ecol. Evol. Syst. 1, 92e120. Jackson, M.L., 1970. Soil Chemical Analysis. Prenstice Hall, Inc, Engle Wood Cliffs.New Jersey. Jetsch, F., Mutton, S.J., Dean, W.R.J., Rooyen, N.V., 1996. Trees spacing and coexistence in semi-arid savannah. J. Ecol. 84, 583e595. Kahi, C.H., Ngugi, R.K., Mureithi, S.M., Ng’ethe, J.C., 2009. The canopy effects of Prosopis Juliflora (Dc.) and Acacia tortilis (Hayne) trees on herbaceous plants species and soil physic -chemical properties in Njemps Flats, Kenya. Trop. Subtrop. Agro-ecosyst 10, 441e449. Karim, B., Mukthar, A., Mukthar, H., Athar, M., 2009. Effect of the tree canopy cover on the organic and inorganic content of soil in Cholistan desert. Pak. J. Bot. 41 (5), 2387e2395. Legare, S., Bergeron, Y., Pare, D., 2001. Comparison of the understory vegetation in boreal forest types of southwest Quebec. Can. J. Bot. 79, 1019e1027. Ludwig, F., Dawson, T.E., de Kroon, H., Berendse, F., Prins, H.H.T., 2003. Hydraulic lift in Acacia tortilis trees on an East African savanna. Oecologia 34, 293e300. Ludwig, F., de Kroon, H., Prins, H.H.T., Berendse, F., 2001. Effects of nutrients and shade on treeegrass interactions in an East African savanna. J. Veg. Sci 12, 579e588. Ludwig, F., de Kroon, H., Berendse, F., Prins, H.H.T., 2004. The influence of savanna trees on nutrient, water and light availability and the understorey vegetation. Plant Ecol. 170, 93e105. Magurran, A.E., 2004. Measuring of Biological Diversity. Blackwell publishing, London, UK, p. 256. Mekuria, W., Veldkamp, E., Haile, M., Nyssen, J., Muys, B., Gebrehiwot, K., 2007. Effectiveness of enclosures to restore degraded soils as a result of overgrazing in Tigray, Ethiopia. J. Arid. Environ. 69, 270e284. Mitchell, R., Keith, A., Potts, J., 2012. Overstory and understory vegetation interact to alter soil community composition and activity. Plant Soil 352, 65e84. Moustakas, A., Kunin, W.F., Cameron, T.C., Sankaran, M., 2013. Facilitation or competition? Tree effects on grass biomass across a precipitation gradients. PLoS One 8 (2), e57025. http://dx.doi.org/10.1371/journal phone 0057025. NRC (Natural Resources Conservation Service), 1996. Soil Survey Laboratory Methods Manual. Soil Survey Investigations. Report No. 42. USDA, Washington, DC. Olsen, S.R., Cole, C.V., Watanable, F.S., Dean, L.A., 1954. Estimation of Available Phosphorus in Soil by Extraction with Sodium Bicarbonate. Circ 939. USDA. Philips, S., 1995. Poacae (gramineae). In: Hedberg, I., Edwads, S. (Eds.), Flora of Ethiopia, vol. 7. The National Herbarium. Addis Ababa University, Addis Ababa and University of Systematic Botany, Uppsala University, Sweden.

Power, I.L., Thorrold, B.S., Balks, M.R., 2003. Soil properties and nitrogen availability in silvopastoral plantings of Acacia melanoxylon in North Island, New Zealand. Agroforest. Syst. 57, 225e237. Prescott, C.E., Grayston, S.J., 2013. Tree species influence on microbial communities in the litter and soil: current knowledge and research needs. Ecol. Manage 309, 19e37. Riegel, G.M., Miller, R.F., 1992. Competition for resources between understory vegetation and Overstory Pinus ponderosa in northeastern Oregon. Ecol. Appl. 2 (1), 71e85. Rietkerk, M., van de Koppel, J., 1997. Alternate stable states and threshold effects in semi-arid grazing systems. Oikos 79, 69e76. Sankaran, M., Ratnam, J., Hanan, N.P., 2004. Tree-grass coexistence in savannas revisited-insights from an examination of assumptions and mechanisms invoked in existing models. Ecol. Lett. 7, 480e490. SAS (Statistical Analysis System) Institute, 2009. Applied Statistics and the SAS Programming Language 2nd Edition Cary, North Carolina. Savadogo, P., Elfving, B., 2007. Prediction models for estimating available fodder of two savanna tree species (Acacia dudgeon and Balanites aegyptiaca) based on field and image analysis measures. Afr. J. Range Forage Sci. 24, 63e71. Scholes, R.J., 1990. The influence of soil fertility on the ecology of Southern African savannas. J. Biogeogr. 17, 415e419. Scholes, R.J., Archer, S.R., 1997. Tree-grass interactions in savannas. Ann. Rev. Ecol. Syst. 28, 517e544. Ter Braak, C.J.F., 1997. CANOCO-aFORTRAN Program for Canonical Community Ordination by [partial] [detrended] Canonical] Correspondence Analysis, Principal Component Analysis and Redundancy Analysis (Version 3.15), vol. 95. Institute of Applied Computer Science, Wageningen, The Netherlands. Tessema, Z.K., de Boer, W.F., Baars, R.M.T., Prins, H.H.T., 2011. Changes in vegetation structure, herbaceous biomass and soil nutrients in response to grazing in semiarid savanna in Ethiopia. J. Arid. Environ. 75, 662e670. Tobler, M., Cochard, R., Edwards, P.J., 2003. The impact of cattle ranching on largescale vegetation patterns in a coastal savanna in Tanzania. J. Appl. Ecol. 40, 430e444. Treydte, A.C., Heitkonig, I.M.A., Prins, H.H.T., Ludwig, F., 2007. Trees improve grass quality for herbivores in African savannas. Perspect. Plant Ecol. Evol. Syst. 8, 197e205. Van Reeuwijk, L.P., 1992. Procedures for Soil Analysis, 3rded. International Soil Reference Centre Wageningen (ISRIC), The Netherlands, Wageningen. Walkley, A., Black, C.A., 1934. An examination of wet digestion method for determining soil organic matter and proposed modification of the chromic acid titration method. Soil Sci. 37, 29e38. Weltzin, J.F., Coughenour, M.B., 1990. Savanna tree influence on under canopy vegetation and soil nutrients in northwestern Kenya. J. Veg. Sci 1, 325e334. Wiegand, K., Saltz, D., Ward, D., 2006. A patch-dynamics approach to savanna dynamics and woody plant encroachment - insights from an arid savanna. Persp. Plant Ecol. Evol. Syst. 7, 229e242.