Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Impacts of livestock grazing on plant species composition in montane forests on the northern slope of Mount Kilimanjaro, Tanzania

Impacts of livestock grazing on plant species composition in montane forests on the northern... International Journal of Biodiversity Science, Ecosystem Services & Management, 2015 Vol. 11, No. 2, 114–127, http://dx.doi.org/10.1080/21513732.2015.1031179 Impacts of livestock grazing on plant species composition in montane forests on the northern slope of Mount Kilimanjaro, Tanzania a, b Imani A. Kikoti * and Cosmas Mligo a b Kilimanjaro National Park, P.O. Box 96, Marangu, Tanzania; Department of Botany, University of Dar es Salaam, P.O. Box 35060, Dar es Salaam, Tanzania (Submitted 10 April 2014; accepted 26 February 2015; edited by Patricia Balvanera) The montane forests on Mount Kilimanjaro are part of the ecologically protected area and provide ecosystem services to the local communities on the lower slopes. Despite their ecological value, the montane forests on the northern slope of Mount Kilimanjaro have been affected by increased livestock grazing pressure due to prolonged drought and expansion of agricultural land. We examined the impacts of livestock grazing on plant species composition in the montane forests on the northern slope of Mount Kilimanjaro. The study area was sub-divided into heavily grazed, moderately grazed, lightly grazed and ungrazed sites. A quadrat method was used for field data collection. A total of 115 plant species distributed within 93 genera and 39 families were identified. Moderately grazed and lightly grazed areas had higher plant species diversities than heavily and ungrazed areas. This observation concurs with the intermediate disturbance hypothesis. However, plant species that contributed to high species richness and diversity in grazed areas were mainly grasses, shrubs and herbs. It was concluded that livestock grazing in montane forests on the northern slope of Mount Kilimanjaro has serious impacts on vegetation community composition. Conservation of montane forest habitat and improvement of rangelands on communal land are necessary. Keywords: grazing ecology; Mount Kilimanjaro; plant species diversity; plant life form; tropical montane forest; vegetation communities 1. Introduction and Stern et al. (2002), livestock grazing leads to changes in floristic composition and structure within grazed areas. Livestock grazing in natural ecosystems results in changes Most studies have documented the impacts of livestock in vegetation structure and composition (Stern et al. 2002). grazing on vegetation in woodlands and grasslands (Hardy Changes in vegetation composition from palatable grasses et al. 1999; Mligo 2006; Sun et al. 2011; Cingolani et al. and sedges to less palatable forbs resulting from heavy 2013; Deng et al. 2014; Koerner & Collins 2014). There grazing have been reported in northwest China (Sun et al. has been little effort to cover knowledge gaps regarding 2011), Libya (Zatout 2014), North America (Bakker et al. the impacts of livestock grazing on plant species composi- 2003; Koerner & Collins 2014), South Africa (Koerner & tion in afromontane forests (Reed & Clokie 2000), one of Collins 2014) and Argentina (Cingolani et al. 2013). The which is the montane forest of Mount Kilimanjaro. The response of plant species richness varies according to northern slopes of Mount Kilimanjaro are inhabited by grazing intensity. For example, in grassland ecosystem, pastoral communities, particularly in the villages of Deng et al. (2014) reported that plant species richness Kamwanga, Irkaswa, Kitendeni and Lerangw’a in the increased with decreasing grazing intensity. The author Longido District. These villages had been sharing land observed the highest plant species richness at light and with wildlife in the former Longido Game Controlled moderate grazing intensities. Random grazing patterns can Area. However, major land-use changes have occurred lead to spatial heterogeneity in light availability, soil nutri- between 1952 and 2001 in the area (Noe 2003). These ent availability and vegetation community dynamics. This changes have been associated with increasing land area can reduce plant competition for environmental resources under cultivation and establishment of settlements, espe- because the vegetation exists in patches (Bakker et al. cially on the lower slopes of the northern side of Mount 2003). At the within-patch scale, light grazing intensity Kilimanjaro. For example, in 1952 there was only one can promote plant diversity, but heavy grazing intensity village in this area (the Kamwanga village); however, by can lead to the exclusion of the species intolerant to graz- 2001 three more villages, Lerangw’a, Irkaswa and ing (palatable), resulting in an increase in grazing tolerant Kitendeni, had been established in 1975, 1982 and 2001, plant species (unpalatable) (Watt & Gibson 1988; Calvert respectively (Noe 2003). Concomitant to the influx of 2001). In North America, light browsing resulted in sup- pastoralists to the area, there was also a migration of pression of dominant shrubs and maintenance of shrub non-pastoral people to the area in the past three decades, diversity (Pekin et al. 2014). According to Mligo (2006) *Corresponding author. Email: ikikoti@gmail.com © 2015 Taylor & Francis International Journal of Biodiversity Science, Ecosystem Services & Management 115 which has contributed to an increase in the amount of land adjacent to the Tanzania – Kenya international boundary, used for cultivation at the expense of pastoral land at latitude 2°45ʹ3°25ʹS and longitude 37°00ʹ37°43ʹE, (Muruthi & Frohardt 2000). Because of the decrease in approximately 300 Km south of the equator (Lambretchs grazing land for pastoralists, the Maasai community has et al. 2002). changed from pastoralism to agropastoralism. The general The climate of Mount Kilimanjaro varies from one practices of the aforementioned activities depend on sea- ecological zone to another based on altitudinal gradient sonal rains; however, there has been lack of short rains that and aspect. The northern slope of Mount Kilimanjaro are usually received between November and January in forms the leeward side, thus being drier than the southern recent years. The decrease and delay of long rains in the and eastern slopes. There are two distinct rainy seasons in study area has also been reported. At present, long rains this forest, a long rainy period occurring between March start in April and end in May instead of starting in March and June and short rainy periods between November and and ending in June as in normal years (Ndikumana 2007). December. Therefore, the climate of the area is bimodal. This has resulted in a shortage of forage for livestock and, The nearest weather station to the study area is in consequently, the expansion of grazing areas into montane Lyamungo, which is at approximately 50 km on the wes- forests at higher altitudes ranging from 1600 to 2600 m a. tern slope and receives a mean annual rainfall of 1714 mm s.l., particularly during dry seasons. In recent years, there with a mean temperature of 19.3°C (Appendix 2). has been an increase in the livestock population and graz- A preliminary survey was carried out to define vegeta- ing pressure in montane forests on the northern slopes of tion segments based on their homogeneity and grazing Mount Kilimanjaro (TANAPA 2006). The most preferred intensity. Random sampling procedures were used in locat- composition of livestock types driven into montane forests ing quadrats within each sampling site. The degree of live- on the northern slopes of Mount Kilimanjaro includes: stock trampling was assessed by measuring the percentage cattle, goats, sheep and donkeys. However, the montane (by area) of trail and bare ground in a 20 m × 25 m quadrat. forests are part of the Kilimanjaro National Park, such that Qualitative estimation of grazing intensity in each quadrat the wild communities have historically co-existed with the based on an assessment of percentage utilisation of vegeta- vegetative communities as habitat without putting one tion by livestock was also conducted in each quadrat community in jeopardy, and therefore, their foraging beha- (Rulangalanga 2000). A total of four grazing intensity viours have been considered natural. classes based on a modified Rulangalanga’s classification Livestock grazing may be affecting plant composition in system were established (Rulangalanga 2000)(Table 1). the montane forest on the northern slope of Mount Within each grazing intensity class, a detailed study on Kilimanjaro. The grazing intensity in this montane forest the plant species composition was carried out. Ten rectan- varies based on accessibility and the protection status of the gular quadrats of 20 m x 25 m were established in each forest. The lower montane forest, at approximately grazing intensity category, making a total of 40 quadrats in 16001800 m, is regarded as a heavily grazed area. In this the entire study area. In each quadrat, the trees with area, a Memorandum of Understanding between local com- Diameters at Breast Height (DBH) equal or greater than munities and the Forest and Beekeeping Department was 10 cm were identified to species, enumerated and measured. signed in 1963 to allow use of this area for livestock grazing Two (2 m × 5 m) subquadrats nested in bigger quadrats (of during the dry season. On the other hand, the upper part of the 20 × 25 m) as recommended by Stohlgren et al. (1995) were montane forest, with an approximate altitude of 18002400 m, established for sampling trees with DBHs less than 10 cm has no legal access for livestock grazing. However, it has been and shrubs. Additionally, grasses and herbaceous species grazed at different levels: moderately, lightly or ungrazed. were determined using three 1 m -quadrats randomly estab- Livestock grazing in the montane forest on the northern lished inside the bigger quadrats (20 m × 25 m), and the slope of Mount Kilimanjaro may be a threat to the conserva- number of individuals and percentage cover for each spe- tion of plant species and the associated economic and ecolo- cies was estimated in relation to the quadrat size. All plants gical values of the mountain. This study investigated the were identified to species level when it was possible in the impacts of livestock grazing on plant composition to under- field. For plant species that were not easily identified in the stand how livestock grazing pressure alters montane forest field, specimens were collected, pressed and taken to the communities on the northern slope of Mount Kilimanjaro. It herbarium in the Department of Botany, University of Dar was hypothesised that livestock grazing has negative impacts es Salaam (DSM) for identification by matching with her- on the plant community in the montane forests of Mount barium specimens and/or keying using relevant flora iden- Kilimanjaro. We compared plant species diversity in sites tification texts, such as Flora of Tropical East Africa and with different grazing intensities in the montane forests on Flora Zambeziaca. the northern slope of Mount Kilimanjaro. 2.1. Data analysis 2. Material and methods Plant species composition in each area was determined in This study was carried out in the montane forests on the terms of species diversity index, species evenness, and northern slope of Mount Kilimanjaro, which is found in species richness and plant density for all recorded plant Northern Tanzania (Figure 1). The study area is located species. 116 I.A. Kikoti and C. Mligo Figure 1. Map showing the location of study area in Mount Kilimanjaro. significance level for all tests was based on a 5% critical Table 1. Grazing intensity classes and their interpretation. limit. Correlation analysis was used to assess the influence of grazing intensity on plant species diversity. The classifi- Grazing intensity class Interpretation cation of vegetation communities was done using Two-Way 0 No grazing Indicator Species Analysis (TWINSPAN) based on live- 11–20%: Little grazed stock grazing disturbances (Hill 1979). 221–60%: Moderately grazed 361–100%: Heavily grazed Note: Modified from Rulangalanga (2000). 3. Results 3.1. Plant species composition in grazed areas The plant species diversity was determined from raw A total of 115 plant species distributed within 39 families and data using the Shannon’s diversity index (H′) (Shannon 93 genera were identified in this study (Appendix 1). The 1948), as shown in the formula below: commonly occurring plant species in ungrazed areas were Teclea simplicifolia, Achyranthus aspera, Asparagus africana and Cassipourea malosana, while in lightly grazed areas H ¼ ðp ln p Þ i i common species were Achyranthes aspera, Olea europaea i¼1 subsp. africana and Teclea simplicifolia (Table 2). In moder- ately grazed areas, Croton megalocarpus, Commelina ben- where p = n /N is the number of individuals found in the i i ghalensis, Setaria homonyma and Teclea simplicifolia were ith species as a proportion of the total number of indivi- represented with high stem densities, and the densely popu- duals found in all the species and ln the natural logarithm lated species in the heavily grazed areas were Euclea divi- base e. norum, Themeda triandra, Carissa spinarum and Rhus Before applying one-way analysis of variance natalensis (Table 2). The most dominant families with regards (ANOVA) to compare plant species diversity and richness to plant species composition were Poaceae, Acanthaceae, among areas with different grazing intensities, tests for Rutaceae, Amaranthaceae, Euphobiaceae, Meliaceae and normality (KolmogorovSmirnov D statistic) and homoge- Rubiaceae. neity of variance (Levene statistic) of richness and diversity Population structure and DBH size class distribution of indices for all recorded plant species were carried out. The trees varied considerably among the four studied areas. The International Journal of Biodiversity Science, Ecosystem Services & Management 117 Table 2. Density of dominant plant species recorded in areas with different grazing intensity (T = Tree, H = Herbaceous, ST = Small Tree, V = Vine and G = Grass). Density (Individuals/ha) Growth Plant species form Heavily Moderately Little Ungrazed Bersama abyssinica Fresen. T 0 31 36 58 Calodendron capense (L.f.) Thunb. T 0 7 10 6 Canthium lactescens Hiern. T 0 44 6 0 Cassipourea malosana Alston. T 0 17 23 72 Cassytha pondoensis var. schliebenii (Robyns & Wilczek) M.A. T0 2 8 0 Diniz Clausena anisata (Willd.) J.Hk. ex Benth. T 0 7 178 12 Croton megalocarpus Hutch. T 0 67 20 0 Diospyros abyssinica Hiern. T 0 24 2 26 Erythrococca bongensis Pax. T 0 6 36 16 Euclea divinorum Hiern T 98 48 29 0 Maytenus senegalensis T5 4 8 5 Maytenus undata T0 2 1 4 Olea europaea ssp. africana T 18 19 138 31 Teclea simplicifolia T 0 78 81 252 Turraea holstii T 0 12 41 46 Turraea robusta T 0 28 89 0 Vangueria infausta T4 5 60 11 Achyranthes aspera L. H 2 70 178 205 Allophyllus africanus P. Beauv. ST 0 4 17 3 Asparagus africanus Lam. V 0 4 11 9 Commelina benghalensis L. H 44 19 0 Crotalaria goodformis Vatke. ST 0 34 33 0 Hibiscus fuscus Garcke H 3 0 18 0 Hibiscus micranthus L.F. H 7 0 0 0 Hyparrhenia filipendula (Hochst.) Stapf G 37 0 0 0 Hypoestes aristata (Vahl) Sol. ex Roem. & Schult. H 0 0 160 69 Justicia flava (Vahl) Vahl H 4 20 19 0 least grazed and moderately grazed areas had large numbers lightly grazed areas and heavily and ungrazed areas of individuals within the DBH size classes of 10–40 cm (p > 0.05). Additionally, the highest plant species diversity relative to ungrazed and heavily grazed areas (Figure 2). (ShannonWiener (H′) index of 2.767) was recorded in However, old stand trees with diameter size classes greater moderately grazed areas, followed by lightly grazed, heav- than 100 cm were highly represented in ungrazed areas, and ily grazed and ungrazed areas (Table 4). The difference in these huge size classes were missing in other areas. Heavily the plant species diversity indices among areas with dif- grazed areas had only 40 stems/ha in the 10–20 cm DBH size ferent grazing intensities was significant based on an classes, while lacking trees with DBHs above 20 cm ANOVA (F = 32.87 and p < 0.05). (Figure 2). 3.3. Vegetation communities in the study area 3.2. Plant species richness and diversity Based on TWINSPAN, three vegetation community types Plant species richness was highest in moderately grazed (denoted as A, B and C) existed, as shown in the dendro- areas (21.3), followed by lightly, heavily and ungrazed gram (Figure 3). At the first level of TWINSPAN division, areas with mean species numbers of 18.70, 12 and 10.9, plant community C emerged with Themeda triandra (+) as respectively (Table 3). The difference among areas with the indicator of division. The eigenvalue of 0.7546 for this different grazing intensities was significant based on an division was relatively high, showing large differences in ANOVA (F = 35.7, p < 0.05). Multiple comparison tests species composition from the rest of the communities in among areas with different grazing intensities using the study area. Community A consisted of samples of Tukey’s Honestly Significant Difference showed that sig- vegetation data from ungrazed areas, which were com- nificant differences exist between moderately vs. heavily monly represented by Clutia abyssinica, Maytenus sene- grazed areas, moderately vs. ungrazed areas, lightly vs. galensis, Achyranthes aspera and Turraea holstii, among heavily grazed areas and lightly grazed vs. ungrazed areas others. The indicators of species composition differences (p < 0.05) (Table 3). Otherwise, no significant variations between species in communities A and B were Commelina in species richness were found between moderately and benghalensis (+), Turraea robusta (+), Prunus africana 118 I.A. Kikoti and C. Mligo Figure 2. DBH size class distribution in areas with different grazing intensities. Table 3. Variations in plant richness among areas with different grazing levels in Northern slopes of Mount Kilimanjaro. a, b, c, d (M = mean ± standard error and indicates significant difference at p < 0.05). a b c d Grazing Heavily Moderately Little grazed Ungrazed level (M = 12.0 ± 0.42) (M = 21.3 ± 1.03) (M = 18.7 ± 1.03) (M = 10.9 ± 0.75) a a a Heavily <0.001 <0.001 0.796095 b b Moderately 0.152067 <0.001 c c Little <0.001 Ungrazed Table 4. Variations in plant species diversity (ShannonWiener diversity index (H′)) among areas with different grazing levels in a, b, c, d montane forest of Northern slopes of Mount Kilimanjaro based on Tukey’s test (M = mean± Standard Error and letters indicates statistical level of significance at p < 0.05). a b c d Grazing Heavily Moderately Little grazed Ungrazed level (M = 2.2940 ± 0.03) (M = 2.7673 ± 0.05) (M = 2.5094 ± 0.08) (M = 1.9166 ± 0.08) a a a Heavily <0.001 0.090117 <0.001 b b b Moderately 0.030508 <0.001 c c Little <0.001 Ungrazed (−) and Cassipourea malosana (−). The presence of Vangueria infausta and Hypoestes aristata. Community Commelina benghalensis (+) and Turraea robusta (+) C was composed of plant species from heavily grazed implied that ecological changes were caused by distur- areas and was represented by Carissa spinarum and bance through livestock grazing. On the other hand, P. Cynodon dactylon. These two species were the character- africana (−)and Cassipourea malosana (−) were sensitive istics of a shrubby community and this implies that live- to disturbances through grazing and hence were indicators stock grazing pressure contributed to the observed change of communities undisturbed by grazing. The measure of of montane forest communities into shrubby communities. variation in species composition (eigenvalue of 0.49) indi- cates minor differences between communities A and B 3.4. The relationship between grazing intensities and from lightly and moderately vs. ungrazed areas. This plant species diversity means that community B was dominated by plant species that were also found in both the lightly and moderately A significant negative correlation was found between grazing grazed areas. The indicator species for moderately grazed intensity and species diversity (r = −0.4871, p = 0.0063) conditions were Clausena anisata, Celtis africana, (Figure 4). Although the correlation coefficient was not strong International Journal of Biodiversity Science, Ecosystem Services & Management 119 uP9 uP1 uP2 uP3 Clutia abyssinica Maytenus senegalensis Turraea holstii uP5 Achyranthes aspera uP7 uP4 uP6 uP10 uP8 Lp7 Prunus africana Lp8 Cassipourea malosana Lp10 Commelina benghalensis Turraea holstii Lp6 Turraea robusta Euclea divinorum Acalypha kilimandsc ica Lp9 Vernonia adoensis Lp3 Clausena anisata Lp4 Celtis africana Lp5 Vangueria infausta Lp1 Hypoestes aristata Lp2 mp3 mp5 Asystasia gangetica mp6 Abutilon mauritianum mp7 Themeda triandra Chloris pycnothrix mp8 mp9 mp10 mp1 mp2 mp4 Gutenbergia spp. hP1 Cynodon dactylon hP2 Conyza pyrrhopappa hP10 Olea europaea ssp. africana hP6 Carissa spinarum Senecio abyssinicus hP7 hP8 hP9 hP3 hP4 hP5 Figure 3. Vegetation community classification using TWINSPAN. (Lp = Little Grazed Plots; hP = Heavily grazed Plots; uP = Ungrazed Plots and mp = Moderately Grazed Plots). enough to explain the significant influence of grazing pressure in plant species composition among sampling sites was on species diversity, based on the montane nature of the forest, contributed by variation in livestock grazing intensities. the impacts of livestock grazing on diversity were very sig- nificant. This indicates that plant species are sensitive to dis- turbance through livestock grazing. Additionally, the montane 4. Discussion forest plant species are not adaptedtoco-exist with livestock 4.1. The influence of grazing pressure on plant species grazing. This implies that at any level of grazing intensity plant diversity species diversity in montane forests will decrease. We observed higher plant species richness and diversity in moderately grazed and lightly grazed than in heavily grazed and ungrazed areas. Ungrazed areas were charac- 3.5. The response of different plant life forms to terised by old growth stands with DBHs ranging from 10 livestock grazing pressure to 310 cm. The low diversity of plant species in ungrazed There were differences in plant life forms and plant size areas may have been caused by a few dominant tree stands structures among areas with different grazing intensities. A that usurp the lion’s share of the habitat resources (nutri- high composition of tree species was found in moderately, ents and light). Many studies have reported that a few ungrazed and lightly grazed areas. Heavily grazed areas had plant species favoured by lack of grazing tend to out- the lowest tree density and averaged 0.8 tree species per plot compete plants with smaller statures (Belsky 1992; Pekin with a 0.19 Shannon’s diversity index (Table 5). Despite the et al. 2014). On the other hand, lower plant species diver- lower tree species density in heavily grazed areas, these sities were recorded in the heavily grazed areas than in areas had higher compositions of shrubs and grass species moderately and lightly grazed areas. This implies that the than ungrazed areas. The heavily grazed areas were com- increase in livestock grazing intensity resulted in decreases monly dominated by Themeda triandra, Setaria sphace- in species diversity and abundances in the montane forests lata, Hyparrhenia filipendula, Heteropogon contortus and of northern slope of Mount Kilimanjaro. High plant spe- Dichanthium annulatum. The herbaceous species were cies diversity in moderately and lightly grazed areas may represented by Senecio abyssinicus, Conyza pyrrhopappa, be due to the effects of livestock grazing that results in Gutenbergiaspp., Ruellia tuberose and Thunbergia alata, opening of the canopy, thus giving opportunity for regen- whereas shrubs were represented by Carissa spinarum, eration by gap opportunistic plant species (Pekin et al. Indigofera volkensii, Leucas deflexa, Lantana trifolia, 2014). Livestock grazing may also have reduced competi- Solanum incanum and Hibiscus micranthus. The variation tion among plant species through selective grazing on 120 I.A. Kikoti and C. Mligo Figure 4. The influence of grazing intensities on plant species diversity in montane forests of Northern slopes of Mount Kilimanjaro. Table 5. Plant life forms recorded in areas with different grazing level in montane forest of northern slopes of Mount Kilimanjaro (SN = Species number, H′ = ShannonWiener diversity index). Trees Shrubs Herbs and grasses Level of grazing Mean SN Mean H′ Mean SN Mean Hʹ Mean SN Mean Hʹ Heavily 0.8 ± 0.29 0.19 ± 0.09 4 ± 0.25 0.19 ± 0.55 8.5 ± 0.42 1.6 ± 0.09 Little 3.7 ± 0.52 0.95 ± 0.14 6.3 ± 0.63 1.55 ± 0.95 9.6 ± 0.66 1.9 ± 0.08 Moderately 5.2 ± 0.57 1.26 ± 0.15 6.3 ± 0.26 1.58 ± 0.26 14.5 ± 0.63 2.3 ± 0.06 Ungrazed 3.7 ± 0.51 1.00 ± 0.11 3.8 ± 0.53 1.02 ± 0.38 5.2 ± 0.53 1.49 ± 0.12 palatable competitors as well as trampling of both unpala- competitively dominant species from excluding other spe- table and palatable plants during grazing in the montane cies from the community (Markey & Currie 2001; Catford forest (Rooney & Waller 2003). This finding is in agree- et al. 2012). This brings about a trade-off between plant ment with the intermediate disturbance hypothesis pro- species ability to compete and tolerate various forms of posed by Connell (1978). Models and metadata analysis disturbance. Species diversity is low at extremely low have indicated that species richness and the Shannon levels of disturbance because only the best competitors Wiener diversity index are strong predictors of the inter- dominate and persist within community (Connell 1978). mediate disturbance hypothesis (Svensson et al. 2012). This concept is in agreement with the findings from this The mechanism underlying the intermediate disturbance study, in which the habitat where disturbance was low hypothesis is centred on a complex interplay between life displayed low plant species diversity. However, in the history, biotic interaction and historical disturbance regime severely and the highly disturbed areas only a few species (Catford et al. 2012). The increased availability of plant persisted or repeatedly colonised after every similar requirements, such as light, following disturbances regime of disturbance, thus resulting in low species diver- through livestock grazing may explain why high diversi- sities (Connell 1978). This concept also applies to the ties were observed in moderately and lightly grazed areas findings from this study in which heavily grazed areas in the montane forests on the northern slope of Mount had lower species diversities than moderately grazed Kilimanjaro. According to Roberts and Gilliam (2003), areas. The overgrazed areas were dominated by plants intermediate disturbance causes changes in local microcli- unpalatable to livestock, including a few shrubs that per- mates by opening up space in the canopy, resulting in the formed under extremely modified habitat conditions. release of resources that would otherwise not be accessible Therefore, the balance between competitive exclusion to understory plants. Physical disturbances prevent and the loss of competitive dominants through disturbance International Journal of Biodiversity Science, Ecosystem Services & Management 121 is attained at intermediate disturbances (Markey & Currie total P. africana regeneration potential due to grazing and 2001). At the highest peak of species diversity, conditions browsing in afromontane forests in southern Ethiopia. favour the coexistence of both the competitive species and Currently, P. africana is included on the IUCN Red List disturbance-tolerant species (Connell 1978). of Threatened Species and is categorised as a vulnerable The purpose of Kilimanjaro National Park is to protect species (WCMC 1998). Another tree species that was Africa’s highest and one of the world’s largest free stand- restricted to ungrazed and lightly grazed areas was ing mountains and to conserve its unique socio-economic, Cassipourea malosana. The indicator plant species of cultural and ecological values and features of the fragile favourable habitat in the lightly and moderately grazed mountain ecosystem (TANAPA 2006). On that basis, areas in cluster B were Celtis africana, Vangueria maintenance of the status quo in the montane forest is of infausta, Clausina anisata and Hypoestes arisata. Some high priority, as it is an important regional water catch- of these species (for example, Celtis africana) are vulner- ment for people outside the park. The plant species diver- able to livestock grazing and browsing. The mortality of sity observed in the grazed areas was mainly due to the C. africana due to grazing observed in the southern dominance of shrubs, grasses and herbs. Therefore, the Ethiopian Afromontane forest cannot be excluded from observed changes in species composition due to cattle the heavily grazed areas of the northern slopes of Mount grazing in the montane forests were contributed by the Kilimanjaro because there was no representation of indi- cover abundances of weeds and shrubs, which were char- viduals of C. africana. Cluster C consisted of plant com- acteristic in the heavily grazed areas. Therefore, grazing munities from heavily grazed areas with indicator species may lead to an increase in the number of plant species and of plant communities that are resilient to livestock distur- the overall plant species diversity of an area, although the bances, such as Carissa spinarum, (shrub), Senecio abys- additional species may be either weeds or early colonisers sinica (herb) and Cynodon dactylon (grass). However, the (Landsberg et al. 2003). Catford et al. (2012) identified presence of remnants of the forest-dependent Olea euro- five main ways in which alien species may reduce local paea subsp. africana indicates transformed/degraded mon- diversity along disturbance gradients. These include: niche tane forest is present on the northern slope of Mount pre-emption, apparent competition, interference competi- Kilimanjaro. Cynodon dactylon is a grass species that tion, exploitative competition and transformation of the performs well in heavily disturbed habitats. The perfor- environment. Despite the increase in diversity of shrubs mance of C. dactylon in heavily grazed areas is possible and weeds in heavily grazed areas of the montane forest, because of its extensive stolon and rhizome system, which these species may alter ecosystem processes (i.e., erosion provide a means of rapid expansion, allowing it to thrive rates and seasonal flows) to which native montane tree well in overgrazed areas. During this study, this species species are adapted. was restricted only to heavily grazed areas in the montane forests. Similar observations indicating a high abundance of C. dactylon in over-grazed areas were reported in the 4.2. Variation in plant species composition in Ethiopian highlands (Mwendera et al. 1997) and Southern vegetation communities with different levels of Maasailand of Kenya (Kioko et al. 2012). Among the grazing dominant shrubs found within overgrazed communities was Carissa spinarum, which is also a drought tolerant From TWINSPAN, three clusters of vegetation commu- species and performs in heavily grazed areas. Bahiru nities, denoted as A, B and C, were clearly observed. Plant (2008) observed a high frequency of C. Spinarum in species communities in ungrazed and lightly grazed areas grazed areas in Ethiopia because it is less preferred by were characterised by indicator species that are not toler- livestock. The avoidance of C. Spinarum by livestock ant of anthropogenic disturbances, particularly livestock grazing. For example, the African cherry (P. africana)is most likely is due its unpalatability due to its high percen- typically restricted to afromontane forest (White 1983). tage of tannins, which range from 9% to 15% in its leaves During fieldwork, this species was only recorded in (Parmar & Kaushal 1982). ungrazed areas of the montane forests on the northern Tannins are polyphenolic compounds with great struc- slope of Mount Kilimanjaro that had been influenced by tural diversities that decrease the digestibility of protein low-intensity human land use for a very long period. This (Giner-Chavez 1996). Due to the physiological difficulties species was not found in grazed areas where sampling was most herbivores experience in their digestion and the carried out, presumably due to livestock-related distur- differences in the chemical reactivity of tannins, they are bances, and can therefore be considered a potentially sen- not preferred by grazing and browsing animals (Aganga sitive plant species and indicator of late succession in the et al. 1997). The thorny characteristics of C. spinarum montane forest. Stewart (2009) observed declines in popu- make it resistant to herbivore grazing and browsing lations of P. africana due to anthropogenic activities impacts. including bark harvesting, wildfires and livestock grazing In heavily grazed areas, there has been a visible change in in Cameroon. Substantial evidence indicates that livestock the vegetation community from montane forest to shrubland. grazing is among the causative agents for regeneration Only three tree species with DBH>10 cm, Euclea divinorum, failure of P. africana in Mount Oku, Cameroon (Stewart Olea europaea ssp. africana and Acacia nilotica, were 2009). Abebe (2008) also observed a loss of 42–61% of observed in heavily grazed areas of the montane forests. 122 I.A. Kikoti and C. Mligo Aganga AA, Tsopito CM, Moroke KM. 1997. Tannin content of Livestock grazing changes plant structure and species com- some indigenous browse of Botswana. In: Proceeding of position in mountainous ecosystems. The observation of XVII International Grassland Congress, ID No 136; only a few woody species and the small diameters of the Winnepeg, Manitoba, Canada. trees in the heavily grazed areas indicate negative impacts of Bahiru KM. 2008. Enclosure as a viable option for rehabilitation livestock grazing on the montane forest on the Northern of degraded lands and biodiversity conservation: the case of Kallu Woreda, Kallu Woreda, Southern Wello [MSc thesis]. slope of Mount Kilimanjaro. Therefore, livestock grazing in Addis Ababa University, Addis Ababa. the montane forest of Mount Kilimanjaro has strong implica- Bakker C, Blair JM, Knapp AK. 2003. Does resource availabil- tions for ecosystem integrity and sustainability. In Central ity, resource heterogeneity or species turnover mediate Argentina, heavy grazing pressure was blamed for convert- changes in plant species richness in grazed grasslands? ing mountain ecosystems into rocky deserts through elimina- Oecologia. 137:385–391. Belsky AJ. 1992. Effects of grazing, competition, disturbance tion of woodlands (Cingolani et al. 2013). and fire on species composition and diversity in grassland communities. J Veg Sci. 3:187–200. doi:10.2307/3235679 Calvert GA. 2001. The effects of cattle grazing on vegetation 5. Conclusion diversity and structural characteristics in the semi arid Livestock grazing significantly influences plant species lands of North Queensland [PhD thesis]. James Cook diversity in areas with different grazing intensities on the University, North Queensland. northern slope of Mount Kilimanjaro. Moderately grazed Catford JA, Daehler CC, Murphy HT, Sheppard AW, Hardesty BD, Westcott DA, Rejmánek M, Bellingham PJ, Pergl J, areas had the highest plant species diversity, complying Horvitz CC, Hulme PE. 2012. The intermediate disturbance with the intermediate disturbance hypothesis. However, hypothesis and plant invasions: implications for species rich- low tree species richness and a lack of advanced aged stands ness and management. Perspect in Plant Eco Evol and Syst. in heavily grazed areas indicated that the impacts of live- 14:231–241. Cingolani AM, Cabido MR, Renison D, Neffa VS. 2013. Combined stock grazing prevented trees from attaining large size effects of environment and grazing on vegetation structure in classes. Grazing pressure is a function of the ecological argentine granite grasslands. J Veg Sci. 14:223–232. variation existing in the montane forest in the study area Connell JH. 1978. Diversity in tropical rain forests and coral and the plant species responses to the aforementioned reefs. Science. 199:1302–1310. anthropogenic disturbance. Conservation of montane for- Deng L, Sweeney S, Shangguan Z-P. 2014. Grassland responses to grazing disturbance: plant diversity changes with grazing inten- ests is important for the protection of habitat for species sity in a desert steppe. Grass and Forage Sci. 69:524–533. with different conservation statuses (i.e., rare, endemic and [FAO] Food and Agriculture Organization of United Nations. threatened species). Despite the fact that moderate grazing 2001. FAOClim 2.0. Agroclimatic database CD-ROM + promoted overall plant diversity in the study area, an endan- users manual (72 pp.) Environment and Natural Resources gered species, African cherry (Prunus africana), was Working paper. No 5. Available from: http://www.fao.org/nr/ climpag/pub/en1102_en.asp. severely affected by all levels of grazing, reducing the Giner-Chavez BI. 1996. Condensed tannins in tropical forages ecosystem services the montane forest provides to the [PhD thesis]. Ithaca (NY): Cornell University. local community. Therefore, it is recommended that the Hardy MB, Barnes DL, Moore A, Kirkman KP. 1999. The montane forest be protected from livestock grazing management of different types of veld. In: Tainton NM, encroachment for conservation of biodiversity, the editor. Veld management in South Africa. Pietermaritzburg: watershed catchment and the soil. The park management University of Natal Press. Hill MO. 1979. TWINSPAN A FORTRAN program for arran- needs to collaborate with pastoralists in the northern part of ging multivariate data in an ordered two-way table by classi- Mount Kilimanjaro to observe the boundaries of the pro- fication of the individuals and attributes. Ithaca (NY): tected area that include the montane forests. This may Ecology and Systematics, Cornell University. ensure a protected water catchment for providing a sustain- Kioko J, Kiringe JW, Seno SO. 2012. Impacts of livestock grazing able water supply for wildlife as a means of biodiversity on a savanna grassland in Kenya. J Arid Land. 4:29−35. Koerner SE, Collins SL. 2014. Interactive effects of grazing, conservation and to satisfy the water requirements of live- drought, and fire on grassland plant communities in North stock outside of the protected area (montane forests) on the America and South Africa. MedMed (Ecology). 95:98–109. northern slope of Mount Kilimanjaro. In this collaboration, Lambretchs C, Woodley B, Hemp A, Hemp C, Nnyiti P. 2002. the park may provide technical assistance in establishing Aerial survey of the threats to Mount Kilimanjaro Forests. Dar es Salaam: UNDP; 33 pp. watering points outside the montane forest to prevent access Landsberg J, James CD, Morton SR, Müller WJ, Stol J. 2003. by pastoralists to the montane forest habitats. Abundance and composition of plant species along grazing gradients in Australian rangelands. J Appl Ecol. 40:1008–1024. Markey RC, Currie DJ. 2001. The diversity – disturbance bal- Disclosure statement ance relationship: is it generally strongly peaked? Ecology. No potential conflict of interest was reported by the authors. 82:3429–3492. Mligo C. 2006. Effects of grazing pressure on plant species composition and diversity in the semi-arid rangelands of References Mbulu district, Tanzania. Agric J. 1:277–283. Abebe GT. 2008. Ecology of regeneration and phenology of seven Muruthi P, Frohardt K. 2000. Study on the development of indigenous tree species in a dry tropical Afromontane forests, transboundary natural resource management areas in Africa. Southern Ethiopia [PhD dissertation submitted at School of Kilimanjaro Heartland case study. Nairobi, Kenya: African Graduate studies]. Addis Ababa: Addis Ababa University. Wildlife Foundation (AWF). International Journal of Biodiversity Science, Ecosystem Services & Management 123 Mwendera EJ, Saleem MAM, Woldu Z. 1997. Vegetation Shannon C. 1948. A mathematical theory of communication. response to cattle grazing in the Ethiopian highlands. Agric Bell Syst Tech J. 27:379–423. Ecosyst Environ. 64:43–51. Stern M, Quesada M, Stoner KE. 2002. Changes in composi- Ndikumana J. 2007. Livestock situation analysis in North tion and structure of a tropical dry forest following inter- East Tanzania. Vol. 1. Nairobi: Pastoralist Bulletin. A- mittent cattle grazing. Int J Trop Biol. 50:1021–1034. AARNET/International Livestock Research Institute. Stewart K. 2009. Effects of bark harvest and other human Noe C. 2003. The dynamic of land use changes and their impacts on activity on populations of the African cherry (Prunus wildlife corridor between Mount Kilimanjaro and Amboseli africana) on Mount Oku, Cameroon. For Ecol Manage. National Park, Tanzania. Land Use Change, Impacts and 258:1121–1128. Dynamics (LUCID) Project Working Paper No. 31; Nairobi: Stohlgren TJ, Falkner MB, Schell LD. 1995. A modified-whit- International Livestock Research Institute. taker nested vegetation sampling method. Vegetatio. Parmar C, Kaushal MK. 1982. Carissa spinarum [Internet]. In: 117:113–121. Wild Fruits. New Delhi, India: Kalyani Publishers; p. Sun DS, Wesche K, Chen DD, Zhang SH, Wu GL, Du GZ, 15–18; [cited 2012 Aug 3]. Available from: http://www. Comerford NB. 2011. Grazing depresses soil carbon storage hort.purdue.edu/newcrop/parmar/04.html through changing plant biomass and composition in a Pekin BK, Wisdom MJ, Endress BA, Naylor BJ, Parks CG. Tibetan alpine meadow. Plant Soil and Env. 57:271–278. 2014. Ungulate browsing maintains shrub diversity in the Svensson JR, Lindegarth M, Jonsson PR, Pavia H. 2012. absence of episodic disturbance in seasonally-arid conifer Disturbance–diversity models: what do they really predict forest. PLoS One. 9:e86288. doi:10.1371/journal. and how are they tested? Proc Royal Soc B: Biol Sci. pone.0086288 279:2163–2170. Reed MS, Clokie MRJ. 2000. Effects of grazing and cultivation [TANAPA] Tanzania National Parks. 2006. General management on forest plant communities in Mount Elgon National Park, plan for Kilimanjaro National Park (2006–2015). Arusha: Uganda. Afr J Ecol. 38:154–162. Tanzania National Parks/UNDP/CAWM. Roberts MR, Gilliam FS. 2003. Response of the herbaceous Watt TA, Gibson WD. 1988. The effects of sheep grazing on layer to disturbance in eastern forests. In: Gilliam FS, seedling establishment and survival in grassland. Vegetatio. Roberts MR, editors. The herbaceous layer in forests of 78:91–98. eastern North America. New York (NY): Oxford Press; p. White F. 1983. The vegetation of Africa: a descriptive memoir to 302–320. accompany the UNESCO/AETFAT/UNSO vegetation map Rooney TP, Waller DM. 2003. Direct and indirect effects of of Africa by F. White. Paris: UNESCO. white-tailed deer in forest ecosystems. For Ecol Manage. [WCMC] World Conservation Monitoring Centre. 1998. Prunus 181:165–176. africana [Internet]. The IUCN Red List of Threatened Rulangalanga ZK. 2000. Obstacles to the sustainable utiliza- Species. Version 2014.3. [cited 2015 Apr 1]. Available from: tion of the vegetation of semi-arid areas of Mbulu district www.iucnredlist.org for animal husbandry. Paper presented at a Workshop on Zatout MMM. 2014. Effect of negative human activities on plant Current Initiative in Sustainable Management of Dryland diversity in the Jabal Akhdar pastures. Int J Bioassays. Biodiversity, Arusha, Tanzania. 3:3324–3328. 124 I.A. Kikoti and C. Mligo Appendix 1. List of plant species recorded in Northern slopes of Mount Kilimanjaro. Plant species Family Growth form 1 Abutilon angulatum (Guill. & Perr.) Mast. Malvaceae shrub 2 Abutilon mauritianum (Jacq.) Medik. Malvaceae shrub 3 Acacia nilotica (L.) Delile Fabaceae Tree 4 Achyranthes aspera L. Amaranthaceae herb 5 Ageratum conyzoides L. Asteraceae herb 6 Allophyllus africanus P. Beauv. Sapindaceae tree 7 Asparagus africanus Lam. Asparagaceae Vine 8 Asparagus falcatus L. Asparagaceae Climbing 9 Asystasia gangetica (L.) T.Anderson Acanthaceae herbaceous 10 Asystasia schimperi T. Anderson Acanthaceae herbaceous 11 Bersama abyssinica Fresen. Melianthaceae Tree 12 Bidens pilosa L. Asteraceae herb 13 Bidens schimperi Walp. Compositae herb 14 Calodendron capense (L.f.) Thunb. Rutaceae Tree 15 Canthium lactescens Hiern. Rubiaceae Tree 16 Capparis tomentosa Lam. Capparidaceae shrub 17 Casearia battiscombei Flacourtiaceae Tree 18 Carissa spinarum Linn. Apocynaceae Shrub 19 Cassipourea malosana Alston. Rhizophoraceae Tree 20 Cassytha pondoensis var. schliebenii (Robyns & Wilczek) M.A.Diniz Lauraceae Tree 21 Celtis africana N.L.Burm Cannabaceae tree 22 Chloris pycnothrix Poaceae grass 23 Clausena anisata (Willd.) J.Hk. ex Benth. Rutaceae Tree 24 Clerodendrum myricoides Lamiaceae Shrub 25 Clutia abyssinica Jaub. & Spach Peraceae tree 26 Coccinia adoensis (A. Rich.) Cogn. Cucurbitaceae herb 27 Coccinia grandis (L.) J.Voigt Cucurbitaceae herb 28 Commelina africana L. Commelinaceae herb 29 Commelina benghalensis L. Commelinaceae herbaceous 30 Conyza pyrrhopappa Sch.Bip. ex A.Rich Compositae Herb 31 Crotalaria goodformis Vatke. Papilionaceae Herb 32 Croton megalocarpus Hutch. Euphorbiaceae Tree 33 Cyathula cylindrica Moq. Amaranthaceae Herb 34 Cynodon dactylon (L.) Pers. Poaceae grass 35 Dactyloctenium germinatum Hack. Poaceae grass 36 Dichanthium annulatum (Forssk.) Stapf Poaceae grass 37 Digitaria abyssinica Hochst. ex A. Rich.) Stapf Poaceae grass 38 Diospyros abyssinica Hiern. Ebenaceae Tree 39 Eragrostis volkensii Pilg. Poaceae grass 40 Erythrococca bongensis Pax. Euphorbiaceae Tree 41 Euclea divinorum Hiern Ebenaceae Tree 42 Euphobia hirta Euphorbiaceae Herb 43 Ficus thonningii Blume Moraceae Tree 44 Galinsoga parviflora Cav. Asteraceae Herb 45 Grewia similis K. Schum. Malvaceae shrub 46 Gutenbergia spp Sch. Bip. Asteraceae shrub 47 Heteropogon contortus (L.) P.Beauv.ex Roem.& Schult. Poaceae grass 48 Hibiscus fuscus Garcke Malvaceae Shrub 49 Hibiscus micranthus L.F. Malvaceae Shrub 50 Hyparrhenia filipendula (Hochst.) Stapf Poaceae grass 51 Hypoestes aristata (Vahl) Sol. ex Roem. & Schult. Acanthaceae Herb 52 Hypoestes forskalii (Vahl) Roem. & Schult. Acanthaceae Herb 53 Hypoestes triflora Acanthaceae Herb 54 Indigofera arrecta Fabaceae. shrubs 55 Indigofera cuneata Baker ex Oliv. Fabaceae. shrubs 56 Indigofera volkensii Fabaceae. shrubs 57 Jasminum sp. Oleaceae Herb 58 Justicia flava (Vahl) Vahl Acanthaceae Herb 59 Lantana trifolia Verbenaceae Shrubs 60 Leonotis nepetifolia (L.) R.Br. Lamiaceae shrub 61 Leucas deflexa Hook. f. Lamiaceae shrub 62 Leucas grandis Vatke Lamiaceae shrub 63 Lippia javanica (Burm.f.) Spreng Verbenaceae shrub 64 Lippia triphylla Verbenaceae shrub (Continued) International Journal of Biodiversity Science, Ecosystem Services & Management 125 Appendix 1. (Continued). Plant species Family Growth form 65 Macrotyloma axillare (E. Mey.) Verdc Fabaceae Herb 66 Malacantha alnifolia (Baker) Pierre Sapotaceae shrub 67 Maytenus senegalensis (Lam.) Excell Celastraceae Tree 68 Maytenus undata (Thunb.) Blakelock Celastraceae Tree 69 Monechma debile (Forssk.) Nees Acanthaceae shrub 70 Nuxia congesta R.Br. ex Fresen. Buddlejaceae Tree 71 Ocimum sp. Lamiaceae herb 72 Olea europaea L. subsp. africana Mill. Oleaceae tree 73 Oplismenus hirtellus (L.) P. Beauv. Poaceae grass 74 Oxalis corniculata L. Oxalidaceae Herb 75 Panicum heterostachyum Hack. Poaceae grass 76 Panicum monticolum Hook. f. Poaceae grass 77 Panicum trichocladum Poaceae grass 78 Paullinia pinnata L. Sapindaceae Tree 79 Pennisetum mezianum Leeke Poaceae grass 80 Prunus africana (Hook.f.) Kalkman Rosaceae Tree 81 Psychotria cyathicalyx Petit Rubiaceae Tree 82 Ptilotrichum elliottii Brassicaceae Herb 83 Rauvolfia sp. Apocynaceae. Tree 84 Rhus natalensis Krauss Anacardiaceae Tree 85 Rhus vulgaris Anacardiaceae Tree 86 Rhynchosia hirta (Andrews) Meikle & Verdc. Fabaceae Herb 87 Rhynchosia minima (L.) DC Fabaceae Herb 88 Rhynchosia resinosa (Hochst. ex A. Rich.) Baker Fabaceae Shrub 89 Rinorea ilicifolia (Welw. ex Oliv.) Kuntze Violaceae tree 90 Rubia cordifolia L. Rubiaeae Tree 91 Ruellia tuberosa Acanthaceae Herb 92 Rytigynia uhligii (K. Schum. & K. Krause) Verdc. rubiaeae tree 93 Scutia myrtina (N. L. Burman) Kurz Rhamnaceae tree 94 Senecio abyssinicus Sch. Bip. Asteraceae Herb 95 Setaria homonyma (Steud.) Chiov. Poaceae grass 96 Setaria sphacelata Poaceae grass 97 Setaria verticillata Poaceae grass 98 Solanum incanum Solanaceae shrub 99 Spondias cytherea Sonn. Anacardiaceae tree 100 Sporobolus fimbriatus (Trin.) Nees Poaceae grass 101 Tarenna graveolens (S.Moore) Bremek. Rubiaceae tree 102 Teclea simplicifolia (Engl.) Mziray Rutaceae tree 103 Themeda triandra Forssk. Poaceae grass 104 Thunbergia alata Bojer ex Sims Acanthaceae Herb 105 Tinnea aethiopica Kotschy ex Hook.f. Lamiaceae Shrub 106 Toddalia asiatica (L.) Lam. Rutaceae wood climber 107 Tragia sp Euphorbiaceae shrub 108 Triumfetta rhomboidea Jacq. Malvaceae/Tiliaceae Shrub 109 Turraea holstii Gürke Meliaceae Tree 110 Turraea robusta Meliaceae Tree 111 Urena lobata L. Malvaceae Tree 112 Urtica mosaica Mildbr. Urticaceae Herb 113 Vangueria infausta Burch. Rubiaceae tree 114 Vernonia adoensis Sch.Bip. ex Walp. Asteraceae Herb 115 Zanthoxylum leprieurii Guill. & Perr. Rutaceae Tree 126 I.A. Kikoti and C. Mligo Appendix 2. Climate diagram of Lyamungu, Tanzania (average of 30 years from 1950 to 1980; FAO, 2001) Appendix 3. Photo showing montane forest of northern slopes of Mount Kilimanjaro International Journal of Biodiversity Science, Ecosystem Services & Management 127 Appendix 4. Site characteristics of study sites. Grazing intensity Altitude range Slope Ungrazed 2072–2094 7–15° Little grazed 1787–1902 7–15° Moderately grazed 1663–1730 10–15° Heavily grazed 1745–1760 7–10° http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png International Journal of Biodiversity Science, Ecosystem Services & Management Taylor & Francis

Impacts of livestock grazing on plant species composition in montane forests on the northern slope of Mount Kilimanjaro, Tanzania

Download PDF
14 pages

Loading next page...
 
/lp/taylor-francis/impacts-of-livestock-grazing-on-plant-species-composition-in-montane-aFrKQwsVr0

References (43)

Publisher
Taylor & Francis
Copyright
© 2015 Taylor & Francis
ISSN
2151-3732
eISSN
2151-3740
DOI
10.1080/21513732.2015.1031179
Publisher site
See Article on Publisher Site

Abstract

International Journal of Biodiversity Science, Ecosystem Services & Management, 2015 Vol. 11, No. 2, 114–127, http://dx.doi.org/10.1080/21513732.2015.1031179 Impacts of livestock grazing on plant species composition in montane forests on the northern slope of Mount Kilimanjaro, Tanzania a, b Imani A. Kikoti * and Cosmas Mligo a b Kilimanjaro National Park, P.O. Box 96, Marangu, Tanzania; Department of Botany, University of Dar es Salaam, P.O. Box 35060, Dar es Salaam, Tanzania (Submitted 10 April 2014; accepted 26 February 2015; edited by Patricia Balvanera) The montane forests on Mount Kilimanjaro are part of the ecologically protected area and provide ecosystem services to the local communities on the lower slopes. Despite their ecological value, the montane forests on the northern slope of Mount Kilimanjaro have been affected by increased livestock grazing pressure due to prolonged drought and expansion of agricultural land. We examined the impacts of livestock grazing on plant species composition in the montane forests on the northern slope of Mount Kilimanjaro. The study area was sub-divided into heavily grazed, moderately grazed, lightly grazed and ungrazed sites. A quadrat method was used for field data collection. A total of 115 plant species distributed within 93 genera and 39 families were identified. Moderately grazed and lightly grazed areas had higher plant species diversities than heavily and ungrazed areas. This observation concurs with the intermediate disturbance hypothesis. However, plant species that contributed to high species richness and diversity in grazed areas were mainly grasses, shrubs and herbs. It was concluded that livestock grazing in montane forests on the northern slope of Mount Kilimanjaro has serious impacts on vegetation community composition. Conservation of montane forest habitat and improvement of rangelands on communal land are necessary. Keywords: grazing ecology; Mount Kilimanjaro; plant species diversity; plant life form; tropical montane forest; vegetation communities 1. Introduction and Stern et al. (2002), livestock grazing leads to changes in floristic composition and structure within grazed areas. Livestock grazing in natural ecosystems results in changes Most studies have documented the impacts of livestock in vegetation structure and composition (Stern et al. 2002). grazing on vegetation in woodlands and grasslands (Hardy Changes in vegetation composition from palatable grasses et al. 1999; Mligo 2006; Sun et al. 2011; Cingolani et al. and sedges to less palatable forbs resulting from heavy 2013; Deng et al. 2014; Koerner & Collins 2014). There grazing have been reported in northwest China (Sun et al. has been little effort to cover knowledge gaps regarding 2011), Libya (Zatout 2014), North America (Bakker et al. the impacts of livestock grazing on plant species composi- 2003; Koerner & Collins 2014), South Africa (Koerner & tion in afromontane forests (Reed & Clokie 2000), one of Collins 2014) and Argentina (Cingolani et al. 2013). The which is the montane forest of Mount Kilimanjaro. The response of plant species richness varies according to northern slopes of Mount Kilimanjaro are inhabited by grazing intensity. For example, in grassland ecosystem, pastoral communities, particularly in the villages of Deng et al. (2014) reported that plant species richness Kamwanga, Irkaswa, Kitendeni and Lerangw’a in the increased with decreasing grazing intensity. The author Longido District. These villages had been sharing land observed the highest plant species richness at light and with wildlife in the former Longido Game Controlled moderate grazing intensities. Random grazing patterns can Area. However, major land-use changes have occurred lead to spatial heterogeneity in light availability, soil nutri- between 1952 and 2001 in the area (Noe 2003). These ent availability and vegetation community dynamics. This changes have been associated with increasing land area can reduce plant competition for environmental resources under cultivation and establishment of settlements, espe- because the vegetation exists in patches (Bakker et al. cially on the lower slopes of the northern side of Mount 2003). At the within-patch scale, light grazing intensity Kilimanjaro. For example, in 1952 there was only one can promote plant diversity, but heavy grazing intensity village in this area (the Kamwanga village); however, by can lead to the exclusion of the species intolerant to graz- 2001 three more villages, Lerangw’a, Irkaswa and ing (palatable), resulting in an increase in grazing tolerant Kitendeni, had been established in 1975, 1982 and 2001, plant species (unpalatable) (Watt & Gibson 1988; Calvert respectively (Noe 2003). Concomitant to the influx of 2001). In North America, light browsing resulted in sup- pastoralists to the area, there was also a migration of pression of dominant shrubs and maintenance of shrub non-pastoral people to the area in the past three decades, diversity (Pekin et al. 2014). According to Mligo (2006) *Corresponding author. Email: ikikoti@gmail.com © 2015 Taylor & Francis International Journal of Biodiversity Science, Ecosystem Services & Management 115 which has contributed to an increase in the amount of land adjacent to the Tanzania – Kenya international boundary, used for cultivation at the expense of pastoral land at latitude 2°45ʹ3°25ʹS and longitude 37°00ʹ37°43ʹE, (Muruthi & Frohardt 2000). Because of the decrease in approximately 300 Km south of the equator (Lambretchs grazing land for pastoralists, the Maasai community has et al. 2002). changed from pastoralism to agropastoralism. The general The climate of Mount Kilimanjaro varies from one practices of the aforementioned activities depend on sea- ecological zone to another based on altitudinal gradient sonal rains; however, there has been lack of short rains that and aspect. The northern slope of Mount Kilimanjaro are usually received between November and January in forms the leeward side, thus being drier than the southern recent years. The decrease and delay of long rains in the and eastern slopes. There are two distinct rainy seasons in study area has also been reported. At present, long rains this forest, a long rainy period occurring between March start in April and end in May instead of starting in March and June and short rainy periods between November and and ending in June as in normal years (Ndikumana 2007). December. Therefore, the climate of the area is bimodal. This has resulted in a shortage of forage for livestock and, The nearest weather station to the study area is in consequently, the expansion of grazing areas into montane Lyamungo, which is at approximately 50 km on the wes- forests at higher altitudes ranging from 1600 to 2600 m a. tern slope and receives a mean annual rainfall of 1714 mm s.l., particularly during dry seasons. In recent years, there with a mean temperature of 19.3°C (Appendix 2). has been an increase in the livestock population and graz- A preliminary survey was carried out to define vegeta- ing pressure in montane forests on the northern slopes of tion segments based on their homogeneity and grazing Mount Kilimanjaro (TANAPA 2006). The most preferred intensity. Random sampling procedures were used in locat- composition of livestock types driven into montane forests ing quadrats within each sampling site. The degree of live- on the northern slopes of Mount Kilimanjaro includes: stock trampling was assessed by measuring the percentage cattle, goats, sheep and donkeys. However, the montane (by area) of trail and bare ground in a 20 m × 25 m quadrat. forests are part of the Kilimanjaro National Park, such that Qualitative estimation of grazing intensity in each quadrat the wild communities have historically co-existed with the based on an assessment of percentage utilisation of vegeta- vegetative communities as habitat without putting one tion by livestock was also conducted in each quadrat community in jeopardy, and therefore, their foraging beha- (Rulangalanga 2000). A total of four grazing intensity viours have been considered natural. classes based on a modified Rulangalanga’s classification Livestock grazing may be affecting plant composition in system were established (Rulangalanga 2000)(Table 1). the montane forest on the northern slope of Mount Within each grazing intensity class, a detailed study on Kilimanjaro. The grazing intensity in this montane forest the plant species composition was carried out. Ten rectan- varies based on accessibility and the protection status of the gular quadrats of 20 m x 25 m were established in each forest. The lower montane forest, at approximately grazing intensity category, making a total of 40 quadrats in 16001800 m, is regarded as a heavily grazed area. In this the entire study area. In each quadrat, the trees with area, a Memorandum of Understanding between local com- Diameters at Breast Height (DBH) equal or greater than munities and the Forest and Beekeeping Department was 10 cm were identified to species, enumerated and measured. signed in 1963 to allow use of this area for livestock grazing Two (2 m × 5 m) subquadrats nested in bigger quadrats (of during the dry season. On the other hand, the upper part of the 20 × 25 m) as recommended by Stohlgren et al. (1995) were montane forest, with an approximate altitude of 18002400 m, established for sampling trees with DBHs less than 10 cm has no legal access for livestock grazing. However, it has been and shrubs. Additionally, grasses and herbaceous species grazed at different levels: moderately, lightly or ungrazed. were determined using three 1 m -quadrats randomly estab- Livestock grazing in the montane forest on the northern lished inside the bigger quadrats (20 m × 25 m), and the slope of Mount Kilimanjaro may be a threat to the conserva- number of individuals and percentage cover for each spe- tion of plant species and the associated economic and ecolo- cies was estimated in relation to the quadrat size. All plants gical values of the mountain. This study investigated the were identified to species level when it was possible in the impacts of livestock grazing on plant composition to under- field. For plant species that were not easily identified in the stand how livestock grazing pressure alters montane forest field, specimens were collected, pressed and taken to the communities on the northern slope of Mount Kilimanjaro. It herbarium in the Department of Botany, University of Dar was hypothesised that livestock grazing has negative impacts es Salaam (DSM) for identification by matching with her- on the plant community in the montane forests of Mount barium specimens and/or keying using relevant flora iden- Kilimanjaro. We compared plant species diversity in sites tification texts, such as Flora of Tropical East Africa and with different grazing intensities in the montane forests on Flora Zambeziaca. the northern slope of Mount Kilimanjaro. 2.1. Data analysis 2. Material and methods Plant species composition in each area was determined in This study was carried out in the montane forests on the terms of species diversity index, species evenness, and northern slope of Mount Kilimanjaro, which is found in species richness and plant density for all recorded plant Northern Tanzania (Figure 1). The study area is located species. 116 I.A. Kikoti and C. Mligo Figure 1. Map showing the location of study area in Mount Kilimanjaro. significance level for all tests was based on a 5% critical Table 1. Grazing intensity classes and their interpretation. limit. Correlation analysis was used to assess the influence of grazing intensity on plant species diversity. The classifi- Grazing intensity class Interpretation cation of vegetation communities was done using Two-Way 0 No grazing Indicator Species Analysis (TWINSPAN) based on live- 11–20%: Little grazed stock grazing disturbances (Hill 1979). 221–60%: Moderately grazed 361–100%: Heavily grazed Note: Modified from Rulangalanga (2000). 3. Results 3.1. Plant species composition in grazed areas The plant species diversity was determined from raw A total of 115 plant species distributed within 39 families and data using the Shannon’s diversity index (H′) (Shannon 93 genera were identified in this study (Appendix 1). The 1948), as shown in the formula below: commonly occurring plant species in ungrazed areas were Teclea simplicifolia, Achyranthus aspera, Asparagus africana and Cassipourea malosana, while in lightly grazed areas H ¼ ðp ln p Þ i i common species were Achyranthes aspera, Olea europaea i¼1 subsp. africana and Teclea simplicifolia (Table 2). In moder- ately grazed areas, Croton megalocarpus, Commelina ben- where p = n /N is the number of individuals found in the i i ghalensis, Setaria homonyma and Teclea simplicifolia were ith species as a proportion of the total number of indivi- represented with high stem densities, and the densely popu- duals found in all the species and ln the natural logarithm lated species in the heavily grazed areas were Euclea divi- base e. norum, Themeda triandra, Carissa spinarum and Rhus Before applying one-way analysis of variance natalensis (Table 2). The most dominant families with regards (ANOVA) to compare plant species diversity and richness to plant species composition were Poaceae, Acanthaceae, among areas with different grazing intensities, tests for Rutaceae, Amaranthaceae, Euphobiaceae, Meliaceae and normality (KolmogorovSmirnov D statistic) and homoge- Rubiaceae. neity of variance (Levene statistic) of richness and diversity Population structure and DBH size class distribution of indices for all recorded plant species were carried out. The trees varied considerably among the four studied areas. The International Journal of Biodiversity Science, Ecosystem Services & Management 117 Table 2. Density of dominant plant species recorded in areas with different grazing intensity (T = Tree, H = Herbaceous, ST = Small Tree, V = Vine and G = Grass). Density (Individuals/ha) Growth Plant species form Heavily Moderately Little Ungrazed Bersama abyssinica Fresen. T 0 31 36 58 Calodendron capense (L.f.) Thunb. T 0 7 10 6 Canthium lactescens Hiern. T 0 44 6 0 Cassipourea malosana Alston. T 0 17 23 72 Cassytha pondoensis var. schliebenii (Robyns & Wilczek) M.A. T0 2 8 0 Diniz Clausena anisata (Willd.) J.Hk. ex Benth. T 0 7 178 12 Croton megalocarpus Hutch. T 0 67 20 0 Diospyros abyssinica Hiern. T 0 24 2 26 Erythrococca bongensis Pax. T 0 6 36 16 Euclea divinorum Hiern T 98 48 29 0 Maytenus senegalensis T5 4 8 5 Maytenus undata T0 2 1 4 Olea europaea ssp. africana T 18 19 138 31 Teclea simplicifolia T 0 78 81 252 Turraea holstii T 0 12 41 46 Turraea robusta T 0 28 89 0 Vangueria infausta T4 5 60 11 Achyranthes aspera L. H 2 70 178 205 Allophyllus africanus P. Beauv. ST 0 4 17 3 Asparagus africanus Lam. V 0 4 11 9 Commelina benghalensis L. H 44 19 0 Crotalaria goodformis Vatke. ST 0 34 33 0 Hibiscus fuscus Garcke H 3 0 18 0 Hibiscus micranthus L.F. H 7 0 0 0 Hyparrhenia filipendula (Hochst.) Stapf G 37 0 0 0 Hypoestes aristata (Vahl) Sol. ex Roem. & Schult. H 0 0 160 69 Justicia flava (Vahl) Vahl H 4 20 19 0 least grazed and moderately grazed areas had large numbers lightly grazed areas and heavily and ungrazed areas of individuals within the DBH size classes of 10–40 cm (p > 0.05). Additionally, the highest plant species diversity relative to ungrazed and heavily grazed areas (Figure 2). (ShannonWiener (H′) index of 2.767) was recorded in However, old stand trees with diameter size classes greater moderately grazed areas, followed by lightly grazed, heav- than 100 cm were highly represented in ungrazed areas, and ily grazed and ungrazed areas (Table 4). The difference in these huge size classes were missing in other areas. Heavily the plant species diversity indices among areas with dif- grazed areas had only 40 stems/ha in the 10–20 cm DBH size ferent grazing intensities was significant based on an classes, while lacking trees with DBHs above 20 cm ANOVA (F = 32.87 and p < 0.05). (Figure 2). 3.3. Vegetation communities in the study area 3.2. Plant species richness and diversity Based on TWINSPAN, three vegetation community types Plant species richness was highest in moderately grazed (denoted as A, B and C) existed, as shown in the dendro- areas (21.3), followed by lightly, heavily and ungrazed gram (Figure 3). At the first level of TWINSPAN division, areas with mean species numbers of 18.70, 12 and 10.9, plant community C emerged with Themeda triandra (+) as respectively (Table 3). The difference among areas with the indicator of division. The eigenvalue of 0.7546 for this different grazing intensities was significant based on an division was relatively high, showing large differences in ANOVA (F = 35.7, p < 0.05). Multiple comparison tests species composition from the rest of the communities in among areas with different grazing intensities using the study area. Community A consisted of samples of Tukey’s Honestly Significant Difference showed that sig- vegetation data from ungrazed areas, which were com- nificant differences exist between moderately vs. heavily monly represented by Clutia abyssinica, Maytenus sene- grazed areas, moderately vs. ungrazed areas, lightly vs. galensis, Achyranthes aspera and Turraea holstii, among heavily grazed areas and lightly grazed vs. ungrazed areas others. The indicators of species composition differences (p < 0.05) (Table 3). Otherwise, no significant variations between species in communities A and B were Commelina in species richness were found between moderately and benghalensis (+), Turraea robusta (+), Prunus africana 118 I.A. Kikoti and C. Mligo Figure 2. DBH size class distribution in areas with different grazing intensities. Table 3. Variations in plant richness among areas with different grazing levels in Northern slopes of Mount Kilimanjaro. a, b, c, d (M = mean ± standard error and indicates significant difference at p < 0.05). a b c d Grazing Heavily Moderately Little grazed Ungrazed level (M = 12.0 ± 0.42) (M = 21.3 ± 1.03) (M = 18.7 ± 1.03) (M = 10.9 ± 0.75) a a a Heavily <0.001 <0.001 0.796095 b b Moderately 0.152067 <0.001 c c Little <0.001 Ungrazed Table 4. Variations in plant species diversity (ShannonWiener diversity index (H′)) among areas with different grazing levels in a, b, c, d montane forest of Northern slopes of Mount Kilimanjaro based on Tukey’s test (M = mean± Standard Error and letters indicates statistical level of significance at p < 0.05). a b c d Grazing Heavily Moderately Little grazed Ungrazed level (M = 2.2940 ± 0.03) (M = 2.7673 ± 0.05) (M = 2.5094 ± 0.08) (M = 1.9166 ± 0.08) a a a Heavily <0.001 0.090117 <0.001 b b b Moderately 0.030508 <0.001 c c Little <0.001 Ungrazed (−) and Cassipourea malosana (−). The presence of Vangueria infausta and Hypoestes aristata. Community Commelina benghalensis (+) and Turraea robusta (+) C was composed of plant species from heavily grazed implied that ecological changes were caused by distur- areas and was represented by Carissa spinarum and bance through livestock grazing. On the other hand, P. Cynodon dactylon. These two species were the character- africana (−)and Cassipourea malosana (−) were sensitive istics of a shrubby community and this implies that live- to disturbances through grazing and hence were indicators stock grazing pressure contributed to the observed change of communities undisturbed by grazing. The measure of of montane forest communities into shrubby communities. variation in species composition (eigenvalue of 0.49) indi- cates minor differences between communities A and B 3.4. The relationship between grazing intensities and from lightly and moderately vs. ungrazed areas. This plant species diversity means that community B was dominated by plant species that were also found in both the lightly and moderately A significant negative correlation was found between grazing grazed areas. The indicator species for moderately grazed intensity and species diversity (r = −0.4871, p = 0.0063) conditions were Clausena anisata, Celtis africana, (Figure 4). Although the correlation coefficient was not strong International Journal of Biodiversity Science, Ecosystem Services & Management 119 uP9 uP1 uP2 uP3 Clutia abyssinica Maytenus senegalensis Turraea holstii uP5 Achyranthes aspera uP7 uP4 uP6 uP10 uP8 Lp7 Prunus africana Lp8 Cassipourea malosana Lp10 Commelina benghalensis Turraea holstii Lp6 Turraea robusta Euclea divinorum Acalypha kilimandsc ica Lp9 Vernonia adoensis Lp3 Clausena anisata Lp4 Celtis africana Lp5 Vangueria infausta Lp1 Hypoestes aristata Lp2 mp3 mp5 Asystasia gangetica mp6 Abutilon mauritianum mp7 Themeda triandra Chloris pycnothrix mp8 mp9 mp10 mp1 mp2 mp4 Gutenbergia spp. hP1 Cynodon dactylon hP2 Conyza pyrrhopappa hP10 Olea europaea ssp. africana hP6 Carissa spinarum Senecio abyssinicus hP7 hP8 hP9 hP3 hP4 hP5 Figure 3. Vegetation community classification using TWINSPAN. (Lp = Little Grazed Plots; hP = Heavily grazed Plots; uP = Ungrazed Plots and mp = Moderately Grazed Plots). enough to explain the significant influence of grazing pressure in plant species composition among sampling sites was on species diversity, based on the montane nature of the forest, contributed by variation in livestock grazing intensities. the impacts of livestock grazing on diversity were very sig- nificant. This indicates that plant species are sensitive to dis- turbance through livestock grazing. Additionally, the montane 4. Discussion forest plant species are not adaptedtoco-exist with livestock 4.1. The influence of grazing pressure on plant species grazing. This implies that at any level of grazing intensity plant diversity species diversity in montane forests will decrease. We observed higher plant species richness and diversity in moderately grazed and lightly grazed than in heavily grazed and ungrazed areas. Ungrazed areas were charac- 3.5. The response of different plant life forms to terised by old growth stands with DBHs ranging from 10 livestock grazing pressure to 310 cm. The low diversity of plant species in ungrazed There were differences in plant life forms and plant size areas may have been caused by a few dominant tree stands structures among areas with different grazing intensities. A that usurp the lion’s share of the habitat resources (nutri- high composition of tree species was found in moderately, ents and light). Many studies have reported that a few ungrazed and lightly grazed areas. Heavily grazed areas had plant species favoured by lack of grazing tend to out- the lowest tree density and averaged 0.8 tree species per plot compete plants with smaller statures (Belsky 1992; Pekin with a 0.19 Shannon’s diversity index (Table 5). Despite the et al. 2014). On the other hand, lower plant species diver- lower tree species density in heavily grazed areas, these sities were recorded in the heavily grazed areas than in areas had higher compositions of shrubs and grass species moderately and lightly grazed areas. This implies that the than ungrazed areas. The heavily grazed areas were com- increase in livestock grazing intensity resulted in decreases monly dominated by Themeda triandra, Setaria sphace- in species diversity and abundances in the montane forests lata, Hyparrhenia filipendula, Heteropogon contortus and of northern slope of Mount Kilimanjaro. High plant spe- Dichanthium annulatum. The herbaceous species were cies diversity in moderately and lightly grazed areas may represented by Senecio abyssinicus, Conyza pyrrhopappa, be due to the effects of livestock grazing that results in Gutenbergiaspp., Ruellia tuberose and Thunbergia alata, opening of the canopy, thus giving opportunity for regen- whereas shrubs were represented by Carissa spinarum, eration by gap opportunistic plant species (Pekin et al. Indigofera volkensii, Leucas deflexa, Lantana trifolia, 2014). Livestock grazing may also have reduced competi- Solanum incanum and Hibiscus micranthus. The variation tion among plant species through selective grazing on 120 I.A. Kikoti and C. Mligo Figure 4. The influence of grazing intensities on plant species diversity in montane forests of Northern slopes of Mount Kilimanjaro. Table 5. Plant life forms recorded in areas with different grazing level in montane forest of northern slopes of Mount Kilimanjaro (SN = Species number, H′ = ShannonWiener diversity index). Trees Shrubs Herbs and grasses Level of grazing Mean SN Mean H′ Mean SN Mean Hʹ Mean SN Mean Hʹ Heavily 0.8 ± 0.29 0.19 ± 0.09 4 ± 0.25 0.19 ± 0.55 8.5 ± 0.42 1.6 ± 0.09 Little 3.7 ± 0.52 0.95 ± 0.14 6.3 ± 0.63 1.55 ± 0.95 9.6 ± 0.66 1.9 ± 0.08 Moderately 5.2 ± 0.57 1.26 ± 0.15 6.3 ± 0.26 1.58 ± 0.26 14.5 ± 0.63 2.3 ± 0.06 Ungrazed 3.7 ± 0.51 1.00 ± 0.11 3.8 ± 0.53 1.02 ± 0.38 5.2 ± 0.53 1.49 ± 0.12 palatable competitors as well as trampling of both unpala- competitively dominant species from excluding other spe- table and palatable plants during grazing in the montane cies from the community (Markey & Currie 2001; Catford forest (Rooney & Waller 2003). This finding is in agree- et al. 2012). This brings about a trade-off between plant ment with the intermediate disturbance hypothesis pro- species ability to compete and tolerate various forms of posed by Connell (1978). Models and metadata analysis disturbance. Species diversity is low at extremely low have indicated that species richness and the Shannon levels of disturbance because only the best competitors Wiener diversity index are strong predictors of the inter- dominate and persist within community (Connell 1978). mediate disturbance hypothesis (Svensson et al. 2012). This concept is in agreement with the findings from this The mechanism underlying the intermediate disturbance study, in which the habitat where disturbance was low hypothesis is centred on a complex interplay between life displayed low plant species diversity. However, in the history, biotic interaction and historical disturbance regime severely and the highly disturbed areas only a few species (Catford et al. 2012). The increased availability of plant persisted or repeatedly colonised after every similar requirements, such as light, following disturbances regime of disturbance, thus resulting in low species diver- through livestock grazing may explain why high diversi- sities (Connell 1978). This concept also applies to the ties were observed in moderately and lightly grazed areas findings from this study in which heavily grazed areas in the montane forests on the northern slope of Mount had lower species diversities than moderately grazed Kilimanjaro. According to Roberts and Gilliam (2003), areas. The overgrazed areas were dominated by plants intermediate disturbance causes changes in local microcli- unpalatable to livestock, including a few shrubs that per- mates by opening up space in the canopy, resulting in the formed under extremely modified habitat conditions. release of resources that would otherwise not be accessible Therefore, the balance between competitive exclusion to understory plants. Physical disturbances prevent and the loss of competitive dominants through disturbance International Journal of Biodiversity Science, Ecosystem Services & Management 121 is attained at intermediate disturbances (Markey & Currie total P. africana regeneration potential due to grazing and 2001). At the highest peak of species diversity, conditions browsing in afromontane forests in southern Ethiopia. favour the coexistence of both the competitive species and Currently, P. africana is included on the IUCN Red List disturbance-tolerant species (Connell 1978). of Threatened Species and is categorised as a vulnerable The purpose of Kilimanjaro National Park is to protect species (WCMC 1998). Another tree species that was Africa’s highest and one of the world’s largest free stand- restricted to ungrazed and lightly grazed areas was ing mountains and to conserve its unique socio-economic, Cassipourea malosana. The indicator plant species of cultural and ecological values and features of the fragile favourable habitat in the lightly and moderately grazed mountain ecosystem (TANAPA 2006). On that basis, areas in cluster B were Celtis africana, Vangueria maintenance of the status quo in the montane forest is of infausta, Clausina anisata and Hypoestes arisata. Some high priority, as it is an important regional water catch- of these species (for example, Celtis africana) are vulner- ment for people outside the park. The plant species diver- able to livestock grazing and browsing. The mortality of sity observed in the grazed areas was mainly due to the C. africana due to grazing observed in the southern dominance of shrubs, grasses and herbs. Therefore, the Ethiopian Afromontane forest cannot be excluded from observed changes in species composition due to cattle the heavily grazed areas of the northern slopes of Mount grazing in the montane forests were contributed by the Kilimanjaro because there was no representation of indi- cover abundances of weeds and shrubs, which were char- viduals of C. africana. Cluster C consisted of plant com- acteristic in the heavily grazed areas. Therefore, grazing munities from heavily grazed areas with indicator species may lead to an increase in the number of plant species and of plant communities that are resilient to livestock distur- the overall plant species diversity of an area, although the bances, such as Carissa spinarum, (shrub), Senecio abys- additional species may be either weeds or early colonisers sinica (herb) and Cynodon dactylon (grass). However, the (Landsberg et al. 2003). Catford et al. (2012) identified presence of remnants of the forest-dependent Olea euro- five main ways in which alien species may reduce local paea subsp. africana indicates transformed/degraded mon- diversity along disturbance gradients. These include: niche tane forest is present on the northern slope of Mount pre-emption, apparent competition, interference competi- Kilimanjaro. Cynodon dactylon is a grass species that tion, exploitative competition and transformation of the performs well in heavily disturbed habitats. The perfor- environment. Despite the increase in diversity of shrubs mance of C. dactylon in heavily grazed areas is possible and weeds in heavily grazed areas of the montane forest, because of its extensive stolon and rhizome system, which these species may alter ecosystem processes (i.e., erosion provide a means of rapid expansion, allowing it to thrive rates and seasonal flows) to which native montane tree well in overgrazed areas. During this study, this species species are adapted. was restricted only to heavily grazed areas in the montane forests. Similar observations indicating a high abundance of C. dactylon in over-grazed areas were reported in the 4.2. Variation in plant species composition in Ethiopian highlands (Mwendera et al. 1997) and Southern vegetation communities with different levels of Maasailand of Kenya (Kioko et al. 2012). Among the grazing dominant shrubs found within overgrazed communities was Carissa spinarum, which is also a drought tolerant From TWINSPAN, three clusters of vegetation commu- species and performs in heavily grazed areas. Bahiru nities, denoted as A, B and C, were clearly observed. Plant (2008) observed a high frequency of C. Spinarum in species communities in ungrazed and lightly grazed areas grazed areas in Ethiopia because it is less preferred by were characterised by indicator species that are not toler- livestock. The avoidance of C. Spinarum by livestock ant of anthropogenic disturbances, particularly livestock grazing. For example, the African cherry (P. africana)is most likely is due its unpalatability due to its high percen- typically restricted to afromontane forest (White 1983). tage of tannins, which range from 9% to 15% in its leaves During fieldwork, this species was only recorded in (Parmar & Kaushal 1982). ungrazed areas of the montane forests on the northern Tannins are polyphenolic compounds with great struc- slope of Mount Kilimanjaro that had been influenced by tural diversities that decrease the digestibility of protein low-intensity human land use for a very long period. This (Giner-Chavez 1996). Due to the physiological difficulties species was not found in grazed areas where sampling was most herbivores experience in their digestion and the carried out, presumably due to livestock-related distur- differences in the chemical reactivity of tannins, they are bances, and can therefore be considered a potentially sen- not preferred by grazing and browsing animals (Aganga sitive plant species and indicator of late succession in the et al. 1997). The thorny characteristics of C. spinarum montane forest. Stewart (2009) observed declines in popu- make it resistant to herbivore grazing and browsing lations of P. africana due to anthropogenic activities impacts. including bark harvesting, wildfires and livestock grazing In heavily grazed areas, there has been a visible change in in Cameroon. Substantial evidence indicates that livestock the vegetation community from montane forest to shrubland. grazing is among the causative agents for regeneration Only three tree species with DBH>10 cm, Euclea divinorum, failure of P. africana in Mount Oku, Cameroon (Stewart Olea europaea ssp. africana and Acacia nilotica, were 2009). Abebe (2008) also observed a loss of 42–61% of observed in heavily grazed areas of the montane forests. 122 I.A. Kikoti and C. Mligo Aganga AA, Tsopito CM, Moroke KM. 1997. Tannin content of Livestock grazing changes plant structure and species com- some indigenous browse of Botswana. In: Proceeding of position in mountainous ecosystems. The observation of XVII International Grassland Congress, ID No 136; only a few woody species and the small diameters of the Winnepeg, Manitoba, Canada. trees in the heavily grazed areas indicate negative impacts of Bahiru KM. 2008. Enclosure as a viable option for rehabilitation livestock grazing on the montane forest on the Northern of degraded lands and biodiversity conservation: the case of Kallu Woreda, Kallu Woreda, Southern Wello [MSc thesis]. slope of Mount Kilimanjaro. Therefore, livestock grazing in Addis Ababa University, Addis Ababa. the montane forest of Mount Kilimanjaro has strong implica- Bakker C, Blair JM, Knapp AK. 2003. Does resource availabil- tions for ecosystem integrity and sustainability. In Central ity, resource heterogeneity or species turnover mediate Argentina, heavy grazing pressure was blamed for convert- changes in plant species richness in grazed grasslands? ing mountain ecosystems into rocky deserts through elimina- Oecologia. 137:385–391. Belsky AJ. 1992. Effects of grazing, competition, disturbance tion of woodlands (Cingolani et al. 2013). and fire on species composition and diversity in grassland communities. J Veg Sci. 3:187–200. doi:10.2307/3235679 Calvert GA. 2001. The effects of cattle grazing on vegetation 5. Conclusion diversity and structural characteristics in the semi arid Livestock grazing significantly influences plant species lands of North Queensland [PhD thesis]. James Cook diversity in areas with different grazing intensities on the University, North Queensland. northern slope of Mount Kilimanjaro. Moderately grazed Catford JA, Daehler CC, Murphy HT, Sheppard AW, Hardesty BD, Westcott DA, Rejmánek M, Bellingham PJ, Pergl J, areas had the highest plant species diversity, complying Horvitz CC, Hulme PE. 2012. The intermediate disturbance with the intermediate disturbance hypothesis. However, hypothesis and plant invasions: implications for species rich- low tree species richness and a lack of advanced aged stands ness and management. Perspect in Plant Eco Evol and Syst. in heavily grazed areas indicated that the impacts of live- 14:231–241. Cingolani AM, Cabido MR, Renison D, Neffa VS. 2013. Combined stock grazing prevented trees from attaining large size effects of environment and grazing on vegetation structure in classes. Grazing pressure is a function of the ecological argentine granite grasslands. J Veg Sci. 14:223–232. variation existing in the montane forest in the study area Connell JH. 1978. Diversity in tropical rain forests and coral and the plant species responses to the aforementioned reefs. Science. 199:1302–1310. anthropogenic disturbance. Conservation of montane for- Deng L, Sweeney S, Shangguan Z-P. 2014. Grassland responses to grazing disturbance: plant diversity changes with grazing inten- ests is important for the protection of habitat for species sity in a desert steppe. Grass and Forage Sci. 69:524–533. with different conservation statuses (i.e., rare, endemic and [FAO] Food and Agriculture Organization of United Nations. threatened species). Despite the fact that moderate grazing 2001. FAOClim 2.0. Agroclimatic database CD-ROM + promoted overall plant diversity in the study area, an endan- users manual (72 pp.) Environment and Natural Resources gered species, African cherry (Prunus africana), was Working paper. No 5. Available from: http://www.fao.org/nr/ climpag/pub/en1102_en.asp. severely affected by all levels of grazing, reducing the Giner-Chavez BI. 1996. Condensed tannins in tropical forages ecosystem services the montane forest provides to the [PhD thesis]. Ithaca (NY): Cornell University. local community. Therefore, it is recommended that the Hardy MB, Barnes DL, Moore A, Kirkman KP. 1999. The montane forest be protected from livestock grazing management of different types of veld. In: Tainton NM, encroachment for conservation of biodiversity, the editor. Veld management in South Africa. Pietermaritzburg: watershed catchment and the soil. The park management University of Natal Press. Hill MO. 1979. TWINSPAN A FORTRAN program for arran- needs to collaborate with pastoralists in the northern part of ging multivariate data in an ordered two-way table by classi- Mount Kilimanjaro to observe the boundaries of the pro- fication of the individuals and attributes. Ithaca (NY): tected area that include the montane forests. This may Ecology and Systematics, Cornell University. ensure a protected water catchment for providing a sustain- Kioko J, Kiringe JW, Seno SO. 2012. Impacts of livestock grazing able water supply for wildlife as a means of biodiversity on a savanna grassland in Kenya. J Arid Land. 4:29−35. Koerner SE, Collins SL. 2014. Interactive effects of grazing, conservation and to satisfy the water requirements of live- drought, and fire on grassland plant communities in North stock outside of the protected area (montane forests) on the America and South Africa. MedMed (Ecology). 95:98–109. northern slope of Mount Kilimanjaro. In this collaboration, Lambretchs C, Woodley B, Hemp A, Hemp C, Nnyiti P. 2002. the park may provide technical assistance in establishing Aerial survey of the threats to Mount Kilimanjaro Forests. Dar es Salaam: UNDP; 33 pp. watering points outside the montane forest to prevent access Landsberg J, James CD, Morton SR, Müller WJ, Stol J. 2003. by pastoralists to the montane forest habitats. Abundance and composition of plant species along grazing gradients in Australian rangelands. J Appl Ecol. 40:1008–1024. Markey RC, Currie DJ. 2001. The diversity – disturbance bal- Disclosure statement ance relationship: is it generally strongly peaked? Ecology. No potential conflict of interest was reported by the authors. 82:3429–3492. Mligo C. 2006. Effects of grazing pressure on plant species composition and diversity in the semi-arid rangelands of References Mbulu district, Tanzania. Agric J. 1:277–283. Abebe GT. 2008. Ecology of regeneration and phenology of seven Muruthi P, Frohardt K. 2000. Study on the development of indigenous tree species in a dry tropical Afromontane forests, transboundary natural resource management areas in Africa. Southern Ethiopia [PhD dissertation submitted at School of Kilimanjaro Heartland case study. Nairobi, Kenya: African Graduate studies]. Addis Ababa: Addis Ababa University. Wildlife Foundation (AWF). International Journal of Biodiversity Science, Ecosystem Services & Management 123 Mwendera EJ, Saleem MAM, Woldu Z. 1997. Vegetation Shannon C. 1948. A mathematical theory of communication. response to cattle grazing in the Ethiopian highlands. Agric Bell Syst Tech J. 27:379–423. Ecosyst Environ. 64:43–51. Stern M, Quesada M, Stoner KE. 2002. Changes in composi- Ndikumana J. 2007. Livestock situation analysis in North tion and structure of a tropical dry forest following inter- East Tanzania. Vol. 1. Nairobi: Pastoralist Bulletin. A- mittent cattle grazing. Int J Trop Biol. 50:1021–1034. AARNET/International Livestock Research Institute. Stewart K. 2009. Effects of bark harvest and other human Noe C. 2003. The dynamic of land use changes and their impacts on activity on populations of the African cherry (Prunus wildlife corridor between Mount Kilimanjaro and Amboseli africana) on Mount Oku, Cameroon. For Ecol Manage. National Park, Tanzania. Land Use Change, Impacts and 258:1121–1128. Dynamics (LUCID) Project Working Paper No. 31; Nairobi: Stohlgren TJ, Falkner MB, Schell LD. 1995. A modified-whit- International Livestock Research Institute. taker nested vegetation sampling method. Vegetatio. Parmar C, Kaushal MK. 1982. Carissa spinarum [Internet]. In: 117:113–121. Wild Fruits. New Delhi, India: Kalyani Publishers; p. Sun DS, Wesche K, Chen DD, Zhang SH, Wu GL, Du GZ, 15–18; [cited 2012 Aug 3]. Available from: http://www. Comerford NB. 2011. Grazing depresses soil carbon storage hort.purdue.edu/newcrop/parmar/04.html through changing plant biomass and composition in a Pekin BK, Wisdom MJ, Endress BA, Naylor BJ, Parks CG. Tibetan alpine meadow. Plant Soil and Env. 57:271–278. 2014. Ungulate browsing maintains shrub diversity in the Svensson JR, Lindegarth M, Jonsson PR, Pavia H. 2012. absence of episodic disturbance in seasonally-arid conifer Disturbance–diversity models: what do they really predict forest. PLoS One. 9:e86288. doi:10.1371/journal. and how are they tested? Proc Royal Soc B: Biol Sci. pone.0086288 279:2163–2170. Reed MS, Clokie MRJ. 2000. Effects of grazing and cultivation [TANAPA] Tanzania National Parks. 2006. General management on forest plant communities in Mount Elgon National Park, plan for Kilimanjaro National Park (2006–2015). Arusha: Uganda. Afr J Ecol. 38:154–162. Tanzania National Parks/UNDP/CAWM. Roberts MR, Gilliam FS. 2003. Response of the herbaceous Watt TA, Gibson WD. 1988. The effects of sheep grazing on layer to disturbance in eastern forests. In: Gilliam FS, seedling establishment and survival in grassland. Vegetatio. Roberts MR, editors. The herbaceous layer in forests of 78:91–98. eastern North America. New York (NY): Oxford Press; p. White F. 1983. The vegetation of Africa: a descriptive memoir to 302–320. accompany the UNESCO/AETFAT/UNSO vegetation map Rooney TP, Waller DM. 2003. Direct and indirect effects of of Africa by F. White. Paris: UNESCO. white-tailed deer in forest ecosystems. For Ecol Manage. [WCMC] World Conservation Monitoring Centre. 1998. Prunus 181:165–176. africana [Internet]. The IUCN Red List of Threatened Rulangalanga ZK. 2000. Obstacles to the sustainable utiliza- Species. Version 2014.3. [cited 2015 Apr 1]. Available from: tion of the vegetation of semi-arid areas of Mbulu district www.iucnredlist.org for animal husbandry. Paper presented at a Workshop on Zatout MMM. 2014. Effect of negative human activities on plant Current Initiative in Sustainable Management of Dryland diversity in the Jabal Akhdar pastures. Int J Bioassays. Biodiversity, Arusha, Tanzania. 3:3324–3328. 124 I.A. Kikoti and C. Mligo Appendix 1. List of plant species recorded in Northern slopes of Mount Kilimanjaro. Plant species Family Growth form 1 Abutilon angulatum (Guill. & Perr.) Mast. Malvaceae shrub 2 Abutilon mauritianum (Jacq.) Medik. Malvaceae shrub 3 Acacia nilotica (L.) Delile Fabaceae Tree 4 Achyranthes aspera L. Amaranthaceae herb 5 Ageratum conyzoides L. Asteraceae herb 6 Allophyllus africanus P. Beauv. Sapindaceae tree 7 Asparagus africanus Lam. Asparagaceae Vine 8 Asparagus falcatus L. Asparagaceae Climbing 9 Asystasia gangetica (L.) T.Anderson Acanthaceae herbaceous 10 Asystasia schimperi T. Anderson Acanthaceae herbaceous 11 Bersama abyssinica Fresen. Melianthaceae Tree 12 Bidens pilosa L. Asteraceae herb 13 Bidens schimperi Walp. Compositae herb 14 Calodendron capense (L.f.) Thunb. Rutaceae Tree 15 Canthium lactescens Hiern. Rubiaceae Tree 16 Capparis tomentosa Lam. Capparidaceae shrub 17 Casearia battiscombei Flacourtiaceae Tree 18 Carissa spinarum Linn. Apocynaceae Shrub 19 Cassipourea malosana Alston. Rhizophoraceae Tree 20 Cassytha pondoensis var. schliebenii (Robyns & Wilczek) M.A.Diniz Lauraceae Tree 21 Celtis africana N.L.Burm Cannabaceae tree 22 Chloris pycnothrix Poaceae grass 23 Clausena anisata (Willd.) J.Hk. ex Benth. Rutaceae Tree 24 Clerodendrum myricoides Lamiaceae Shrub 25 Clutia abyssinica Jaub. & Spach Peraceae tree 26 Coccinia adoensis (A. Rich.) Cogn. Cucurbitaceae herb 27 Coccinia grandis (L.) J.Voigt Cucurbitaceae herb 28 Commelina africana L. Commelinaceae herb 29 Commelina benghalensis L. Commelinaceae herbaceous 30 Conyza pyrrhopappa Sch.Bip. ex A.Rich Compositae Herb 31 Crotalaria goodformis Vatke. Papilionaceae Herb 32 Croton megalocarpus Hutch. Euphorbiaceae Tree 33 Cyathula cylindrica Moq. Amaranthaceae Herb 34 Cynodon dactylon (L.) Pers. Poaceae grass 35 Dactyloctenium germinatum Hack. Poaceae grass 36 Dichanthium annulatum (Forssk.) Stapf Poaceae grass 37 Digitaria abyssinica Hochst. ex A. Rich.) Stapf Poaceae grass 38 Diospyros abyssinica Hiern. Ebenaceae Tree 39 Eragrostis volkensii Pilg. Poaceae grass 40 Erythrococca bongensis Pax. Euphorbiaceae Tree 41 Euclea divinorum Hiern Ebenaceae Tree 42 Euphobia hirta Euphorbiaceae Herb 43 Ficus thonningii Blume Moraceae Tree 44 Galinsoga parviflora Cav. Asteraceae Herb 45 Grewia similis K. Schum. Malvaceae shrub 46 Gutenbergia spp Sch. Bip. Asteraceae shrub 47 Heteropogon contortus (L.) P.Beauv.ex Roem.& Schult. Poaceae grass 48 Hibiscus fuscus Garcke Malvaceae Shrub 49 Hibiscus micranthus L.F. Malvaceae Shrub 50 Hyparrhenia filipendula (Hochst.) Stapf Poaceae grass 51 Hypoestes aristata (Vahl) Sol. ex Roem. & Schult. Acanthaceae Herb 52 Hypoestes forskalii (Vahl) Roem. & Schult. Acanthaceae Herb 53 Hypoestes triflora Acanthaceae Herb 54 Indigofera arrecta Fabaceae. shrubs 55 Indigofera cuneata Baker ex Oliv. Fabaceae. shrubs 56 Indigofera volkensii Fabaceae. shrubs 57 Jasminum sp. Oleaceae Herb 58 Justicia flava (Vahl) Vahl Acanthaceae Herb 59 Lantana trifolia Verbenaceae Shrubs 60 Leonotis nepetifolia (L.) R.Br. Lamiaceae shrub 61 Leucas deflexa Hook. f. Lamiaceae shrub 62 Leucas grandis Vatke Lamiaceae shrub 63 Lippia javanica (Burm.f.) Spreng Verbenaceae shrub 64 Lippia triphylla Verbenaceae shrub (Continued) International Journal of Biodiversity Science, Ecosystem Services & Management 125 Appendix 1. (Continued). Plant species Family Growth form 65 Macrotyloma axillare (E. Mey.) Verdc Fabaceae Herb 66 Malacantha alnifolia (Baker) Pierre Sapotaceae shrub 67 Maytenus senegalensis (Lam.) Excell Celastraceae Tree 68 Maytenus undata (Thunb.) Blakelock Celastraceae Tree 69 Monechma debile (Forssk.) Nees Acanthaceae shrub 70 Nuxia congesta R.Br. ex Fresen. Buddlejaceae Tree 71 Ocimum sp. Lamiaceae herb 72 Olea europaea L. subsp. africana Mill. Oleaceae tree 73 Oplismenus hirtellus (L.) P. Beauv. Poaceae grass 74 Oxalis corniculata L. Oxalidaceae Herb 75 Panicum heterostachyum Hack. Poaceae grass 76 Panicum monticolum Hook. f. Poaceae grass 77 Panicum trichocladum Poaceae grass 78 Paullinia pinnata L. Sapindaceae Tree 79 Pennisetum mezianum Leeke Poaceae grass 80 Prunus africana (Hook.f.) Kalkman Rosaceae Tree 81 Psychotria cyathicalyx Petit Rubiaceae Tree 82 Ptilotrichum elliottii Brassicaceae Herb 83 Rauvolfia sp. Apocynaceae. Tree 84 Rhus natalensis Krauss Anacardiaceae Tree 85 Rhus vulgaris Anacardiaceae Tree 86 Rhynchosia hirta (Andrews) Meikle & Verdc. Fabaceae Herb 87 Rhynchosia minima (L.) DC Fabaceae Herb 88 Rhynchosia resinosa (Hochst. ex A. Rich.) Baker Fabaceae Shrub 89 Rinorea ilicifolia (Welw. ex Oliv.) Kuntze Violaceae tree 90 Rubia cordifolia L. Rubiaeae Tree 91 Ruellia tuberosa Acanthaceae Herb 92 Rytigynia uhligii (K. Schum. & K. Krause) Verdc. rubiaeae tree 93 Scutia myrtina (N. L. Burman) Kurz Rhamnaceae tree 94 Senecio abyssinicus Sch. Bip. Asteraceae Herb 95 Setaria homonyma (Steud.) Chiov. Poaceae grass 96 Setaria sphacelata Poaceae grass 97 Setaria verticillata Poaceae grass 98 Solanum incanum Solanaceae shrub 99 Spondias cytherea Sonn. Anacardiaceae tree 100 Sporobolus fimbriatus (Trin.) Nees Poaceae grass 101 Tarenna graveolens (S.Moore) Bremek. Rubiaceae tree 102 Teclea simplicifolia (Engl.) Mziray Rutaceae tree 103 Themeda triandra Forssk. Poaceae grass 104 Thunbergia alata Bojer ex Sims Acanthaceae Herb 105 Tinnea aethiopica Kotschy ex Hook.f. Lamiaceae Shrub 106 Toddalia asiatica (L.) Lam. Rutaceae wood climber 107 Tragia sp Euphorbiaceae shrub 108 Triumfetta rhomboidea Jacq. Malvaceae/Tiliaceae Shrub 109 Turraea holstii Gürke Meliaceae Tree 110 Turraea robusta Meliaceae Tree 111 Urena lobata L. Malvaceae Tree 112 Urtica mosaica Mildbr. Urticaceae Herb 113 Vangueria infausta Burch. Rubiaceae tree 114 Vernonia adoensis Sch.Bip. ex Walp. Asteraceae Herb 115 Zanthoxylum leprieurii Guill. & Perr. Rutaceae Tree 126 I.A. Kikoti and C. Mligo Appendix 2. Climate diagram of Lyamungu, Tanzania (average of 30 years from 1950 to 1980; FAO, 2001) Appendix 3. Photo showing montane forest of northern slopes of Mount Kilimanjaro International Journal of Biodiversity Science, Ecosystem Services & Management 127 Appendix 4. Site characteristics of study sites. Grazing intensity Altitude range Slope Ungrazed 2072–2094 7–15° Little grazed 1787–1902 7–15° Moderately grazed 1663–1730 10–15° Heavily grazed 1745–1760 7–10°

Journal

International Journal of Biodiversity Science, Ecosystem Services & Management – Taylor & Francis

Published: Apr 3, 2015

Keywords: grazing ecology; Mount Kilimanjaro; plant species diversity; plant life form; tropical montane forest; vegetation communities

There are no references for this article.