PLANT CHECK LIST BY JOSHUA MUASYA AND TRUMAN YOUNG Joshua M. Muasya & David N. Okebiro
East African Herbarium, National Museums of Kenya
P. O. Box 40658, Nairobi, Kenya

Truman P. Young
Louis Calder Center
Fordham University, Box K Armonk NY 10504, USA

The Ol Ari Nyiro/Mukutan Gorge area is one of the most diverse non-forest areas in East Africa. It is home to The Gallmann Memorial Foundation, which has strived to keep it in as natural a state as possible, and which sponsors ecological research on the semi-arid ecosystems of north-central Kenya (Young & Francombe 1991; Oloo et al., manuscript).

Ol Ari Nyiro is a 360 km2 (approximately 90,000 acres) ranch located on the Laikipia plateau in the Republic of Kenya. It lies between longitudes 36° 17' and 36° 29' E and between latitudes 00° 29' and 00° 42'N. Elevations range from 1260m at the base of the gorge to 2200m in the hills to the southwest. The diverse topography supports a variety of vegetation types from dry marginal forest to semi-arid bushland. Mean annual rainfall over the past 30 years has been 790mm at the ranch headquarters (Young & Francombe 1991). Mean yearly maximum and minimum temperatures are approximately 30° and 14°C, respectively (Odinga 1971).

Ol Ari Nyiro is a valuable remnant of the diverse flora and fauna that once covered vast areas of the Laikipia Plateau and the eastern Rift Valley escarpment. Our goal was to produce a preliminary plant checklist and a description of the vegetation of Ol Ari Nyiro/Mukutan Gorge to assist in future conservation, management, and development plans for the area.

This survey comprised the entire Ol Ari Nyiro Ranch, including the upper Mukutan Gorge, and those parts of the lower Mukutan Gorge that lie outside the boundaries of the ranch. The Laikipia plateau was the result of uplift 20-25 million years ago, producing a broad high plain between the future sites of Mount Kenya and the Aberdare Mountains. The highest point on the ranch is 2200 m above sea level, and most of the ranch lies between 1800 and 2000 m.

Faulting and folding of rocks during the formation of the Rift Valley is thought to have produced faultlines in the Mukutan area, leading to weaknesses and sinking. The Mukutan River was possibly redirected along such lines of weakness, further deepening and widening them by erosion. There are a number of side gorges running perpendicular to the main gorge, and parallel to exposed fault lines. The Mukutan River flows through the gorge from east to west, ultimately reaching Lake Baringo 20 km to the west. The mouth of the gorge is 1260 m above sea level, rising to 1780 m at the eastern end. At the deepest point, the gorge walls are ~600 m high.

PLANT CHECKLIST:

This plant list is a compilation based on collections made between 1986 and 1991 by several individuals:

Ann Robinson - AR - 15 specimens
Collin Leakey - CL - 20 specimens
David Okebiro - DO - 214 specimens
Joshua Muasya - JM - 1223 specimens
Timothy Oloo - TO - 100 specimens
Truman P. Young - TPY - 268 specimens

The vascular plants listed here include 711 species and subspecies; four Fungi, ten Pteridophyta, one Gymnosperm, 568 Dicotyledons, and 128 Monocotyledons. No attempt was made to collect non-vascular plants. The species on this list represent ten percent of the estimated flora of the Republic of Kenya (Beentje, pers. comm.), making Ol Ari Nyiro one of the most diverse non-forest areas in East Africa.

This list is not exhaustive, it records probably not more than 80% of plant species occurring study area. Areas like the Mukutan Gorge and much of the western aspect of the ranch are inaccessible due to their steep slopes, though further collecting trips by Museum personnel are planned. An estimated total flora of over 800 species is not unreasonable. One new species was found, later named Aloe francombei. Identifications were confirmed at the Herbarium of National Museums of Kenya, with the help of Flora of Tropical Africa (Polhill 1957-1989), Agnew (1974; UKWF hereafter) and Dale & Greenway (1954). The sequence and numbering of families is that of Hutchinson (1934). This is the sequence used in The Flora of West Tropical Africa, and in the East African Herbarium.

For each species entry, one collector1s number (except for sight records - SR) is given for reference to a collection in the East African Herbarium, although more than one specimen may have been collected. In the list below, most of the collections by DO were made at the base of the Mukutan gorge.

FOOD PLANTS OF THE BLACK RHINOCEROS (DICEROS BICORNIS) ON OL ARI NYIRO RANCH, LAIKIPIA, KENYA

Published by Afr. J. Ecol. 1994, Volume 32, pages 142-157

Tim Oloo, Rob Brett & Truman P. Young
The Gallmann Memorial Foundation
P. O. Box 45593, Nairobi, Kenya

ABSTRACT:

Daily indirect observations were made on the diet and feeding habits of the black rhinoceros (Diceros bicornis) on Ol Ari Nyiro ranch, Laikipia, Kenya over a six month period. Individual rhinos were followed along their feeding tracks, plants consumed by tracked animals were identified, and herbivory quantified.

In total, 9665 individual feeding points were recorded at 1967 feeding stations. At least 103 plant species from at least 37 families were identified as rhino food plants. The diet of black rhinos on Ol Ari Nyiro was at least as species-rich as that in bushland habitats in Tsavo National Park, which are considered to be ideal habitats for black rhinoceros, and considerably more species-rich than the diet of rhinos in Masai Mara Reserve. Black rhino ate selectively, and showed a marked preference for Acacia species and Phyllanthus fisheri.

They apparently fed less on each plant in the dry season than in the wet season. This may be due to decreased palatability of food plants, and implies that rhinos may travel further per day in the dry season than in the wet season.

INTRODUCTION:

Black rhinoceros are browsers, feeding mainly on woody vegetation, and have an ability to feed on coarser material than any other herbivore. The feeding behaviour and ecology of black rhinoceros in East Africa have been studied in Ngorongoro and Olduvai (Goddard, 1968), Tsavo National Park (Goddard, 1970) and Masai Mara Reserve (Mukinya, 1977). These studies were carried out in relatively open habitats where visibility was good, and involved direct observations of individual animals feeding on particular plant species at a given time and place. However, the feeding ecology of black rhinos had not been studied previously in the dense bushlands that comprise their preferred habitats.

Rhinoceros species are endangered throughout their range, largely due to heavy poaching pressure. Conservation strategies for black rhinoceros in Kenya include translocations of animals from areas of low rhino density and from unprotected areas to a series of small, protected rhino sanctuaries. At least one of these sanctuaries (Nakuru National Park) has not historically been host to a large resident black rhino population. On-going studies of the suitability of these areas for translocation depends on detailed knowledge of rhino food plants and feeding behavior in their preferred natural habitats. Ol Ari Nyiro ranch is home to Kenya's largest indigenous population of Black Rhinoceros (Diceros bicornis), estimated at 45-50 animals. Dense vegetation and intensive anti-poaching activities by ranch personnel have helped protect this population. Their continued survival is therefore central to the future of the species in Kenya. Information on the ecology and behaviour of this population is essential for conservation and management of black rhinoceros.

STUDY SITE METHODS:

Study Area: Ol Ari Nyiro ranch is located on the western edge of Laikipia District, Kenya, between 36° 15' and 36° 30' East and between 0° 30' and 0° 45' North. Mean annual rainfall is 700 mm. The ranch covers approximately 37,000 hectares, although the rhino population uses only about a third of this area. The major drainage is the Mukutan River, which flows west through a deep gorge into the Rift Valley and Lake Baringo. A series of man-made dams and the springs feeding the Mukutan River provide water for wildlife, domestic stock, and ranch employees. The area varies from wooded grasslands and precipitous scrubby gorges on the west to bushlands and open grass plains in the east (Muasya et al., 1993). The major vegetation types in the study area are as follows.

1. COMBRETUM WOODEN GRASSLAND:
Wooded grassland covers the areas on the top of most ridges and hills. The vegetation is characterised by scattered trees of Combretum molle and Acacia hockii. Other common plants are Solanum incanum, Grewia similis, Bersania abyssinica, Indigofera schimperi, Justicia verticillaris and J. diclipteroides.

2. GORGE SCRUB:
The dense scrub vegetation on the steep gorge walls is strikingly different from the wooded grassland on the ridge tops. It is dominated by Acacia brevispica, with large numbers of Euphorbia kibweziensis, E. candelabrum, and other small Euphorbia species. Other species include Phyllanthus fisheri, Plectranthus zatarhendi, Croton dichogamus, Ecbolium revolutum, Tinnea aethiopica, Ruttya fruticosa, Grewia spp. and Sanseveria spp.

3. TARCONANTHUS (LELESHWA) BUSHLAND:
This habitat covers most flat parts of the study area, with a uniformly dense cover of leleshwa (Tarconanthus camphoratus). Other species include Euclea divinorum, Rhus natalensis, Lantana trimera, Lippia javanica, Tephrosia emeroides, Euphorbia crotonoides, Tinnea aethiopica, Ruttya fruticosa, and Grewia spp.

4. EUCLEA-CARISSA BUSHLAND:
This vegetation, dominated by Euclea divinorum and Carissa edulis, was found on level areas not dominated by leleshwa. Other major plant species include Rhus natalensis, Acacia hockii, Croton dichogamus, Lantana trimera, Lippia javanica, Ferula communis, Tinnea aethiopica, and Grewia spp.

5. RIVERRINE WOODLAND:
This vegetation is found mainly along drainage lines leading into the Mukutan Gorge, and along the river itself. It forms a narrow strip of tall dense cover in most areas, although some parts are open wooded grasslands, dominated by Acacia xanthophloea. Acacia gerrardii and Phyllanthus fisheri are common. Additional species include Hibiscus aponeurus, H. lunarifolius, Conyza sumatrensis, Psiadia punctulata, and Aspilia mossambicensis.

METHODS:
This study was carried out from June 1987 to January 1988. Rainfall records from the ranch headquarters show that the period from April to August 1987 was relatively wet (mean 99.6 mm/mo), and the period from September 1987 to January 1988 was relatively dry (mean 32.5 mm/mo). The study period was divided into "wet" from June to August 1987, and "dry" from September 1987 to January 1988. During the "dry" period, the conditions were noticeably drier and most of the annual plants died back.

For two reasons, the rhinos on Ol Ari Nyiro were far less visible than in sites of previous feeding studies (Goddard, 1968, 1970; Mukinya, 1977). First, the ranch vegetation was very dense. Second, the rhino themselves were extremely shy, rarely venturing into the open except at night. Under these conditions it was not possible to use the direct observational techniques of previous studies.

We therefore developed a practical technique of indirect observation that quantified aspects of rhino feeding ecology. This tracking technique was designed to approximate as closely as possible the methods of Goddard (1968, 1970) and Mukinya (1977). Individuals rhino were identified by their unique tracks, which differed in size and wrinkle patterns. This was essential to avoid confusion when the tracks of different rhino crossed.

The tracks of a rhino were located early in the morning, and the path of feeding was followed until the rhino was located, usually sleeping in mid-morning. Rhinos typically feed during the early morning and late afternoon. If rhinos were located before they had stopped to sleep, it was possible to follow them at a distance as closely as 100 m without disturbing them, depending on bush cover and wind direction. Rhinos were followed at distances ranging from 100 m to 1000 m . Other large browsing species in the study area were eland, giraffe, and elephant. However, by following closely behind rhino and considering only fresh herbivory, it was possible to limit observations to rhino feeding.

The rhino browsed vegetation in a very distinct manner, clipping off twigs and shoots cleanly. As rhino feeding tracks were followed, freshly browsed plants were identified along the track. Each place the rhino stopped to feed was called a "station", forming a rough semi-circle in front of the browsing rhino, whose front legs were stationary (after Goddard, 1968).

Feeding on particular plant species was quantified by counting the number of freshly browsed stem tips, called "cuts". A given cut stem could be due to a single bite or to several bites, and bite size could not be quantified (as in previous studies). Although this made most interspecific comparisons inappropriate, it did allow broad estimates of relative importance of various food plants, and allowed intraspecific comparisons between seasons.

At each station, data were collected on the number of cuts on each species of food plant, the part(s) eaten, and the time and date of the observation. The locations of feeding tracks were recorded on a 1:50,000 map of the ranch. Food plants were identified on the spot if possible. Other food plants were collected in plastic bags and later pressed and dried for identification by the Herbarium of the National Museums of Kenya. Several food plants were not found in a reproductive state and not identified by the Herbarium. These species were assigned local (Kimeru) names provided by the experienced trackers that accompanied the observer. Additional food plant species were recorded when encountered outside of normal data collection periods.

Data were analyzed separately for wet and dry periods. Food plant diversity was calculated using the Shannon index. The number of cuts per station was calculated for each species. All twenty food plant species that were represented by ten or more stations in both time periods were included in a t-test comparing the number of cuts per station between wet and dry periods.

RESULTS:

At least 103 plant species from at least 37 families were eaten by black rhinos on Ol Ari Nyiro, either during specific data collection periods (Table 1) or incidentally (Table 2). Latin names were not determined for 14 of these plants. Families having at least four representatives were: Acanthaceae (9), Papillionaceae (8), Compositae (7), Euphorbiaceae (6), Mimosaceae (4), Verbenaceae (4), Anacardiaceae (4), Rhamnaceae (4). In total, 9665 individual cuts were recorded at 1967 stations. As in previous studies at other sites, Acacia species were important food plants, comprising 27% (wet period) and 36% (dry period) of all stations recorded.

Similar numbers of plant species were eaten during the wet (N=64) and dry (N=63) periods. However, the diversity of food plants eaten was ~15% greater during the wet period than during the dry period (Table 3). This was despite the fact that sample sizes were twice as great during the dry period, which would tend to increase measured species richness. There was a more even distribution of commonly eaten plant species during the wet period than during the dry period. In the wet period there were 21 species that each accounted for at least 1% of the diet, but in the dry season only 16 species. The two plant species most commonly eaten during the wet season (Acacia hockii & A. brevispica) accounted for 26% of the wet period stations, whereas the two plant species most commonly eaten during the dry period (Acacia brevispica & Phyllanthus fisheri) accounted for 46% of the dry period stations (Figure 1).

The relative importance of food plants differed between wet and dry seasons (Table 1 & Figure 1). Eighteen species were eaten only during the wet period, and seventeen species were eaten only during the dry period. Staple plants species eaten commonly during both wet and dry periods were Acacia brevispica, A. hockii, Phyllanthus fisheri, Carissa edulis, Tinnea aethiopica, and Euclea divinorum. Phyllanthus fisheri and Acacia brevispica represented much greater proportions of rhino diet during the dry season than during the wet season. Phyllanthus fisheri was favored early in the dry period, and Acacia brevispica late in the dry period (Figure 2). During the wet period, several seasonally available plant species were favored: Ferula communis, Indigofera schimperi, and Asparagus buchananii. Ferula communis was the most eaten plant early in the wet period, but was not eaten at all during the dry period, when this species was completely dry (Figure 2). Herbaceous species (which included some semi-woody herbs) accounted for 18% of the cuts and 27% of the stations in the wet period, but only 10% of the cuts and 13% of the stations in the dry period (cuts: C2=128.0, p<.001; stations: C2="62.7," p<.001). Different plant growth forms were not represented equally well. We conducted an intensive botanical survey of one diverse site within the study area. Of the species on that list, 67% (24/36) of the trees and shrubs, 9% (6/64) of the perennial herbs, and 0% (0/15) of the annual herbs were listed as rhino food plants (C2="33.5," d.f.="2," p<.001). Although it is possible that rhinos do not eat many annuals or herbs in general, it is more likely that the presence of these plants in the diet is cryptic, with rhinos eating often entire plants and leaving no evidence behind.

Goddard (1970) and Mukinya (1977) noted that black rhinoceros ate all parts of individual plants more often than they ate leaves only, stems only, or inflorescences only. On Ol Ari Nyiro, a similar pattern was observed. Rhinos usually ate stems, leaves, inflorescences, and shoot tips of the same plants, and often in the case of small herbs, the entire plant.

However, there were exceptions where some parts of some plant species were selected more than other parts by rhino. For example, only the stems of young Ferula communis plants were eaten, and the feathery leaves rejected. Entire plants of the herbs Hypoestes verticillaris and Justica cordata were uprooted, but only the above-ground parts eaten, leaving the roots behind. Rhino ate the bark of Boscia angustifolia during the height of the dry season; in many cases rhino tore off the bark of trees with their teeth and horns.

In general, rhino ate the same parts of particular plant species during both wet and dry periods. Exceptions were the shrubs Euclea divinorum and Phyllanthus fisheri, in which rhinos at both stems and leaves during the wet period, but only stems during the dry period. Phyllanthus fisheri is deciduous and had no leaves during the dry period. In contrast, the leaves of Euclea divinorum were present year-round, but not eaten during the dry period.

The number of cuts per station can be considered an approximate measure of the amount eaten on particular plants. In most cases, the number of cuts counted was apparently equal to the number of stems eaten. However, Acacia hockii was unusual in that rhino often ate plants extensively to down near the ground below where the stems branched, so that the number of cuts remaining was likely to be less than the number of stems eaten.

There were twenty plant species for which there was considerable data (at least four stations) in both the wet and the dry periods. Of these, nineteen had fewer cuts per station during the dry period than during the wet period (Signs test, p<.001). The lone exception was A. hockii (see above). Overall, the number of cuts per station among these species was significantly lower (by an average of 30%) during the dry period than during the wet period (Table 3, t="5.99," p<.001).

RESULTS:

Discussion: Black rhinoceros are browsers with broad diets wherever they have been studied in East Africa. The list of 103 food plant species identified for Ol Ari Nyiro rhinos falls within the range of studies in Masai Mara (70 spp., Mukinya, 1977), Tsavo (102 spp., Goddard, 1970), and Ngorongoro (191 spp., Goddard, 1968). All of these lists are probably incomplete. Rhino apparently have the ability to eat a wide variety of plants, at least in small quantities.

We did not collect data on the absolute frequencies of plants in the study area, and so did not calculate feeding "preferences". However, it is striking that leleshwa (Tarconanthus camphoratus) was never recorded as a food plant of rhino during this study, even during the driest conditions, despite being the dominant plant in much of the study area (Young & Francombe, 1991, Muasya et al., 1993). Even cattle will occasionally eat leleshwa (Young & Francombe, 1991). Given the ability of rhino to feed on some of the most unpalatable species (Datura stramonium, Euphorbia spp., Sanseveria spp., Euclea divinorum) the absence of leleshwa from the rhinos1 diet is interesting. Leleshwa is rich in aromatic oils such as camphor (Peter Waterman, pers. comm.). The ability of leleshwa to form monospecific stands that exclude most other plant species (Young & Francombe, 1991) may be due in part to the competitive advantage provided by their unpalatability to rhino (and elephant).

Conversely, the availability of particular food plants can effect the movements of rhino. For example, scattered Euphorbia candelabrum trees, although making up a small proportion of the diet, were occasionally highly sought after. Frequently, rhino tracks led straight to individual E. candelabrum trees over considerable distances. Similarly, rhinos would make visits to abandoned boma sites where Datura stramonium was abundant.

The most common potential competitors of black rhino in East Africa are probably elephant and giraffe. Acacia and Phyllanthus spp. are preferred food plants of giraffe (Pellew, 1984; Young & Isbell, 1991), but giraffes tend to avoid dense scrub vegetation favored by rhino. Elephant also prefer Acacia spp., but eat considerable grass when available (Buss, 1961; Napier & Sheldrick, 1963; Laws & Parker, 1968).

Goddard (1970) suggested that the relative abundance and availability of legume species may be the best indicator of an optimal habitat for black rhinoceros. Similarly, on Ol Ari Nyiro, Acacia was the most commonly eaten genus of plants, and rhino fed most in areas with a high density of low acacias, such as the gorge slopes. Ol Ari Nyiro Ranch has the largest natural population of black rhinoceros in Kenya, and this is apparently due to the occurrence of appropriate food, thick cover, and vigorous anti-poaching activity.

There was a tendency in this study for rhinos to feed in the vicinity of water. In the wet period, rhino fed closed to wallow sites and seasonal streams and waterholes. These feeding areas were mainly on the tops and sides of the gorges. In the dry season, when standing water and seasonal streams had dried up, rhinos fed near man-made reservoirs and in the riverine habitat which retained more water and had a longer persistence of palatable herbs. This led rhino to be more concentrated during the dry period than during the wet period. Counting cuts and feeding stations cannot measure accurately the absolute amount of material consumed from each plant species (Goddard, 1968, 1970). However, these data do reflect the composition of the diet, and allow for comparisons within plant species between wet and dry periods. The indirect method developed here allowed for the first time studies of rhino feeding behaviour in the dense scrub that is their preferred habitat.

There are several possible explanations for seasonal changes in rhino diets. First, some food plants are only available during wet periods. These include Ferula communis, Commelina spp., Euphorbia crotonoides, Oxygonum sinuatum, and Malva verticillata. As mentioned above, our list of small and annual herbs may well be underestimated relative to woody species. Second, as more favored plant species become less available during dry periods, rhinos may shift to less palatable species. Third, some species may provide key resources during critical times, such as succulents during drought. It is difficult to distinguish between the latter two effects.

Lastly, it appears that plants differed in their relative palatability between wet and dry periods (c.f. Young, 1985). Rhinos ate different plant parts of some plants during different periods. During (wet) periods of plant growth, nutrients may be translocated for development of shoots, leaves, and flowers and fruits. In dry periods, nutrients may be translocated to the roots and bark. This may explain the tendency for rhinos to eat the bark of Boscia angustifolia and the stems of Euclea divinorum during the dry season, and for the leaves of E. divinorum to be eaten more during the wet period than during the dry period.

Common rhino food plants exhibited fewer cuts per station during the dry period than during the wet period. This could be due to either fewer palatable plant parts available or to a general decrease in palatability of all parts. If this pattern means that rhino fed less on each individual plant, it would imply that rhino travelled farther each day during the dry period and take more time to eat the same amount of food.

MANAGEMENT IMPLICATIONS:

Although there appears to be no shortage of "staple" rhino food plants on the ranch, the distribution and seasonal use of particular food plants should be considered in the future management of these rhinos. The movements of rhino off the ranch into vulnerable areas in time of drought, and the influence that food shortage may have on these movements, are of critical importance to the future of rhino, on Ol Ari Nyiro, and perhaps elsewhere. In 1984, during a severe drought, as many as 20 rhino wandered off the ranch and were poached. It is not known whether the increase in the rhino's; movements was due to lack of food, although this is likely, particularly considering the partiality of rhinos for succulent plants (Euphorbia and Sanseveria spp.) in dry periods.

Due to the reductions in rhino numbers on the ranch, this wandering in search of food may not be a problem at present, with rhinos only occupying the central third of the ranch. But if continued successful protection of the population and breeding success lead to a larger population, these animals might be under increased pressure to wander in search of sufficient food. The management may wish to consider providing dry season food, perhaps in the form of the palatable Euphorbia turicalli, around waterholes and salt licks. This would act as lures to reduce the tendency for rhinos to leave the ranch during drought, and could also serve to concentrate the population within the safer centre of the ranch.

The majority of Kenya's rhino occur on private ranches, which have served as reserves for rhino removed from vulnerable areas, and as sources for introduction programmes. The fates of these rhinos hinge on the ability of these ranches to effectively manage and protect them. Further research into game movements in these areas, and into effective and affordable barriers to these movements, are a vital necessity.