A RECONNAISSANCE SURVEY OF THE VEGETATION OF MUTINONDO WILDERNESS AREA

Report on the study carried out in April 2003 Paul P. Smith PhD., Royal Botanic Gardens, Kew, Wakehurst Place, Ardingly, W. Sussex RH17 6TN, U.K.

SUMMARY

The aims of the reconnaissance survey of the vegetation of Mutinondo Wilderness Area (MWA) were as follows: to carry out a botanical survey and inventory of MWA; to identify unusual habitats and plants; to map the vegetation of MWA; to assess the plant diversity of MWA; to identify useful plants; to assess the potential for carbon sequestration in MWA's miombo; and to provide training to Zambians in botanical survey techniques. The fieldwork and training component of the survey was carried out in April 2003, and plant identification, data interpretation and analyses were carried out from May-December 2003.

During the fieldwork phase, plot-based and plotless methods were used to characterise the vegetation of MWA, and plant specimens were collected for identification. The study identified four major vegetation types in MWA: miombo woodland (covering approx. 80% of MWA); riverine forest/Mushitu (3.1%); grasslands (9.4%); and inselbergs (7.5%). Detailed descriptions of each vegetation type are given, and for miombo woodland, plot data recording species density, biomass and frequency are analysed. These analyses suggest that MWA miombo is typical, mature plateau woodland, comparable with old growth miombo elsewhere in Zambia and the region. A preliminary vegetation map of MWA, based on Landsat ETM+ imagery, and showing the four major vegetation types is presented.

The plant species recorded in MWA are listed in Annex 4 and comprise ca. 400 species, indicating that the botanical diversity of MWA is very high. More plant inventory work is needed to add more species to this list. In Annex 5, common woody species are assessed for their utilitarian value, and of the 100 species analysed, 67 were reported in the literature as being sources of non-timber products and 39 were useful timber species.

Carbon biomass in MWA was measured at between 18.7 and 66.1 tonnes per hectare depending on which biomass equation was used. The relevance of old growth miombo carbon sinks to the Clean Development Mechanism of the Kyoto Protocol is discussed.

Finally, recommendations are made for the continued study, protection and management of the vegetation of MWA

1. INTRODUCTION

The following study was carried out in response to a request made by Mike and Lari Merritt, managers of the Mutinondo Wilderness Area (MWA) concession in a letter to Dr Paul Smith at the Royal Botanic Gardens, Kew, dated 20th of July 2002. The terms of reference for this study arising from this request included the following:
-  Carry out a botanical survey and inventory
-  Identify unusual habitats and plants
-  Map the vegetation of MWA
-  Identify useful plants
-  Assess the plant diversity of the MWA
-  Assess the potential for carbon sequestration in the MWA
-  Train Zambian counterparts in botanical survey techniques
It was envisaged that the data generated by this survey would have great utility to conservation initiatives in the MWA, including the adjacent, community based Chintu Mukulu project. The fieldwork and training component of the survey took place in April 2003, and plant identification, data interpretation and analysis was carried out from May-December 2003.

2. MATERIALS AND METHODS

2.1. STUDY AREA

Mutinondo Wilderness Area (MWA) is a 10,000 hectare concession located between 12º 22'S to 12º 30'S and 31º15'E to 31º21'E in Mpika District, Northern Province, Zambia. MWA falls within Chief Mpumba's jurisdiction.

2.1.1. Geology
MWA is situated on the Muchinga Escarpment west of the Luangwa valley, which is made up of igneous (e.g. granite) and metamorphic (e.g. gneiss and quartzite) rocks. This basement complex corresponds to the Kibaran tectogenic province made up of pre-Katangan rocks in excess of 1300 million years old. The Muchinga mountains are part of the Irumide fold belt which extends over a width of about 120 km along the line of the Great North Road between Kapiri Mposhi and the Zambia-Tanzania border (Drysdall et al., 1972). The inselbergs of the MWA are isolated erosional remnants - mountain ranges that have eroded away.

2.1.2. Topography and hydrology
MWA is more or less flat, with a gently undulating landscape at around 1450-1500 m. This landscape is punctuated by granite inselbergs, the highest of which, Mayense peak, reaches 1684 m. MWA is centred around the perennial Musamfushi river and its associated dambos. The Musamfushi is a tributary of the Mutinondo river, which eventually discharges into the Luangwa river, a tributary of the Zambezi river.

2.1.3. Climate
MWA lies in a region of unimodal rainfall, with the wet season extending more or less from November to April. On the Muchinga escarpment and plateau mean rainfall (Mpika) is 1065mm. Temperatures are: mean max. 25.2°C; mean min. 13.8°C, with frost occurring from time to time during June/July (Mpika figures).

2.1.4. Soils
Mutinondo soils are primarily utisols and oxisols from non-basic rocks, weathered, strongly leached with low cation exchange capacity, high base saturation, low phosphorus availability, low aggregate stability and micro-nutrient deficiencies.

2.2.5. Vegetation
The vegetation of the MWA is predominantly miombo (Fanshawe, 1971), a deciduous woodland dominated by the leguminous tree genera Brachystegia, Julbernardia and Isoberlinia. Important non-leguminous genera in miombo are Uapaca, Protea and Faurea. Miombo woodland typically takes the form of a two or three storeyed woodland in which the canopy attains a height of 15 m or more, largely depending on soil factors. The herbaceous layer is usually sparse, but tall grass miombo does occur in the MWA. Miombo is generally more easily defined floristically than physiognomically, and this applies to the MWA where miombo woodland ranges from single storey scrub on shallow soils to tall woodland with a canopy height of 20 metres or more.
Miombo ecosystems also encompass the grassy, seasonally waterlogged depressions called dambos (Vesey-Fitzgerald, 1963) which are characterised by poorly drained acid vertisols and are dominated by the nutritionally poor grasses Loudetia simplex and Hyparrhenia spp.
Associated with the Musamfushi river and its tributary streams is riverine forest known locally as 'mushitu' (Fanshawe, 1971). This is a three storeyed forest found flanking rivers and streams, and sometimes extending to swampy areas. Mushitu is dominated by large trees such as Syzygium cordatum, Agauria salicifolia and Uapaca lissopyrena, with climbers and understorey species forming more or less dense thicket.
Mutinondo's inselbergs have their own vegetation, dominated by Xerophyta equisetoides and shrubs such as Vernonia bellinghamii, Ozoroa reticulata subsp. insignis and Iboza riparia. Characteristic species of the rock surface are Myrothamnus flabellifolius, the mat-forming sedge Coleochloa setifera, and tussock grasses such as Microchloa indica and Loudetia simplex. Inselberg seepage sites include ferns, sedges and semi-aquatics such as Drosera spp. and Xyris spp.

2.2. SITE SELECTION AND DATA COLLECTION

2.2.1. Vegetation structure and composition
Plot and plotless methods were used to characterise the structure and composition of the vegetation of MWA. In the miombo woodlands, the dominant vegetation type of the MWA, a total of 20 belt transects were set up. Placement of transects was not random as an effort was made to select homogenous areas of vegetation and to avoid ecotones and atypical landscape features such as tracks and roads. Another consideration was accessibility, with most transects located within sight of a road to ensure that they could be easily found again and revisited in future studies. Transect positions were determined using a portable GPS (Garmin 12XL). Standard transects measured 50 x 10 m (= 500 m2), rectangular plots being chosen for ease of sampling and to maximise species diversity (Condit et al., 1996). Transects incorporated a minimum of 50 trees/shrubs or at least 15 specimens of a dominant species. Where these criteria were not met, transects were broadened or lengthened accordingly (Taylor & Walker, 1978). Within each transect, all woody plants were identified and measured for diameter at breast height (DBH) and height. Presence of herbaceous species was also recorded. An example of a belt transect field data form is reproduced in Annex 1.
Using the plotless methodology all four major vegetation types in MWA (miombo, riverine, dambo and inselberg habitats) were assessed at stratified sampling sites using a 'Vegetation type field data form' (Annex 2). Vegetation structure and composition; disturbance; nature of threat; land form; soil type; lithology; slope; aspect and exposure were recorded. In addition, all fertile species were recorded and/or collected systematically at the selected sites.

2.2.2. Aerial survey
A partial scene (90 km x 90 km) LANDSAT ETM+ satellite image (September 28, 1999), georeferenced and geocoded to UTM map projection was supplied by the University of Maryland Global Landcover Facility (http://glcf.umiacs.umd.edu/index.shtml) as digital data. Mutinondo Wilderness Area's boundaries, taken from the 1: 50,000 ordnance survey series, were superimposed on to the aerial scene, and vegetation boundaries delineated using a supervised classification based on the ground survey data (see below).

2.3. DATA ANALYSIS AND INTERPRETATION

2.3.1. Ground survey data
The plot based and plotless ground study data provided detailed information on vegetation structure and composition. Thus, for each vegetation type defined by the Landsat imagery and ground survey, it was possible to describe vegetation structure (after White, 1983), characteristic woody species, associated grass and herb species, soils, lithology, topography and area covered. In addition, for the miombo woodlands of the MWA, based on the plot data collected, it was possible to:

-  List plant species.
-  Calculate tree/shrub density, expressed as n ha-1 (Brown, 1954)
-  Calculate woody biomass, expressed as t ha-1 using the following formulae developed in similar habitats.

In Sengwa, Zimbabwe (Guy, 1981):

Tree biomass (Kg) = 0.0549 x (diameter at breast height)2.5101

In Central Zambia (Chidumayo, 1991, 2002):

Woody biomass (Kg) = -14.01 + 0.391dbh2

-  Determine dominant species by calculation frequency of occurrence, expressed as % occurrence in the plots.

2.3.2. Map production
The vegetation map of MWA (scale 1:50,000) was based on upon a partial scene Landsat image from which vegetation boundaries were designated in a supervised classification based on the ground survey results presented below. The satellite imagery was processed as follows:

-  The imagery was imported into ERDAS Imagine 8.5
-  The image was cropped to the area of interest
-  A maximum likelihood supervised classification was applied to this area using the ground survey data

The end result is a map incorporating 4 broad vegetation types. This map will form the basis of a Geographical Information System (GIS) for the MWA. More detailed vegetation data, as well as information about geology, topography, water relations, plant-animal interactions, etc. can all be incorporated into this database in the future in order to build up a more complete picture of the MWA ecosystem.

2.4 TRAINING

Informal training in botanical survey and inventory techniques was carried out by the author. Trainees were Mr Clement Chishala of the Forestry Herbarium, Kitwe, and Mr David Chomba of the MWA staff. Subjects covered included:
-  Plotless and plot-based vegetation survey techniques
-  Herbarium specimen collection
-  Data collection
-  Plant identification

3. RESULTS

3.1 DESCRIPTION OF MIOMBO WOODLAND IN MWA

Miombo woodland is the dominant vegetation type of the upper Muchinga escarpment and covers 80% (81.15 km2) of the Mutinondo Wilderness Area. It takes the form of a two- or three-storeyed woodland with an open to lightly closed canopy of semi-evergreen trees 15-20m high and an understorey of trees and shrubs of 2-10 m. The grass layer is typically sparse.
Characteristic canopy trees include Brachystegia longifolia, Julbernardia paniculata, Brachystegia spiciformis, B. utilis, B. boehmii, B. floribunda, Isoberlinia angolensis, Parinari curatellifolia, and Pericopsis angolensis. Common smaller trees are Uapaca kirkiana, U. sansibarica, Brachystegia stipulata, Cratersiphon quarrei, Phyllocosmos lemaireanus, Memecylon flavovirens, Dalbergia nitidula, Combretum zeyheri, Anisophyllea pomifera, Pseudolachnostylis maprouneifolia and Monotes spp. Frequent shrubs are Protea angolensis, P. welwitschii, P. gaguedii, Rothmannia engleriana, Ochna schweinfurthiana and Rourea orientalis. Climbers are infrequent in this vegetation type, and the grass layer is sparse and generally restricted to scattered clumps of predominantly tall grasses. Common species are Andropogon chinensis, A. schirensis, Anthephora elongata, Heteropholis sulcata, Loudetia simplex and Sporobolus sanguineus.

On the edges of dambos, trees such as Brachystegia taxifolia, B. gossweileri, Uapaca nitida, Parinari curatellifolia and Syzygium guineense subsp. guineense are characteristic.

The miombo woodland of the MWA is primarily associated with deep, colluvial, stoneless sandy loams or sandy clays. These soils are slightly acid (pH 6.6 - 7.0) and often contain laterite nodules and mica aggregates in the B horizon.

Mutinondo miombo is dominated by the species in Table 1, which are listed according to their density, biomass and frequency. Full listings of all species recorded in the miombo plots are given in Annex 1.

Total number of trees/shrubs per hectare for the MWA miombo plots was 2045 stems ha-1. Stem area was measured at 12.65 m2 ha-1.

Woody biomass was calculated to be 37.366 t ha-1 using the Sengwa formula (Guy, 1981) and 132.23 t ha-1 using the central Zambia formula (Chidumayo 1991, 2002).

Table 1: Woody species in miombo plots ordered according to density (no. ha-1); biomass (Kg ha-1); and frequency (%).

3.2 DESCRIPTIONS OF MUSHITU, DAMBO AND INSELBERG VARIATIONS VEGETATION

3.2.1. Riverine forest/Mushitu
The rivers and streams of Mutinondo Wilderness Area are bounded by dense, usually three-storeyed evergreen forest (known as 'Mushitu' in chiBemba) which may extend to adjacent swampy areas. The canopy is closed and the trees from which it is formed may be 20m or more in height. Mushitu covers 3.1% (3.27 km2) of MWA.
Characteristic tall trees in this vegetation type include Syzygium cordatum, Agauria salicifolia, Gardenia imperialis, Sapium ellipticum, Syzygium guineense subsp. guineense, Uapaca lissopyrena, Uapaca sansibarica, Diospyros natalensis (also a shrub) and Ficus spp. Characteristic understorey species are Bequaertiodendron magalismontanum, Craterispermum schweinfurthii, Faurea saligna, Bersama abyssinica and Annona senegalensis. Shrubs include Diospyros natalensis, Rhus longipes, Tricalysia coriacea, Oncoba spinosa, Dissotis denticulata, Dombeya whytei,Psychotria succulenta, Rhamnus prinoides and Erythroxylum emarginatum. Characteristic subshrubs are Kotschya strigosa, K. strobilantha and Tinnea aethiopica. Climbers associated with this habitat include Artabotrys stoltzii, Smilax anceps, Keetia venosa and Psydrax livida. Grasses are not a significant component of this habitat - see 'Grasslands' below.
The igneous/metamorphic geology of the Muchinga Escarpment means that many of its rivers pass through rocky terrain (granite, quartzite) and waterfalls are common. Soils associated with this vegetation type are very variable in texture and tend to be alluvial on the valley bottoms and colluvial on the valley sides.

3.2.2. Grasslands
On the Muchinga escarpment the rivers and streams do not have a well developed meander belt and, as a result, the herbaceous riverside vegetation is less extensive than in the Luangwa valley. However, the numerous dambos and drainage channels associated with these watercourses do have a characteristic grass component. Grasslands cover approximately 9.4% (9.52 km2) of MWA.
Early in the rainy season, the dambos of MWA are characterised by species such as Loudetia simplex, Setaria sphacelata, Setaria pumila, etc. Later on, in April/May, Hyparrhenia species are prevalent: H. diplandra, H. nyassae, H. rufa, H. filipendula, H. cymbaria, H. variabilis and H. bracteata may all be found at this time. Other late season dominants are Andropogon chinensis, Diheteropogon filifolius, Monocymbium ceresiiforme and Pennisetum unisetum. Common herbs associated with the dambos of the upper escarpment include Gnidia chrysantha, Aeollanthus engleri, Dissotis debilis, Justicia elegantula and Epilobium salignum. In waterlogged sites, aquatic species such as the carnivorous Utricularia and Genlisea spp. are common, and semi-aquatics such as Xyris spp., Eriocaulon spp. and sedges such as Ascolepis protea are frequently found.
Dambos are usually associated with escarpment streams and occur on flat terrain where drainage is poor. Dambo soils are poorly drained and compacted. They are typically black, dark grey or dark brown, and acid (pH 5-6).

3.2.3. Inselbergs
The inselbergs and granite kopjes of the MWA support a distinctive vegetation type and although many of the miombo species listed above may occur on deep soil pockets, many additional species are distinctive on the thin soils associated with these rocky features. Inselbergs cover about 7.5% (7.7 km2) of the MWA.
Characteristic shrubs and small trees include Ozoroa reticulata subsp. insignis, Bersama abyssinica, Vernonia bellinghamii, Vangueriopsis lanciflora, Hymenodictyon floribundum, Schrebera alata, Iboza multiflora, Combretum psidioides subsp. psidioides, Erythrina abyssinica, Vitex mombassae and Vitex madiensis. A distinctive subsite found on MWA's inselbergs are the fibrous mats formed by the sedge species Coleochloa setifera. These mats provide a footing for Xerophyta equisetoides, Myrothamnus flabellifolius and succulents such as Aloe, Euphorbia and Kalanchoe spp. Ferns such as Actiniopteris dimorpha may also be present. Inselberg seepage sites contain carnivorous species such as Drosera burkeana, Genlisia africana and Utricularia spp. Sedges also characterise these sites, and include Bulbostylis abortiva, Lipocarpha nana, Scleria pulchella, Scleria pergracilis and Scleria flexuosa. Other water loving plants found here are Eriocaulon maronderanum, Polygala africana, and Xyris straminea. The flatter, seasonally wet depressions are populated by Aeollanthus spp., Hibiscus meeusii, Chamaecrista mimosoides, Polygala petitiana, Crotalaria teretifolia, Hirpicium gracile, Indigofera spp., Tephrosia lepida, Antherotoma naudinii, and Craterostigma plantagineum. A number of grass species, e.g. Stereochlaena cameronii, Heteropogon contortus, Trachypogon spicatus, also occupy this niche.
Inselberg soils are colluvial or residual humic sands derived from the rock itself. They are acidic and very thin, the deeper pockets supporting woody vegetation and the thinner soils supporting herbaceous species.

3.4 CARBON SEQUESTRATION
A hectare of miombo woodland in the MWA is measured as containing 37.366 tonnes of biomass, using the Sengwa formula (Guy, 1981). This is equivalent to 18.683 tonnes of carbon per hectare. Using the central Zambian formula of Chidumayo (2002), a hectare of MWA miombo woodland contains 132.23 tonnes of biomass, equivalent to 66.115 tonnes of carbon. Implications for the miombo woodlands of MWA being credited as a carbon sink under the Clean Development Mechanism of the Kyoto Protocol are discussed below in Section 4.4.

4. DISCUSSION AND RECOMMENDATIONS

4.1. BOTANICAL DIVERSITY AND BIOMASS

4.1.1. Botanical diversity
All four of the major vegetation types of MWA are high in plant diversity. A full list of recorded plant species is given in Annex 4. Species on the list without voucher citations should be regarded as provisional records - the only way to confirm the presence of these plants is to collect herbarium voucher specimens and get them identified. The total number of species recorded for MWA currently stands at ca. 400 but this number will increase with more survey work.
Certain habitats in MWA are under-collected and will require more attention, e.g. Riverine forest/Mushitu. In addition, plant groups like the monocotyledons (particularly grasses, sedges and geophytes) and ferns have not been systematically collected. Nevertheless, for an area of only 100 km2 the plant diversity of MWA is exceptional.
In terms of rare species, this survey and previous collecting expeditions have produced a number of plants that can't be matched with known species at Kew (e.g. PPS1907 Polygala sp.). Some of these may turn out to be new species. MWA is home to one of the rarest plants in Zambia, Encephalartos schmitzii, a cycad, which should be given special protection. Cycads are particularly vulnerable to plant collectors (see Recommendations below).

4.1.2. Density and biomass of miombo woodland in MWA
Table 2 shows the density and biomass figures for this study and other, similar miombo studies. Biomass figures for MWA and Sengwa (Guy, 1981) are comparable, and stem density differences are probably down to the fact that the Sengwa study did not measure any stems below 6 cm in diameter. Stem areas are smaller in Sengwa probably for the same reason. Stem numbers and areas are also smaller in Marondera (Campbell et al., 1995), probably because, in this case, stems below 9 cm in diameter were not measured. Both Zimbabwe sites are drier than MWA.
Biomass figures for the North Luangwa National Park (Smith, 1998) are double those of MWA, although stem densities are similar. The discrepancy in biomass between the two sites is probably due to the fact that the NLNP upper escarpment miombo plots included plateau and upper escarpment deep soil sites, the latter bearing very large trees in areas with no cultivation history. This in comparison to the MWA miombo, which occurs on thinner, plateau soils with a history of cultivation.
In central Zambia (Lusaka Rural, Kabwe Rural and Mumbwa) Chidumayo (1991) measured woody biomass in old growth miombo on similar sized plots at 141.75 t ha-1. In this study, the author measured actual woody biomass by cutting and weighing wood from the sample sites, and this was later used to develop a formula for estimating woody biomass from diameter at breast height (DBH) (Chidumayo, 2002). If this formula is applied to the DBH figures obtained for MWA miombo woodland, a woody biomass measurement of 132.23 t ha-1 is obtained for MWA. This figure is not far off that of the old growth miombo woodland in the central Zambian study, indicating that the MWA miombo has been undisturbed for many years, and is nearing maturity.
Clearly a major limitation with the current data is that the formulae used for measuring biomass in MWA is not calibrated for local miombo woodland. The huge variation between the figures calculated for woody biomass using the two different formulae is testament to this. Both the formulae are useful for making comparisons with other studies, but not for calculating actual woody biomass in the MWA. Of the two, the central Zambian formula is probably the most appropriate for application to the MWA in lieu of a cutting and weighing calibration.
A further limitation of the current study is the small number of sample sites - in addition, as the vegetation map in Figure 1 shows, the sample site localities are concentrated in the west of MWA. Further ground work therefore needs to be carried out in the east for a fuller picture to emerge.

Table 2: Comparison of density and biomass measurements in different miombo studies

4.2. UTILITARIAN SPECIES
A preliminary literature search was carried out with 100 of the woody species identified in this study (Annex 5). Of these 100 species, 67 were reported in the literature as being useful sources of browse, food, medicines, fibre etc. and 39 were useful timber species. From these results it is clear that most species can be used sustainably by local communities, particularly for harvesting of non-timber forest products such as fruits, browse and medicines. In addition, with such a wide range of timber trees available regulated offtake need not be detrimental in the long term.
The uses recorded in Annex 5are taken from the literature (Storrs, 1979; Storrs 1982; Malaisse, 1987; Smith & Allen, 2004). These publications cover all of Zambia and some neighbouring countries. It would be more useful if we had information about local uses of these plants in order to build up a picture of the value of such natural resources to local people (see 4.5 Recommendations, below).

4.3 CARBON SEQUESTRATION
Despite the fact that 20% of anthropogenic CO2 emissions are produced by land use change, particularly tropical deforestation, maintenance of natural forest is not accredited under the United Nations Framework Convention on Climate Change (UNFCCC) and the Kyoto Protocol. The main reason for this is that many environmental NGO's argue that the main cause of global climate change is the emission of greenhouse gases by industrialised nations. Under article 6 of the Kyoto Protocol, the Joint Implementation mechanism (JI), industrialised countries will be able to offset their carbon emissions by trading emission reduction units (ERUs) with other industrialised nations. Similarly, under the Clean Development Mechanism (CDM) of article 12, industrialised countries will be able to purchase Certified Emissions Reductions (CERs) from a project in a developing country, or 'removal units' (RMUs) if the project concerns carbon sequestration through afforestation or reforestation. This means that the industrialised nations will be able to avoid reducing emissions in their own countries, and this is why, as a political compromise, the use of old forest carbon sinks has been limited under the CDM.

A second problem is the technical and operational difficulties related to measuring and monitoring carbon sequestration in forests - this applies equally to afforestation, reforestation and maintenance of old forests. Year zero under the UNFCCC is 1990, and in many developing countries there is inadequate baseline forest extent data for this year. In addition, it is argued that the extent of carbon sequestration by any given forest needs to be measured initially, but then monitored in perpetuity because carbon sequestration rates will change over time. This is potentially technically difficult and expensive. In addition, a mechanism needs to be worked out as to who is responsible for any carbon losses after a CDM project is finished. With evolving assessment and monitoring technologies, particularly in remote sensing, many argue that these problems are not a serious impediment to the inclusion of natural forest maintenance in the carbon sequestration trade.

At present, developing countries such as Zambia have no specific emission reduction targets under the Kyoto protocol. However, there are many potential opportunities for mitigating atmospheric carbon in sustainable land management in developing countries, particularly through reforesting degraded lands, implementing sustainable agricultural practices on existing lands and slowing tropical deforestation. Of these three mitigation strategies only the first, reforestation, is currently eligible for financing under the Kyoto Protocol.

Although not necessarily covered by the Kyoto Protocol yet, land based opportunities to mitigate carbon emissions include the following:

For forests:
-  Protecting secondary and other degraded forests to allow them to regenerate naturally
-  Restoring native forests through assisted and natural regeneration
-  Maintaining existing forest-carbon stocks and sink processes by avoiding deforestation
-  Establishing plantations on non-forested lands
-  Managing forests sustainably to provide biomass energy

On agricultural lands:
-  Adopting zero- or minimum tillage practices on arable land
-  Improving rangeland management
-  Using green manures to cover crops
-  Amending soil with straw and manures
-  Increasing tree cover on agricultural or pasture lands with agroforestry

The Kyoto Protocol will evolve with use, and many of the above activities may be credited in the future. Critics of the Kyoto Protocol as it is currently stands argue that:

1. Under the Clean Development Mechanism (CDM) by restricting allowable forestry measures to afforestation and reforestation, and excluding protection of threatened native forests, this may create incentive for clearing native forests in developing countries.
2. The Kyoto Protocol, as currently interpreted, could cause certain commercial forestry operations to relocate to developing countries where they would be unencumbered by GHG liabilities, leading to biodiversity loss.

The Kyoto Protocol has not yet entered into force, largely because a number of influential countries (e.g. USA, Russia, Australia) have not ratified the treaty. However, the negotiations continue, and if projects under the Kyoto protocol do get going and the current set of conditions is found to be detrimental to biodiversity, management of old growth natural forests may well be included in the future. In addition, if developing countries are given emission reduction targets, as seems likely, there will be additional pressure on the treaty to include the maintenance of old forest carbon sinks.

In summary, as far as Zambia is concerned, management and maintenance of old growth miombo woodland currently counts for nothing under the Kyoto Protocol. However, if in the future it was considered equivalent to afforestation or reforestation, at a price of $5-10 per tonne of carbon, then maintaining a hectare of miombo woodland would be worth the following depending on how it was measured:

1) According to the Sengwa formula (Guy, 1981) the MWA miombo contains 37.366 tonnes of biomass per hectare (equivalent to 18.683 tonnes of carbon) and would be worth US$93.42 to US$186.83 per year under a CDM project.
2) According to the central Zambia formula (Chidumayo, 2002) the MWA miombo contains 132.23 tonnes of biomass per hectare (equivalent to 66.115 tonnes of carbon) and would be worth $330.58 to $661.15 per year through the CDM.

Under the current conditions of the Kyoto Protocol, opportunities for carbon sequestration projects do still exist for Chintu Mukulu in the field of forest reforestation or afforestation, but these would require significant planning and investment.

4.4. RECOMMENDATIONS

The following recommendations are made for the continued study, protection and management of the vegetation of MWA:

-  Systematically record plant species and collect voucher specimens to add to the checklist;

-  Systematically collect specimens in Mushitu, dambos and inselbergs;

-  Systematically collect under-collected groups, e.g. grasses, sedges, geophytes and ferns;

-  Collect local information about the utility of species;

-  Monitor boundaries and intactness of minority vegetation types, especially Mushitu;

-  If possible, obtain up to date aerial photographs of MWA

-  Expand the network of ground plots in the MWAmiombo, and set up similar plots in the other vegetation types.

-  Avoid excessive disturbance of delicate habitats, e.g. trampling on inselberg mats or seepage sites.

-  Map the cycads, protect their populations and keep their localities confidential

5. ACKNOWLEDGEMENTS

Thanks to my field assistants Clement Chishala and David Chomba, without whose help I couldn't have carried out the ground survey work. My thanks also go to Mike and Lari Merritt for their hospitality and help during the survey work, and to Quentin Allen for lending us his vehicle, his company and providing advice on Latin pronunciation (!). In the GIS laboratory I am grateful to Justin Moat at the Royal Botanic Gardens, Kew for processing the Landsat imagery and delimiting vegetation boundaries. Thank you also to Mr F.M. Malaya, Chief Forestry Research Officer, and to Mr L. Mulongwe of the Forestry Research Department for issuing the permits and facilitating this study. Finally, the author would like to acknowledge the Bentham Moxon Trust for sponsoring the ground survey work.