Previous PageTable Of ContentsNext Page

Trees outside Forests: Sudan

William M. Glen, Consultant
Bonshaw, Prince Edward Island, Canada

Photo 58. Agroforestry stand of Acacia senegal (gum-tree). (© Cossalter/Cirad)


Sudan covers an area of 2.5 million km2 and is characterized by three vegetation zones: desert in the north, savannah and dry grassland in the central region, and the large swamp called the Sudd with some rain forest in the south. According to the 1995 Forest Resources Assessment, forests cover 17 percent of the territory and wooded land 10 percent. Sudan, though not a heavily forested country, is heavily dependant on tree resources for its basic supplies of fuelwood and building materials (FAO, 1995b).

Maintaining the productivity of the land is crucial in such a context, and this implies the existence of land use and land cover data. The 1995 National Forest Inventory (NFI) had as its prime objective the assessment of plant cover, and was designed to inventory both land cover and land use.


Sudan's forest service, the Forests National Corporation (FNC), established no legal or other type of definition for `forest', and, accordingly, none for Trees outside forests. The data on land distribution indicate the following division: 27 percent forest and wooded land, 29 percent desert, and the remaining 44 percent upon which off-forest trees can be found.

Patterns and extent of change

The 1995 Forest Resources Assessment showed a 0.8 percent deforestation rate for 1990-95. A review of old maps of Sudan at the 1:250 000 scale revealed a significant decrease in treed area since 1900.

There is no real acknowledgement of the importance of trees growing in non-forested areas, with the exception of Acacia Senegal (L.) Willd in the savannah area. There is a growing and worldwide demand from the food and pharmaceutical industries for the gum Arabic extracted from it. Sudan is the world's top producer of gum Arabic. Exports in 1995 amounted to US$ 56 million, i.e., almost 11 percent of all Sudanese exports for that year.

The people of Sudan are partly dependant on Trees outside forests for the wood to build their homes, for charcoal, and especially for livestock forage. To some extent they also depend on these trees for their livelihood. Fodder and wood for charcoal have not been quantified, but their utilization is clearly exerting growing pressure on trees (FNC, 1995). These trees are also clearly beneficial for soil maintenance. Erosion, especially wind erosion, is twice as severe in treeless areas. Tree-planting to stabilize dunes and check the encroachment of the desert onto farmland is a common practice in northern Sudan.


Forest law and the guidelines for action of the Forests National Corporation (under the Ministry of Environment and Tourism) clearly recognize the importance of forests. The FNC established that a 20 percent expansion of wooded areas was essential to meet national needs. It is carefully tilting management strategies toward a more balanced ratio of forestland to farmland and rangeland. Dry grasslands are now the focus of increasing attention. This contrasts with an earlier lack of interest in these areas because of their low woody biomass, despite their importance as a source of supply for local populations (FAO, 1988; von Maydell, 1990; FNC, 1994 and 1995).

Assessment and planning

FNC's mandate includes forest resource assessment. Most inventories have focussed on areas of known tree density. Moreover, the 1983-84 Blue Nile Province inventory, which used aerial photography combined with randomly laid out ground plots, was not repeated due to the excessive cost.

Some inventory projects had focussed on areas with low tree densities. The work done by a team from Lund University of Sweden from 1987 to 1990 covered 5 800 000 ha in the Eastern Region using Landsat Thematic Mapper imagery, and the Sudan Reforestation and Anti-desertification /Sudan Resource Assessment and Development projects (SRAAD) between 1987 and 1993 covered 3 60000 000 ha, using a combination of satellite imagery and systematically laid out ground plots.

The main lessons learned from these two projects were that satellite imagery without good ground verification can produce misleading results, and that there is a need for ground plots to supply details on volume, stems/ha, species, site conditions and land use. Global Positioning System (GPS) units used in the SRAAD projects demonstrated the feasibility of sampling grid networks. The review of the procedures used for these inventories plus the expertise acquired by the staff facilitated the 1995 implementation of the NFI.

National forest inventory

The primary purpose of the NFI project was to provide area and volume statistics for regional and national planning, thus defining this as a reconnaissance survey. The data on plots and species distribution could facilitate assessment of nonwood products such as gums, fruits and nuts.

Methodology. The inventory design was flexible enough to combine with more detailed inventories, such as the national parks survey, or be used as a working model for regional inventories. The field work was done in 1995 and 1996, following a 1994 methodology preparation stage and staff training sessions in early 1995. Only ground plots were selected as saving on both costs and time (there were only three crop seasons for inventorying 77 000 000 ha). The lack of current inventories and current base maps (the available maps dated back more than 40 years), were further constraints.

The inventory covered 2 608 sampling plots representing some 26 000 000 ha in Central Sudan. These are typical dry grassland and savannah zones, which would not be considered forest land under the FAO definition that sets the tree cover threshold at 10 percent. The inventory project was based on the measurement of fixed area plots on a systematic grid throughout the inventory area. The use of the GPS units meant that the sampling locations could be mapped using their co-ordinates without reference to any base mapping.

The sampling grid used was 10 km by 10 km with a field plot being established at each grid intersection. Because the sampling was systematic, the measured plot was assumed to be representative of the sampling cell (i.e., 10 km or 10 000 ha). The chances of the plot falling into any particular land use or forest type were proportional to the occurrence of that type in a given area (FAO, 1981; SRAAD project, 1990b).

The plot size used was 20 by 100 m, with an east/west orientation, the same as the SRAAD projects. This type of plot had been used in most inventories in the Sudan since 1983, and the FNC field staff was familiar and comfortable with this plot design (Groupe Poulin, 1984; Vink, 1987; SRAAD projects 1990 a, 1990b; Ali, 1993; Vogt, 1994). Rectangular plots are more difficult to accurately establish than circular plots, however, and so the latter would have been preferable.

A number of data categories were devised as follows, each corresponding to multiple choice questions, excepting the items `slope percent and aspect' and `year of felling':

Photo 59. Camel browsing on one of the few trees in the region (© Braatz/FAO).

All trees or shrubs in the plot with a root collar diameter greater than five cm were tallied. Each plant was recorded as to species, live or dead, root collar diameter, total height, crown diameter measured at right angles to the plot's long axis, diameter at breast height, bole height and percent of cull. Shrub measurements were only taken in accordance with form and accessibility.

A regeneration subplot of 1 x 10 m was located in the south-western corner of the plot. The species and number of stems in it were tallied for all woody plants with a height greater than 15 cm and less than 130 cm.

The field data was entered into three databases, one for the plot observations, and one for the tree measurements and the third for the regeneration plot. Plot location information which included the 1/250 000 map sheet references, the Universal Trans-Mercator (UTM) co-ordinates and the administrative region were added to the plot file. The information in these databases combined with volume equations were used to produce a summary database which included statistics for each plot. This plot summary file was used to produce aggregations of area or other criteria.

Results. The above results (Glen, 1996) concern 16 730 000 ha and 1 673 plots in the country south of Khartoum straddling the two Niles. The land cover results were as follows: bare ground 30.2 percent, grass 34.7 percent, shrubs 14.1 percent, and trees 19.3 percent and water (the two Niles) 1.6 percent. The land use for the same area was: cultivation 41.4 percent, grazing 21.0 percent, forestry 18.7 percent, population centres 2.8 percent and 16.1 had no evident use. On these 16 730 000 ha, wind erosion had damaged 500 000 ha (three percent), and water erosion 2 080 000 ha (12 percent), confirming that the impact of erosion is in direct proportion to land cover - the more exposed the soil the more severe the erosion.

Out of a total wooded area10 of 4 680 000 ha, only 1 160 000 ha, or some 7 percent of the area, met the FAO definition of forest. TOF represented 21 percent of the wooded areas measured, and thus a little over 75 percent of the total wooded areas, in addition to the potential represented by areas currently without trees, such as pasture, estimated at 35 percent of the entire inventory area.

Table 15 shows the estimated volume of woody biomass for the inventory areas for all diameters and species lumped together. The per hectare volume of wood is low in most places, but it must be remembered that this vegetation is certainly not negligible in terms of local wood supplies for Smallwood and other miscellaneous products, not to mention soil protection.

The inventory identified 33 species and their distribution can be mapped. Acacia, Balanites, Combretum and Terminalia were the main genera in treed areas outside the forest. The overall statistics from the 177 field plots established in the TOF area were as follows: average crown cover, 4.43 percent (including shrubs); average number of stems with a root collar diameter greater than 5 cm, 13; average live volume 4.6 m3/ha; average dead volume 0 m3/ha; average basal area 0.24m2/ha; and number of regenerating seedlings 4.61/ha.


The methodology of this primarily reconnaissance survey, with its sampling rate of 0.002 percent, enabled us to obtain the desired information at low cost and in a short timeframe. It could be adapted for various situations by modifying the sampling rate of the grid, and the size and type of parcels as well as the data-gathering criteria. Bearing in mind that the plots were located with the use of GPS co-ordinates, they can be revisited to produce data on size and any changes. Only ground plots were used for this national forest inventory. Aerial photographs could be helpful in areas where access is a problem and aerial photography is available. The combination of both ground plots and aerial photographs is another possible option (CFIC, 1998).

Sudan now has data to take up issues of sustainable land use. The data on land use and erosion damage could underpin a land use planning effort to address the scope and location of soil conservation problems.

Table 15: Average per hectare volume and area by diameter class


Area by diameter class (`000 ha.)

Diameter classes

Average volume (cm.)





20 et plus























Wad Medani







Aba Island




























Er Roseires















3 410






Source: Glen (1996).
Note: These measurement were taken only on `wooded' plots, i.e., with woody vegetation, out of a total area of 4 680 000 ha, excluding plots with no woody vegetation.
The total volume is the volume and the aboveground portion as of a diameter of 5 cm
The woody biomass can be calculated when the conversion equation is available.


Ali, A.M. 1993. On the Use of Suitable Sample Plot Size and Tariff System for Efficient Inventory. M. Sc. Forestry Thesis. Department of Agriculture. University of Khartoum. Khartoum.

CFIC (Canadian Forest Inventory Committee). 1998. Developing the National Forest Inventory Design. Canadian Forest Inventory Committee. Secretariat: Canadian Forest Service. Natural Resources Canada. Pacific Forestry Center. Victoria, British Columbia.

FAO. 1981. Manual of Forest Inventory with special references to mixed tropical forests. Rome.

FAO. 1988. Non-timber Uses of selected arid zone trees and shrubs in Africa. Conservation Guide. Rome.

FNC (Forests National Corporation). 1994, Studies on Consumption of Forest Products in the Sudan (1993/1993), Preliminary Results Findings, Summary Tables. Forests National Corporation Forestry Development in the Sudan. Khartoum.

FNC (Forests National Corporation). 1995. Forest Products Consumption in The Sudan. Final Report. Forest Development project FAO/GCP/SUD/047/NET. Forests National Corporation. Ministry of Environment and Tourism. Khartoum.

Glen, W.M. 1996. National Forest Inventory for the Sudan. FAO/GCP/SUD/047/NET. Forests National Corporation. Ministry of Environment and Tourism. Khartoum.

Groupe Poulin, Theriault Ltee Consultants. 1984. Forest Inventory and Market Demand Study Project, Blue Nile Province, Democratic Republic of the Sudan, Forest Inventory Report Volume 1, Project Report. Québec, Canada. Khartoum.

SRAAD (Sudan Reforestation and Anti-Desertification). 1990a. Pilot Project Procedures Handbook. United States Geological Survey, USDA Forest Service and Sudan Survey Department. Khartoum

SRAAD (Sudan Reforestation and Anti-Desertification) 1990b. Kazgail Woody Vegetation Mapping and Inventory Report. Preliminary Report. Khartoum.

Vink, A.T. 1987. Forest Inventory of Rawashda Forest Reserve, Kassala Province, Eastern Region. FAO/ GCP/SUD/033/NET. Field Document 21, Fuelwood Development for Energy in Sudan. Khartoum.

Vogt, K. 1994. El Ain Natural Forest Management Project, Management Plan for 1994-2001. SOS Sahel -UK, London, England or SOS Sahel, El Ain Natural Forest Management Project. El Obeid. Sudan.

Von Maydell, H. 1990. Trees and Shrubs of the Sahel, Their Characteristics and Uses. GTZ. Verlag Josef Margraf Scientific Books. Weikersheim, Germany.

10 The 4 680 000 ha were measured on 468 wooded plots containing woody vegetation.

Previous PageTop Of PageNext Page