Inventory of the browse potential of bushes, trees and shrubs in an area of the Sahel in Mali

Method and initial results**

Pierre Hiernaux

Ecologist to the ILCA/Mali Programme

Introduction

2. Plant formations as basic units in compiling a regional browse potential inventory

2.1 Identification and mapping of plant formations

2.2 The description of ligneous communities

3 annual foliage production by species

3.1 Annual production as a function of height of trunk, and height and area of the crown

3.2 Seasonal distribution of browse potential

4.Towards a functional expression of the feed potential of browse populations 

4.1 The integration of production data with population descriptions

4.2 Assessment of the feed potential of perimeter on the Niono ranch

4.3 Regional assessment of feed potential

5. Discussion: problems arising from the characteristics of browse consumption with regard to the functional expression of feed potential

5.1 Supplementary feed or basic feed

5.2 Palatability: specific and seasonal selection according to animal species and management method

5.3 Accessibility

5.4 Effect of grazing and pruning and trimming practices on browse potential

6. Conclusion

Bibliography

Introduction

This inventory forms part of the study on animal production systems carried out by a multidisciplinary team from ILCA. The area involved covers the inner delta of the Niger and its western and northern fringes, an area of 71,000 km².

The method followed for compiling the inventory, illustrated in the initial results, is presented in three parts. The identification, survey and mapping of browse plant formations are discussed in the first part, plant formations being considered as the basic units for evaluating browse potential. In the second part specific foliage production measurements are briefly presented, since they are more fully set out in a more detailed publication (Cissé, 1980 a,b). In the third part, the integration of production data with population descriptions is demonstrated first by means of an acacia species of the Sahel, secondly with a perimeter on the livestock ranch at Niono, and finally for an area as a whole, comprising the first attempt at a regional assessment for the ILCA study area in Mali. After the explanation of the inventory method, the discussion moves on to the practical value of the results, taking into account the peculiar features of browse consumption.

2. Plant formations as basic units in compiling a regional browse potential inventory

2.1 Identification and mapping of plant formations

Plant formations are identified by a combination of phytoecological criteria. The physical and floristic structure of the vegetation and the principal edaphic and biotic characteristics are taken into account, following the methods advocated by CEPE ecologists (Godron et al, 1968).

Apart from inundated grazing such as are found in the inner delta of the Niger, the vegetation of the plant formations in the Sahel and Sudano-sahel zones is primarily defined by the floristic and structural characteristics of ligneous populations. The grass cover, which is dominated by annuals, fluctuated very considerably from one year to the next, as regards both flora and the production level (Hiernaux, Cissé and Diana, 1979), reducing its indicative value. As an example a description of one of the plant formations found on the ranch at Niono is given, as represented on the vegetation map for the Niono ranch (Hiernaux, 1977).

Formation (code 2):

Plant community with Pterocarpus lucens, Combretum micrantum, Grewia bicolor, Acacia ataxacantha, Lannea acida.

The population is regular and contagious, and the stratification is more or less regular: beneath a high shrub stratum there is a bush stratum with low shrubs; the vegetation occurs in clumps.

The grass cover is discontinuous (especially where there are clumps) and very diverse, forming a mosaic in which the following dominate: Loudetia togoensis, Diheteropogon hagerupii, Borreria chaetocephala, Andropogon pseudapricus, Panicum walense, Pennisetum pedicellatum. Bottom-land with a very humped microrelief, with the clumps forming on the ridges; alluvial plain; hydromorphic soil with a sandy loam texture interspersed with clay seams. Two degrees of degradation, coded 2' and 2" are identified for this formation.

Furthermore as regards survey or mapping work, it is the ligneous formations which show up most of all on the photographs currently used.

Ligneous populations are thus the major components in the identification of plant formations. In the ILCA study area 75 types of ligneous community were identified (to which should be added 25 degrees of degradation). They are currently being mapped.

2.2 The description of ligneous communities

For each of the plant formations identified a methodical survey of the physical and floristic characteristics of the ligneous community was carried out. The site from which the sample is taken is determined on aerial photographs, and consists of 160 m² (i.e. an area of 2.56ha). Without going into details of the method used (Hiernaux, 1976), we may state that it involves:
a) Counting the number of stocks per species;
b) Describing each of the stocks, including number and height of trunks, overall cover and cover for each stratum.

The following observations are then carried out:
a) Plot by plot observations on 16 adjacent square plots 10 M² each, forming a strip 10 m wide by 160 m long;
b)Extra surveys are then carried out for new species and for species with a forage value which were not adequately represented on the plots previously inventoried, using adjacent strips which are 10 m, then 20 m, then 40 m and finally 80 m wide.

2.2.1 Floristic description

The first component of a description is a list of flora. The list is drawn up so that the habitat-species curve can be systematically identified. This allows the floristic homogeneity of the samosample to be checked a posteriori. It also allows the optimum phytosociological range for each population to be calculated. For example, Figure 1 gives the range-species curves using both arithmetical and gaussian logarithmic coordinates for an acacia community dominated by Acacia laeta (relevé no. 25).

In the latter case linear adjustment confirms the floristic homogeneity of the population studied and allows the optimum phytosociological range to be calculated (Poissonet and César, 1972). It is approximately 12,000 km², a lower value than average values usually found, which lie between 2500 and 600 m². It should be noted that for grass cover in the Sahel the range is usually between 100 and 200 m².

Moreover, the observations made enable the list of flora to be compiled in accordance with the share of each species in the total number of stocks or in the total cover (see Table 1). Frequently it turns out that, as in the case of the example selected, these two classification are different.

Table 1.Classifications of species within an acacia community (rp 25) as function of the number of stocks per unit area and total cover.

Share in stock numbers (out of 1612 total)	No. of stocks ha	%	Share of canopy cover (overall, 16.5%)	No. of stocks ha	%
Boscia senegalensis	869	53.9	Acacia laeta	11.6	68.6
Acacia laeta	400	24.8	Boscia senegalensis	2.5	14.8
Acacia seyal	131	8.1	Acacia seyal	2.1	12.4
Boscia angustifolia	31	1.9	Balanites aesyptiaca	0.2	1.2
Cadaba farinosa	31	1.9	Cadaba farinosa	0.1	0.6
Dischrostachys glomerata	31	1.9	Boscia angustifolia	0.1	0.6
Balanites aegyptiaca	25	1.5	Acacia nilotica	0.1	0.6
Acacia nilotica	19	1.2	Commiphora africana	0.1	0.6
Other species	75	4.6	Other species	0.1	0.6
Total	1612	99.8	Total	16.9	100

2.2.2 Description of physical structure

Observations made during the survey enable the number of stocks per hectare, the cover of each stratum and the distribution of trunks as a function of height to be calculated for each species. For the plant taken as an example, Table 2 shows this distribution in absolute and relative values for the species with the greatest forage value.

Table 2.Distribution of trunks as a function of their diameter for the main browse species in a Sahel acacia colony (rp 25)

Trunk diameter at 10 cm (in cm)	Acacia Laeta		Acacia seyal		Acacia nilotica		Boscia senegalensis		Balanites aegyptiaca
	/ha	%	/ha	%	/ha	%	/ha	%	/ha	%
0–2.5	181	32.2	6	3.8	–	–	1588	90.7	6	25
2.5–5	131	23.3	44	26.9	3	14.3	113	6.4	13	50
5–10	181	32.2	50	30.8	5	21.4	44	2.5	6	25
10–15	63	11.1	50	30.8	8	36.7	6	0.4	–	–
15–20	6	1.1	13	7.7	3	14.3	–	–	–	–
20– 25	–	–	–	–	3	14.3	–	-	–	–
Total	562	100	163	100	22	100	1751	100	25	100

3.1 annual foliage production by species

The production measurements carried out by Cissé (1980) and Hiernaux et al (1979), took the following into account separately: annual total foliage production as a function of physical characteristics of the tree, and production curve.

3.1 Annual production as a function of height of trunk, and height and area of the crown

Cissé took direct measurements by felling and stripping carried out at the optimum vegetation stage on 5 to 10 individuals per circumference, plus for the main browse species of the zone under study. These were: Pterocarpus lucens, Acacia seyal, Acacia albida, Ziziphus mauritiana, Balanites aegyptiaca. Acacia Senegal had already been measured in a Sahel area of Senegal (Poupon, 1976). Production measurements have also been carried out on bushes, such as Combretum aculeatum, Cadaba farinosa, Boscia senegalensis, Ferethia apodanthera.

In trees, Cissé found a highly significant correlation between the circumference or height and annual foliage production (Cissé, 1980). In both cases and for each species he established an exponential relationship.

The relationship between production and canopy cover was not systematically established. However, for Acacia seyal, Pterocarpus lucens, Acacia albida and Ziziphus mauritania, the relationship is of the same kind as above. For bushes the data given by Cissé are currently limited to the average values which can be used for bushes over 50 cm in height.

3.2 Seasonal distribution of browse potential

This is established by means of direct measurements carried out every 15 or 30 days on whole stocks for bushes and on a sample of graded branches for trees. The available foliage is expressed as a percentage of total production per month or fortnight. Table 3 gives two examples.

Table 3. Distribution of foliage biomass over time for two Sahel species. Foliage biomass is expressed as a percentage of annual production.

	June	July	Aug.	Sept.	Oct.	Nov.	Dec.	Jan.	Feb.
Combretum aculeatum	i(f10	13	37	73	94	63	40	15	15FI0
Commiphora africana	i(f10	37	68	98	23	0	0	0	0

4. Towards a functional expression of the feed potential of browse populations

4.1 The integration of production data with population descriptions

This is done by means of survey work. A Sahel acacia colony (Relevé no. 25) will be taken as an example. The first stage consists of evaluating the annual forage potential. In order to do this it is necessary:

c) to add up the specific foliage products.

This process is carried out in Table 4, for an acacia colony taken as an example. The second stage involves an attempt to establish this evaluation for each season. For this purpose the seasonal distributions of foliage biomass established for each species are used (see para. 3.2), which are simply multiplied by the specific annual products calculated per unit area. This is done in Table 5 for the acacia colony taken as an example.

Table 4. Distribution of numbers and foliage production as a function of species and trunk height For a Sahel acacia colony (relevé 25)

Table 5.Contribution of species to available browse per season, using a typical Sahel colony as an example (relevé 25)

^a Seasonal foliage biomass as % of annual production

^b Seasonal foliage biomass in kg of DM/ha

The figures for available foliage per season for those species with a forage value provide an evaluation of the feed potential of browse populations which can be easily brought into focus by adding the corresponding values for forage units and digestible crude protein.

4.2 Assessment of the feed potential of perimeter on the Niono ranch

All that has to be done is to estimate the feed potential of the ligneous population of each component formation, then to correct these values for the area occupied by each formation in the perimeter under study. This operation is carried out in Table 6 for pasture on the Niono ranch, for which a map of the ligneous populations was compiled (Hiernaux and Haywood,, 1977); see also Boudet (1970). The very considerable fluctuation of the feed potential from one formation to another, ranging from 3.5 t of foliage production per year (in DM/ ha) for a dense Pterocarpus lucens thicket to 21 kg for an open tree formation with Seleroyarya birrea and Guiera senegalensis, should be noted. A further verification of the feed potential of Growse populations on the ranch (0.992 t DM/ha) puts the figure at almost half the product of the grass cover, which on the ranch lies at between 1.5 and 2 t of DM/ha, depending on the year.

Table 6. Plant formations on tide mono ranch, main features, area and forage production of ligneous populations

4.3 Regional assessment of feed potential

Although the data are still very incomplete, a first approximation of a table of annual feed potential has been drawn up for ligneous populations of the Sahelian and Sudano-Sahelian sectors of the ILCA study area in Mali. Table 7 gives the main results. The regional feed potential results for ligneous populations are a little below half the corresponding grass production, except in the case of the delta.

Table 7. Annual browse potential per ecological sector of the ILCA study area in Mali

Ecological regions	Live delta	Dead Delta		Continental shelf		Whole areaa
and sectors	(Farimaké de de Péroudji	Sahelian	South Sahelian	Sahelian	South Sahelian
Average per sector in t. of DM/ha	0.27	0.50	0.75	0.70	0.86	0.52
Per sector estimate in '000 t of DM	100	425	560	763	1035	2784

'North Sudanian sectors excluded.

5. Discussion: problems arising from the characteristics of browse consumption with regard to the functional expression of feed potential

The biomass of available feed for each season can be taken as a very rough expression of the feed potential of the ligneous population, when the particular characteristics of the consumption of these feeds are taken into consideration, and especially the following four points.

5.1 Supplementary feed or basic feed

An experiment on the Niono ranch in late 1978 confirmed that goats are satisfied with a feed entirely made up of tree leaves, even those of only one variety (in this case Acacia seyal and Pterocarpus lucens see "Rapport d'Activite" 1978–79) at a rate of 3.6 kg and 2.56 kg of DM/100 kg liveweight respectively, in a closely controlled experiment involving stall feeding with weight gains of 2% and 6%per month for goats weighing 25 kg on average). On the other hand, the observations made by Mme Dicko-Toure illustrate the limited part which can be played by these feeds, which is extremely variable from one season to the next, under 1% to 25% of grazing time depending on the month, for the sedentary system of the Office du Niger (Dicko-Touré, personal communication).

Similarly, for the transhumant Macina production system (Diallo,1978) only 7 days of browse consumption were recorded out of 23 days distributed over the annual cycle. The average share of browse was thought to be 3.9%, but with very considerable variations, since Diallo recorded a browse proportion of 81% at the daily rate in an observation carried out during the migration of animals into the Sahel, while other readings were lower than 3%.

Leaving aside the case of goats, browse availability should thus be regarded only as playing a potential role as a supplementary feed; it should therefore not be considered in isolation from the available grass feed.

5.2 Palatability: specific and seasonal selection according to animal species and management method

This is the most confusing aspect of browse consumption. The idea of palatability is far more relative here than with grass feeds. Observations gleaned from various trials on the Niono ranch lead to the impression that palatability depends on:

a) plant species and its phenological stage; this is well known, but further factors are:

b) animal species, together with sex and age, but more especially

c) feeding habits as linked to management method.

In particular, since controlled grazing is the exception, there is a great deal of difference between the behaviour of animals in sedentary and transhumant systems: the first are usually reduced to following the same grazing route every day, at least for some period of the year.

In this case the animals select their menu with a full knowledge of the facts, so to speak. And in this case it was possible to note the importance which the animals attached to the variety of their menu. If the grazing route covers pastures which vary very little, the animal will search out areas where "new" feeds can be found, even if these are elsewhere alleged to be less palatable or even refused altogether. This is the case, for example, for Guiera senegalensis and also for Combretum ghazalense, and even for Callotropis procera, for herds grazing on the periphery of the ODN rice plots.

On the other hand, the selection processes of animals accustomed to changing their grazing area every day are more complex. Possibly the influence of the herdsman is greater here and it would seem, at least for zebus, that the contribution of browse is less than in the preceding case. The method used in the estimates, i.e. that of establishing the list of palatable species, then using it to work out the list of species which are left untouched, is therefore a simplified one which could lead to misunderstandings.

5.3 Accessibility

Only a small proportion of "available" browse is directly accessible to animals. If 2 m is taken as the upper height limit for direct access, Table 8 shows the share of foliage biomass directly accessible for three species of the Sahel, both in absolute values and as a percentage of the overall foliage biomass of the tree.

Table 8 shows that this proportion varies very considerably according to the height of the tree, but also according to the species. From these values it should be possible to evaluate this proportion for each of the sample areas were it not for the fact that the herdsman's habit of cutting and trimming the branches, with the result of placing the non-accessible parts of the tree within reach of the animals, renders this exercise somewhat vain. In the acacia colony taken as an example it may be noted that of the Acacia seyal stocks (two-thirds of shrubs higher than 1 m), as well as Balanites, Acacia nilotica, Ziziphus and even Acacia laeta plants (20% of shrubs 2 m high) were all cut.

Table 8. quantity of foliage directly accessible to animals (under 2 m high) as a function of tree height for three browse species of the Sahel.

Pterocarpus lucens			Ziziphus mauritiana		Acacia seyal
			Foliage biomass over 2 m
Height of tree	In kg of DM	% of total b.	In kg of DM	% of total b.	In kg of DM	% of total b.
0–1	–	–	–	–	0.08	100
1 – 2	0.11	100	0.04	100	0.10	100
2 – 3	0.35	81	0.13	37	0.35	59
3 – 4	0.83	42	0.16	16	0.24	15
4 – 5	1.10	27	0.15	7	0.28	8
5 – 6	0.28	8	0.09	3	–	–
6– 7	0.40	6	0.11	5	0.24	7
7– 8	0.45	7	0	0	–	–

5.4 Effect of grazing and pruning and trimming practices on browse potential

Obviously browse, and especially the processes associated with browsing, have important consequences for ligneous forage production.

a) In the short term, over and above the process by which non-accessible parts are turned into accessible ones, there are aftereffects on the trees subject to browse processes during the following years. These aftereffects concern both quantity and quality since as a general rule the new shoots are mainly vegetative with a phenological cycle more stretched out in time than that of natural growths. A complex experimental procedure has been launched to quantify these effects for a number of species (Acacia seyal, Acacia albida, Pterocarpus lucens). Measurements are under way. From the first results (see compte-rendu d'Activité 1978/79, pp. 37/44), it may be seen that trimming has an effect which varies very considerably according to species, amounting to 24.4% for Acacia albida but 70.3% for Acacia seyal and 71.82% for Pterocarpus lucens, in the case of trimming carried out at the end of the growth season.

The effect of browsing as such has not been measured; on the other hand, data have been collected by Cissé on the effects of various types of stripping intended to simulate the more extreme cases of browsing. These trials (Cissé, 1980) were carried out on a number of browse bushes, and it has been found that the effect of stripping on the production for that year need not always be depressing and that it varies enormously according to the species. The aftereffects on production the following year also vary and are primarily linked with the intensity of stripping.

b) In the long term, changes in production may occur through alterations in the population. Usually this tendency develops to the detriment of those species most valued in terms of feed, since the underutilized species take the place of those which die out as a result of over-intensive utilization. This is the case for the formation coded 2" which is frequently found on the edge of the rice plots managed by the ODN. This is a degraded form caused by cutting and overgrazing, especially by goats, of the formation coded 2, which is described as an example in the first section. Table 9, comparing the floristic composition of two populations, illustrates this development. The browse species, grouped on the left of the table, decline in terms of overall cover, while rejects such as Combretum niorontum and Acacia ateracantum increase; Grewia bicolor is in an intermediate position.

Table 9. Comparison of the floristic structure of a browse colony in the southern Sahel with the same type of colony in a degraded state caused by overgrazing and cutting

	Plant formation	Pterocarpus lucens	Combretum aculeatum	Acacia seyal	Grewia bicolor	Combretum micrantum	Acacia ataxacantha	Whole colony
Cover as % of total area	Original	15.2	1.3	1.5	0.7	2.4	0.4	12.4
	Degraded	1.2	0.1	0.3	0.2	6.8	3.7	14.2
Cover as % of total cover	Original	67.9	5.8	6.7	3.3	10.7	1.9	100
	Degraded	8.1	1	2.2	1.8	48.6	26.4	100
No. of areas per ha	Original	87	168	50	25	100	31	618
	Degraded	37	50	12	25	350	150	1000
No. of stock as % of total no.	Original	12.8	24.8	7.3	3.7	14.7	4.6	100
	Degraded	3.7	5	1.2	2.5	35	15	100

However, not all degradation takes the form of a decline in browse potential. It can happen that the secondary formation consists of varieties which are more valuable in teens of feed than the original formation (for example, a formation dominated by Combretum glutinosum was improved by the addition of Balanites aegyptiaca, Acacia Senegal and Acacia seyal).

6. Conclusion

A somewhat laborious method has been advanced for reaching an estimate of the seasonal feed availability provided by browse populations of the Sahel. It combines a precise description of populations with a series of production measurements. Since the latter can be obtained on a one-off basis, the general application of the method becomes easier. However, the peculiar features of browse consumption described in the last few paragraphs effectively prevent any automatic application of the results obtained. Since browse consumption depend as much on the type of animal as on the herd management method, the calculation of a carrying capacity as a function of browse potential expressed in this way could only be undertaken for a clearly defined animal production system.

Bibliography

Boudet, G. (1970). 'Etude agrostologique pour la création d'une station d'embouche dans la région de Niono (Rép. du Mali). Maisons-Alfort, IEMVT, Et. Agrost No. 29, 268 p.

Cissé, M.I. (1980). 'Effet de divers régime d'effeuillage sur la production foliaire de quelques buissons fourragers de la zone soudano-sahélienne'. Acte du colloque sur les Fourrages Ligneux en Afrique, Addis Abéba, CIPEA.

Cissé M.I. (1980). 'Production fourragère de quelques arbres sahéliens relation entre la biomasse foliaire maximale et divers paramètres physiques. Acte du colloque sur les Fourrages Ligneux en Afrique, Addis Abéba, CIPEA.