Iv. Biology and primary production

Measuring the primary palatable production of browse plants**

J.C. Bille

Senior Ecologist, ILCA, Nairobi, Kenya.

Primary palatable production:A definition

Measurement methods

Biomass from individual browse plants

Analysis of measurements

Interannual variations

Ligneous populations

Pluri-annual productivity variations

Browse productivity and climate

Trees in the ecosystem

Bibliograp[hy

Interest in browse production and its utilization by domestic and wild animals is a recent phenomenon. Although traditional herdsmen have for centuries recognized the value of browse, scientific work on the subject in tropical Africa is generally less than 15 years old, and the best known works describe less than a dozen sites (Figure 1), mostly those studied as part of the International Biological Programme. In this paper the results will simply be related to the country in which the measurements were carried out.

Figure 1.Sites of main research projects quoted in this paper, with year of measurements or beginning of measurements.

Primary palatable production: A definition

Doubt as to the quantity of edible ligneous matter is partly related to the lack of a clear definition of browse products. According to the climate and the authors, total biomass may be taken into account, or else net total productivity, which is higher than biomass if the foliage is strictly deciduous, but which can be much lower for foliage which is relatively persistent, or for other plant elements such as young branches or bark.

Alternatively, the accessible mass can be considered, usually located less than 2 m from the ground, but in this case camels and many wild animals are not taken into account. Sometimes there is a tendency to include in biomass only the young leaves which are sought after by animals, whereas the older and tougher elements are left aside under observation conditions. The results obtained in different cases are scarcely comparable. Similarly, usually the quantity of leaves on the tree is measured as well as the quantities of flowers and fruits falling to the ground, but foliage production is also estimated on the basis of leaf litter. Finally, interannual or pluriannual variations compound the difficulties of interpreting the results.

A further problem can also be raised straight away: that of individual variations in accordance with the form and general appearance of the sample tree. Thus for example, in Mali, the foliage mass of young trees varies from 10 to 90 g per individual, while for older plants the biomasses quoted are 1000 ± 600 g and 3.6 ± 1.6 kg, with averages established on the basis of five measurements. Again these are total masses to which no subjective coefficient has been applied to express accessibility or utilization.

Measurement methods

Net primary production results from apparent photosynthesis and the most logical way of measuring it would be to determine the gaseous interchange within the plant. In practice it would not be easy to establish the quantities of carbon absorbed by a whole tree and measurements rely on a few leaves (but where does one stand with the chlorophylleous branches of Balanites?) The results are converted by using the leaf area index or some other means, but the relationships needed for arriving at real production are far from simple. Using radioisotopes has the advantage of allowing samples to be multiplied and events to be observed without, disrupting them: however, apart from technical methods which rarely accessible, the method still requires a rather large number of hypotheses before the quantity of material produced in various elements can be deduced.

As a result most work uses direct harvesting of the production and determination of the biomass. This process is more or less destructive according to the method adopted: litter harvesting, total destruction of a few individuals with establishment of relationships between product and some other more accessible parameter, or sampling of parts of individuals, implying the subsequent use of two successive correlations simply to estimate the biomass per unit tree. Even after a host of subsampling exercises accuracy remains an illusion, and to be convinced of this one has only to examine what is involved in measuring the foliage mass of a baobab, a process which includes the following stages: establishing the quantity of leaves per secondary branch (destructive), the number of secondary branches per primary (usually from a photograph), the volume of primary branches as a function of trunk circumference, and examining the population of baobabs.

The result is that measuring browse plants discourages investigators with a taste for accuracy, and that work in which all the components necessary for a proper estimate are explicit is rare. If a standard procedure based on the most commonly applied practices could be put forward it would doubtless consist of the following:

1. Destructive measurement of biomasses per individual as a function of stem diameter;
2. Inter-and pluri-annual variations of these biomasses on a standard branch;

3. Description of specific stands by unit area on the ground and plant community.

Biomass from individual browse plants

Describing measurement methods would be of limited interest if unrelated to the results obtained, which are the only indisputable criteria of the value of the methods used. The elements selected for their browse relevance will be reviewed here, quoting a few results.

a) Bark: this is of little or no interest to domestic livestock, but a constant food source for rodents and elephants. Measurements are rare but include, for example, 430 kg of DM in Senegal for an epigeal ligneous biomass of 1850 kg/ha, or in other words nearly a quarter of the above-ground biomass. Interspecific variability is high, and pyrophytes have thick bark which is unpalatable. Some barks are used in small-scale local crafts (Grewia, and Ficus for example).

b)Branches: new growth is always liable to be consumed and is often measured together with the leaves. The branches and branchlets (less than 5 cm in diameter), again in Senegal, represent from 10 to 15% of the total biomass in Balanites, 8 to 10% for Commiphora, 15 to 25% for Acacia senegal 30 to 45% for Guiera and 15 to 30% for Grewia bicolor. The biomass of branches can thus be linked with the circumference of the tree, or its height, using relationships similar to those described for total mass as a function of the same parameters.

In Zambia, the sum of the diameters of the branches appears to be fairly constant for an average individual in a given species: 200 mm in Brachystegia which had been cut back, 160 mm for Julbernardia or 100 mm for Isoberlinia. These results were for branches produced during a given year, and these values might represent an interesting characteristic.

c)Flowers and fruit: although the feed value of flowers and accessory elements is hardly satisfactory, the fruit, and to an even greater extent the seeds, are important in this respect and sometimes have a by-no-means insignificant biomass. In Acacia the fruit may equal a quarter of the annual deciduous product, and the proportion is even higher for Balanites; however, Balanites fruit is often used as a human foodstuff. Strychnos, Annona, and Solanum shrubs also produce a fine quantity of fruit; however, it must be admitted that the fruit often represents only 10% or less of the deciduous biomass and that the quantities of 400 to 600 kg/ha of pods quoted for acacia are exceptional.

d)Leaves: although the measurement of leaf production is as complex as could be, it is one of the most frequently carried out measurements. Tables 1 to 3 show examples of leaf measurement, and it will be noted that in the arid and semiarid zones the leaf biomass is fairly low. In almost all cases the values were obtained by completely stripping a number of individuals, precluding any subsequent measurement of the same sample, whether the trees were actually felled or whether growth has been irreparably disrupted.

Table 1. Foliage and fruit biomass per shrub (Senegal, rainfall 250 mm).

Table 2.Foliage mass per individual (Upper Volta, rainfall 440 mm).

Diameter	0	5	10	15	20	25	30
Acacia laeta	150	500	2200	2500	3100	3500
A.seyal	60	520	1300	1700	3800	6900	8000
A. tortilis	50	250	300	700	1100	1200	1600
Balanites	50	80	150	2500	6000	8800	10500
Guiera	40	(430)	(850)	(900)

Table 3.Foliage mass per individual (Mali, rainfall 600 mm)

Diameter	1.6	4.8	9.5	14.3	20.7	25.4	30.2
A. albida	40	340	1400	3100	6450	9600	13000
Pterocarpus Lucens	50	260	1600	3980	5050	6450
Ziziphus m.	50	220	1130	1420	4700
Acacia	50	340	1780	3610

Variability: 50 g ± 40 g (80%), 1000 g ± 600 g (60%), 3600 g + 1600 g (45%).

In the humid zone and forest areas leaf litter has been more frequently used. For example, an average value of 5.7 kg of DM per tree was obtained in Ivory Coast, in an area with 1250 mm of rain.

These processes make it unlikely that the mineral cycles described for forest zones can be applied to the savanna, and the enriched soil quality found beneath trees is probably better explained by the partial renewal of the roots each year, which produce almost as much as the leaves. The foliage on the ground probably undergoes two periods of rapid degradation, the first immediately after the leaves fall and the second at the outset of the following rainy season, with a relatively stable period without decomposition between the two.

Analysis of measurements

The results obtained in the field are processed using regression equations or curves as shown in Figures 2 to 4. The sampling parameters most commonly used are the stem circumference (or diameter) and the height of tree, but it can be easily demonstrated that these two values are connected for a given species, and the stem circumference is clearly the easiest measurement to take, especially since the ligneous density often influences the habit of trees and shrubs.

Figure 2. Leaf biomass as a function of trunk diameter for two Sahel species.

Figure 3. Comparative leaf / biomasses of various species as a function of trunk diameter.

Figure 4. Examples of a relationship between the log of foliage mass ( in g DM) and the log of stem diameter ( in cm).

The type of regression equation which can be used is still open to discussion. Sometimes a first degree ratio is convenient, as also are second or third degree polynomials. Nevertheless, it is almost always possible to express the logarithm of the biomass lineally as a function of the logarithm of stem circumference, i.e. log m. = a log c + b.

It will also be noted that the coefficients a and b do not vary very much for leaves, with b often lower than 1 and a approximately 2 or 2.5. The mean equation log m = 2 1og (diameter), + 1, corresponding to 250 g of leaves for a stem diameter of 5 cm, 2.2. kg for d =10 cm and 6.2 kg for d =15 cm probably constitutes an acceptable approximation in the Sahel zone. Thus an average tree in the Sahel probably barely produces a kilogramme of leaves per year (and 250 g of fruit, plus 4 to 5 kg of branches).

In dry areas the duration of the period in leaf is 5 to 7 months for species with annual foliation, and the fruit outlives the leaves sometimes by 2 or 3 months (Commiphora, Acacia).

What happens when the fruit falls to the ground is still little understood: in Senegal various consumers, especially termites, use between 5 and 20% of it, depending on the microenvironment. Its chemical composition is altered by partial destruction of the protein content and by mineral leaching, while the biomass is scattered and redistributed by the wind.

Interannual variations

Interannual variations are relevant as regards quasi-persistent foliage. Table 4 shows the fluctuations measured in Zambia, but it should be noted that the trees were browsed by livestock and that, given a defoliation level of 50% or more, the yields were seriously affected and also many trees did not survive the treatment.

Table 4. Leaf biomass on a 10 mm branch (Zambia, rainfall 1250 mm) in g/dm

Date	Julbernardia paniculata		Isoberlinia angolensis	Brachystegia spiciformis	Brachystegia longifolia	Baphia bequaerti
1970 XII	16.9		26.7	24.7	26.6	9.8
I-II	19.5		20.0	15.5	21.0	13.1
III	18.5		16.8	9.7	16.9	2.0
IV-V	18.3		23.5	10.5	12.9	0.0
VI	12.5		9.8	7.2	19.0	dead
VII-VIII	19.8		23.1	11.3	6.1
X	26.5		21.1	11.4	29.8
1971
XII		27.4	29.1	10.8	27.1
I-II		5.9	29.5	28.6	30.3
III		0.4	6.2	1.8	16.9
IV-V		11.3	14.1	0.0	25.3
VI		7.0	11.7	.0	11.9
VII-VIII		12.0	21.8	0.0	0.0
XII		19.8	24.9	2.4	26.4
Defoliation intensity
		50	20	60	20	100
Mortality over two years
		27	0	97		100

Ligneous populations

Although describing ligneous populations does not require lengthy or difficult measurements, interest in doing so has only recently emerged, apart from a few cautious estimates of the total numbers of individual ligneous plants. The total number of trees per unit area can be obtained by simply counting, using sample areas or more elaborate statistical methods, of which the best known is the PCQ method (Point Centred Quarter), whereby the distance is measured from the centre of a cross to the nearest tree in each of the four right angles it forms, an operation repeated between 30 and 100 times. However, for a unit area in a given environment the best form of characterization consists of determining the number of trees and shrubs in different circumference or diameter classes so as to draw diagrams either of the overall population or of a particular species.

Table 5 shows the browse population in the case of Senegal, while Figures 5 and 6 show examples of the diagrams obtained.

Table 5.Ligneous populations: distribution by circumference class (Senegal)

Circumference	Guiera	Boscia	Grewia	Acacia	Balanites	Commiphora	Total
0–10	34.2						36.5
10–15	22.4	4.0					28.7
15–20	9.1	3.3	4.7	4.0	5.2		19.4
20–25	2.8	1.0	2.8				8.9
25–30		0.7	2.3	2.6	4.0	0.3	9.9
30–40			2.8	2.0	2.9	0.7	8.4
40–50			1.3	1.0	2.4	1.5	6.2
50–60				0.4	1.1	2.7	4.2
60–80					1.0	4.0	5.0
Over					0.7	1.6	2.3
Total	68.5	9.0	13.9	10.0	17.3	10.8	129.5

Figure 5. Diagrams of total ligneous populations

( no. of individuals per diameter of circumference class)

Figure 6.Population structure of some species

Figures 5 and 6 give example of diagrams of this kind but they require several points to be made:

a) Younger individuals are usually more numerous and they die off gradually over successive years; this is why two forms of regression are put forward to describe these stands; the first is a negative exponential and expresses the various classes, while the second is a geometrical progression with a ratio of less than one, the terms of which can be recategorized in order to reestablish the circumference classes.

b) Regression equations enable theoretical stands to be considered from which actual stands will diverge to a greater or lesser degree, revealing favourable or unfavourable periods for trees or for a specific species; in particular it should be noted that the initial slope of the curve shows the degree of aggression of the stand-in these examples it was very high in Ivory Coast but negative in Senegal.

c) It would also be noted that some species are apparently on the way to extinction (Cussonia barteri) and that regeneration is highly irregular, possibly requiring a number of conditions which are rarely combined. In Kenya almost all the actual curves show a deflection at the point in time when recent ranches were set up and may thus be used as management indicators.

Thus the diagrams and their comparison in time at the same point provide varied and important information on the dynamism of the ligneous stratum (and sometimes as to its probable causal agents), the reactions of species to environmental variations (including their resulting palatability), and in a general way on the balance of eco-systems. Moreover, the total number of trees and shrubs in many cases probably gives an adequate indication of ligneous productivity, if the description of the high stratum had to be restricted to a single parameter, the parameter chosen should be population study.

Pluri-annual productivity variations

The combination of information on tree populations and on feed biomass per tree should enable the biomasses per unit area to be put forward. Independently of the reservations already expressed on the relationships between biomasses and productivity, the validity of the results should be examined as a function of the climatic context of the measurements. The ligneous productivity as a whole is highly dependent on rainfall.

Table 6. Leaf and fruit biomasses/ha: pluri-annual variations (Senegal) in kg/DM.

Species	1971	1972	1973	Average 1970–73
Balanites	22.0	6.3	15.2	16.5
Commiphora	8.0	3.2	9.3	7.9
Acacia	16.1	6.2	10.3	11.6
Guiera	21.2	8.3	17.9	16.8
Grewia	35.6	6.6	17.1	24.0
Total	102.9	30.6	70.5	76.9

The enormous variations observed here are due to the great drought of 1972, during which the zone received only 30 mm of rain. It is important to note that productivity in 1973 is still far from reaching its usual level, for two reasons:

a) Many trees are dead, especially in their upper parts, or have been replaced by young Boscia, which are unproductive, on slopes. Table 7 compares the stands before and after the drought.

b) A dry year has a definite aftereffect on the following year. The aftereffect is also reflected in wood production and can be identified by analysing radial growth of the tree, although growth rings are not always very easy to see in the tropics and they are difficult to situate in time. Nonetheless it is true that in areas with a very pronounced dry season the thickness of growth rings varies in a ratio of one to three, and the same ratio could be applied in particular to leaf production since leaves are photosynthetic.

Table 7.ligneous populations pluri-annual variations (Senegal)

Circumference	Crests		No. of individuals on Slopes		Dips
	1970	1975	1970	1975	1970	1975
0–10	29	1	27	31	81	53
10–20	41	17	44	148	109	190
20–30	15	10	18	30	36	38
30–40	8	10	7	6	5	9
40–50	5	5	6	4	3	1
50–60	3	3	5	2	2	1
60–80	3	2	1	3	0	0
80	2	0	0	0	0	0
Total	106	48	108	224	238	292

A last cause of variations is the intensity of utilization by domestic or wild animals, not only during the year of measurement but also during previous years, as a function of the above. Complete measurements fail dismally in this respect besides the Zambia experiment cited above, work is also in progress in Tanzania, taking into account a high sampling level (60 to 80% for acacia). It is still too early to describe the long term consequences. In Mali also this aspect is one of the major subjects under study. Despite the possibility of intensive utilization over a short period, it many well be sensible to recommend a cautious approach to the utilization of browse plants.

Browse productivity and climate

The values most frequently quoted are biomasses per unit area. Figure 7 summarises estimates for leaves in a number of different places, and it would seem that foliage biomass increases very rapidly when rainfall is above 200 to 1000 mm per year. Beyond this point variations are highly important (from 300 kg/ha in Cameroon to over 2 t/ha for subforest savanna), but the average results imply that often the limit lies around 1500 kg/ha, and this is doubtless linked with the substratum.

Figure 7.Productivity as a function of climate

The increase in productivity is partially explained by the variations in the number of trees and shrubs when we move from the arid zone to the humid zone. However, in some cases it has been possible to establish leaf production for the "average" tree; the regression curve between these values and annual rainfall appears more like a straight line, and the establishment of a relationship of this kind would be highly desirable. Provisionally the following orders of magnitude can be selected: 400 g of DM in the arid zone, 700 g in the semiarid zone and 1 kg or more for the others but in the latter case production would only be very partially accessible or palatable.

Trees in the ecosystem

An examination of the methods for evaluating the ligneous stratum and of the results obtained from these methods has frequently led to the drawing of inferences, or to consequences arising from the information about trees. These indications generally relate to the whole ecosystem under consideration, they concern relationships between the high and low strata, the significance of the structure of ligneous stands, the reactions of the tree to utilization, to name only the most important points. To this should be added the problems arising through spontaneous or accidental deforestation, and the renewed interest in the ligneous stratum, a renewed interest often linked to overall study of natural biological system.

The general conclusion will thus appear to be that trees have an essential role to play in natural savanna. The tree would appear to be at once an indispensable component in savanna and the best indicator for pastoral management methods: the state of health of ligneous communities, their equilibrium, their productivity—and here it must be understood that too many trees or a production which is disproportionate with that of grass are just as much be condemned as the scarcity of ligneous elements—are thus the touchstone for evaluating the utilization of savanna by man. As a result the easier forms of monitoring the tree and shrub strata, and in particular the description of ligneous stands, should be systematically implemented.

Bibliography

Attiwill, P.M. (1966). 'A method for estimating crown weight in eucalyptus and some implications of relationships between crown weight and stem diameter. Ecology 47, 796

Bille, J.C. (1971). Observations préliminaires sur quelques arbres du Sahel sénégalais Dakar, ORSTOM.

Bille, J.C. (1971). Etude de la production primarie nette d'un écosystème sahélien Paris, Travoux et Doc. ORSTOM.

Braun, H.M.H. (1973). 'Primary production in the Serengeti' Ann. Univ. Abidjan, E, 6–2.

Bray, R.G. et Gorham, E. (1964). 'Litter production in forests of the world'. Adv. Ecol. Res. 2, 100–158.

Cissé, MI (1979). 'Production des ligneux fourragers' Rapport annuel d'activité 1978-79. Mali-CIPEA.

Duvigneaud, P. (1971). Concepts sur la productivité des écosystèmes forestiers Paris, UNESCO.

Hood, R.J. (1972). 'The development of a system of beef production for use in the Brachystegia woodlands of northern Zambia'. Ph.D. Thesis, University of Reading.

Hopkins, B. (1968). 'Vegetation of the Olokemeji Forest Reserve, Nigeria'. J. Ecol. 56, 97–115.

Lubke, R.A. (1976). 'A reassessment of the woody vegetation of the Nylsvley study area'. Report to Nat. Com. for Envir. Sc., S. Africa.

Menaut, J.C. (1971). 'Etude de quelques peuplements ligneux d'une savane guinéenne de Côte d'Ivoire'. Thése de spécialité. Paris: Faculté des Sciences de Paris.

Menaut, J.C. (1976). La production primaire des écosytèmes pâtures tropicaux, Paris, CNRS.

Menaut, J.C. (1977). 'Analyse quantitative des ligneux dans une savane arbustive préforestière de Côte d'Ivoire'. Geo. Eco. Trop. 1 (2), 77–94.

Misra, R et al. (1968). 'Dry matter production in sun and shade leaves, and a simple method for measurement of primary productivity'. Curr. Sc. 37–11.

Nebout, J.P. et Toutain, B. (1978). 'Etude sur les arbres fourragers dans la zone sahélienne (Oudalan voltaique)'. Paris, CIFT-IEMVT.

Newbould, D.J. (1967). I.B.P. Handbook No. 2. Primary production in forests. Oxford, Blackwell.

Pellew, R. (1980). 'Etudes en cours sur la productivité des ligneux au Serengeti, Tanzanie', communication personnelle.

Piot, J. (1969). 'Végétaux ligneux et pâturages des savanes dans 1'Adamoua au Cameroum'. Rev. IEMYT 22, 4, 541–59.

Poupon, H. (1976). 'La biomasse et 1'évolution de sa repartition au tours de la croissance d'Acacia Senegal dans une savane sahélienne'. Rev. Bois et For. Trop. 166.

Poupon, H. (1979). 'Structure et dynamique de la strate ligneux dune staffe sahélienne au nord du Sénéga1'. Thése Doct. Paris.

Rutherford, M.C., Carr, P.S. (1976).Woody species productivity within South African savanna ecosystem-project'. Report to Nat. Com. for Envir. Sc. of S. Africa.

Schuster, J.L. (1965). 'Estimated browse twig and stem measurements'. J. Range Manag. 18, 220.

Unesco, (1971). 'Productivité des écosystèmes forestiers'. Actes du colloque de Bruxelles Ecologie et Conservation No. 4 Paris, Unesco.

Whittaker, R. H., and Woodwell, G. M. (1971). 'Measurements of ner primary production of forests'. Paris, Unesco.