Review of discussions
1. Evaluations through mapping with computer assistance
2. Surveying the environment
3. Simulation models
4. Vegetation surveys
5. Collection data and sampling techniques
The discussions can be grouped into five main topics:
1) Evaluations through mapping with computer assistance;
2) Survey of the environment;
3) Simulation models;
4) Vegetation surveys;
5) Gathering of data and sampling techniques.
The technique shown by Dr. Volger is noteworthy. Can he specify the advantages of it?
This technique makes photo-interpretation easier and permits estimating the ratio of the different types of vegetation in one limited area. This is a quick method whose results can be recorded on the computer and finally give the exact number of hectares of the different vegetation types in a zone under study.
Are the computer capabilities suited to Africa, to the mapping techniques, and to the management method of pasturelands?
The proposed system can in fact function in these regions, since the government of Sumatra is going to use mini-computers to gather different data and information into a bank to be used in planning.
How is the sampling set up to result in the creation of a data bank? Is the work carried out through a collection of all the available components, or the systematic study of one aspect on which the different parameters are surveyed?
The method of sampling has been different for the 98 categories of information or data gathered.
What is the size of the sampling? Is it 1/100 or 1/10 of the area studied?
The number of sections studied for areas of different sizes can be very large, but in practice it's around 5,000 to 10,000. If an area is 4,000 to 6,000 square kilometres, the size of one section is one square kilometre. The soil control covers 5 percent of the total surface along a line that cuts through 15 different geographic settings that have been defined.
How is all this information gathered without the help of a computer?
The gathering of the data can be done by hand. Their processing can be much more difficult. The use of minicomputers in the field permits solving this problem.
What efforts have been taken by the researchers to associate the local people with their work?
The proposed system is very simple and can be used by non-specialists in information. The programs destined for large computers have also been rewritten to be used in small computers, which are always available near the places being studied and which allow for processing the information.
What Dr. Maiga says is true for all research. After beginning, it is necessary to train men from the area and in each field.
It is necessary to underline the importance of the shifting of nutritive elements from one zone to the next. For some years, it has been determined that the fertility of the Nile Valley has been decreasing, since the sediment has been blocked up. According to certain Egyptian researchers, the Nile Valley also receives deposits from the Sahara, and this point should not be overlooked.
What are the effects of this change in alluvial movement? What action do we have to take?
Dr. Lundholm is interested from the standpoint of the problems concerning pollution and the displacement of particles such as cadmium or lead. But have the losses by the Sahel in nutritive elements such as nitrogen, potash and phosphorous been measured?
There is very little information on these losses. Nevertheless, you could obtain some information from the study of the sedimented deposits.
There is a significant carrying of nutritive elements by dust. This has been studied in India and has already been the subject of a talk in Nairobi in 1974.
Dr. Penning De Vries' and Dr. Van Heemt's model has been sucessful if you consider the change in amount of rainfall and the biomass as well as the rainfall necessary for germination. However, the growth periods in experimentation are longer than those observed in the field. What are the reasons for this? What species have been used? Would it be possible to have some more details on this simulation model?
F. PENNING DE VRIES:
The simulation model is used as a measurement of the productivity of the vegetation in the Sahel. All kinds of annual species must be used in the formation of this model. The production of nutritive elements and the sampling and use of water are some of the factors that are simulated. Nitrogen is in low quantity and is quickly used up. That reduces the length of the growing season and the maximum amount of biomass obtained. The model works in this way: at any given time the amount of biomass present and the amount of water in the soil are measured; the rainfall and the evaporation rate of the plant are also surveyed. Thus, the process of water loss and plant growth are simulated from data obtained on a chronological basis day after day.
The simulation models are good due to the data and hypotheses that are included. For example, if rainfall is considered, it is not enough to take into consideration the average rainfall in a given period. It would be useful as well if the models were able to go beyond the prediction level of production under given soil and climatic conditions and allow for a projection of the estimated production in terms of particular management conditions of grazing intensities, etc. The probable productivity of a given system could be estimated and compared with the productivity following an imposed change in the system. From the viewpoint of those working in the field, it would be very useful if the inventors of the models asked them for the data they need. The men in the field would be very happy to be able to help them expand on their model.
C. DE WIT:
The simulation models take into account:
1) the processes that take place in the soil;
2) the parameters that are a result of the physical properties of the soil, and the physiology and ecology of the plants.
It is not possible to put every existing condition into model form. On the other hand, it is possible to represent either the potential production levels (for example, the increase in fodder-plants) or the limiting factors, water and soil, or the nitrogen cycle, although here there cannot be an exact replica of natural conditions. In considering the quantity and the types of data necessary for a simulation model, it is useful to begin the model with as little data as possible. It's not easy to know which data are necessary, and which are difficult to gather. The basic data include average rainfall data. The intensity of rainfall indicates, for example, the degree of water absorption and its loss. Concerning the soil, the data are usually limited to the topography and the principal physical properties which can be determined in an ordinary laboratory. It is necessary to take into account that simulation models have limits. For example, it is not possible to simulate the effects on grazing.
The model expansion permits obtaining the complement of what happens in the field. Our goal is to reduce the experimentation by 25 percent and to reduce as well the costs and efforts assumed. It is essential that the experimentation in the field continue simultaneously with that of the model and that those responsible for the model expansion go into the field and stay in contact with what is going on and what is being researched. The availability of qualified personnel and of computers should no longer pose problems in Africa.
The simulation techniques for annual species and perennial woody species are different. This is due in part to the long period that the perennial species must go through to survive from one year to the next. It is important that the data destined for simulation be deduced on the hectare or square metro basis, especially for the perennial species. An estimation of the harvests ought to be done at least once on foot.
It is nearly impossible to differentiate between living and dead roots. Thus, measurements of biomass are very difficult. It is also difficult to estimate the productivity of woody perennials. The technique used is to take a plant sample, a branch, etc., at random and to establish a sample unit that can be reasonably measured in terms of biomass, leaves produced, woody material, etc. These units can then be classified in order to give an estimation per hectare.
This estimation of the productivity can be compared through measurement of plant height, stem width, etc., and a relationship between these dimensions and the biomass can be researched. The development simulation would therefore imply an important sampling in the field. The productivity, phenology, production of fodder, climate and climatic variations would then be the principal parameters implied in simulation models of this type.
A lot of time will probably be necessary before the type of complex simulation model of which we are talking is perfected. The most difficult problem is to gather the exact data necessary. As these data are furnished by long-term experiments, the developers of the simulation model will need a lot of time in order to fully use them.
The possibilities for simulation models, in particular those that study the interactions inside a complete system between the plants, animals, environment and men, are appreciated by many of us. Attention must be given to social data as well as biological and environmental data to introduce into these models. Although there is still a lot to do, a relatively low level of precision in the simulation models already permits a better understanding of the implied workings.
The value of the precise details in several years through the use of simulation models will be limited by the imperfections in the climatic models. In the United States, some attempts have been made to overcome this problem by using results from species in terms of climatic variations studied in past years. This historical account has been reclassified into a growth index which is itself tied to a utilization or deterioration factor. A nearly exact simulation of fodder productivity over 30 or 40 years has been carried out.
C. DE WIT:
It is essential that the expansion of models be related as directly as possible to experimentation taking place in the field. The "modelization" of woody species is in fact a long-term research. But fortunately, the herbaceous "modelization" on a fodder plant balance can be carried out rather quickly.
- What benefit can the Sahelian countries realize from the results of this simulation model?
- Concerning the species used to determine the biomass, it should be necessary to take the floral composition into account. In fact, the perennial species have a much more important biomass than the annual species. For, working with annual species, how can 9 tons per hectare be obtained? What is the explanation for it?
F. PENNING DE VRIES:
Actually, the perennial species are studied so as to include them in this model. As for the high yields observed, they are not extraordinary because 7 to 8 tons per hectare per year are obtained in Israel with only 250 millimetres of rainfall.
In his simulation model does Dr. Penning de Vries make the distinction between the different annual plants or does he consider them as one and the same species?
F. PENNING DE VRIES:
Different studies have been carried out on the transpiration and growth of wheat. We consider there to be no difference between the different species of pastureland plants.
A relationship is given concerning the evolution of the biomass. The only variable component is rainfall. Is this an oversimplification or an approximation? Can such a complex factor be summed up in rainfall?
Perhaps it is a question of oversimplification. It's the simplest factor actually available to us, and in our present state of knowledge one cannot make a greater error than to take only this primary factor into consideration.
Is the calculation of biomass carried out by Bille based on the total biomass, namely aerial bodies plus root mass, or even a palatable biomass? What is the ratio of this last point with relation to the total biomass?
It's a question of total biomass.
Mr. Bille has mowed the pastureland, and therefore he must not consider his calculation as translating total biomass.
It is necessary to make the distinction between productivity and biomass. The biomass is the quantity of organic matter collected at the level of the ecosystem. The productivity is the production speed of organic matter. In what period of the rainy season are the greatest speeds of production or organic matter, in other words, the greatest productivity, recorded?
Couldn't we replace rainfall by the dampness of soil, which is a factor equally as variable?
F. PENNING DE VRIES:
The interpretation of rainfall data is practically impossible if they are not combined in a meaningful way with what is known about the soil, its retention capacity, and its loss by evaporation. Precision is greatly improved if the soil is taken into account.
To obtain the maximum benefit from the surveys and experimentation which are taking place in Africa, an existing international African organization ought to develop a common data bank. It would be useful to discuss here the characteristics of basic data.
Another seminar with data banks as the topic would be very useful.
The storage and standardization of data, the methods of collection, the men who have to collect them, the conditions and the goals make the whole thing very complex and demand a lot of time. Another question not discussed is the question of sampling procedure, use of samples, ways to identify them, their size, methods of sampling, etc. All that depends heavily on the objectives of the study. To help avoid useless measurements, I.L.C.A. could easily put together a book on sampling methodology.
Three books have already been written on sampling methods: by a Frenchman, an American and an Australian.
The common point in all our discussions is the need for dialogue between the users and the researchers and between the researchers in Africa and those outside. An international body could act to improve these relations.
What are the types of evaluations that lead in time to practical planning and a management method? We have no great need to know the research techniques, but rather to define the type necessary for immediate planning.
A lot of information is available. How can it be put at the level of the people on the land and in the bush, and how can it be passed from one researcher to another? The OAU or ILCA, or both, could be asked so many questions. To determine what can be useful and applicable in the field, it is necessary that all the results of the research reach the OAU and ILCA so that they can look them over. This could be one of the conclusions of the seminar.
That need not be a prerogative of ILCA, but rather a dialogue through the Scientific, Technical and Research Commission of the OAU - for example, between the researchers themselves, and between them and the people on the land.