An e-publication by the World Agroforestry Centre
METEOROLOGY AND AGROFORESTRY
|
|
An e-publication by the World Agroforestry Centre |
|
METEOROLOGY AND AGROFORESTRY |
|
|
SECTION I : GENERAL Some principles and problems of collecting and interpreting climatic data for the description of tree and shrub species P.G. von Carlowitz
Abstract The development of climatic profiles for woody perennials used in agroforestry is indispensable. Without them any description of species would be incomplete and render species selections for identified sites impossible. Based on the experience of several years of data collection for ICRAF's Multipurpose Tree & Shrub (MPTS) Data Base the considerations and problems of obtaining climatic data are outlined. One of the considerations is associated to the objectives of the data base for which climatic data are collected and used. ICRAF's MPTS Data Base deliberately includes as wide a range of species as possible, most of which are little known. It must, therefore, be restrictive in its selection of parameters. A carefully considered balance between data that are desirable and those that are indispensable for a reasonably reliable climatic species description must be maintained. Even if restricted to a minimum set of climatic data, the problems of obtaining them are manifold. They range from inavailability of even the basic data in the field, to the absence or insufficiency of climatic data in relevant literature. Consequently it is recommended to promote the importance of climatic data, particularly among non-climatology informants. It is further advocated that species trials should be equipped with climatological instruments. As a means of closing gaps in climatic species descriptions, informations relevant to plant growth should be calculated from easily obtainable actual data.
In comparing conventional agriculture with agroforestry systems, the distinctive feature of the latter is the deliberate introduction and management of woody perennials for their expected beneficial environmental and economic effects. From the topoclimatic and microclimatic point of view, the presence of trees and shrubs, so different in stature from most crops, is bound to make a considerable impact. This is well demonstrated by the effect of shelterbelts, on which a substantial body of research results exists. Also the negative or positive tree/crop interface, often observed but not yet well understood or researched, demonstrates microclimatic implications associated with the introduction of woody perennials. Unlike the situation in forestry with a comparatively small number of species to consider, restricted production objectives and a clear lead with regard to related research, the woody perennials considered for use in agroforestry systems encompass a wide array of species, most of which are little known. Therefore one is faced with a situation in which are site-specific choices of most crops based on a rather high level of knowledge whereas for most multipurpose trees and shrubs, in the absence of such knowledge, decisions are largely left to speculation. In order to close this wide-open gap, substantial research is required. Data need to be collected, on topics such as, taxonomy, morphology, phenology, response to tree manipulation and management, environmental requirements in general and climatic information in particular. Considering that we are dealing with a previously neglected part of the plant community, there is a wide range of problems. Information on climate as an important prerequisite for plant growth is one of the most crucial. Based on several years of experience in designing and operating a multipurpose tree and shrub data base it is the intention of this paper to expose some of the principles, considerations and problems of data collection with particular respect to climate.
Not only in agroforestry situations but also in general terms, the description of woody perennials, their properties, uses and limitations would be incomplete, if not to say of little value for all practical purposes, should environmental requirements and tolerances not be included. Without the questions answered as to where and under which soil, hydrological and climatic conditions any given species can grow, any further qualifications and properties of a species are of anecdotal rather than practical value. Therefore due consideration has to be given, among other factors, to the collection of climatic data. Basically, climatic aspects in connection with woody perennials in an agroforestry situation have to be approached from two different angles. One set of information deals with the specific requirements and tolerances of tree and shrub species, another set with the topoclimatic and microclimatic impact which the tree/shrub component makes on its environment. Although both aspects deserve an equal ranking, only the one dealing with species-specific requirements is discussed here. With this restriction in mind and considering that relevant information has to be obtained from various sources which include non-specialist informants, the following principles are to be observed:
Obviously the level of detail with which climatic data are collected, interpreted and used for the assessment of climatic requirements and tolerances of plants depends on the following considerations: The above-mentioned considerations are indeed reflected in various data bases and their respective conceptual approaches. Hackett (1983), in his modular system of describing the qualities of plant species, not only allows for a description of both woody perennials and non-woody annual plants, but also goes to greater length in his choice and number of descriptors, as is the case with data bases which are restricted either in the kind of plant material dealt with (e.g. woody perennials only), or in the ecological zone to which description of the plant material is limited. In the latter case certain climatic aspects, as, for example, amount of precipitation and Koeppen classes, are narrowed down already; and consequently the climatic description may dispense with some of the otherwise necessary descriptors. SEPASAT (Wickens 1986), a data base which originally had been limited to arid and semiarid environments, provides an example. Climatic descriptors are restricted to total rainfall, rainfall distribution, daily and seasonal temperature variation, and occurrence of frost. The INSPIRE Data Base (Webb et al. 1984) is another example in as far as the self-imposed limitation lies in the exclusion of all non-woody plant material. Climatic descriptors are reduced to mean annual rainfall, seasonality of rain, number of dry months, hottest and coldest months and mean annual temperature, obviously in an attempt to provide an overview which allows a comparison of many species, rather than to extend into detailed climatic description of a few well-known and better researched species. In view of the pros and cons of detailed climatic description of plants and of the various arguments provided in this connection, the multi-level approach, as adopted by Young (1985) in his concept of an Environmental Data Base for agroforestry, appears to be particularly attractive. In climatic terms the lowest or summary level confines itself to a generalized climatic class, as for example, humid tropics, and to Koeppen classes. The highest, level 2, adds to these the following descriptors: altitude zone, rainfall regime, mean annual temperature, mean annual rainfall, number of dry months, mean monthly rainfall, hottest and coldest months, driest month, frost occurrence, evaporation, humidity index and growing period. Because this method caters for the whole range of possible levels of data availability, it is certainly less exposed to criticism of being either too elaborate for certain purposes or too vague for others. However, from a practical point of view and considering that an output programme may be designed aimed at comparability with a view to species selection for certain site specifications, this method also has its disadvantages.
ICRAFs multipurpose tree and shrub data base is an inventory-type data base. Its objectives can be summarized as follows:
Within the boundaries of these objectives the use of Koeppen climate classes proved be reasonably satisfactory for the following reasons:
Apart from determining the Koeppen class by processing the data on rainfall and temperature by means of a computer subroutine programme, some additional information relevant to plant growth and to assessing the climatic requirements of individual species can also be generated from the information contained in the data base. Day length, an important aspect with regard to such phonological criteria as flowering and fruit-setting, can be determined by the latitude obtained as part of the geographic site definition. Lengths of dry periods, wherever applicable, can be extracted from the values of mean monthly rainfall that are provided. Based on these monthly means, contained in the data base as part of the description of any one species in a given location, the Koeppen class-specific range of rainfall can be narrowed down to the actual rainfall reported, thus refining the assignment of Koeppen classes to individual species.
As mentioned earlier, the number of descriptors by which climatic data are collected is restricted and comparatively simple. Nevertheless, in the course of the actual data collection a number of drawbacks and problems have been experienced.
Site-specific records, as loaded into the computer, are formatted descriptions of a tree or shrub species growing at a particular site. Such data, obtained from informants in the field, are original and authentic and are particularly suitable for gauging the biophysical range of individual species described at various sites. Records of this kind form the backbone of the data base. As the bulk of the species concerned are not well known and not yet subject to species trials, they are described in either their natural range or in farmland and pasture situations where the availability of meteorological equipment is the exception rather than the rule. Consequently the climatic data provided by informants are often incomplete, based on estimates with an undetermined degree of reliability or obtained from meteorological stations at varying distances from the growing site of a species described. Anticipating the problem of unavailability of climatic data on site, informants are requested to name the location of the nearest meteorological station from which data are quoted and to state its distance from the growth site. This may allow a rough assessment of the applicability of data provided, but the position remains doubtful as considerable differences in climatic conditions may occur over short distances if the area is marked by a heterogeneous topography. Even with descriptions obtained under experimental conditions there is often a certain degree of unreliability of data, with the exception of meteorologically well-equipped research stations. Species field trials maybe equipped with at least a minimum set of instruments, such as rain gauges and minimum-maximum thermometers, and monitoring may be done regularly. However, due to limited project durations the observation period with regard to climate is often too short to take account of long-term variability, particularly of rainfall. Means obtained from short-period observations may deviate from the actual average condition and could, particularly in borderline cases, lead to a wrong Koeppen class assignment and subsequently to inappropriate definitions of species-specific climatic requirements.
The site-specific records mentioned above are complemented by species records generated from the literature. Most contain relevant information on taxonomy, morphology, phenology and uses which constitute valuable additions to the data base. However, this information is usually not accompanied by a description of climatic requirements and tolerances. If in some cases reference is made to climate, it often suffers from so broad a generalization that it does not permit derivation of Koeppen classes for the species.
Those publications which represent already a summary description of species and are based both on references to competent secondary literature and/or research results usually prove to be a valuable source of information. Publications like the ones by the National Academy of Sciences (1980,1983), Webb et al. (1984) and Little (1983) are formatted in a way that provides guidance for species selection by various criteria and therefore include environmental information, although at varying levels of detail. A common feature, often observed in screening literature of this kind, is the assignment of altitude ranges to individual species that is not fully crompehensible. If a wide altitudinal range of, for example, 0 to 1800 m asl is combined with a wide latitudinal range of 30° and more, this summarized information on altitudinal tolerance of a species is reduced to an almost anecdotal status as it ignores the correlation between altitude and latitude with regard to temperature. However, if a specific altitude related to species tolerance is mentioned in combination with a specific latitude it may offer an opportunity not only to support temperature data which for whatever reason suffer from a low degree of reliability, but also to estimate temperature ranges, within tolerable margins when data are not available at all.
It has to be acknowledged that the effect of certain climatic events on the growth performance of woody perennials is not as yet always fully understood and researched. For example a tree species which grows well in one location, described by annual rainfall, seasonality of rains, length of dry period and mean annual temperature, may be unsuccessful at another site although these specific climatic characteristics are the same. The reason may be water stress, recognizing that defining a dry season only by its length is insufficient; with higher temperatures and wind velocity during the dry period, evapotranspiration may cause an earlier and more intensive water stress at one site than at another, thus exceeding the drought tolerance of a species at one site but not at the other. The conclusion is that more climate-oriented research is needed to assess not only the site-specific climatic conditions, but also the impact that different climatic parameters make on individual species. However, until considerable progress has been achieved in this vast field of agroclimatic research, ICRAF's multipurpose tree and shrub data base is faced with the problem of categorizing a large number of mostly lesser-known species of woody perennials by comprehensive, broad climate types to which the species can be assigned. Considering the generally low level of knowledge on most species that still prevails, such a strategy probably provides the best summary assessment of multipurpose tree species presently available. The assignment of a large number of species to standard climatic types allows a quick comparison of these species, at least in broad terms, and a shortlisting of candidate species for defined climatic conditions. With regard to the lack of availability of even simple climatic data in 'field' conditions, it could be argued that the density of the network of meteorological stations is sometimes insufficient. However, it is likely that financial constraints will prohibit the establishment of many more. Furthermore, it is doubtful whether a greater density of stations would result in a high degree of reliability of climatic data at the growing sites of trees. What has to be done is to promote awareness among non-climatology researchers engaged in data collection that climatic data are indispensible components of plant description and that special efforts to obtain them are therefore justified. It also should be stressed that species trials should be equipped, as a matter of routine, with climatological instruments. For arriving at more and better climatic description, another possibility which needs to be explored is to derive information relevant to plant growth by means of calculation from actual data that are easy to obtain. One such example is to assess the possibility of developing an equation or formula to establish the interdependence between latitude and altitude in determining site-specific temperature data. The practical relevance of the successful establishment of such an equation is obvious for reasons which have been mentioned earlier. National meteorological institutions and their staff and all climatologists should accept the challenge to help and cooperate in this field. Such cooperation is certainly of mutual interest and is likely to be beneficial for both parties.
Bluethgen, 1.1980. Allgemeine Klimageographie. (revised 3rd Edition by W. Weischet) Berlin : W. de Gruyter. Hackett, C. 1983. A draft set of core modules and three supplementary modules for a tabular method of describing the qualities of plant species. Technical Memorandum 83/26. Canberra, Australia: CSIRO. Little, E. L. 1983. Common fuelwoodcrops —A handbook for their identification. Morgantown, West Virginia: Communi-Tech Associates. National Academy of Sciences. 1980. Firewood crops, shrub and tree species for energy production. Washington D.C. National Academy of Sciences. 1983. Firewood crops, shrub and tree species for energy production, Vol. 2. Washington D.C. von Carlowitz, P. 1986. ICRAF's multipurpose tree and shrub data base — concepts, contents and objectives. (Prepared for the Forestry and Fuelwood Research and Development Workshop in Bangkok, Sept. 1986). Webb, D., P. Wood, J. Smith and G. Henman. 1984. A guide to species selection for tropical and sub-tropical plantations. Commonwealth Foresty Institute, Tropical Forestry Papers No. 15. 2nd edition. Univ. of Oxford, U.K. Wickens, G. 1986. Personal communication and data sheet for SEPASAT data base. Young, A. 1985. An environmental data base for agroforestry. Working Paper No. 5. Nairobi; ICRAF. |