Michel Grouzis and Marcel Sicot
Office de la Recherche Scientifique et Technique Outre-Mer, Ouagadaugou, Haute Volta
3. Application to study of variables: influence of ecological factors on phenology
The work described here is part of a multidisciplinary study carried out at the Mare d'Oursi in northern Upper Volta and organized by the Comité Lutte contre l'Aridité en Milieu Tropical of the Dé1égation Générale à la Recherche Scientifique et Technique. The work consists of making an inventory of the resources of biological and physical environments, describing their utilization by man and their development as a function of environmental factors. More specifically, it forms part of the investigations carried out by the botanical and agronomic section of the Office de la Recherche Scientifique et Technique d'Outre-Mer (ORSTOM), into the study of productivity in ecosystems of the Sahel as a function of ecological variables (Grouzis, 1976), especially with regard to water resources.
Against this background the study on the phenology of browse species was carried out for various reasons, including:
a) the relative lack of previous work in this field, although the part played by ligneous plants in animal feeds in the Sahel is considerable, in terms of both quantity (Le Houérou, 1979) and quality (Bille, 1978);
b) the need to gain a more precise understanding of active growth periods in order to measure productivity;
c) the characterization of relationship between vegetation periodicity and environmental factors, with a view to identifying the factors affecting the phenological behaviour of species and gauging their reaction to fluctuating ecological variables. The aim of the present paper is not to report in detail on the phenological cycles of the nine species observed since July 1977 at Oursi, since the regional interest of this study is limited in a meeting of this kind, but rather to explain the method used and to discuss it, illustrating a few characteristics of the variability of phenomena observed and the effects of environmental factors in a few specific examples.
Works on phenology in the tropical environment are few and far between. Some authors (Granier and Cabanis, 1975; Duranton, 1978; Grouzis, 1979) have dealt with grass formations, using methods which are not suited to the ligneous stratum (sampling, relevé areas, number of stocks, etc). The most complete studies on the phenology of ligneous plants in the Sahel have been carried out in the Ferlo desert in Senegal by Poupon (1979a). The observations at Oursi by the Centre Forestier Tropical (Anon, 1980) should also be cited, together with those of Delwaulle (1976) and Traoré (1978).
An analysis of all these papers reveals the following:
a) the number of samples is often low (1 to 6 individual plants for Delwaulle, op. cit., 5 individuals for Davies, 1975), or unequal (1 individual to over 200 in the case of Poupon, 1979a);
b) the samples are not always representative of the population of the species considered;
c) the nature of observations and the criteria are not always precisely determined.
The method utilized represents an attempt to combine ideal conditions as set out by Frankie et al (1974) for a phenological study, i.e. a station which has not been subject to disruption, a high number of samples, and observations made over several years.
The chief feature of the method is that it applies to the population of the species in the station under consideration. Since the numbers of a species population on a standard hectare are sometimes high (Grouzis, 1979), a stratified sample (Gounot, 1969) was taken on the basis of population structure, i.e. of the frequency histogram per class of diameter at base of trunk.
It should be noted that the determination of this structure is equally necessary for the population study and for productivity measurements. The size of the sample (30 to 36 individuals) is subsequently determined in proportion to the number for each class. It should be noted that the entire population for the sample hectare is investigated when the number is lower than 30.
a) Nature and criteria
The observations referred to foliation, flowering and fructification stages. The work of Floc'h (1969) provided the basis for morphological characterization of the various phenological stages; for flowering, the following stages were selected:
V1: Swelling buds, no leaf development,
V2: leaf buds and open buds (over 10% and less than 50% of these organs in each individual),
V3: leaves mostly open,
V4: leaves and dry leaves, or leaves which have changed colour (over 10% but under 50%),
V5: over 50% in each individual with dry leaves and falling leaves. This is a difficult stage to monitor since it may extend over several months according to the species (Guiera senegalensis, for example).
For flowering the following stages were monitored: fl: Floral buds only,
f2: Floral buds and open flowers (over 10% and less than 50%),
f3: Over 50% of organs carrying open flowers,
f4: Open flowers and dry flowers (over 10% and less than 50%),
f5: A majority of dry flowers and shedding of floral elements.
Fructification was characterized by the following stages:
Fl: Early setting stage,
F2: Development of fruit to normal size,
F3: Maturity,
F4: Ripe fruit and onset of dissemination (opening of pods or fall of fruit),
F5: Fruits dried and fallen
Stage 1 corresponds to the beginning of the phase and stage 5 to its culmination. Stages 2, 3 and 4 represent a specific phase for each individual; each of the three is characterized by the following degrees of intensity: low, optimum and declining.
b) Frequency of observations
During the active growth season observations are carried out every 10 days. During the dry season survey frequency is about once a month.
c) Presentation of results
The frequency variations for foliation, flowering and fluctuation as a function of time (Mooney et al, 1974) are represented instead of the classical phenograms, for reasons which will be seen later.
d) Survey of environmental factors
It should briefly be noted that temperature and relative air humidity are logged under standard screening conditions (Bernard et al, 1978 and 1980; Claude et al, 1980). Precipitation is recorded in a cumulative raingauge (diameter 15 cm) situated 1 m above ground level in the middle of the sample hectare.
The moisture content of the soil to a depth of 1.50 m was measured using a neutron humidimeter (Solo type, CEA). The methodology is given in detail by Sicot (1978).
The classical phenograms for ten individuals are shown in Figure la. Note that the ten individuals selected belong to the modal type of histogram as regards structure and therefore had every likelihood of being considered as average individuals, since the distribution of numbers per diameter class is normal for this species (Grouzis, 1979). The average phenogram (b) of the sample (N = 35) and the phenological spectrum (c) of the same sample are also reproduced in Figure 1.
Figure 1. Acacia nilotica var. adansonii (Guill. and Pert.) O. Ktze Phenograms for a few individuals (a), average phenogram (b), and phenological population spectrum (c) for the 1978 cycle. The variations for the foliation (v), flowering (f) and fructification (F) frequencies, together with the leafless phase (vo), are represented on the diagram as a function of time. The arrow on the phenograms for individuals 282, 246 and 245 indicates that stage fl was observed.
Examination of the phenogram of the different individuals (a) reveals a high variability between different phases. The duration of the stage at which the leaves appear, for example, varies from 10 days (no. 345) to 30 days (no. 245), while that of flowers ranges from 10 days (no. 318) to 20 days (no. 1). Similarly, the duration of flowering varies from 30 days (no. 95) to 110 days (no. 378). As regards fructification, only four individuals out of the ten observed actually bore fruit. There is therefore no point in wasting time, pretending that a single individual can ever reflect the phenology of the species as a whole.
The phenological spectrum (Figure lc) shows the same characteristics as the classical phenogram (Figure lb) as regards the onset of phases and their duration. The advantages of representation by our method lie in the following:
The preceding example (Figure la) is a good illustration of the importance of individual variations in behaviour, independent of age. Intrapopulation variations may be the response of the organism (within the limits of its genotype) to fluctuations in environmental factors (such as spatial diversity in the substratum); or, on the other hand, they could be the result of a differentiation in the genotype.
Individual variations may also depend on age. Thus, for example, Poupon (1977) has shown that Commiphora africana does not begin to bear flowers until the trunk circumference has reached 28 to 30 cm, corresponding to 8-9 age-rings. In 1979 the same author showed that these limits vary with the relief and the year. It should be noted that age does not act only on the acquisition of the reproductive function, but also affects the plant cycle together with the budding stage, as has been shown by Irgens Moller (1967, in Nienstaedt, 1974).
To sum up, it has to be acknowledged that a large number of individuals have to be monitored in order to compensate for individual variability, the latter reflecting the adaptation of species to contrasting ecological conditions.
Figures 2a A and 2b B summarize the results for Combretum aculeatum Vent, in two different situations, for which the solid and moisture conditions are outlined in Table 1.
Table 1. Soil characteristics of the two sites
Geomorphological units |
Pediment site |
Sloping bottom land | |
Characteristics |
Tropical ferruginous soil |
Brown subarid vertisol | |
|
Clay (% DM) |
9.234.6 |
37.247.6 |
Fine loam (1% DM) |
2.56.9 |
9.112.7 | |
Coarse loam (% DM) |
2.50.8 |
2.0 5.1 | |
Fine sand (% DM) |
46.715.4 |
16.127.9 | |
Coarse sand (% DM) |
38.136.7 |
13.622.5 | |
Organic matter (% DM) |
0.4 |
1.1 | |
|
Density (g/g) |
1.54 2.15 |
1.241.61 |
Total porosity (% vol) |
24.532.3 |
25.518.3 | |
Humidity at depth of- | |||
2.5 2.5 (% DM) |
7.216.7 |
||
3.0 (0/b DM) |
5.014.2 |
17.521.2 | |
4.2 (% DM). |
2.710.2 |
11.013.9 | |
Figure 2a. Intersite variability. Phenological spectrum for Combretum aculeatum Vent. in hydromorphic bottom land. Relation to ecological factors: MC = moisture content of soil at 1.50 m; F = frequency of phenophases; P =10-day precipitation; AP = annual precipitation AH= 10-day average relative air humidity at standard screening, 6;00 am; t: 10-day average mean daily temperatures. Inset: population structure; n = no. of samples; N= no. of population on standard sample hectare.
Figure 2b. Intersite variability. Phenological spectrum for Combretum aculeatum Vent. on pediment site. Relation to ecological factors. (for symbols, see 2 a above).
First it should be noted that the productions are comparable, as is shown in the structure histograms given in Figure 2.
An examination of the frequency variation for the different phases shows that:
there is a period of inactivity on the bottom land (Winde station, Figure 2a), as contrasted with the pediment site (Gountouré station, Figure 2b);
foliation occurs early at Gountouré; this phase did not occur until the second 10-days period in July at Winde, a time by which over 60% of the population at the Gountouré station were already in flower;
flowering occurs early at Gountouré; more than 50 days elapsed between the appearance of flowers at the two stations;
flowering lasted over very unequal periods; these were 68 days at Winde and 120 at Gountouré, not counting the residual flowering occurring during the cold season. It should be noted that the latter has been a constant characteristic for this species in this station, throughout 3 years of observations. Similar behaviour was also found by Poupon (1979 b) for Acacia senegal, and it is therefore not exceptional;
flowering is characterized by three successive waves at Gountouré (30%, 40% and 50%), but by a single peak (80%) at Winde. In the latter site the single flowering stage is immediately followed by abundant fructification, whereas the first flowers at Gountouré were mostly abortive.
These behavioural differences may be interpreted in terms of the variations in environmental factors. In the bottom land station (Winde) macroclimatic factors apparently do not play a crucial role on the phenology of Combretum aculeatum. The appreciable increase in air humidity (30 to 60%), the current variations in temperature and the first 40 mm of rainfall do not appear to influence the occurrence of phenophases directly.
The close relationship between variations in the moisture content of the soil and in the frequency of foliation and flowering (Figure 2a) suggests that in fact an important part is played by moisture conditions in the phenology of this species in bottom land areas. Such a discovery should cause no surprise for this type of soil, which has a high degree of retention and in which there is a great deal of competition for water between the plant and the soil at the beginning of the rainy season. These results show the importance of influence of water content in the soil on plant phenology and corroborate those of Ackerman et al (1974).
The physical characteristics of the soil at Gountouré (Table 1) promote a rapid effect of rainfall on variations in the soil moisture content. The result is a more rapid and easier utilization of water by the plant in this type of soil, compared with the previous example. Although correlation between the moisture content and foliation and flowering variations (15/6 to 10/7 and the third 10-day period in July, in Figure 2) is high, there is also a degree of parallel development between foliation and relative air humidity. Moreover, the simultaneous nature of the effects of soil moisture, rainfall and relative air humidity do not enable a dominating role to be attributed to soil moisture in this case, as was possible for the bottom land station.
The differences observed in the phrenology of C. aculeatum in the two ecological situations emphasize the importance of intersite variation and illustrate the effects of certain factors pertaining to stations.
The phenological spectrum of C. aculeatum (1977, 1978 and 1979 cycles), together with the variations in a number of environmental factors over the same period, are shown in Figure 3.
The graphs show that foliation occurs during the first 10 days of July in 1978 and during May in 1979. In the latter year 19% of the population were already in leaf by the first 10 days of June.
Table 2 summarizes the characteristic values for the frequency curves for flowering (Figure 3). Substantial differences emerge in the dates of onset, the durations, and the extent of flowering between one year and the next. It should be noted that the extent of variations between date of onset of fruiting and that of full fruiting stage are equally large.
Table 2 characteristic flowering values for combretum aculeatum vent on bottom land (winde).
Feature |
Year |
Value |
Extent of variation |
Onset |
1977 |
2nd 10-day period, July |
*40 days |
1978 |
3rd 10-day period, July |
||
1979 |
2nd 10-day period, June previous late not observed |
||
Duration |
1977 |
97 d |
*31 days |
1978 |
68 d |
||
1979 |
99 d at least |
||
Form and extent |
1977 |
2 waves, 20 and 40% |
* 40% * |
1978 |
1 peak, 80% |
||
1979 |
3 waves, 20, 66, 80% |
Figure 3. Interannual variability. Phenological spectrum for Combretum aculeatum Vent. and variation of ecological factors for the 1977, 1978 and 1979 cycles on hydromorphic bottom land. (for symbols, see 2a above).
These interannual differences can largely be explained by the ecological conditions prevailing at the beginning of the various vegetation cycles studied. The importance of the effect of moisture content has already been dealt with as regards the foliation phase during the 1978 cycle. The lack of phenological surveys between 5.5 and 5.6.79 preclude the possibility of specifying the crucial role of this factor on foliation in 1979. Given the clustered distribution of rainfall it is difficult to disentangle the effects of these various factors. This distribution also has a considerable influence on phenology. The rainfall totals for 1978 and 1979 were similar (309 mm and 302 mm); however, the period for which the whole population bore leaves differed by 50 days (67 days in 1978 and 118 in 1979). In general terms the defoliation stage coincides with the drop in relative air humidity and soil moisture (Figure 3).
The effect of soil moisture is most clearly shown in relation to flowering; there are two main reasons for this:
a) limited duration of flower life;
b) considerable sensitivity of the plant to a shortfall in moisture during this phase.
Although the humidimeter broke down, preventing monitoring of the soil moisture content in 1977, the variations in the 1978 and 1979 cycles show that the development of flowering is strongly correlated with fluctuations in moisture content: each flowering wave corresponds to a peak in soil moisture (Figure 3). It should be noted that the absence of a second flowering wave in 1978 (Figure 3) can be explained by the highly favourable soil moisture conditions, which gave rise to a sudden burst of flowering activity in a very short period, leading rapidly to abundant fructification. The growth of the fruit drains almost the entire stock of photo-synthetates, preventing the occurrence of a new wave of flowering when the profile has been restocked.
It should be noted that the reproductive capacity of the population is approximately twice as high in 1978 and 1979 as it was in 1977. This can be partly attributed to the effect of prohibited grazing. Individuals protected from domestic animals during the last 2 years have shown considerable vigour.
This is an observation which relates to the influence of biotic factors on phenology. In particular it should be noted that the interannual variation may be due to the action of biotic factors, but the effect is primarily shown in terms of extent, and will therefore be more measurable in terms of biomass. However, to illustrate this effect, the frequency variations for foliation in Acacia raddiana (Savi) and Balanites aegyptiaca (I.) Del. are indicated in Figures 4a and b.
Figure 4. Influence of biotic factors on foliation in Acacia raddiana (Savi) (a) and Balanites aegyptiaca (L.) Del (b). L = Locust; B = beetle.
Acacia raddiana reacts quickly to locust attack (Figure 4a), since the maximum foliation percentage is reached by the population during the second 10-day period following attack.
The appreciable differences between the 1978 and 1979 cycles for Balanites aegptiaca (Figure 4b) may be attributed to repeated attacks by locusts (June and September) and beetles (mid-August). It should be noted that, at the Gountouré station, Balanites aegyptiaca reacts fairly slowly to external constraints, since individuals suffer from a high degree of parasitic infestation (50% of the population have leaf gall, bugs or scale).
By studying the phenology of Combretum aculeatum over three plant cycles we were able to see the influence of environmental factors and to show the variations in responses.
The present study concerns the phenology of browse species populations in the Sahel (Mare d'Oursi, Northern Upper Volta). Stratified sampling on the basis of populations structure involved 30 to 35 individuals.
The results obtained demonstrate the importance of variation between populations, sites and years which have to be taken into account if the phenological cycles of forage species in this area are to be accurately characterized.
The variations observed are largely explained by the diversity and fluctuation of ecological conditions. Although photoperiodism has been cited as responsible for certain phenomena (Njoku, 1963), from this study it appears that its influence is secondary, especially in terms of flowering. Over 50 days separated the occurrence of flowering in C. aculeatum in the two stations studied in 1978, and at least 40 days separated the onset of flowering in the 1977 and 1979 cycles. However, this factor has not been studied in detail and it is possible that its effect has been masked by other factors. Similarly, there does not seem to be a relationship between current variations in temperature and those in phenological cycles. However, it should be noted that the influence of total temperature E (t - to), (Franquin, 1976) has not been researched since biological zero (to) for these species is not known. On the other hand, emphasis should be placed on water supplies, which are the most serious limiting factor. Soil moisture content can play a dominating role in determining phases, a result which agrees with those of Sauer and Uresk (1976), and Ackerman et al (1974). However, in some cases the simultaneous effects of precipitation, relative air humidity and moisture content of the soil do not enable the latter to be attributed an exclusive role.
The plants studied present two behavioural types in relation to fluctuating ecological factors. Some species are relatively insensitive to differences in climatic conditions Guiera senegalensis and Ziziphus mauritiana (Grouzis, unpublished results) and Z. mauritiana and Boscia senegalensis (Poupon, 1973) while others mirror the variations in environmental conditionsCombretum aculeatum, Grewia tenaxand Cadaba farinosa (Poupon, 1979a). This second type of behaviour is a special adaptation to the erratic rainfall of this area. However, it should be recognized that for these species little is known on the periodicity induced by environmental factors, quite independently of the genotype concerned.
Acherman, T.L. Bamberg, S.A. (1974). 'Phenological studies in the Mojave Desert at Rock Valley (Nevada Test Site)'. In: Phenology and seasonality H. Lieth ed., Berlin, Springer-Verlag.
CTFT (1980). 'Etude quantitative de Putilisation des ligneux saheliens par les herbivores domestiques dans la zone sud de la mare d'Oursi (Oudalan Voltaïque)'. (Sons presse). (s.l.).
Bille, J.C. (1978). 'le rôle des arbres et des arbustes en tant que sources de protéines dans la gestion des pâturages de l'Afrique tropicale'. 9ème Congrés Forest. Mondial. Jakarta.
Bernard, A. et al (1978). 'Observations climatologiques d la station météorologique de Djalafanka, Mare d'Oursi. Juin 1976/Decembre 1977. A.C. Lutte contre l'aridité dans l'Oudalan. (Haute Volta). DGRST-ORSTOM.
Bernard, A., Claude J., Bardin E., (1980). 'Observations climatologiques à la station météorologique de Djalafanka, Mare d'Oursi'. (sous presse). (s.l., s.n.)
Claude J., et al (1979). Observations climatologiques a la station météreologique de Djalafanka, Mare d'Oursi, aneannee 1978 DGRST-ORSTOM. (s.l.)
Davies, J.J.F. (1975). 'Studies of the flowering season and fruit production of some arid zone shrubs and trees in western Australia'. J. Ecol., 64, (2), pp. 665687.
Delwaulle, J.C., (1976). 'Rapport de mission d'un mois'. A.C.c. Lutte contre l'aridité dans l'Oudalan (Haute Volta), DGRSTCTFT.
Duranton, J.F. (1978). 'Etude phenologique de groupements herbeux en zone tropicale semi-aride'. 1. Methodologie. Adansonia, sér. 2, 18 (2), pp. 183197.
Franckie, G.W. et al (1974). 'Tropical plant phenology: applications for studies in community ecology'. In: Phenology and seasonality modelling, H. Lieth ed., Berlin, Springer Verlag.
Franquin, P. (1976). 'Formulation des phenomenes apparents de photopriodisme en conditions naturelles. Des modeles progressifs'. Physiol. Veg., 14, (1), 179191.
Godron, M. (1966). Une application de la thorie de l'information d l'etude de l'homogénite t de la structure de la végétation. CEPE/CNRS, Doc. 24, (s.p.).
Gounot, M. (1969). Méthodes d'etudes quantitative de la végétation. Paris, Masson et Cie.
Granier, P., Cabanis Y. (1975). 'Note sur la phénologie des graminées de savane'. Rev. Elev. Méd. Vét. Pays Trop., 28, (1), 7982.
Grouzis, M. (1976). 'Propositions pour une étude de la production et de la dynamique du couvert herbacé au niveau du périmétre de la Mare d'Oursi; relations entre la composition floristique, la structure de la vegetation, la biomassed et les termes dubilan hydrique'. ORSTOM. (s.l.).
Grouzis, M. (1979). 'Structure, composition floristique et dynamique de la production de matiere seche de formations vegetales saheliennes (Mare d'Oursi; Haute Volta)'. A.C.C. Lutte contre l'aridité dans l'Oudalan (Haute Volta), DGRST-ORSTOM.
Irgens-Moller, H. (1967). 'Patterns of height growth initiation and cessation in Douglas-fir'. Silvae Genetica 16, 5658.
Le Floc'h, E. (1969). 'Caractérisation morphologique des states et des phases phénologiques dans les communautés végétales'. CEPE/CNRS, Doc. 45.
Le Houérou, H.N. (1979). 'Le rôle des arbres et des arbustes dans les pâturages sahéliens'. C.D.I.
Mooney, H.A., Parsons D.J. (1974). 'Plant development in Mediterranean climates'. In Phenology and seasonality modelling. H. Lieth ed., Berlin, Springer-Verlag.
Nienstaedt, H. (1974). 'Genetic variation in some phenological characteristics of forest trees'. in Phenology and seasonality modelling. H. Lieth ed., Berlin, Springer-Verlag.
Njoku, E. (1963). 'Seasonal periodicity in the growth and development of some forest trees in Nigeria'. J. Ecology; 51, 617624.
Poupon, H. (1973). 'Influence de la sécheresse de l'année 1972-1973 sur la vegetation d'une savane sahélienne du Ferlo septentrional, Sénégal'. In: La désertification au Sud du Sahara, Coll. de Nouakchott, 17-19 Décembre 1973.
Poupon, H. (1971). 'Recherches écologiques sur une savane sahélienne du Ferlo septentrional. Sénégal: premieres données sur Commiphora africana (Rich.) Engl'. La terre et la vie, 31, 127162.
Poupon, H. (1979a). 'Etude de la phénologie de la strate ligneuse à Fété-Olé de 1971 à 1977'. Bull IFAN, sér. A, (sous presse).
Poupon, H. (1979b). 'Structure et dynamique de la strate ligneuse d'une steppe sahélienne au nord du Sénégal'. Thése Doctorat, Université Paris Sud.
Sauer, R.H., Uresk D.W. (1976). 'Phenology of steppe plants in wet and dry years'. North West Sci., 50, (3), 133139.
Sicot, P. (1978). 'Cycle de l'eau et bilan hydrique dans les écosystems types du bassin versant de la Mare d'Oursi. Analyse préliminaire des données receuillies en 1977: critique des principes de base du bilan hydrique en milieu sahélien'. A.C.C. Lutte contre l'aridité dans l'Oudalan (Haute Volta). DGRST-ORSTOM.
Traore, B. (1979). 'Observations sur la phénologie de quelques especes herbacées et ligneuses sahéliennes'. A.C.C. Lutte contre l'aridité dans l'Oudalan (Haute volta). DGRST-ORSTOM-ISP.
