Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Afrique du Sud → The prediction of woody productivity in the Savanna biome, South Africa

University of the Witwatersrand (1997)

The prediction of woody productivity in the Savanna biome, South Africa

Shackleton CM

Titre : The prediction of woody productivity in the Savanna biome, South Africa

Auteur : Shackleton CM

Université de soutenance : University of the Witwatersrand

Grade : Doctor of Philosophy (PhD) 1997

This thesis reports upon the findings of a four year study into the nature and management of woody plant productivity in the savanna biome, South Africa. The key components were to : (1) quantify the range of woody productivity from a variety of sites throughout the biome ; (2) seek relationships between the measured productivity and site and environmental variable ; and (3) elucidate the impact of selected management actions on woody productivity, Each of these was pursued through a combination of field surveys and empirical experimentation. The basic tenet of the study was to highlight the range of variability and to sample across it, rather than the pursuit of central tendencies and mean responses. Absolute woody plant productivity, as indexed through mean annual increment in stand basal area at 51 sites over four years, was spatially and temporally variable, ranging from 0.2 m2/ha/yr to 0.65 m2/ha/yr (corrected for death). Significant correlates were rainfall and soil texture. Mean relative woody plant productivity ranged from 2.0 % to 6.4 %. The primary correlate was stand basal area, but a significant relationship was also evident with mean annual rainfall. Whilst most sites indicated an increase in woody basal area from year to year, high mortality rates resulted in 34 % of sites having a net decrease in basal area over the four year period. Stem mortality averaged 4.4 % per year on a density basis. In comparison, seedling recruitment was not prevalent across all sites, but where it was recorded, the number of recruits was considerably higher than required to compensate for mortality. Inter-site recruitment was positively related to rainfall. Densitometry analysis from stem cross sections failed as a means to determine past growth rates. The addition of fertiliser and water increased basal area increment and shoot length. Conflicting results were evident with respect to the role of competition. Nearest-neighbour analysis from a number of sites suggested that competition was not important at most, but not all, sites. Yet, the strong negative relationship between relative growth rate and stand basal area, and growth enhancement from a thinning study, indicated that competition is a significant factor affecting individual tree and stand productivity. Most of annual production is allocated to leaves, other than in large trees, where the increase in the wood component is the largest contributor to total production. Production of deadwood was consistent from year to year, at approximately 1.7 % of standing biomass. Most litterfall occurs during the first months of winter (April - June), and occurs during a shorter period in toplands than bottomlands. Annual litterfall ranged from 90 - 360 g/m2 and was strongly related to stand basal area and biomass. Over 80 % of litterfall was leaves. Fruit production varied between years, both in the amount per tree, and the proportion of trees bearing fruits. In more arid situations and years, only the larger trees bear fruit. Trees start to bear fruit, given adequate rainfall, at approximately 20 % of their maximum potential size, and all individuals of a given species fruit once larger than two-thirds of their maximum potential size. At 1.5 - 2.0 % of stand biomass, extrapolated fruit production was higher than has been recorded in other studies. Wood harvesting has a marked impact on woody community structure and biomass and hence productivity. Relative growth rates are increased because of the dominance by smaller, faster growing stems. Absolute yields are reduced because of the lowered basal area and biomass. All species examined demonstrated strong coppicing ability. Within a species, the number of coppice shoots was positively related to cutting height. Between species, the density of coppice shoots per stump was negatively related to the maximum potential height of the species. Pruning of regrowth of Terminalia sericea significantly accelerated regrowth of the remaining coppice shoots, although this was interactive with mean shoot number. Turn around times to the next harvest ranged from 3-9 years depending on original stem size. Fire had complex impacts on woody community structure and biomass, and hence productivity. The impact of fire at the scale of an individual tree was dependent upon tree size and fuel load. The smaller the stem and the higher the fuel load, the greater was the proportion of stems killed during a fire. Over 50 % of stems less than 5 cm in circumference were killed during burning. Overall, approximately 6.5 % of stems were killed, and approximately 23 % were unaffected by fire. The rest experienced some degree of damage and resprouted either from the roots, from aerial portions, or both. Overall community structure is significantly altered by fire, and a predictable pattern is evident in relation to fire frequency. The mean biomass, height and circumference decreases with increasing fire frequency, whilst stem density increases. The change to a site dominated by smaller stems with less biomass per unit area results in a higher relative basal area increment, but a lower absolute basal area increment. The effects of fire on germination of woody plant seeds appears to be species specific, with some being stimulated and some being inhibited by fire. Consequently, fire may also facilitate a change in community species composition.

Présentation (NRF)

Page publiée le 12 septembre 2017, mise à jour le 3 janvier 2019