The influence of tree thinning and subhabitat differentiation on the reproductive dynamics of acacia mellifera subsp. detinens
Hagos, Mesghena Ghilay
MetadataShow full item record
The study was conducted in an area described as "Kalahari Thomveld" in the vicinity of the towns of Bray and Pomfret in the North-West Province, where Acacia mel/itera subsp. detinens is the dominant woody species. The soils of the area are deep sand to loamy soils described as Kalahari sand with an extremely low organic matter and mineral element content. The study area consisted six 0.5 ha plots (50 m x 100 m), where the trees were thinned during November 1989 to different densities, ranging from a totally cleared plot (0%) to plots thinned to the equivalent of 10%, 20%, 30%, and 50% of the tree density of a control plot (100%) of about 2 000 tree equivalents (TE) ha-l. The plots were located adjacently on a homogeneous area of 3.0 ha. Treatments were allocated randomly to the plots and not in numerical order. A permanent transect of 20 m x 100 m was established in the center of each plot for phenological observations and harvesting of seeds. The remainders of the experimental plots were used for soil sample collection and to determine seed distribution within the defined subhabitats. The spatial canopies of all rooted live A. mel/itera trees encountered in the fixed transects (5 m x 100 m), located in the middle of each of the experimental plots, were measured and the number of Evapotranspiration Tree Equivalents (ETTE) ha-l calculated, using the BECVOLmodel. Estimates of the browsing capacities were also made from the leaf dry matter estimates. For the study of the phenology, seed harvesting and leaf biomass estimates of individual trees, fifty (50) A. mel/itera trees (10 sample trees/plot) were randomly selected. Only one phenological observation was done at the onset of flowering (August 2000) and all the pods and seed from the marked trees were harvested during late November 2000. Tree thinning brought about ear1yflowering of A. mel/itera in the lower tree density plots, possibly as a result of reduced inter-tree competition. However, the mean seed production over the tree density gradient did not follow a specific trend, and differences were observed between the seed production of individual trees. Significant correlations between ETTE ha" and leaf dry mass (kg ha") and total seed production ha" were established. Although the correlation coefficients were low, there are indications that seed production of individual trees increase with an increase in ETTE tree-land leaf dry mass tree". For the determination of seed mass and seed dimensions ripe A. mellifera seeds were randomly selected from each sample tree (n = 10 and 50, respectively) after which they were weighed and measured. Seeds harvested from the high tree density plot were longer, wider and thicker than those from the lower tree density plots. The average seed dry mass production tree-1 showed no distinct pattern and was not correlated with the tree density, but the total seed dry mass ha-l increased in relation to an increase in tree density. Leaf biomass of the A. mellifera trees per unit area, expressed as ETTE ha" and leaf dry mass ha", was highly correlated with the seed dry mass ha-l . There is no carry over of A. mellifera seed from one season to another and thus no persisting seed bank exists. The percentage of viable fresh A. mellifera seeds was observed to decrease with the increase of the time and exposure to high temperature and moisture fluctuations. For the study of seed distribution within defined subhabitats (stem base area, under canopy spread and open area), six A. mellifera trees in each tree density plot were randomly selected. Small rectangular plastic containers were inserted. in the three subhabitats of all marked trees (in two opposing directions). In all subhabitats a larger number of seeds were blown toward the prevailing wind direction. The number of seeds accumulated in the open subhabitats in both wind directions was also higher in comparison to the canopied subhabitats. A very low, non-significant correlation was observed between ETTE ha-l and seed distribution along the tree density gradient of the three subhabitats. For the assessment of germination potential, 50 normal and 20 bruchid beetle infested seeds were randomly selected from each plot. Germination tests were conducted at the facilities of the Department of Agronomy, UFS, and root and coleoptile lengths and growth rates were also measured. Thinning of A. mellifera trees had no effect on the germination potential of the seeds from the various tree density plots and an extremely high germination potential of the fresh A. mellifera seeds were found. Though some damage was caused, bruchid beetle infested seeds exhibited a low but fast germination rate, possibly due to faster imbibition of moisture. No marked difference in root length and root growth rates were observed between seeds of the various plots. However, seeds harvested from the lower tree density plots developed a higher coleoptile length with a faster coleoptile growth rate than the control (100%) plot. Soil from the canopied and uncanopied subhabitats were analyzed for soil nutrient status in order to evaluate seedling growth. Soils excavated from the canopied subhabitats were more acidic and displayed higher concentrations of P, total Nand organic matter than the uncanopied (open) subhabitat. However, no marked differences in exchangeable cautions were demonstrated between the subhabitats, except Ca that displayed higher concentrations in the canopied subhabitat. No allelopathic effect that inhibits the growth of seedlings was found to be present in soil from the canopied subhabitats of A. mellifera. Marked differences in growth parameters of the seedlings grown in the soils from around the stem base area in comparison to the other two subhabitats were observed. The relatively higher soil nutrient status observed in soils of this subhabitats is considered responsible for this increased growth. In general, A. mellifera seedlings grown in the soil from the lower tree density plots exhibited higher growth rates than those from the control (100%) plot. A denser root system, as reflected by higher total root lengths and root dry mass, was observed in seedlings grown in soil from the low tree density plot and declined linearly in soils collected along the tree density gradient. In general, the subhabitat differentiation had a more pronounced effect on seedling growth than soil differences associated with the tree density gradient.