Species and provenance trial of Acacia aneura, A. farnesiana, A. nilotica, A. senegal and A. tortilis
The best provenances of A. nilotica, A. senegal and A. tortilis had produced a biomass of 8-10 t ha-1, which corresponds to an annual production of 1.6 – 2 t ha-1 during the first five years at a spacing of 3x4 metres. Compared to the Prosopis trials at Bebedouro this is a relatively high production. Brazil2 (P. juliflora), the best producing provenance of Prosopis at Bebedouro in the trials included in this series, had an annual production of biomass of 1.6 t ha-1. It appears that a range of Acacia species, including at least the three species mentioned above, could have a satisfactory production at the site.
In the univariate analyses there were only few signs of differences between species, and when the tests were corrected for multiple comparisons by the sequential Bonferroni tablewide method, all differences became non-significant. In the multivariate approach, it seemed that the provenances of A. farnesia and A. aneura were far apart from the main group of the three other species. However, judging the performance of a species by testing only one provenance is tricky, since other provenances within the same species could be better adapted to the site. Thus conclusions have to be given at the provenance level rather than at the species level.
There were several highly significant differences between the provenances in the trial. The survival was quite variable, which in turn influenced many of the other variables. Even though the variation in basal area of the mean tree was modest, the differences in survival meant that the total basal area and total dry weight varied with a factor of 10 from the smallest to the largest provenance.
From the trial it is clear that certain provenances have a poor performance at the site. These provenances include Australia2 (A. aneura), Mexico01 (A. farnesiana), the A. nilotica provenances from Rajasthan, and Sudan15 of A. tortilis. On the other hand, the provenances Senegal15 and Sudan09 (A. nilotica), Senegal22 (A. senegal) and Senegal35 (A. tortilis) had an impressive growth and should be considered for further testing. An unclear factor in the choice of provenances is the attacks by insects. Even though there were significant differences between the provenances, the damage was not concentrated on certain groups of provenances. There are theoretical difficulties in the analysis of the damage to trees as it is difficult to make an objective evaluation of degree of impact the damage has to a tree, and the subject deserves more attention.
Analysing differences within the species revealed two interesting facts. In A. nilotica, both the univariate and the multivariate analyses indicated that there were important differences between the provenances. Since each variety was represented only once, the interpretations should be cautious, but a few things deserve mention. First, the provenances from Rajasthan were collected at the same site and did not separate clearly from each other even though representing two different varieties (indica cupressiformis and indica vediana). In the multivariate analysis, the provenances came out separately, but still in the same end of the diagram (Fig. 20). Second, the provenances Sudan09 (var. adstringens) and Senegal15 (var. adansonii) were placed together in the multivariate analysis even though being separated by a wide distance geographically. According to newer taxonomy (Brenan 1983) the two varieties are considered as being the same (see introduction). Third, these two provenances were slightly separated from the third provenance from Africa, Senegal12, of the variety tomentosa. This variety tolerates inundation and appear to be restricted to habitats along rivers and seasonally flooded areas, whereas var. adstringens is found predominantly in wooded grassland, on deep sandy-loamy soils (von Maydell 1986, Ross 1979, Fagg & Barnes 1990). Fourth, the African provenances seemed to perform better than the provenances from India, although Ahmedabad1 was close to Senegal12. The material is sparse, but this could be a hypothesis worth further testing.
In A. tortilis it was clear that the provenance Sudan15 of the variety spirocarpa was different from the provenances of the variety raddiana, even though one these provenances were also of Sudanian origin. This points to the variability within the species.