Soil Quality and Microbial Responses to Tree Species and Land Terracing in Rwanda

Soil conservation measures, including forest plantations and land terracing, have been
implemented worldwide to restore degraded soils and/or counter land degradation
processes. However, these anthropogenic actions may have contrasting effects on soil
quality and ecosystem functioning, depending on climate and biophysical characteristics.
In Rwanda, afforestation and land terracing are the two major land-use forms commonly
implemented, not only for restoring the country’s severely degraded soils, but also as a
means to provide for wood-derived goods and services and to enable the cultivation of its
steep terrains. In this thesis, we assessed the responses of soil quality, through the
measurement of physical, chemical and microbiological indicators, to commonly planted
tree species and agricultural terracing in southern Rwanda.
We investigated the long-term effects on soil quality of 3 eucalyptus, 3 agroforestry, 2
native species and native species mixed in a self-regenerated plot in the Ruhande
arboretum, Rwanda. Potential effects were measured in the upper soil layers at 0-5 cm
and 5-10 cm depth. Our results indicate that significantly higher values and more
pronounced effects of tree species on most soil properties and microbial processes were
restrained in the upper 0–5 cm layer, highlighting the importance of this thin layer for soil
quality and ecosystem functioning under these forest plantations. Planting native tree
species (i.e., Entandrophragma excelsum and Polyscias fulva) improved soil quality via
alleviation of soil acidity, increasing concentrations of exchangeable base cations, and
promoting higher microbial biomass and activity. Eucalyptus species acidified the soil, but
also significantly increased soil organic matter contents and did not adversely affect
microbial biomass and activity. For example, results showed a significant increase in
microbial biomass under Eucalyptus grandis and increased N mineralization under
Eucalyptus maidenii, despite reports on detrimental effects of eucalyptus species on
growth and activity of soil microorganisms, due to their soil acidifying effects and
secretion of allelopathic compounds. This study therefore suggests that we cannot
generalize the effects of planting Eucalyptus on soil quality in general and, in particular,
on soil microbial biomass and activity.
Labile fractions of soil organic matter, particularly those extracted with hot water, were
the main drivers of differences in soil microbial activity between tree species, indicating
that they would better indicate tree-induced changes in substrate availability and soil quality than total soil organic matter. Further, results suggest that combining analysis of
these labile C and N fractions with that soil microbial biomass and activity would give an
early indication of management-induced changes in soil conditions.
This study also evaluated the effects of planted tree species on the abundance of
ammonia-oxidizing archaea (AOA) and bacteria (AOB), as well as their contribution to the
rates of soil nitrification, which are important indicators of N cycling in terrestrial
ecosystems. Abundance of the amoA gene (ammonia monooxygenase–subunit A) of AOA
and AOB and their activity demonstrated the numerical and functional dominance of AOA
over AOB in terms of amoA gene copies and potential soil nitrification rates across tree
species. These results are consistent with reports indicating higher abundance and
activity of AOA under low pH and limited substrate availability. Soil pH and labile nitrogen
were found to influence the differences in abundance and activity of nitrifiers between
tree species. Generally, Polyscias fulva, Eucalyptus grandis, Grevillea robusta, and Cedrela
serrata showed highest potential nitrification rates both by AOA and AOB.
The influence of land terracing was investigated in three paired terraced – unterraced
agricultural plots. Land terracing did not affect most soil physico-chemical properties,
which were mostly influenced by hillslope position both in terraced and unterraced fields.
The results from this study contradict our hypothesis about the effects of land terracing
on decline of total SOM and associated soil properties. The reduction of SOM was expected
following the construction of terraces which disrupts soil structure through excavation,
leading to vertical soil redistribution and thus oxidation of SOM once stored in deeper soil
layers. The results, however, supported our hypothesis in which soil quality increases in
lower hillslope position as a result of long-term erosional movement and sedimentation
of fertile topsoil downwards. Despite the increase in labile C and N fractions as well as soil
microbial parameters, especially downslope of terraced land, the overall results did not
allow us to draw an explicit conclusion on soil quality restoration by land terracing in the
studied sites.