Influence of riparian land use and streambed topography on decomposition
rates of Syzygium guineense and Eucalyptus saligna leaf litter in a
tropical montane stream, the Kamweti River, Kenya
Abstract
Leaf litter decomposition is a key ecosystem process that serves as a
source of energy in stream ecosystems. However, land use change from
forest to agricultural land has been reported to negatively affect
stream ecosystem functions. Streams exhibit heterogeneity in terms of
physical, chemical, biological characteristics and human-related
influences. This may in turn affect stream ecosystem functions (e.g.,
leaf litter decomposition). However, information on the effects of land
use and streambed topography on the functioning of many tropical stream
ecosystems is still limited. This study assessed the influence of land
use (i.e., forest vs agriculture) and streambed topography (i.e.,
riffles vs pools) on the decomposition rates of Syzygium
guineense and Eucalyptus saligna leaf litter in the Kamweti
River, Kenya. A total of 400 coarse-mesh litter bags were used to
enclose 5 (± 0.05) g of each plant species leaf litter. The leaf litter
was incubated in the selected study sites and litter bags were retrieved
after 1, 3, 8, 14, 21, 28, 35, 42, 49 and 56 days of incubation.
Physico-chemical parameters were also measured in all the sites.
Decomposition rates were estimated using the negative exponential decay
model. Linear Mixed-effect Models were used to evaluate the effect of
land use, streambed topography and plant species on leaf litter
decomposition rates. Physico-chemical water parameters differed
significantly between the two land uses (all p < 0.05),
except for pH and total nitrogen concentration ( p>
0.05). Forested sites ( Syzygium = 0.03 ± 0.004,
Eucalyptus = 0.04 ± 0.004) had relatively higher mean leaf litter
decomposition rates than agricultural sites ( Syzygium = 0.02 ±
0.004, Eucalyptus = 0.03 ± 0.006), although not significantly
different. On the other hand, riffle habitats had significantly higher
decomposition rates ( p<0.05) than pool habitats across
the two land uses. Streambed topography had a significant effect on
decomposition rates of leaf litter, as opposed to land use. Therefore,
local scale factors are more important in the decomposition process than
catchment scale factors in the Kamweti River. Stream conservation and
management efforts should be directed to the local scale factors as
opposed to only riparian and catchment factors.