Rodent population dynamics during the study periods in all habitats indicated that high birth rates accounted for the rapid population growth and turnover. Regular control and sustainable operations are thus essential if rodent pest populations are to be kept within tolerable limits.

We investigated the composition of the diet of the multi-mammate mouse, Mastomys natalensis, within irrigated rice and fallow field habitats at set time periods related to rice crop growth stages. In both habitats, vegetative plant material, i.e. leaves, stems and seeds, were the most abundant components of the rodent's diet, while other food types (invertebrates, fruits) were observed only in low quantities. We conclude that vegetative plant material and seeds were the main types of food consumed not only due to their relatively higher abundance in the environments under study but also because of the highly specialised herbivorous/granivorous nature of the dominant rodent species, M. natalensis. Thus, the introduction and expansion of continuous rice-cropping using irrigation in Tanzania is likely to be severely constrained by the presence of M. natalensis. In our opinion, field hygiene, including the removal of alternative food resources and nesting sites for M. natalensis near cropping areas, may help to both lower rodent population numbers and reduce immigration potential. Nonchemical rodent control methods such as trap barrier systems developed for lowland irrigated rice in south-eastern Asia should, we argue, be evaluated for their effectiveness under African conditions.

A 2-year capture-mark-recapture study was conducted to estimate home ranges and weekly travel distance of Mastomys natalensis (Smith 1834) in an irrigated rice ecosystem and fallow fields. We found that adults have larger home ranges than subadults in fallow fields but not in rice fields, indicating that fallow fields are more suitable for breeding. Travel distances were larger in rice fields, especially in the transplanting stage, during which rice fields are flooded and provide less food, causing movements into neighbouring fallow fields that then temporarily experience higher population density. A decrease in travel distance was observed in rice fields during the maturity stage, which can be explained by higher food availability and a more suitable, nonflooded situation. Movement of M. natalensis in rice-fallow mosaic landscapes thus seems to be driven by food availability and flooding status of the rice fields, which can be attributed to land use practices. R esum eUne etude de deux ans a et e r ealis ee par capturemarquage-recapture pour estimer le domaine vital et la distance de d eplacement hebdomadaire de Mastomys natalensis (Smith 1834) dans un ecosyst eme de rizi eres irrigu ees et de champs en jach ere. Nous avons constat e que les adultes avaient un domaine vital plus etendu que les subadultes dans les jach eres mais pas dans les rizi eres, ce qui indique que les jach eres sont plus appropri ees pour la reproduction. Les distances parcourues etaient plus grandes dans les rizi eres, sp ecialement au moment du repiquage, quand les rizi eres sont inond ees et procurent moins de nourriture, ce qui entraîne des d eplacements dans les jach eres proches qui connaissent temporairement une densit e de population plus elev ee. On observe une diminution des distances parcourues dans les rizi eres a maturit e, ce qui peut s'expliquer par la plus grande disponibilit e de nourriture et une situation non inond ee plus adapt ee. Les d eplacements de M. natalensis dans des paysages en mosa€ ıque de rizi eres-jach eres semblent donĉ etre d etermin es par la disponibilit e en nourriture et par l' etat, inond e ou pas, des champs de riz, qui peuventêtre attribu es aux pratiques d'utilisation des terres.

15An understanding of the dispersion patterns of a pest is an important pre-requisite for developing an 16 effective management programme for the pest. In this study, rodents were trapped in two rice fields 17 and two fallow fields for three consecutive nights each month from June 2010 to May 2012. 18Mastomys natalensis was found to be the most abundant rodent pest species in the study area, 19 accounting for >95% of the trapped rodent community. habitat with the least number of rodents. It was concluded that such spatio-temporal patterns could 32 serve as a tool for developing stratified biodiversity sampling plans for small mammals and decision 33 making for rodent pest management strategies. 34 35

: We investigated the recruitment and survival of the multimammate mouse, Mastomys natalensis , within irrigated rice and fallow field habitats at different time periods related to rice crop growth stages. Capture-MarkRecapture data were collected for M . natalensis each month from June 2010 to May 2012, and both recruitment and survival were estimated in relation to land use (irrigated rice or fallow field) within the agro-ecosystem. Higher recruitment and survival were observed in rice fields than in fallow fields suggesting the relationship was compensatory when there was a higher quality food resource. In terms of management, farmers in the study area should implement management strategies in rice fields at both transplanting and maturity stages of crop growth in order to maintain recruitment and survival at low levels.