Soil carbon dioxide emissions in a cereal cropping system under nitrogen application in an inland valley wetland in Uganda

Abstract

Utilization of inland valley wetlands for agriculture is on the increase, in effect threatening the values associated with an intact wetland ecology. Moreover, it has become inevitable that some form of agriculture must be undertaken on these fragile ecosystems, since the hitherto recommended upland areas are no longer favourable for cultivation due to soil moisture variability, increased human populations and non-alternative agricultural uses. It is, therefore, important that strategies are adopted to minimise the damage to the environment, as well as preserve productivity to levels closer to the natural wetland ecology. This presupposes understanding of the status-quo and immediate conversion of the system to cultivation using the most popular crops. In Uganda, the most common wetland crop is rice (Oryza sativa L), (both lowland and upland varieties). In more recent years, paddy rice growing communities in wetlands in Uganda are increasingly cultivating maize (Zea mays L), sometimes side by side. Owing to differences in the distribution of soil moisture in inland valley wetlands, these wetlands have been divided into three zones, namely (i) the Fringe, which is the nearest to the summit side, (ii) the Middle just after the Fringe downwards, and (iii) the Centre zone where the streams are located. It is hypothesized that maize or rice is grown in either of these zones at different times of prevailing soil moisture conditions. Evidently the productive capacity quickly wanes with subsequent harvests of rice and maize; and this has been attributed, among other things, to declining soil fertility (especially N) and soil moisture deficits associated with climate change. What is yet to be understood is the effect of wetland conversion and continuous use of inputs such as organic or inorganic N, on soil CO2 emissions and how this will be expected to vary with crop productivity and hence overall carbon intensity.

The objectives of this study were to determine (i) the effect of converting an inland valley wetland into either rice or maize cropping system on soil CO2 emissions; and (ii) effect of N application on carbon intensity from rice and maize cropping system in an inland valley wetland. The study was laid out in two parts, the first part investigated the effect of conversion of inland valley wetland in to rice and maize production; and the second one on wetland cultivation with application of different N sources. The study was conducted at National Crops Resources Research Institute (NaCRRI) during two rainy seasons in the case of rice and two dry seasons for maize, between 2015 and early 2017. Treatments for the first study part included three hydrological zones (Fringe, Middle and Centre) and intact natural vegetation as the control. In part two of the study, besides the treatments in the first part of the study, N sources (organic and mineral) were included. Data were collected on soil CO2 emissions using LI-8100A emissions system; while data on soil moisture and soil temperature were simultaneously measured, using theta soil moisture and temperature probes, respectively. Grain yield data for each of the two test crops were collected. The soil CO2 emissions from crop production periods were analysed using R statistical software version 3.4.3. A two-way repeated measures analysis of variance (RM-ANOVA) was performed to assess the influence of wetland conversion into crop production under different management options and hydrological zones, on soil CO2 emissions rate throughout the experimental period. Means of significant treatments were separated using Tukey Kramer’s multiple comparison tests at P< 0.05 significance level. Carbon intensity was used to evaluate the amount of carbon lost per kilogram of grain yield produced in an inland valley wetland used for cereal production under different N sources. The effect of conversion of the inland valley wetland to rice production without external inputs (N sources) was inconsistent, with the highest increment in soil CO2 emissions being in the Fringe zone (20% increase), a depression of (12%) in the Middle, and minimal effect in the Centre hydrological zone. On the other hand, there was a marked reduction in soil CO2 emissions in the Fringe (5-12%) and Middle (19-45% reduction) zone resulting from cropping of inland valley wetland with maize. In contrast, the Centre zone had an increase in soil CO2 emissions (15-26%). Conversion of an inland valley wetland to rice or maize production, without external inputs, made no distinct effect in terms of soil CO2 emissions across the production periods. However, without external inputs, maize was generally less carbon intense (higher grain yield with lower soil CO2 emissions) than rice (lower grain yield with higher soil CO2 emissions) in this wetland. Irrespective of crop type, there was no measurable effect of applied N sources (mineral or organic) on soil CO2 emissions. Nevertheless, rice production using organic N source was less carbon intense compared to production without inputs. As for maize, use of organic N had no significant effect on carbon intensity compared to production without inputs. Use of inorganic N (120kg N ha-1) for production of either cereal crop was less carbon intense compared to when production was done with use of organic N (120kg N ha-1). Although, irrespective of hydrological zone use of inorganic N (120kgN ha-1) reduces carbon intensity in both rice and maize.