The Lusaka site had generally higher soil moisture status, even though the precipitation was much lower than the Kabwe site. This was largely reflecting differences in the water holding capacities of the two different soils. CO2 emissions were positively correlated with soil moisture in both Lusaka and Kabwe sites. This positive relationship was previously reported in sub-Saharan African soils and suggests that soil moisture is an important factor controlling microbial activities [4, 24, 34]. Organic amendments also significantly increased soil moisture in both Lusaka and Kabwe sites. Contrastingly, organic amendments stimulated CO2 emissions only in the Kabwe site. The litter bag decomposition rates and the soil bacterial and archaeal abundance were also positively influenced by organic amendments only in Kabwe soils. This suggested that the positive impacts of organic amendments on soil microbial condition and then decomposition process appeared stronger in the Kabwe site. According to Thomsen et al. , soil texture indirectly affected soil microbial decomposition through water holding capacity. Organic amendments are therefore critical to maintaining soil microbes’ ability to decompose organic matter. At the same time, this could have been relevant only in sand-rich soils with limited soil moisture than in clay-rich soils, similar to the Kabwe site.
The PM amendment did not increase the CO2 emission rates compared to the control in the Kabwe site. Poultry manure used in the current research had a relatively higher C concentration than other organic materials and the C might be more recalcitrant than other organic materials, reducing the decomposition rates and the consequent C emissions. Another reason could be due to the smaller positive moisture retention effects on the PM treatment, compared to other organic amendments. We adjusted the organic material application rates based on their C content. Therefore, the relatively smaller amount of poultry manure applied could have less impacts on soil microbes and did not result in the increased CO2 emission rates, in the Kabwe soils.
In the clay-rich Lusaka site, soil moisture was more important in controlling soil respiration rather than fertilizer treatments. Soil faunal abundance collected using pit traps was higher in the Lusaka site compared to the Kabwe site (data not shown). Soil fauna could have played a key role in physical breakdown of organic materials and ultimately led to higher decomposition rates in the Lusaka site . This possibility suggests that the applied C in the Lusaka site was less recalcitrant than the Kabwe site, and it is more susceptible to loss from soil through decomposition. Thus, the organic amendments might not be expected to be a sole method to improve soil C content at the Lusaka site. Further studies are needed to understand interactions among soil biological community (faunal-microbial interaction) and soil C dynamics after organic amendments.
The soil bacterial and archaeal phyla was mainly dominated by Chloroflexi and Actinobacteria in both Lusaka and Kabwe sites. These dominant phyla show typical soil microbial community structures in nutrient-limited soils in tropics [3, 36]. There were also clear differences between the two sites in relative abundances of soil bacteria, mainly derived from some oligotrophs. Edaphic factors, in particular soil pH, soil C content, and moisture condition, are largely associated with soil microbial community structures . Lusaka soils had a higher abundance of Acidobacteria but lower abundances of Firmicutes and Verrucomicrobia compared to Kabwe soils. These bacteria phyla are strongly associated with soil C contents and moisture. Acidobacteria can be described as slow-growing oligotrophs and their abundance is negatively correlated with soil C contents . Higher relative abundance of Verrucomicrobia was previously observed in the subsurface (>10 cm depth) soil, suggesting an oligotrophic life strategy dependent on lower C availability [39–41]. Members of the Firmicutes are Gram-positive bacteria and should be favored by limited soil water holding capacity [3, 36, 42].
Within individual experimental sites, the temporal changes of bacterial and archaeal communities were highly variable in all fertilizer treatments. De la Cruz-Barrón et al.  also reported that bacterial community structure variation was more significant temporally (over time) and less influenced by fertilizer treatments in Zimbabwean soils largely due to short-term effect of organic materials. However, in the Lusaka site, soil microbial community structures clearly showed two different clusters throughout the experimental period, with one cluster having the CM and NF treatments and the other having the treatments that received both organic and inorganic fertilizers. Therefore, the combined use of chemical and organic N fertilizers largely affects the soil microbial community structure, compared to single use of organic materials or no fertilization.
The temporal variation of bacterial and archaeal diversity indices in Kabwe soils was higher and values were significantly lower than in Lusaka soils. This strongly reflects that the magnitudes of stability and response of soil microbial community structure to edaphic factors and soil physical disturbance events (e.g. weeding management and rainfall fluctuation) rather than organic amendments. Clay particles protect soil microbes from environmental stress such as predation, drought, and heat [43–45]. Also, the soil C content to limited levels could decrease both microbial biomass and diversity . Therefore, it is possible that Kabwe soils are more vulnerable to environmental stress and such conditions consequently destabilize microbial community structure due to enhanced microbial facilitation and niches overlapping . The vulnerability of microbial ecosystems in Kabwe soils could potentially be reduced by using organic amendments. Increased bacterial and archaeal abundance due to organic amendments was observed only in the Kabwe soils. A similar trend was observed for soil respiration and litter bag decomposition rates. These results are consistent with a previous study conducted in loamy sand soils, in which 16S rRNA gene abundance increased in organic farm soils compared to conventional farm soils in India . Organic amendments altered soil bacterial and archaeal communities to those more suitable to decompose complex organic substances. Previous studies stated that long-term organic amendments in sandy soils in the tropical arid regions have a potential to increase soil C and microbial diversity [10, 11, 48]. Therefore, organic amendments can contribute to maintaining the abundance of soil microbiome, and consequently increase C sequestration in the Kabwe soils. However, it should be emphasized that the rate of C sequestration depends on the capacity of the soil to protect the C.