4.1. Effect of plant cover on soil CO2 efflux
Plant cover can highly influence soil CO2 efflux (Raich & Tufekciogul, 2000) and removal of native forests and land use changes during mining activities accelerate the TC losses by increasing CO2 efflux (Lorenz & Lal, 2007; Polglase et al ., 2000; ). Planting trees is one of the most effective activities in recovering TC stocks of degraded areas (Moghiseh et al., 2013). The present study identified that CO2 efflux tends to increase with the planting of forest cover. Higher diversity in plant species, together with favorable climate conditions, may result in abundance and diversity of soil microbiological community (Langeet al ., 2015). Higher microbial activity may affect the decomposition of tree’s residues and SOM, and consequently, the production of soil CO2.
Planting different types of forest cover led to an increase in CO2 efflux due to the benefits of increasing SOM. Forest species favor soil microbiological activity by contributing organic material from the shoots and roots that contribute to nutrient cycling, speeding up the ecosystem recovery (Shrestha & Lal, 2006), and which may reflect in the increase in soil CO2 efflux. While the treatment with no plant covers presented the lower values of CO2 efflux, the Woodland showed higher values (Fig. 2). Soils of mining areas pose a limitation on plant growth, but our results show that planting forest species is the right way to achieve similar soil CO2 efflux compared with the unmined Woodland.
The forest types Euc, Ap and Nat showed higher values for CO2 efflux compared to NCov. Different forest covers can influence the soil chemical and biological properties by the input and composition of organic material from the trees (litter and roots), resulting in an increase of TC and soil biological activity. Different tree species contribute with organic material of different characteristics (Craine & Gelderman, 2011; Martin et al ., 2009), which may also affect the soil microbiological activity (Barretoet al ., 2008) and the increase of TC. Therefore, differences in root biomass, soil organic matter content, and the spatial arrangement of the trees may contribute to soil CO2 efflux variations (Katayama et al ., 2009; Gomes et al ., 2016).
Optimum soil moisture and temperature conditions combined with higher values for TC, TN, LC, C-MB, N-MB and root density (Table 1), may explain the greater values for CO2 efflux in Woodland (Figure 3). With favourable conditions of temperature and moisture, soil microorganisms act more intensely and produce more CO2. In addition, the higher root density in Woodland is associated with the higher soil CO2 efflux for the same climate conditions. Tree roots are responsible for a large flow of C and nutrients in soil, with respiration accounting for a third or more of the total soil CO2 efflux in tropical forests (Högberg & Read, 2006). In addition, the amount and quality of available C and photosynthetic activity also influence soil CO2 concentrations and consequently the efflux (Kuzyakov & Cheng, 2001). Therefore, increasing the diversity of forest species, promotes in time the development of soil microorganisms and roots, resulting in higher rates of CO2 efflux.

4.2. Autotrophic and heterotrophic respiration

The contribution of soil heterotrophic respiration to total respiration in the forest species was on average 55% during the driest period, and 72% during the wettest period (Figure 3). The relative contributions of heterotrophic and autotrophic respiration are specific context-dependent, since are influenced by local soil temperature and moisture conditions (Liu et al. , 2009; Hanson et al ., 2000), and by intrinsic plant factors, such as the photoassimilates allocation to roots (Epron et al ., 2011; Högberg & Read, 2006). The Ap forest cover showed the highest values for the relative contribution of HR (77.4%) during the rainy season. This result may be related to the high decomposition rate of the leaflets from this type of forest cover compared to the others (Gama-Rodrigues et al. , 1997) (Figure 4). Wei et al . (2010) analysed a global data of forest ecosystems and also found increases in HR in rainy seasons. Soil microbiota is sensitive to soil moisture and may explain the higher HR contribution of HR to total soil respiration during the rainy season (Carbone et al ., 2011). Variations in AR are more dependent on plant photosynthetic activity (Gomes et al ., 2016), responding slowly to variations in soil temperature and soil moisture when compared to HR (Bond-Lamberty et al ., 2004; Högberg et al ., 2008, 2009).
The specific partitioning of heterotrophic respiration showed that the soil CO2 produced originates mainly from the mineralization of plant residues (Figure 7). Plant residues are relatively easier to be degraded by microorganisms compared to the SOM, which is protected by clay-mineral associations and has been passed by biological degradation. In areas in an initial process of recovery, the microorganisms have a preference to use the most recent added forms of C (Novara et al ., 2014).

4.3. Influence of biotic and abiotic factors on CO2flux

The principal component analysis shows that the planted forests are placed between NCov and Woodland, and follow the same pattern of soil CO2 efflux (Figure 2). This indicates that the planted forests are improving the soil properties comparing with the treatment with no ground cover and are on the way to the soil quality presented in the native forest.
Soil CO2 efflux is a complex process and is influenced by biotic and abiotic factors. Soil temperature and moisture are the main abiotic factors that influence these processes (Fenn et al. , 2010; Wu et al ., 2010), affecting directly the biotic factors, such as microorganisms activity and plant cover. An increase in soil temperature leads to an increase in microorganisms activity (Atkinet al. , 2000) and root respiration (Schindlbacher et al ., 2011), leading to increases in soil CO2 efflux. Plant cover with tree species creates a microclimate beneath the canopy that reduces variations in soil temperature, and therefore their influence on soil CO2 efflux (Gomes et al ., 2016). However, in the present study, soil temperature did not vary much during the measurements (Figure 1) and a significative response was not found (Figures 9 and 10). Soil moisture is a very important factor in the maintenance and development of microbial activity (Liu et al ., 2009). Our results demonstrated that soil moisture was the most important variable and associated positively with soil CO2 efflux. While low values of soil moisture at the end of the dry season leading to low values of soil CO2efflux (Figure 2), the increase of precipitation in the rainy season promoted the increase of soil CO2 efflux values. These fluctuations of soil CO2 efflux are directly linked with microbial activity during the dry and rainy seasons.

Conclusions

In this study, we found that afforestation increases total CO2 efflux in RMDL compared to those with no vegetation cover.
In afforested RMDL, heterotrophic respiration contributes more to total soil respiration and litter decomposition is responsible for the highest CO2 efflux.
The variation in moisture is the main factor responsible for the soil CO2 efflux changes in the types of forest cover studied.
The plantations of Euc, Ap and Nat forest cover were efficient in recovering the mined soils, with similar CO2 efflux values found under the Atlantic Forest remnant.