Hydrological effects of forest implantation
After 2016, the predominant land use in the C2 and C3 catchments was modified, respectively, to native vegetation (restoration through the planting of native tree species) and to commercial Eucalyptusplantation, as shown in Figure 2. Table 3 shows the precipitation, flow, evapotranspiration and BFI data recorded in the following two water years (2017/2018 and 2018/2019) in the C2 and C3 catchments, and in the C1 catchment, which did not show a change in the predominant land use. Based on the mean values, it can be noted that C3 had higher flow and lower evapotranspiration, while C2 had the lowest value of BFI; it is important to highlight that the annual precipitation of the two years analyzed (1796 mm and 2051 mm) was higher than the average annual precipitation of the region (1319 mm).
Due to the rapid growth of planted forests (West & Mattay, 1993; Whitehead & Beadle, 2004) and the increase in leaf area index in the first years (Almeida et al., 2007), the initial hypothesis was that the C3 catchment would have the highest evapotranspiration rates and lowest values of flow, as already evidenced in other studies with eucalyptus forests (Calder, Rosier, Prasanna, & Parameswarappa, 1997; Farley et al., 2005; Scott, 2005). However, there is evidence that peak consumption of planted forests probably occurs after the third year of growth (Forrester, Collopy, & Morris, 2010; Scott & Lesch, 1997), which suggests that the moment of greatest effect of the planted forest was not analyzed.
Another important point to be considered is the presence of native vegetation in 44% of the C2 catchment near the stream and that was not modified in the years analyzed. This proportion of forest cover could explain the water use in this catchment, which may be related to the presence of a developed root system (Christina et al., 2011) and its proximity to the water table (Salemi et al., 2012). Considering the contribution related to the area occupied by each cover to the average ET observed in the catchment and assuming that the ET/P of the native vegetation near the watercourse is equal to 1; in C3 (ET/P = 0.83), the 13% of native vegetation (ET/P=1) would indicate an ET/P of 0.80 in the rest of the area (87%) under eucalyptus plantation. In C2 (ET/P = 0.94), with 48% of native vegetation (ET/P=1), the value would indicate an ET/P of 0.87 in the rest of the area (52%) under native vegetation. In this case, it is reasonable to think that the proportion of native forest vegetation in C2 may be influencing the ET rates found, as observed in other studies (Salemi et al., 2012; Zhang, Sun, Shi, & Feng, 2012).
On the other hand, it is important to observe that the C2 catchment had the lowest values of BFI compared to the others, and C1 had the highest values. Since BFI reflects the physical characteristics of the catchment (e.g., geomorphology) (Price, 2011), it is possible to assume that C2 has different geological and physical characteristics, which could influence the observed results, so forest cover is not the only factor influencing the hydrological processes of this catchment (Brogna et al., 2017; Chandler, 2006; Price, 2011).
Through the flow duration curves of the daily flow data (Figure 6), it is possible to observe that the C3 catchment stands out from C1 and C2. The daily flow values of C3 are higher along the entire length of the flow duration curve, and the little pronounced difference between the minimum and maximum values of flow gives the curve a flatter shape, which would indicate a smaller variation in the flow data in this catchment. On the other hand, the C2 catchment has the lowest values of flow and a more angled shape for the flow duration curve, due to the more pronounced difference between the minimum and maximum values of flow (Figure 6). These characteristics of FDC are being influenced by both the physical properties of the catchment and land use (Smakhtin, 2001; Zhang et al., 2012). Studies conducted using the ratio of maximum and minimum FDC indicated lower effects on hydrological processes in catchments that had lower rates between maximum and minimum, such as C3 and C1 (Brogna et al., 2017; Strauch, MacKenzie, Giardina, & Bruland, 2015).
Thus, the comparison between the indicators Q10, Q50 and Q90 obtained for the three catchments (Figure 6) demonstrates that the greatest differences were detected between C2 and C3, with highlight for the Q90value of C3, which was 75% higher than that of C2. Greater relative effect on minimum flow rates (Q90) has also been observed in other studies, which found a greater relative difference between the minimum flows for changes in land use of the catchment (Brown, Western, McMahon, & Zhang, 2013; Brown et al., 2005; Lane, Best, Hickel, & Zhang, 2005).
In relation to the catchment with mosaic (C1), it is important to note that it had maximum values lower than those of C3 (47%) and a smaller difference in minimum values (38%) compared to C2. The mosaic as an alternative of forest management was discussed in greater detail in item 3.2; however, it is worth mentioning that the results of this item also indicate that, despite a lower water yield, the catchment would be maintaining the hydrological processes (BFI and Q90), as demonstrated in other studies with mosaic (Ogden et al., 2013).
The monthly values of the Q/P ratio obtained for the three catchments show that C3 (0.16) differs significantly (p<0.05) from C1 (0.10) and C2 (0.10) (Figure 7). This significant difference in the Q/P ratio reaffirms the results found of higher water yield in C3 compared to the other two catchments; despite the implementation of the eucalyptus forest, there were no negative effects in the first years.
Similarly, the catchments differ significantly (p<0.05) from one another in relation to the concentration of total suspended solids, and C3 has the highest concentration (5.3 mg L-1) and C1 has the lowest concentration (2.3 mg L-1) (Figure 7). These differences between the catchments may be related to the management carried out in C3 (plantation), but also the higher precipitation in the period and water yield may have resulted in a higher concentration of solids. Forest management, such as harvesting and planting, has the potential to influence the concentration of solids (Baillie & Neary, 2015; Grace, 2005), but the high water yield in C3 may have contributed to the increase in solids. The concentrations found in the catchments are similar to those reported in studies carried out in planted forests (Câmara & Lima, 1999), and the trend is the decrease in concentrations with the growth of forests (C2 and C3) (Feller, 2005).
Differently, in relation to nitrate, the C2 catchment has the highest concentrations (0.49 mg L-1) and differs significantly (p<0.05) from C1 (0.40 mg L-1) and C3 (0.43 mg L-1) (Figure 7). The highest NO3concentration in C2 may be the result of both lower value of flow, since the lower the amount of water, the higher the concentration, and the demand for this nutrient by the forests planted in C1 and C3 (Laclau et al., 2010).