Scenario Outcomes

smvdhsvm

What are the impacts of changes in landuse practices from the scenarios on water balances?

For both the Jucu and SMV basins, we demonstrated that increased highland agriculture caused higher unregulated annual water yields especially during the low flow season (November - January) for both the Jucu and SMV river basins (Figure 3.10 and Table 3.2). Dry season flows increased by 29% in the SMV and 31% in the Jucu while wet season flows increased by 16% in the SMV and 14% in the Jucu. Overall, crop expansion increased annual flows by 22% in the SMV and by 30% in the Jucu. Forest to crop conversion reduces the transfer of precipitation to the ground due to lower evapotranspiration rates, thus increasing discharge. Agriculture in this region of Brazil has also caused soil compaction, lowering infiltration rates and hydraulic conductivity, and causing excess in overland flow (Hamza & Anderson, 2005). Furthermore, since the Jucu has less croplands than the SMV, an higher increase in runoff is to be expected.

When existing agriculture was converted to forests (baseline to Scenario I), annual water yield decreased but evapotranspiration increased (Jucu - Figure 3.11, SMV – Figure 3.12). In the Jucu, average flows were decreased by 30% and in the SMV, annual flows were decreased by 36%. Based on future scenario analyses, effects of land use change on seasonal and annual water yields are a net balance of change in basin moisture storage size, vegetation-soil interaction, and flow regulation. Tropical forests have higher evaporation from rainfall interception and transpiration than other landcover types as evidenced in our study and corroborated by Worden et al. (2007). The basin hydrology is particularly sensitive to changes in landcover attributes, with a general pattern of increasing runoff with migration from trees to crops due to decreasing evapotranspiration. Upland crop expansion may lead to higher peak flow and higher seasonal and annual yields. This is due to increased water yields resulting from reduced evapotranspiration (Table 3.3). Since the increased water percolates through well-drained soils, primary implications are for downstream main channel flooding rather than increased overland flow. Extractions for irrigation may also reduce downstream impacts and will need to be investigated further. 


Table 3.2. Comparison of observed and simulated annual (January to December) hydrologic compositions for Jucu and SMV

Average annual values (2009 – 2011)
Annual yield m3/s High flow m3/s Low flow m3/s Annual ET a mm Runoff ratio
Jucu
Observed 30.1 17.4 39.4 1120a  0.36
DHSVM 28.2 15.7 31.4 1410 0.28
SMV
Observed 22.6 33.2 13.1 1070a 0.35
DHSVM 20.3 28.6 11.8 1325 0.32
a Value from Satellite-derived Global Record of Monthly Land Surface Evapotranspiration (1983-2006)

Increasing eucalyptus areas decreased discharge (from the IEMA2007 baseline) by an average of 26% in the Jucu and 30% in the SMV (Figure 3.10). The results from increased agriculture and increased eucalyptus scenarios are consistent with studies by Viola et al., 2014 that showed that an increase in eucalyptus areas in Southeastern Brazil would decrease discharge while an increase in agriculture would cause the opposite effect. However, the difference in discharge between Euc.+25% and Euc.+50% was indiscernible in both basins. Smethurst et al., 2014 suggested that eucalyptus plantations would not have any significant impacts on streamflow especially if natural forests buffer the eucalyptus plantations.  The effect of eucalyptus and natural forests are similar in some respects, and this may help explain why the difference between the two different eucalyptus buffer values was minimal; both study basins have a high concentration of native forests interspersed with and surrounding eucalyptus plantation.
Table 3.3. Simulated basin hydrology from different landcover scenarios

Landcover Scenarios Annual yield m3/s High flow m3/s Low flow m3/s Annual ET mm
Jucu
GEOBASE1996 19.3 51.7 29.8 1120
Prim+ 16.7 45.1 27.1 982
Euc. + 25% 18.5 50.0 29.5 897
Euc. + 50% 18.4 49.5 29.3 915
Ag. + 25% 32.7 86.3 55.2 785
SMV
GEOBASE1996 14.5 19.1 9.5 975
Prim+ 12.4 16.5 8.7 795
Euc. + 25% 13.7 18.5 9.4 825
Euc. + 50% 13.1 20.6 6.4 892
Ag. + 25% 23.9 30.8 16.5 692


Longer-term observations of streamflow would be needed in order to fully understand the mechanisms that govern eucalyptus effects on the water uptake and evapotranspiration in these particular basins. Effects of land use on streamflow in the SMV also seem to be governed by underlying soils. Higher soil moisture (Figure 3.11) corresponds to areas containing latosols, a very old type of clay found mostly in the tropics that have a high moisture holding capacity. The effects of this soil type may also account for the differences in the amount of streamflow decrease between the SMV and Jucu. As the soil retains more moisture, an increase in evaporation due to increase eucalyptus areas would reduce the water available for overland runoff.


The results of this study further show that runoff increased substantially from 1997 to 2007. It is surmised that this could be due to increased lowland agriculture. This implies that there needs to be a radical rethinking or development of incentives to change current agricultural practices in this region. The type of monoculture used and the types of landcover replaced differentially impacts delivery of water to the coastal zone. For example, if eucalyptus were to replace primary forest, differences would likely be small. If low-lying vegetation were replaced with eucalyptus, discharge would be reduced and ET increased. Alternatively, if the existing landcover were replaced with crops, discharge would increase. Results from the modeling efforts show that forest cover decreases downstream streamflow. Increasing forest cover also reduces the amount of exposed soil susceptible to erosion since forest covers have deeper roots that stabilize soil and that reduce the rainfall through the canopy layer. Further studies would be useful in investigating contributions of agricultural irrigation to changes in streamflow.

Figure 1: Average monthly discharge comparisons from the different landcover scenarios for the (a) SMV (top) and (b) Jucu (bottom)

    We have shown that land cover scenario analysis using a hydrological model is useful analytical tool in helping stakeholders and water managers understand the basin dynamics and the interactions between crops and forest cover and the effects on evapotranspiration and soil moisture. The model was able to replicate hydrographs for both basins with a high Nash-Sutcliffe efficiency even though data was particularly sparse. This modeling approach can be useful in assessing the influence of spatial configuration or fragmentation of land covers.

Land use changes in the Jucu and SMV river basins are largely due to increases in farming and expansion in eucalyptus plantations. I have shown here, using hypothetical landcover scenarios, that agricultural crops change the hydrologic regime of the river basins far more than eucalyptus growth, particularly in the SMV basin where there exist many manmade private irrigation dams and the forest cover is slightly more fragmented. In the SMV, soil types also play an important role in modulating the hydrologic regime.
The primary take away from this study however, is a more practical one – helping the local Espírito Santo state government form effective policies to address land use change issues that would ultimately impact biodiversity and water resources of the region. On a global scale, the results of this study and the methods discussed herein would be easily transferable to tropical regions and other areas with similar climatology and governance needs as Espírito Santo.

Figure 2: Comparisons of evaporation for the different landcover scenarios in the SMV river basin: a) IEMA2007 b) GEOBASE1996 c) Increase in Primary Forests d) Euc. + 25% e) Euc. + 50% and f) Ag. + 25%
Figure 3: Comparisons of soil moisture for the different landcover scenarios in the SMV river basin: a) IEMA2007 b) GEOBASE1996 c) Increase in Primary Forests d) Euc. + 25% e) Euc. + 50% and f) Ag. + 25%
Figure 4: Comparisons of evaporation for the different landcover scenarios in the Jucu river basin: a) IEMA2007 b) GEOBASE1996 c) Increase in Primary Forests d) Euc. + 25% e) Euc. + 50% and f) Ag. + 25%
Figure 5: Comparisons of soil moisture for the different landcover scenarios in the Jucu river basin: a) IEMA2007 b) GEOBASE1996 c) Increase in Primary Forests d) Euc. + 25% e) Euc. + 50% and f) Ag. + 25%

    Land use change in the Jucu and SMV river basins are largely due to increases in highland farming and expansion in eucalyptus plantations. We have shown here, using hypothetical landcover scenarios, that agricultural crops change the hydrologic regime of the river basins far more than eucalyptus growth, particularly in the SMV basin where there exists many manmade private irrigation dams and the forest cover is slightly more fragmented.

    We also showed that DHSVM is useful analytical tool in helping stakeholders and water managers understand the basin dynamics and the interactions between crops and forest cover and the effects on evapotranspiration and soil moisture. The model was able to replicate hydrographs for both basins with a high Nash-Sutcliffe efficiency even though data was particularly sparse. This modeling approach can be useful in assessing the influence of spatial configuration or fragmentation of landcovers.

    The basin hydrology is sensitive to changes in landcover attributes, with a general pattern of increasing runoff with migration from trees to crops due to decreasing evapotranspiration. Highland crop expansion may lead higher peak flow and higher seasonal and annual yields. This is due to increased water yields resulting from reduced evapotranspiration.  Since the increased water percolates through well-drained soils, primary implications are for downstream main channel flooding rather than increased overland flow, and extractions for irrigation may reduce downstream impacts.