Final Report Summary - MED-AFFOREST (Afforestation management in Mediterranean mountain areas: Evaluating the environmental consequences)
Land abandonment and subsequent revegetation processes (due to secondary succession and afforestation practices) are global issues with important implications in Mediterranean mountain areas. The general aim of the MED-AFFOREST project is to apply an interdisciplinary approach to understanding the effect of afforestation on the hydrological dynamics, soil properties, carbon dynamics and landscape structure at different spatial scales (e.g. micro-reliefs, experimental catchments). This objective includes a comparison between different land uses and land covers (bare areas, meadows, secondary succession, afforestation with Pinus sylvestris and Pinus nigra and natural forests) using experimental sites in a sub-humid Mediterranean mountain area (Central Spanish Pyrenees).
To gain insight into the effects of land abandonment and afforestation practices, we have applied landscape metrics (including dominance, fragmentation, diversity, evenness, and landscape geometry indices) for 1956, 1981 and 2009 landscapes by using FRAGSTATS free software. The results of the landscape metrics analyses provide us a global understanding of key trends of landscape structure and its transformation processes. The general increment in the number of patches, including patch density, and the decline of the largest patch index suggested that the present landscape composition is more fragmented (and consequently connectivity tended to decrease). However, important changes in Shannon’s diversity and evenness indices were not registered in the case study, indicating that both diversity and distribution of areas have not changed over the study period.
Several publications have reviewed the impact of land abandonment and revegetation processes on the soil properties dynamics but, so far, limited attention has been paid to Mediterranean sub-humid mountain areas. In order to address this research gap, different land covers were examined. We observed that changes in soil properties after 50 years of land abandonment are limited, even when afforestation practices were carried out, and no differences were observed between both revegetation processes. A well-developed soil quality index -based on statistical analysis- suggested that P. nigra microsites received a higher soil quality rating (no significant different with P. sylvestris and secondary succession). Our general results also demonstrated that the impact of disturbance by afforestation techniques was difficult to discern and no significant differences were observed between the microsites created with the afforestation practices.
Although a large amount of soil carbon is stored in subsurface soils, most studies of C soil dynamics focus on the upper layers. We have carried out a second research to assess the effects of land abandonment and revegetation processes on soil organic carbon stocks (SOC) in the soil profile (0-50 cm). The results indicated that the percentage of SOC in the first 20 cm averaged 37%, 54%, 62% and 48% for meadows, secondary succession, P. sylvestris and P. nigra respectively, reflecting the importance of subsurface soils in C soil dynamics. The total SOC concentrations down to a depth of 50 cm was larger at the P. nigra sites as compared with the other land uses. Afforestation, especially the ones with P. nigra , resulted in a constant C accumulation in the first 10 centimetres.
50 years after land abandonment, aggregate distribution, SOC fractions, and soil respiration were measured. Our study revealed that the interplay between aggregation processes, and the formation of mineral associated SOC was important to understand changes in the amount and stability of SOC in the different land covers. The main changes occurred in the top soil layers (0-10 cm) where: (i) coarser, and larger, and more stable aggregates were observed in the afforestation sites, (ii) the labile fractions increased in afforestation sites compared to the different land covers; and (iii) soil respiration was higher in P. nigra sites. All these changes promoted the formation and stabilization of SOC in the afforestation sites. Aggregation processes were significantly greater in the afforestation soils and, the relative contribution of macro-aggregates to the total SOC was higher at P. nigra sites and tend to increase with depth (suggesting the increase in physical protection of SOC).
Furthermore, the contribution of various density fractions to total SOC differed between land uses in the first 10 cm. These differences (higher labile proportions in the afforestation) were a result of the greater amount of organic inputs from conifers. The respiration tests showed that P. nigra yielded the highest soil CO2 emissions, with a decrease of soil CO2 emissions with depth. The highest C mineralization rates were recorded in the afforestation soil samples.
High input of plant litter, as well as root exudates, have led to increases of SOC concentration in P. nigra afforestation sites, although no differences in SOC stocks were observed (only with bare areas). The first step of SOC stabilization after afforestation is the formation of macro-aggregates promoted by large inputs of organic C rich topsoil with a high contribution of labile organic matter. Besides, the P. nigra showed the highest lignin content following the common trend of higher lignin content in the top soil and lower lignin content in the deepest soil.
The hydrological response recorded in the afforestation catchment showed that afforestations: (i) produced moderate peak flows, storm flows, and long response times; (ii) no one single variable was able to explain the hydrological response, although antecedent moisture conditions did play an important role, and (iii) the effect on peak flows became less important as the size of the hydrological event increased. At catchment scale, we compared annual storm flow coefficients related to different land covers: (i) the highest coefficient (0.54) was obtained in the more degraded badland areas; (ii) in the old cultivated catchment, now abandoned with secondary succession processes, the coefficient was 0.27; (iii) in the natural non-disturbed forest environment the lowest value was obtained (0.19); and (iv) in the afforestation area the coefficient was 0.21 suggesting a reduction of water yield related to bare areas and secondary succession areas. Annual rainfall is almost the same, therefore, the differences in storm flow production must be related to the vegetation cover (water consumption by plant cover, interception), soil properties and soil development.
The temporal dynamics of groundwater was also investigated in the afforestation catchment. The analysis in five locations demonstrated that the groundwater dynamics markedly varied over the year. If we compared the water table information under the different land covers, important differences can be described: (i) in the natural forest, the water table dynamics showed a high variability, with very fast responses due to the high presence of roots and macro-pores. However, there were no saturation periods, suggesting that runoff was generated essentially by subsurface processes; (ii) in the natural revegetation area, a marked seasonal cycle was observed with a saturation period during winter and spring (runoff is generated by a combination of different processes); and (iii) in the afforestation area, the water table dynamics showed a high variability, with really fast responses and rapid oscillations. Consequently, runoff was generated by a combination of different processes: infiltration excess runoff, saturated excess runoff and subsurface flow.
In general, afforestation polices in Mediterranean mountain areas pursued economic and environmental purposes: (i) achieve self-sufficiency in the supply of pulp and paper, and (ii) regulate the hydrological cycle, in order to reduce flood frequency and magnitude, and reduce erosion processes and soil degradation. Regarding economic purposes, results were scarce, because so far, wood material was not obtained. With respect to water, this project demonstrated that afforestation reduced the water yield and the number of floods compared to bare areas and abandoned lands. Moreover, afforestation practices failed in the control of extreme events. The present study has also demonstrated that after 50 years of land abandonment, soil properties and consequently soil quality is similar under secondary succession in comparison to afforestations. Only, small differences were recorded in the SOC concentrations, aggregation and aggregate stability, soil organic pools and C mineralization (needing further exploration). The results of this research can be useful for forest management and environmental planners in order to decide the best practices after land abandonment. However, given the described context, an emergent research dilemma can be easily identified: How to proceed at policy-making level in the longer-term: secondary succession processes or fostering afforestation practices?
To gain insight into the effects of land abandonment and afforestation practices, we have applied landscape metrics (including dominance, fragmentation, diversity, evenness, and landscape geometry indices) for 1956, 1981 and 2009 landscapes by using FRAGSTATS free software. The results of the landscape metrics analyses provide us a global understanding of key trends of landscape structure and its transformation processes. The general increment in the number of patches, including patch density, and the decline of the largest patch index suggested that the present landscape composition is more fragmented (and consequently connectivity tended to decrease). However, important changes in Shannon’s diversity and evenness indices were not registered in the case study, indicating that both diversity and distribution of areas have not changed over the study period.
Several publications have reviewed the impact of land abandonment and revegetation processes on the soil properties dynamics but, so far, limited attention has been paid to Mediterranean sub-humid mountain areas. In order to address this research gap, different land covers were examined. We observed that changes in soil properties after 50 years of land abandonment are limited, even when afforestation practices were carried out, and no differences were observed between both revegetation processes. A well-developed soil quality index -based on statistical analysis- suggested that P. nigra microsites received a higher soil quality rating (no significant different with P. sylvestris and secondary succession). Our general results also demonstrated that the impact of disturbance by afforestation techniques was difficult to discern and no significant differences were observed between the microsites created with the afforestation practices.
Although a large amount of soil carbon is stored in subsurface soils, most studies of C soil dynamics focus on the upper layers. We have carried out a second research to assess the effects of land abandonment and revegetation processes on soil organic carbon stocks (SOC) in the soil profile (0-50 cm). The results indicated that the percentage of SOC in the first 20 cm averaged 37%, 54%, 62% and 48% for meadows, secondary succession, P. sylvestris and P. nigra respectively, reflecting the importance of subsurface soils in C soil dynamics. The total SOC concentrations down to a depth of 50 cm was larger at the P. nigra sites as compared with the other land uses. Afforestation, especially the ones with P. nigra , resulted in a constant C accumulation in the first 10 centimetres.
50 years after land abandonment, aggregate distribution, SOC fractions, and soil respiration were measured. Our study revealed that the interplay between aggregation processes, and the formation of mineral associated SOC was important to understand changes in the amount and stability of SOC in the different land covers. The main changes occurred in the top soil layers (0-10 cm) where: (i) coarser, and larger, and more stable aggregates were observed in the afforestation sites, (ii) the labile fractions increased in afforestation sites compared to the different land covers; and (iii) soil respiration was higher in P. nigra sites. All these changes promoted the formation and stabilization of SOC in the afforestation sites. Aggregation processes were significantly greater in the afforestation soils and, the relative contribution of macro-aggregates to the total SOC was higher at P. nigra sites and tend to increase with depth (suggesting the increase in physical protection of SOC).
Furthermore, the contribution of various density fractions to total SOC differed between land uses in the first 10 cm. These differences (higher labile proportions in the afforestation) were a result of the greater amount of organic inputs from conifers. The respiration tests showed that P. nigra yielded the highest soil CO2 emissions, with a decrease of soil CO2 emissions with depth. The highest C mineralization rates were recorded in the afforestation soil samples.
High input of plant litter, as well as root exudates, have led to increases of SOC concentration in P. nigra afforestation sites, although no differences in SOC stocks were observed (only with bare areas). The first step of SOC stabilization after afforestation is the formation of macro-aggregates promoted by large inputs of organic C rich topsoil with a high contribution of labile organic matter. Besides, the P. nigra showed the highest lignin content following the common trend of higher lignin content in the top soil and lower lignin content in the deepest soil.
The hydrological response recorded in the afforestation catchment showed that afforestations: (i) produced moderate peak flows, storm flows, and long response times; (ii) no one single variable was able to explain the hydrological response, although antecedent moisture conditions did play an important role, and (iii) the effect on peak flows became less important as the size of the hydrological event increased. At catchment scale, we compared annual storm flow coefficients related to different land covers: (i) the highest coefficient (0.54) was obtained in the more degraded badland areas; (ii) in the old cultivated catchment, now abandoned with secondary succession processes, the coefficient was 0.27; (iii) in the natural non-disturbed forest environment the lowest value was obtained (0.19); and (iv) in the afforestation area the coefficient was 0.21 suggesting a reduction of water yield related to bare areas and secondary succession areas. Annual rainfall is almost the same, therefore, the differences in storm flow production must be related to the vegetation cover (water consumption by plant cover, interception), soil properties and soil development.
The temporal dynamics of groundwater was also investigated in the afforestation catchment. The analysis in five locations demonstrated that the groundwater dynamics markedly varied over the year. If we compared the water table information under the different land covers, important differences can be described: (i) in the natural forest, the water table dynamics showed a high variability, with very fast responses due to the high presence of roots and macro-pores. However, there were no saturation periods, suggesting that runoff was generated essentially by subsurface processes; (ii) in the natural revegetation area, a marked seasonal cycle was observed with a saturation period during winter and spring (runoff is generated by a combination of different processes); and (iii) in the afforestation area, the water table dynamics showed a high variability, with really fast responses and rapid oscillations. Consequently, runoff was generated by a combination of different processes: infiltration excess runoff, saturated excess runoff and subsurface flow.
In general, afforestation polices in Mediterranean mountain areas pursued economic and environmental purposes: (i) achieve self-sufficiency in the supply of pulp and paper, and (ii) regulate the hydrological cycle, in order to reduce flood frequency and magnitude, and reduce erosion processes and soil degradation. Regarding economic purposes, results were scarce, because so far, wood material was not obtained. With respect to water, this project demonstrated that afforestation reduced the water yield and the number of floods compared to bare areas and abandoned lands. Moreover, afforestation practices failed in the control of extreme events. The present study has also demonstrated that after 50 years of land abandonment, soil properties and consequently soil quality is similar under secondary succession in comparison to afforestations. Only, small differences were recorded in the SOC concentrations, aggregation and aggregate stability, soil organic pools and C mineralization (needing further exploration). The results of this research can be useful for forest management and environmental planners in order to decide the best practices after land abandonment. However, given the described context, an emergent research dilemma can be easily identified: How to proceed at policy-making level in the longer-term: secondary succession processes or fostering afforestation practices?