This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and

Transcription

1 This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier s archiving and manuscript policies are encouraged to visit:

2 Catena 79 (2009) Contents lists available at ScienceDirect Catena journal homepage: Investigating the effects of afforestation on soil erosion and sediment mobilisation in two small catchments in Southern Italy Paolo Porto a,b,, Des E. Walling a, Giovanni Callegari c a Department of Geography, University of Exeter, United Kingdom b Dipartimento di Scienze e Tecnologie Agro-Forestali e Ambientali, Università degli Studi Mediterranea di Reggio Calabria, Italy c C.N.R. Istituto per i Sistemi Agricoli e Forestali del Mediterraneo, Sezione Ecologia e Idrologia Forestale, Rende (Cs), Italy article Keywords: Caesium-137 Erosion rates Forest harvesting Lead-210 Sediment yield Southern Italy info abstract Annual soil losses in southern Italy can exceed t ha 1 year 1. Where erosion on agricultural land is particularly severe, land use change and afforestation are frequently seen as the most appropriate means of reducing erosion risk. However, the overall effectiveness of afforestation in reducing soil erosion remains uncertain, due to the poor development of the forest cover in some areas, leading to significant areas with sparse tree cover, and the erosional impact of forest harvesting, which commonly involves clearcutting. The study reported here addresses this uncertainty and focuses on two small catchments (W2 and W3) located in Calabria, southern Italy, for which measurements of suspended sediment yield are available. Both the catchments originally supported a rangeland vegetation cover and they were planted with eucalyptus trees in Currently, only catchment W3 supports a continuous forest cover. In catchment W2 the forest cover is discontinuous and there is a significant area of the catchment (ca. 20%) where the tree cover is sparse and the vegetation cover is dominated by natural grasses. Two additional erosion plots were established within catchment W2 in 1991, in order to explore the effect of the density of the tree cover on soil erosion. Information on the sediment yields from the two catchments and the plots for 10 storm events that occurred during the period December 2005 December 2006 and associated information on the 137 Cs and excess 210 Pb of the sediment, have been used to investigate the effectiveness of afforestation in reducing sediment mobilisation and net soil loss from the catchments involved. The results demonstrate that the areas of greatest soil loss are associated with the slopes where the tree cover is discontinuous, and that forest harvesting by clearcutting causes significant short-term increases in sediment mobilisation and sediment yield. These findings, which are consistent with previous work undertaken within the same area, emphasize the importance of vegetation cover density in influencing rates of soil loss in the study catchments. The study also provided a useful demonstration of the potential for using measurements of the 137 Cs and 210 Pb ex content of sediment, in combination with more traditional sediment monitoring, to investigate sediment sources and to compare the sediment dynamics of catchments subjected to different land management practices Elsevier B.V. All rights reserved. 1. Introduction The on-site and off-site impacts of soil erosion are increasingly seen as key concerns for the sustainable management of soils and river basins, since they can result in reduced soil productivity and the degradation of river water quality and aquatic ecosystems, as well as introducing problems for water resource development through reservoir sedimentation. For Southern Italy, Sorriso-Valvo et al. (1995) suggested that soil erosion rates as high as 10 mm year 1 (ca t ha 1 year 1 ) can occur in the mountainous Calabrian region. In such contexts, there has been a drive to reduce soil erosion Corresponding author. University Mediterranea of Reggio Calabria, Località Feo di Vito, Reggio Calabria, Italy. Tel.: address: paolo.porto@unirc.it (P. Porto). by implementing improved management practices and erosion control measures. In some agricultural areas, where soil erosion is particularly severe, land use change and afforestation have been seen as the most appropriate means of reducing erosion risk. Some uncertainty does, however, exist regarding the effectiveness of afforestation programmes in reducing soil loss and catchment sediment yields, including the most appropriate forest species, particularly when the climate and soil conditions mean that the resulting forest cover may be sparse and not well-developed. Eucalyptus trees have, for example, been widely planted, but the canopy cover provided by such trees is generally less dense than that associated with other tree species. Furthermore, the need to obtain some economic return from the afforested land means that forest harvesting, frequently involving clearcutting, is an essential feature of the changed land management /$ see front matter 2009 Elsevier B.V. All rights reserved. doi: /j.catena

3 182 P. Porto et al. / Catena 79 (2009) Table 1 The characteristics of the study catchments. Catchment Drainage area Ha Mean altitude m a.s.l. Mean slope % Mean soil content Measured annual sediment yield Sand Silt Clay Min Max Mean % % % t ha 1 year 1 tha 1 year 1 W W W Estimated long term sediment yield Forest harvesting, which results in complete clearance of the tree cover prior to regrowth, can clearly be expected to increase the sediment yield from a forested catchment, relative to a catchment with a long-term undisturbed forest cover, but it would not necessarily be expected to cause the sediment yield to greatly increase above that of a catchment with natural rangeland cover. To meet a need for empirical evidence regarding the effectiveness of afforestation in reducing soil erosion and catchment sediment yields, a number of catchment experiments were initiated in southern Italy in the latter part of the 20th century. These included a study implemented in 1978 within the framework of the National Research Council of Italy (CNR) Soil Conservation Project, that investigated the hydrological response and sediment yield of three small catchments, ranging in size from 1.38 ha to 1.65 ha, located near Crotone in Calabria (see Avolio et al., 1980; Iovino and Puglisi, 1991). Detailed information on two of these catchments and the measurements undertaken are provided later in this paper. In essence, however, the study comprised three catchments, one (catchment W1) that was left as original rangeland with grass and scrub vegetation, one with a good cover of eucalyptus trees (Eucalyptus occidentalis Endl.) planted in 1968 (catchment W3) and one with a patchy (ca. 80%) cover of eucalyptus trees developed from trees again planted in 1968 (catchment W2). A summary of the main physiographic characteristics of the three catchments and the available estimates of annual sediment yield from these three catchments, over the period 1978 to 1994, which included episodes of forest harvesting, is presented in Table 1. Although it is not possible to make precise comparisons between the sediment yields of the three catchments, due to minor difference in catchment area, altitude, slope and soil characteristics between the catchments, the sediment yield data provided in Table 1 emphasize that, in this location, afforestation may not necessarily reduce the mean annual sediment yield. The sediment yield from catchment W3, which is characterized by a good forest Fig. 1. The study area.

4 P. Porto et al. / Catena 79 (2009) cover, is the lowest for the three catchments and this suggests that a good forest cover will reduce the sediment yield relative to the rangeland vegetation cover. However, the sediment yield from catchment W2, with a less uniform forest cover is substantially higher than those of the other two catchments. The results presented in Table 1 suggest that the presence of an incomplete forest cover, with only sparse grass and sometimes bare areas in the areas without good cover, is the primary cause of the relatively high sediment yield documented for catchment W2. In order to explore further the importance of the areas of poor forest cover within catchment W2 as a sediment source and the impact of forest harvesting and associated clearcutting on sediment mobilisation, additional sampling was undertaken within catchments W2 and W3 during the period December 2005 to December Catchment W3 was clearcut for forest harvesting in March Attention focused on combining information on sediment yield from the two catchments, obtained using traditional monitoring, with measurements of the fallout radionuclides caesium-137 ( 137 Cs) and excess lead-210 ( 210 Pb ex ) undertaken on sediment samples from the study catchments. These sediment samples included samples collected at the catchment outlets and also from the outlets of two plots with contrasting vegetation cover, located within catchment W2, to provide additional information on sediment sources. 2. The study catchments The study focuses on two small (1.38 and 1.65 ha) basins (W2 and W3) located near Crotone ( N, E) in Calabria, southern Italy (see Fig. 1). The catchments are located in the ephemeral headwaters of the larger Crepacuore basin (Fig. 1), which are incised into the Upper Pliocene and Quaternary clays, sandy clays and sands underlying the local area (Sorriso-Valvo et al., 1995). These catchments have never been cultivated and originally supported a rangeland vegetation cover. More recently, however, they were afforested and they are now characterized by a forest cover consisting of eucalyptus trees originally planted in These trees were cut in 1978 and 1990 in catchment W2 and in 1986 and 2006 in catchment W3, respectively, with the tree cover being subsequently restored by natural regrowth. Logging operations took place across the entire area of the catchments and involved the removal of ca m 3 of biomass during each cutting campaign. The harvested timber was extracted from the catchments manually, in order to minimize the impact on the soil surface. In catchment W3 the forest cover is currently almost continuous and only about 2 3% of the area is characterized by a grass cover. Within catchment W2, the tree cover is less uniform and about 20% of its area, located on south facing slopes, supports a cover of sparse discontinuous trees and natural grass and other herbaceous plants. The climate of the area is typically Mediterranean, with a mean annual precipitation of ca. 670 mm at Crotone (10 km distant), most of which falls during the winter season extending from October to March. In 1978 these catchments were selected for instrumentation within the framework of the National Research Council of Italy (CNR) Soil Conservation Project, in order to monitor the effects of afforestation on their hydrological response and sediment yield (Avolio et al., 1980; Iovino and Puglisi, 1991). Each catchment has been instrumented to provide records of rainfall, runoff and sediment yield. Precipitation has been recorded using a tipping bucket raingauge. Runoff has been measured at the outlet of each catchment using an H-flume structure (Brakensiek et al., 1979) equipped with a mechanical stage recorder. The sediment load passing the gauging structure is measured using a Coshocton wheel sampler installed below the H-flume. This sampler diverts a representative fraction of the runoff and sediment output from the catchment to a tank. After each storm event, the sediment load collected in the tank is well mixed and several 1 L suspended sediment samples are collected from different depths within the tank. The sediment concentrations associated with such samples are determined by oven drying at 105 C and the mean sediment concentration of the samples is calculated. The event sediment yield from each catchment, is then calculated as the product of the mean sediment concentration and the water volume measured in the tank, taking into account the fraction of the total flow diverted to the storage tank by the Coshocton wheel. Due to periodic malfunctioning of the sampling device, full Fig. 2. The location of plot P1 and plot P2 in the W2 catchment.

5 184 P. Porto et al. / Catena 79 (2009) records of sediment yield at the catchment outlets are not available for the years 1995 to In December 2005 the sampling equipment was refurbished and the monitoring activity recommenced, with no loss of data to date. In 1991 two large plots (Fig. 2), covering ca m 2, were established within catchment W2, in order to investigate the effects of variations in the density of the tree cover on soil erosion and sediment mobilisation (see Callegari et al., 1994; Porto et al., 2005). These plots have been included within this study, because they can be considered to be representative of the different vegetation covers in the catchment. Plot P1, with a mean slope of 48% and located on a south facing slope, has very few trees and is characterized by a sparse grass cover, which is representative of the original vegetation of the study area. This plot, with a canopy cover ranging from 0 to 10%, can also be considered to be representative of the south facing slopes of the catchment. Plot P2, which has a mean slope of 35% and is located on a north facing slope, supports a cover of eucalyptus trees and is representative of the afforested area. This plot, with a canopy cover ranging from ca. 50 to 60%, can also be considered representative of the north facing slopes of the catchment. Both plots have been instrumented to collect runoff and soil loss at their outlets. A sheet metal cutoff wall, extending 30 cm into the soil and protruding 30 cm above the ground surface, was installed around each plot in order to isolate the plots hydrologically. On the downslope side of each plot, the runoff and sediment output are intercepted by the cutoff wall and diverted into a pipe, which leads into 4 storage tanks of different size. There is currently some uncertainty as to whether the entire output of runoff and sediment from the plots is collected by the storage tanks and malfunctioning of the monitoring equipment during some of the events that occurred during the study period meant that reliable records of the runoff and sediment output were not available for those events. As a result, no attempt has been made to use the data obtained from the plots to compare the magnitude of sediment fluxes from the two plots. However, observations made during storm events indicated that the tanks were able to provide representative samples of the sediment mobilised from the plots and that comparison of the sediment concentrations associated with the collected runoff provided a meaningful measure of the contrasts in sediment output from the outlets of the two plots. These two plots have therefore been used in this study to compare the relative sediment outputs from the areas of catchment W2 characterized by contrasting cover types and to provide representative samples of the sediment mobilised from the contrasting areas of the catchment for further analysis. 3. Sediment sampling and laboratory procedures Several field campaigns were undertaken during the period December 2005 to December 2006, in order to collect sediment samples for laboratory analysis. A total of 20 (10 from each catchment) suspended sediment samples, representative of 10 storm events, were collected from the storage tanks associated with the Coshocton wheel samplers at the outlets of catchments W2 and W3, shortly after the events. Assessment of the mass of sediment accumulated in these tanks also provided an estimate of the total sediment yield associated with each of the 10 events. On the same occasions, representative suspended sediment samples were also collected from the tanks located at the outlets of plots P1 and P2, within catchment W2. Samples representative of all 10 storm events were collected from plot P1; due to absence of runoff for four events only six events were sampled for plot P2. The suspended sediment samples collected from both the catchment outlets and the two plots were air dried and sieved to b2 mm prior to analysis by gamma spectrometry, to determine the mass activity density (Bq kg 1 ) of caesium-137 ( 137 Cs) and unsupported lead-210 ( 210 Pb ex ). Both radionuclides were measured simultaneously, using a high-resolution coaxial HPGe n-type coaxial detector. Count times were typically ca. 80,000 s, providing results with an analytical precision of ca. ±10% at the 95% level of confidence. The total 210 Pb activity of the samples was measured using the 46.5 kev gamma-ray for 210 Pb, and the 226 Ra activity, required to calculate the supported component, was measured using the kev gamma-ray for 214 Pb, a short-lived daughter of 226 Ra. The 137 Cs activities in the samples were obtained from the counts at the 662 kev peak. 4. Results 4.1. Sediment outputs from catchments W2 and W3 The sediment yield data used in this study relate to 10 events documented for catchments W2 and W3 during the period extending from December 2005 to December The sediment outputs from both the catchments for the 10 events are summarised in Table 2, together with information regarding the total rainfall for each event. The information provided by Table 2, together with the cumulative sediment yield during the study period from both catchments presented in Fig. 3, indicates that for the events that occurred during the study period there is a clear difference in sediment yield between the two catchments. The cumulative sediment yield for the 10 events is 9.8 t ha 1 for catchment W2 and 15.6 t ha 1 for catchment W3. This situation contrasts with past findings (Table 1) which indicated that the longer-term mean annual sediment yield was appreciably greater from catchment W2, which had the less complete forest cover. This apparent change in the sediment output from catchment W3, relative to that from catchment W2, reflects the clearcutting that occurred in catchment W3 during the latter part of The clearcutting commenced in March 2006 and Fig. 3 indicates that for the set of storm event that occurred after mid September 2006, the sediment output from catchment W3 was significantly greater than that from catchment W2. The increase is particularly marked for the high magnitude event that occurred on December 2006, Table 2 Sediment output and associated radionuclide activities for individual storm events in the two study catchments. Date Code Rainfall Sediment output (t ha 1 ) Cs-137 (Bq kg 1 ) Pb-210 (Bq kg 1 ) mm W2 W3 W2 W3 W2 W /12/05 EV ND /12/05 09/01/06 EV ND ND /01/06 EV and 28/02/06 EV ND /04/06 EV /06/06 EV /07/06 EV /09/06 EV /12/06 EV ND /12/06 EV Tot Tot Mean Mean

6 P. Porto et al. / Catena 79 (2009) Fig. 3. Cumulative sediment yield from the two study catchments associated with the 10 events during the study period. which accounted for ca. 80% of the total sediment yield for the year from catchment W3. In this case, the sediment yield from catchment W3 (11.3 t ha 1 ) was more than double that from catchment W2 (4.7 t ha 1 ). The effects of tree cutting in increasing erosion and sediment yield had been previously documented in catchment W2 during the period , and emphasize the importance of vegetation in protecting the soil (Cantore et al., 1994). Information on the concentrations of 137 Cs and unsupported 210 Pb ex associated with the suspended sediment collected from the outlets of the two catchments is provided in Table 2. The mean values for catchment W2 (3.1 and 16.8 Bq kg 1, for 137 Cs and 210 Pb ex, respectively) are lower than those listed for catchment W3 (4.2 and 24.8 Bq kg 1 ). This contrast in the 137 Cs and 210 Pb ex activity of sediment sampled at the outlets of the two study catchments is further demonstrated by the plots of radionuclide loss versus sediment yield presented in Fig. 4. The generalised relationships fitted to the two plots indicate that for both radionuclides, the event-based output flux associated with a given sediment yield is less for catchment W2 than for catchment W3. The reduced 137 Cs and 210 Pb activities associated with the sediment output from catchment W2 reflect a number of controls. Firstly, they are likely to reflect the greater longer-term erosion rates within catchment W2, as evidenced by the greater estimate of longterm sediment yield provided in Table 1. Since the depth distribution of both 137 Cs and 210 Pb ex in the uncultivated soils of the study area is characterized by a rapid exponential reduction with increasing depth (see Porto et al., 2001, 2003), an increased erosion rate will remove a greater depth of soil and the newly exposed soil surface, and thus freshly eroded sediment, will be characterized by a reduced activity. Furthermore, an increased erosion rate will reduce the opportunity for fresh fallout to accumulate at the soil surface and the activity of the surface soil mobilised by erosion will be reduced. Secondly, they may reflect the increased importance of rill erosion, relative to sheet erosion, within catchment W2. By virtue of its deeper source, sediment mobilised by rill erosion will be characterized by a lower 137 Cs and 210 Pb ex activity and will cause the overall activity in the mobilised sediment to be reduced. To date no specific measurements of rill erosion have been undertaken within the catchments but an increased incidence of linear transport pathways within catchment W2 might be expected, due to the presence of areas with reduced forest cover and the reduced amounts of litter found in these areas. The differences in the behaviour of the 137 Cs content of the sediment output from the two catchments, as compared to that for 210 Pb ex, seen in Table 2, are likely to reflect the contrasting temporal distribution of the fallout of the two radionuclides. Since the fallout of 137 Cs effectively ceased in the 1970s, erosion will have progressively reduced the radiocaesium activity at the soil surface, and differences in the 137 Cs activity of surface material between rapidly eroding and less rapidly eroding or stable areas, will have intensified through time. However, 210 Pb ex fallout can be viewed as essentially constant from year to year and the occurrence of fresh fallout each year will replenish the radionuclide activity in the surface soil and thus reduce the progressive intensification of the contrast between eroded and noneroded or less severely eroded surfaces. The marked reduction in the 210 Pb ex activity of the sediment output from catchment W3, apparent towards the end of the study period and therefore after clearcutting of the catchment, could be seen as reflecting the increased erosion rates and possibly an increased incidence of rill erosion, both of which can be expected to mobilise sediment from lower in the soil profile. However, the lack of a clear change in 137 Cs activity over the study period suggests that there was no major change in the incidence of rill erosion, since an increase in the importance of rill erosion could be expected to result in some reduction in 137 Cs activity. By virtue of its accumulation at the surface, as a result of the ongoing fallout input, 210 Pb ex can be expected to be particularly sensitive to changes in the intensity of sheet erosion after clearcutting Sediment and radionuclide output from the two plots in catchment W2 Information on the sediment outputs from the two plots associated with the 10 storm events documented for plot P1 and the six events documented for plot P2 are provided in Table 3, which lists the mean sediment concentration associated with the runoff collected in the tank during the event and the radionuclide content of that sediment. For all events, the sediment concentrations listed for plot P1 are substantially greater than those from plot 2. This difference further confirms the increased erosion rates associated with areas with little or no tree cover in catchment W2, as represented by plot P1, and therefore the importance of the forest canopy in reducing erosion. The values for the 137 Cs and 210 Pb ex mass activity density for the sediment samples collected from the outlets of the plots again show a

7 186 P. Porto et al. / Catena 79 (2009) Fig. 4. Relationships between radionuclide loss and sediment yield for the two study catchments. very clear difference between plots P1 and P2, with the mean value of 137 Cs mass activity density recorded for the sediment samples from plot P2 being nearly an order of magnitude greater than from plot P1. The contrast between the mean 210 Pb ex activities associated with the samples collected from plots P1 and P2 is not as great as that shown by 137 Cs, but the mean value for plot P2 is still more than three times greater than that for plot P1. The contrasts in 137 Cs and 210 Pb ex activity in sediment demonstrated by the two plots reflect contrasts in their erosion dynamics and therefore provide important evidence of the influence of the forest cover in influencing both the magnitude and Table 3 The measurements of sediment concentration and radionuclide activity of sediment samples obtained for the two plots in catchment W2. Date Code Sediment concentration (g l 1 ) Cs-137(Bqkg 1 ) Pb-210(Bqkg 1 ) P1 P2 P1 P2 P1 P /12/05 EV /12/05 09/01/06 EV ~0 a /01/06 EV ND ~0 a ND ND and 28/02/06 EV /04/06 EV ND 0.22 ND ND 1 12/06/06 EV ND 1.68 ND ND 08/07/06 EV /09/06 EV /12/06 EV ND ~0 a ND ND 21 23/12/06 EV Mean a These values were below the detection limit.

8 P. Porto et al. / Catena 79 (2009) the nature of erosion within the areas represented by the two plots. The reduced 137 Cs and 210 Pb ex activities associated with the sediment output from plot P1 can be related to the increased erosion rates associated with that plot. However, the reduced contrast between the two plots evident for the 210 Pb ex content of the eroded sediment suggests that rill erosion is not a major contributor of sediment. 5. Discussion The results presented above have a number of implications related to the use of measurements of the 137 Cs and 210 Pb ex activity of sediment to derive additional information on the sediment dynamics of a catchment (Campbell et al., 1988; Ritchie and McHenry, 1990), as well as for the effectiveness of afforestation in reducing soil erosion in the study area and areas in similar environments (Walling, 1998). A comparison of the 137 Cs and 210 Pb ex activity of sediment from catchments W2 and W3 with that from the two plots (see Tables 2 and 3) reveals several important contrasts, which provide further insight into the erosion processes operating within the two study catchments. The 137 Cs and 210 Pb ex activity of sediment collected from the two plots will primarily reflect the operation of sheet and rill erosion processes across the surface of the plots, whereas the activity of sediment collected from the outlets of the two catchments is likely to be also influenced by the addition of sediment mobilised by linear erosion associated with concentrated flow along drainage lines. Such sediment is likely to be characterized by low 137 Cs and 210 Pb ex activity, due to its source lower in the soil profile (He and Walling, 1997). This situation is clearly evident for 210 Pb ex activity, which, in the case of both catchments, is significantly less than that from the two plots. In the case of 137 Cs, the contrast in activity between the sediment from the plots and that from the two catchments is less marked, because of the progressive reduction in the 137 Cs activity of surface materials due to the lack of significant 137 Cs fallout inputs since the early 1970s and the continuing loss of material from the surface by erosion. Whereas the mean 137 Cs activity of sediment from the two catchments is less than that from plot P2, it is greater than that from plot P1. This is consistent with the findings reported above, which indicate that 137 Cs is more sensitive to current contrasts in erosion rates between the two plots. The values of mean 137 Cs activity associated with the two plots effectively reflect two extremes in the intensity of sheet and rill erosion, associated with the good forest cover (Plot P2) and the poor forest cover (Plot P1). The 137 Cs activity of sediment sampled at the outlet of the two catchments is likely to fall between the two extremes, although, as noted above, it may be further depressed by contributions of sediment mobilised by linear erosion along the drainage lines. The fact that the mean 137 Cs activity of the sediment collected from the outlets of the two study catchments is still greater than that associated with plot P1 suggests that linear erosion along the drainage lines is of limited importance in the two study catchments. If it was a significant sediment source, the 137 Cs activity of the sediment collected at the catchment outlets could be expected to be lower. As shown above, measurements of the 137 Cs and 210 Pb ex activity of sediment can provide additional information on the sediment dynamics of catchment subjected to land use change. In the afforested catchments studied here, for example, measurements of sediment flux at the catchment outlets confirmed the results of existing work in demonstrating the impact of harvesting operations on sediment mobilisation and delivery (see Olive and Reiger, 1985; Anderson and Potts, 1987; Cornish and Binns, 1987). However, these measurements revealed little about the source of the eroded material, or about the specific impact of harvesting or the effects of spatial variation in forest cover on sediment mobilisation within the upstream catchment (see Croke et al., 1999; Wallbrink et al., 2002). Although empirical models, such as the Universal Soil Loss Equation (USLE) and its derivatives have been used to predict the spatial distribution of soil erosion in similar environments, their use is highly dependent on the accurate calibration of several parameters (El-Swaify, 1990), requiring longterm measurements of soil loss, which are very difficult to obtain (Loughran, 1989). Furthermore, such models were developed to represent rill and inter-rill erosion and are unable to take account of linear erosion along drainage lines and related channel erosion. In this context, the use of measurements of fallout radionuclide activity in transported sediment can provide further insights into the sources involved and changes in both erosion rates and sediment sources caused by changing vegetation cover and land use (Loughran and Campbell, 1995; Walling and Woodward, 1992). The conjunctive use of 137 Cs and 210 Pb ex offers considerable potential for elucidating the erosional history of a site. As noted above, the cessation of 137 Cs fallout in the 1970s means that the current 137 Cs activity of soils and sediment reflects the longer-term redistribution of the remaining fallout input (Walling and Quine, 1993). In contrast, the 210 Pb ex activity of soils and sediment is more sensitive to recent changes in the spatial distribution and intensity of erosion, since it is directly influenced by contemporary fallout. The results obtained from this study have highlighted the potentially important effects of discontinuities in forest cover in increasing sediment yields from afforested catchments and the effects of forest harvesting activities in changing sediment sources and increasing sediment mobilisation. This study has emphasized that clearcutting across a whole catchment can cause significant increases in sediment mobilisation and that greater care is arguably required in planning such activities in the future Conclusions This contribution reports the results of a study of the effects of afforestation of two small catchments (catchments W2 and W3) with eucalyptus trees and, more particularly, of variations in forest coverage between the two catchments and the impact of forest harvesting by clearcutting, on their sediment dynamics. Information on sediment output from the catchments and from two small plots established within the catchment W2 characterized by a patchy canopy cover and a good canopy cover have provided the basis for evaluating the effectiveness of afforestation strategies in soil protection. Sediment load data and associated information on the 137 Cs and excess 210 Pb ex activity of sediment collected at the outlet of each catchment, were assembled for 10 storm events that occurred during the period December 2005 to December Contrasts in sediment output from the two catchments have been related to contrasts in the density of the forest cover between the two catchments and to recent clearcutting operations in catchment W3. The results emphasize the important influence of the density of the forest cover on erosion and sediment mobilisation, with erosion rates increasing significantly within areas of sparse forest cover, and the role of forest harvesting by clearcutting in causing temporary increases in sediment yield. The findings therefore highlight the importance of the density of the forest cover in influencing rates of soil loss in the study area and indicate the need for greater attention to developing a good forest cover to provide effective soil protection. The study also provides a useful demonstration of the potential for using measurements of the 137 Cs and 210 Pb ex content of sediment, in combination with more traditional sediment monitoring, to investigate sediment sources and to compare the sediment dynamics of catchments subjected to different land management practices. Acknowledgements The study reported in this paper was supported by grants from MIUR RdB The assistance of Helen Jones and Sue Rouillard in producing the figures and of Jim Grapes in undertaking the gamma spectrometry measurements are gratefully acknowledged.

9 188 P. Porto et al. / Catena 79 (2009) References Anderson, B., Potts, D.F., Suspended sediment and turbidity following road construction and logging in western Montana. Water Resourses Bulletin 23, Avolio, S., Ciancio, O., Grinovero, C., Iovino, F., Mirabella, A., Raglione, M., Sfalanga, M., Torri, D., Effetti del tipo di bosco sull'entità dell'erosione in unità idrologiche della Calabria modelli erosivi. Annali Istituto Sperimentale Selvicoltura 11, (in Italian). Brakensiek, D.L., Osborn, H.B., Sheridan, J.M., Field manual for research in agricultural hydrology. Agriculture Handbook no US Department of Agriculture Science and Education Administration, Washington. Callegari, G., Cinnirella, S., Iovino, F., Erosione e trasporto solido in piccoli bacini interessati da piantagioni di eucalitto. Quaderni di Idronomia Montana 13, (in Italian). Campbell, B.L., Loughran, R.J., Elliott, G.L., A method for determining sediment budgets using cesium-137. Sediment Budgets. Proceedings of the Porto Alegre Symposium, IAHS Press, Wallingford, UK. IAHS Publ. No. 174, pp Cantore, V., Iovino, F., Puglisi, S., Influenza della forma di governo sui deflussi liquidi e solidi in piantagioni di eucalitti. L'Italia Forestale e Montana 5, (in Italian). Cornish, P.M., Binns, D., Stream water quality following logging and wildfire in a dry sclerophyl forest in south east Australia. Forest Ecology and Management 22, Croke, J., Hairsine, P., Fogarty, P., Sediment transport, redistribution and storage on logged forest hillslopes in south-eastern Australia. Hydrological Processes 13, El-Swaify, S.A., Research needs and applications to reduce erosion and sedimentation in the tropics. IAHS Publ. No. 192, pp He, Q., Walling, D.E., The distribution of fallout 137 Cs and 210 Pb in undisturbed and cultivated soils. Appllied Radiation Isotopes 48, Iovino, F., Puglisi, S., L'aménagement des reboisements de protection. Un cas d'étude. 10th World Forestry Congress, Paris. Loughran, R.J.,1989. The measurement of soil erosion. Progress in Physical Geography 13, Loughran, R.J., Campbell, B.L., The identification of catchment sediment sources. In: Foster, I.D.L., Gurnell, A.M., Webb, B.W. (Eds.), Sediment and Water Quality in River Catchments. Wiley, Chichester, pp Olive, L.J., Reiger, W.A., Variation in suspended sediment concentration during storms in five small catchments in south-eastern New South Wales. Australian Geographical Studies 23, Porto, P., Walling, D.E., Ferro, V., Validating the use of caesium-137 measurements to estimate soil erosion rates in a small drainage basin in Calabria, southern Italy. Journal of Hydrology 248, Porto, P., Walling, D.E., Ferro, V., Di Stefano, C., Validating erosion rate estimates by caesium-137 measurements for two small forested catchments in Calabria, Southern Italy. Land Degradation and Development 14, Porto,P.,Walling,D.E.,Callegari,G.,2005.Investigatingsedimentsourceswithinasmall catchment in southern Italy. Sediment Budgets I (Proceedings of symposium S1 held during the Seventh IAHS Scientific Assemblyat Fozdo Iguaçu,Brazil,April 2005),vol.291. IAHS Publ., pp Ritchie, J.C., McHenry, J.R., Application of radioactive fallout 137 Cs for measuring soil erosion and sediment accumulation rates and patterns: a review. Journal of Environmental Quality 19, Sorriso-Valvo, M., Bryan, R.B., Yair, A., Iovino, F., Antronico, L., Impact of afforestation on hydrological response and sediment production in a small Calabrian catchment. Catena 25, Wallbrink, P.J., Roddy, B.P., Olley, J.M., A tracer budget quantifying soil redistribution on hillslopes after forest harvesting. Catena 47, Walling, D.E., Use of 137 Cs and other fallout radionuclides in soil erosion investigations: Progress, problems and prospects. Use of 137 Cs in the Study of Soil Erosion and Sedimentation, IAEA-TECDOC International Atomic Energy Agency, Vienna, pp Walling, D.E., Quine, T.A., The use of fallout radionuclides in soil erosion investigations. Nuclear Techniques in Soil Plant Studies for Sustainable Agriculture and Environmental Preservation. Publication ST1/PUB/947. International Atomic Energy Agency, Vienna, pp Walling, D.E., Woodward, J.C.,1992. Use of radiometric fingerprints to derive information on suspended sediment sources. In: Bogen, J., Walling, D.E., Day, T. (Eds.), Erosion and Sediment Transport Monitoring Programmes in River Basins, IAHS Publication no IAHS Press, Wallingford, pp