Search

ABSTRACTAlthough Chamaecyparis obtusareduces soil pH and cation levels, limited information is available regarding their alterations after clear-cutting. We examined the soil pH and base cations on the upslopes and downslopes of individual trees and stumps of C. obtusaat two distances (0.5 m and 1.0 m) and three depths (0–5, 5–10, and 10–30 cm) in intact- and cut-tree plots. The soil pH was higher in the cut-tree plot than in the intact-tree plot, indicating the recovery of soil pH after clear-cutting. The recovery was more evident in the surface soils (0–5 cm) of downslope areas (0.5 m) than in other soil depths, which implies that the recovery of soil pH was caused by the lack of H+supply from stumps after clear-cutting. The results of H+(in H2O: water-soluble and KCl: exchangeable) suggested a lack of soluble H+in the cut-tree plot. Soil Ca2+and Mg2+levels in the surface soil (0–5 cm) were 1.86-fold and 1.93-fold higher in the cut-tree plot than in the intact-tree plot, respectively, implying that the recovery of divalent cations after clear-cutting was caused by a reduction in cation loss through cation exchange with H+derived from the stemflow. Collectively, soil pH and divalent cations were recovered after clear-cutting because the stemflow had ceased.

ABSTRACTBamboos tend to become abundant associated with human activities and possibly affect ecosystem dynamics including vegetation water cycles. We examined the effects of thinning on canopy transpiration (Ec) for a dense, unmanaged (i.e. unthinned) Moso bamboo stand based on sap flow measurements. We unselectively thinned 45% of the culms during summer in a study plot (i.e. the first thinning). During the next spring, many new culms were generated in the plot and a comparable number of culms were thinned (i.e. the second thinning). After the first thinning, mean stand sap flux (Js) increased immediately by approximately 60% at a given potential evaporation (Ep); therefore, the Ecof the thinned stand reached 90% that of the pre-thinning condition. During the next year, after the second thinning, the Jsof the current year culms was generally more than twice that of the pre-thinning stand at a given Ep, whereas the Jsof the older culms decreased again to the level of the pre-thinning condition. The overall Ecwas 78% that of the pre-thinning condition. Our results suggest that moderate thinning of dense Moso bamboo stands may not have major effects on Ec; thus, intensive thinning is required when the aim is to control water loss from bamboo stands. The relative insensitivity of water use in bamboo stands after thinning resulted from dynamic changes in water use in the individual culms. These processes might be related to the clonal nature of bamboo where culms are structurally and physiologically interconnected.

ABSTRACTStanding trees influence the redistribution of soil water around the trunk and alter water movement under the canopy on hillslopes. We investigated soil properties around trees to clarify the effect of trees on soil hydraulic properties of soil surface layers on a slope. The intact tree and cut tree plots of two coniferous tree species, Cryptomeria japonicaand Chamaecyparis obtusa, were considered in the northern part of Kyushu, Japan. The soil properties on the upslope and downslope of individual trees or stumps were compared. The soil near-saturated hydraulic conductivity (kns) on the upslope of individual trees and stumps was higher than that on the downslope for both species. Although the distribution patterns of knswere similar in all plots, the determinant factors of knsseemed to be different. In the Cr. japonicaintact tree plot, more litter tended to accumulate on the upslope of individual trees, and it seemed to indirectly enhance kns. However, the results of the cut tree plot suggested that knswas affected by factors other than the litter layer biomass. In the Ch. obtusaintact plot, high water repellency of soil on the downslope of individual trees reduced kns. Organic matter tended to accumulate on the upslope of Ch. obtusaindividual stumps and in combination with water repellency appeared to affect kns. Our findings showed that soil hydraulic conductivity was affected by the distance and direction from a tree on a slope and that the effect was maintained for a few years after cutting.

ABSTRACTMany coniferous plantations in Japan have not been managed because of the impact of economic recession on forestry. Few studies have examined rainfall partitioning (RP) in these unmanaged plantations with a stand density ≥ 2500 stems ha−1, which is a typical planting density in Japan. The aims of this study were to examine RP in an unmanaged coniferous plantation and to clarify the differences in RP with other coniferous plantations in Japan. We recorded gross rainfall (GR), throughfall (TF) and stemflow (SF) for one year in an unmanaged 32-year-old Japanese cypress (Chamaecyparis obtusaEndl.) plantation with a high stand density (2500 trees ha−1) in Fukuoka, Japan. The GR, TFand SFwere 2248.6 mm, 1068.4 mm (47.5% of GR) and 425.3 mm (18.9% of GR), respectively, which resulted in interception loss (IL) of 754.9 mm (33.6% of GR). The present data were compared with the published data for all components of RP from 36 previous studies of coniferous plantations in Japan. Compared with these plantations, the present ratio of TF/GRwas the lowest and the SF/GRwas the highest, resulting in the highest IL/GR. This study highlights that not only TFand IL, but also SFaccount for a high proportion of RP in unmanaged coniferous plantations.

Previous studies have demonstrated a clear relationship between diameter at breast height (DBH) and tree transpiration (QT) in multi-specific broadleaved forests. However, these studies were conducted with a limited range of tree sizes and species, and thus many multi-specific broadleaved forests fall outside these conditions. Therefore, this study examined the relationship between DBH and QTin a warm-temperate multi-specific broadleaved forest (n= 12 species) with a wide range of tree sizes (5.0–70.0 cm DBH) using the Granier-type heat dissipation method. The results showed that, although sap flow density varied between individual trees and species, there was a significant relationship between log QTand log DBH (r2= 0.66, P< 0.001) because of the strong dependence of sapwood area on DBH. This study confirmed the applicability of the relationship for the stand transpiration (EC) estimates even in a multi-specific broadleaved forest with a wide variation in DBH. Our results also revealed that selecting the sample trees in descending order of DBH effectively reduced potential errors in ECestimates for a specific sample size, as larger trees contribute more to EC. This information should be useful for future studies investigating the transpiration of multi-specific broadleaved forests, reducing errors during the scaling-up procedure.

AbstractStand transpiration (E) estimated using the sap flux methods is affected by the azimuthal, radial, and tree-to-tree variations of sap flux. Although several studies have examined the relative importance of the three variations in estimating E, the seasonality of the three variations remains unknown. In the current study, we attempted to clarify whether the relative importance of these three variations could show seasonal changes. Using sap flux data measured in a subtropical cloud forest from August 2010 to July 2011, we calculated the differences resulting from omitting the three variations in estimating E. The effects of the three variations in estimating Eshowed seasonality. The azimuthal and tree-to-tree variations were more pronounced during winter, whereas the radial variation was more pronounced during summer. However, the effect of tree-to-tree variation was consistently much larger than the other two variations throughout the study period. The tree-to-tree variation is more important in estimating Emonthly, seasonally and annually than both the azimuthal and radial variations, although all three variations have shown seasonality. In addition, the sensor allocation for summer would be acceptable for the practical estimation of Eif aiming at the long time scale.

AbstractPrevious studies have demonstrated a clear relationship between diameter at breast height (DBH) and tree transpiration (QT) in multi-specific broadleaved forests. However, these studies were conducted with a limited range of tree sizes and species, and thus many multi-specific broadleaved forests fall outside these conditions. Therefore, this study examined the relationship between DBH and QTin a warm-temperate multi-specific broadleaved forest (n= 12 species) with a wide range of tree sizes (5.0–70.0 cm DBH) using the Granier-type heat dissipation method. The results showed that, although sap flow density varied between individual trees and species, there was a significant relationship between log QTand log DBH (r2= 0.66, P< 0.001) because of the strong dependence of sapwood area on DBH. This study confirmed the applicability of the relationship for the stand transpiration (EC) estimates even in a multi-specific broadleaved forest with a wide variation in DBH. Our results also revealed that selecting the sample trees in descending order of DBH effectively reduced potential errors in ECestimates for a specific sample size, as larger trees contribute more to EC. This information should be useful for future studies investigating the transpiration of multi-specific broadleaved forests, reducing errors during the scaling-up procedure.

Stand transpiration (E) estimated using the sap flux methods is affected by the azimuthal, radial, and tree-to-tree variations of sap flux. Although several studies have examined the relative importance of the three variations in estimating E, the seasonality of the three variations remains unknown. In the current study, we attempted to clarify whether the relative importance of these three variations could show seasonal changes. Using sap flux data measured in a subtropical cloud forest from August 2010 to July 2011, we calculated the differences resulting from omitting the three variations in estimating E. The effects of the three variations in estimating Eshowed seasonality. The azimuthal and tree-to-tree variations were more pronounced during winter, whereas the radial variation was more pronounced during summer. However, the effect of tree-to-tree variation was consistently much larger than the other two variations throughout the study period. The tree-to-tree variation is more important in estimating Emonthly, seasonally and annually than both the azimuthal and radial variations, although all three variations have shown seasonality. In addition, the sensor allocation for summer would be acceptable for the practical estimation of Eif aiming at the long time scale.

Canopy transpiration (EC) varies among forest stands with different structures. To evaluate different ECfor Japanese cypress we observed ECfor 5 months, by use of the sap flow technique, in two adjacent stands of different ages with contrasting structures. Mean diameter at breast height was 13.5 and 44.6 cm and stem density was 2100 and 350 trees ha−1for the Sakuta and Hiwada plots, respectively. Mean ECmeasured was 1.3 mm day−1for the Sakuta plot and 0.7 mm day−1for the Hiwada plot (or 53 % of that for the Sakuta plot). This difference between ECwas caused by differences between both mean stand sap flux density (JS) and stand sapwood area (AS__stand): JSfor the Hiwada plot was 69 % of that for the Sakuta plot and AS_standfor the Hiwada plot was 75 % of that for the Sakuta plot. The difference between JSwas primarily caused by different reference JSfor given meteorological conditions, not by the different response of JSto meteorological conditions. Previous studies of coniferous plantation forests in Japan reported that differences between ECamong stands with different structures were mainly caused by different AS_stand. This study revealed this is not always true, and that differences between JSshould also be considered when predicting differences between ECamong stands with different structures.

AbstractCanopy transpiration (EC) varies among forest stands with different structures. To evaluate different ECfor Japanese cypress we observed ECfor 5 months, by use of the sap flow technique, in two adjacent stands of different ages with contrasting structures. Mean diameter at breast height was 13.5 and 44.6 cm and stem density was 2100 and 350 trees ha−1for the Sakuta and Hiwada plots, respectively. Mean ECmeasured was 1.3 mm day−1for the Sakuta plot and 0.7 mm day−1for the Hiwada plot (or 53 % of that for the Sakuta plot). This difference between ECwas caused by differences between both mean stand sap flux density (JS) and stand sapwood area (AS__stand): JSfor the Hiwada plot was 69 % of that for the Sakuta plot and AS_standfor the Hiwada plot was 75 % of that for the Sakuta plot. The difference between JSwas primarily caused by different reference JSfor given meteorological conditions, not by the different response of JSto meteorological conditions. Previous studies of coniferous plantation forests in Japan reported that differences between ECamong stands with different structures were mainly caused by different AS_stand. This study revealed this is not always true, and that differences between JSshould also be considered when predicting differences between ECamong stands with different structures.

Japanese cedar (Cryptomeria japonica) and Japanese cypress (Chamaecyparis obtusa) plantations account for approximately 30% of the total forested area in Japan. Both are arbuscular mycorrhizal trees that leach more NO3−in response to nitrogen (N) deposition than do forests of ectomycorrhizal trees. However, little information is available about the size of N exports from these plantations. The aim of this study was to evaluate nonpoint source N exports from a N‐saturated plantation. We collected stream water samples in base‐flow (25 samples) and storm‐flow conditions (20 events) in a watershed (2.98 ha) where Japanese cypress and Japanese cedar were planted in 1969 (41 yr old). The annual NO3−export was calculated from load–discharge relationships. Atmospheric N deposition was also determined. The stream water contained high NO3−concentrations (160 and 165 μmol L−1during base flow and storm flow, respectively), indicating N saturation in the watershed. High bulk atmospheric N deposition (16.5 kg N ha−1yr−1) could explain the N saturation. There were only small variations in NO3−concentrations in stream water in response to discharge volume, because of the N saturation of the forest ecosystem. Consequently, there were only small errors in estimating annual NO3−exports from the studied watershed. The annual NO3−export was high (36.1 kg N ha−1yr−1), comparable to values reported for agricultural and urbanized areas. These results suggest that N‐saturated forest plantations can become important nonpoint N sources. Our results also suggest that N exports from forest plantations across Japan should be quantified to evaluate nonpoint source N accurately.

Modern ecohydrologic science will be critical for providing the best information to policy makers and society to address water resource challenges in the 21st century. Implicitly, ecohydrology involves understanding both the functional interactions among vegetation, soils, and hydrologic processes at multiple scales and the linkages among upland, riparian, and aquatic components. In this paper, we review historical and contemporary ecohydrologic science, focusing on watershed structure and function and the threats to watershed structure and function. Climate change, land use change, and invasive species are among the most critical contemporary issues that affect water quantity and quality, and a mechanistic understanding of watershed ecosystem structure and function is required to understand their impacts on water quantity and quality. Economic and social values of ecosystem services such as water supply from forested watersheds must be quantified in future research, as land use decisions that impact ecohydrologic function are driven by the interplay among economic, social, political, and biological constraints. Future forest ecohydrological research should focus on: (1) understanding watershed responses to climate change and variability, (2) understanding watershed responses to losses of native species or additions of non‐native species, (3) developing integrated models that capitalize on long‐term data, (4) linking ecohydrologic processes across scales, and (5) managing forested watersheds to adapt to climate change. We stress that this new ecohydrology research must also be integrated with socio‐economic disciplines. Published in 2011. This article is a US Government work and is in the public domain in the USA.

AbstractSap flow techniques are practical tools for estimating tree transpiration. Though many previous studies using sap flow techniques did not consider azimuthal variations of sap flux density (Fd) on xylem trunk to estimate tree transpiration, a few studies reported that ignoring the azimuthal variations in Fdcould cause large errors in tree transpiration estimates for some tree species. Therefore, examining azimuthal variations in Fdfor major plantation tree species is critical for estimating tree transpiration. Using the thermal dissipation method, we examined azimuthal variations in Fdin six trees of Japanese cypress Chamaecyparis obtusa(Sieb. et Zucc.) Endl., which is one of the most common plantation tree species in Japan. We recorded considerable variations among Fdat four different azimuthal directions. The Fdvalue for one aspect was more than 100% larger than those for the other aspects. We calculated differences between tree transpiration estimates based on Fdfor one to three azimuthal directions and those based on Fdfor four aspects. The differences relative to tree transpiration estimates based on Fdfor four aspects were typically 30, 20, and 10% in accordance with the Fdfor one, two, and three measurement aspects, respectively. This finding indicates that ignoring azimuthal variations could cause large errors in tree transpiration estimates for Japanese cypress.

Abstract: Sap flow techniques are practical tools for estimating tree transpiration. Though many previous studies using sap flow techniques did not consider azimuthal variations of sap flux density (F d) on xylem trunk to estimate tree transpiration, a few studies reported that ignoring the azimuthal variations in F d could cause large errors in tree transpiration estimates for some tree species. Therefore, examining azimuthal variations in F d for major plantation tree species is critical for estimating tree transpiration. Using the thermal dissipation method, we examined azimuthal variations in F d in six trees of Japanese cypress Chamaecyparis obtusa (Sieb. et Zucc.) Endl., which is one of the most common plantation tree species in Japan. We recorded considerable variations among F d at four different azimuthal directions. The F d value for one aspect was more than 100% larger than those for the other aspects. We calculated differences between tree transpiration estimates based on F d for one to three azimuthal directions and those based on F d for four aspects. The differences relative to tree transpiration estimates based on F d for four aspects were typically 30, 20, and 10% in accordance with the F d for one, two, and three measurement aspects, respectively. This finding indicates that ignoring azimuthal variations could cause large errors in tree transpiration estimates for Japanese cypress.

Abstract: Black locust (Robinia pseudoacacia) is a major reforestation species in the semiarid region in the Loess Plateau of China. There has been increasing concern about the sustainability of the plantations because of their possible high water-use. This study was, accordingly, undertaken to quantify the stand-scale water use of a middle-aged black locust plantation in the region. The thermal dissipation probe method was applied to 27 trees to measure sap flux densities in an experimental plot during the growing season of 2008. The monoculture stand has a basal area of 23.3 m2 ha−1 and a maximum plant area index (PAI) of 2.89. Sapwood areas were estimated by use of a regressive relationship with the diameter at breast height (DBH) for scaling up of stand transpiration. The results showed that DBH could be a good predictor of sapwood area of individual trees. The diurnal cycles of average sap flux densities differed among DBH classes. Daily transpiration can be predicted from mean daily daytime vapor pressure deficit (VPDm) using a fitted exponential saturation model. Model variables were different among seasons, probably owing to different soil water conditions and leaf phenology. By using the derived model for each month, stand canopy transpiration over the growing season was estimated to be 73.8 mm, with an average daily value of 0.41 mm day−1 and a maximum of 0.89 mm day−1. The relatively small estimates of stand transpiration might be attributed to low PAI and sap wood area of the middle-aged stand.

AbstractBlack locust (Robinia pseudoacacia) is a major reforestation species in the semiarid region in the Loess Plateau of China. There has been increasing concern about the sustainability of the plantations because of their possible high water-use. This study was, accordingly, undertaken to quantify the stand-scale water use of a middle-aged black locust plantation in the region. The thermal dissipation probe method was applied to 27 trees to measure sap flux densities in an experimental plot during the growing season of 2008. The monoculture stand has a basal area of 23.3 m2ha−1and a maximum plant area index (PAI) of 2.89. Sapwood areas were estimated by use of a regressive relationship with the diameter at breast height (DBH) for scaling up of stand transpiration. The results showed that DBH could be a good predictor of sapwood area of individual trees. The diurnal cycles of average sap flux densities differed among DBH classes. Daily transpiration can be predicted from mean daily daytime vapor pressure deficit (VPDm) using a fitted exponential saturation model. Model variables were different among seasons, probably owing to different soil water conditions and leaf phenology. By using the derived model for each month, stand canopy transpiration over the growing season was estimated to be 73.8 mm, with an average daily value of 0.41 mm day−1and a maximum of 0.89 mm day−1. The relatively small estimates of stand transpiration might be attributed to low PAI and sap wood area of the middle-aged stand.

Converting coniferous plantations to broad‐leaved forests is believed to result in increased runoff in Japan. This study examined low‐flow changes in runoff in the Tatsunokuchi–Minami catchment, located in a region with summer precipitation, caused by the conversion of a coniferous plantation to a broad‐leaved forest. We compared the flow duration curves for the coniferous plantation period with those for the broad‐leaved forest period. We observed no evidence of increased low flows according to the change in forest type. Thus, the commonly held belief was not supported by the results from a summer precipitation region with respect to low‐flow rates. Copyright © 2009 John Wiley & Sons, Ltd.

Abstract: In recent years, moso bamboo (Phyllostachys pubescens) forest areas in Japan have rapidly expanded, and bamboo is now invading nearby natural or plantation forests. To date, only one study has examined the rainfall interception of a moso bamboo forest. In that study, it was reported that the interception ratio (interception/rainfall) of the bamboo forest did not exceed the interception ratios of other natural and plantation forests (n = 4) in Japan. To expand the current state of knowledge about rainfall interception of bamboo forests, we measured throughfall and stemflow at another bamboo forest site. Annual rainfall (Rf), throughfall (Tf), and stemflow (Sf) during the measurement period were 2,105, 1,556, and 322 mm, respectively. Annual rainfall interception at the plot (I) was 228 mm. Tf/Rf, Sf/Rf, and I/Rf were 73.9, 15.3, 10.8%, respectively. I/Rf was less than 20% throughout the year except in October, the month with lowest rainfall. We also summarized rainfall interception data from 19 other natural and plantation forests. The I/Rf value of our site did not exceed the I/Rf values of these natural and plantation forests (n = 19). Our data will be useful for assessing changes in water resources that result from replacement of natural or plantation forests by bamboo forests.

AbstractIn recent years, moso bamboo (Phyllostachys pubescens) forest areas in Japan have rapidly expanded, and bamboo is now invading nearby natural or plantation forests. To date, only one study has examined the rainfall interception of a moso bamboo forest. In that study, it was reported that the interception ratio (interception/rainfall) of the bamboo forest did not exceed the interception ratios of other natural and plantation forests (n= 4) in Japan. To expand the current state of knowledge about rainfall interception of bamboo forests, we measured throughfall and stemflow at another bamboo forest site. Annual rainfall (Rf), throughfall (Tf), and stemflow (Sf) during the measurement period were 2,105, 1,556, and 322 mm, respectively. Annual rainfall interception at the plot (I) was 228 mm. Tf/Rf, Sf/Rf, and I/Rfwere 73.9, 15.3, 10.8%, respectively. I/Rfwas less than 20% throughout the year except in October, the month with lowest rainfall. We also summarized rainfall interception data from 19 other natural and plantation forests. The I/Rfvalue of our site did not exceed the I/Rfvalues of these natural and plantation forests (n= 19). Our data will be useful for assessing changes in water resources that result from replacement of natural or plantation forests by bamboo forests.

Abstract: We investigated the sapling leaf display in the shade among trees of various leaf lifespans co-occurring under the canopy of a warm-temperate conifer plantation. We measured leaf-area ratio (aLAR) and morphological traits of saplings of evergreen broadleaved tree species and a deciduous tree species. Although we found large interspecific and intraspecific differences in aLAR even among saplings of similar size in the homogeneous light environment, we did not find a consistent trend in aLAR with leaf lifespan among the species. While deciduous trees annually produced a large leaf area, some evergreen broadleaved trees retained their leaves across years and had aLAR values as high as those of deciduous trees. Among leaf-level, shoot-level, and individual-level morphological traits, aLAR was positively correlated with current-year shoots mass per aboveground biomass in deciduous trees, and with the area of old leaves per aboveground mass in evergreen broadleaved trees. Thus, tree-to-tree variation in the degrees of annual shoot production and the accumulation of old leaves were responsible for the interspecific and intraspecific variations in aLAR.

AbstractWe investigated the sapling leaf display in the shade among trees of various leaf lifespans co-occurring under the canopy of a warm-temperate conifer plantation. We measured leaf-area ratio (aLAR) and morphological traits of saplings of evergreen broadleaved tree species and a deciduous tree species. Although we found large interspecific and intraspecific differences in aLAR even among saplings of similar size in the homogeneous light environment, we did not find a consistent trend in aLAR with leaf lifespan among the species. While deciduous trees annually produced a large leaf area, some evergreen broadleaved trees retained their leaves across years and had aLAR values as high as those of deciduous trees. Among leaf-level, shoot-level, and individual-level morphological traits, aLAR was positively correlated with current-year shoots mass per aboveground biomass in deciduous trees, and with the area of old leaves per aboveground mass in evergreen broadleaved trees. Thus, tree-to-tree variation in the degrees of annual shoot production and the accumulation of old leaves were responsible for the interspecific and intraspecific variations in aLAR.

Abstract??The behavior of dissolved and particulate components of stream nitrogen (N) and phosphorus (P) were measured for 2 years in a small mountainous watershed covered primarily with a plantation forest of Japanese cypress (Chamaecyparis obtusa). The load of dissolved N and P bore a consistent relationship to discharge while the load of particulate N and P varied by up to two orders of magnitude at a given discharge level. Most N was exported in a dissolved form (DN) while most P was exported in a particulate form (PP), which bears similarity to loads from agricultural watersheds. Owing to the different behaviors of DN and PP, changes in the total nitrogen (TN) load were primarily attributed to variations in discharge unlike changes in the total phosphorus (TP) load. High flow conditions, resulting from heavy rainfall, displayed PP release significantly larger than expected. The TP load in high flow conditions was severely underestimated using a regression equation expressed as a function of discharge, which was based on the weekly sampling data biased toward low flow conditions. In addition, the TN load during peak discharges in heavy rain events was underestimated by the regression equation because of unpredictable increases in the particulate component. Our study shows that the particulate component ratio determines whether discharge can explain changes in load regardless of chemical species. The results suggest that plantation forests in rainy regions can be a diffuse source of particulate nutrients depending on soil surface conditions.

AbstractThe behavior of dissolved and particulate components of stream nitrogen (N) and phosphorus (P) were measured for 2 years in a small mountainous watershed covered primarily with a plantation forest of Japanese cypress (Chamaecyparis obtusa). The load of dissolved N and P bore a consistent relationship to discharge while the load of particulate N and P varied by up to two orders of magnitude at a given discharge level. Most N was exported in a dissolved form (DN) while most P was exported in a particulate form (PP), which bears similarity to loads from agricultural watersheds. Owing to the different behaviors of DN and PP, changes in the total nitrogen (TN) load were primarily attributed to variations in discharge unlike changes in the total phosphorus (TP) load. High flow conditions, resulting from heavy rainfall, displayed PP release significantly larger than expected. The TP load in high flow conditions was severely underestimated using a regression equation expressed as a function of discharge, which was based on the weekly sampling data biased toward low flow conditions. In addition, the TN load during peak discharges in heavy rain events was underestimated by the regression equation because of unpredictable increases in the particulate component. Our study shows that the particulate component ratio determines whether discharge can explain changes in load regardless of chemical species. The results suggest that plantation forests in rainy regions can be a diffuse source of particulate nutrients depending on soil surface conditions.

In Sri Lanka, irrigation reservoirs (tanks) are usually connected sequentially and form cascades along the landscape. A study was carried out in the Anuradhapura District in the dry zone of Sri Lanka to understand the role of return flows in such tank cascade systems. The water balance of a tank cascade system was estimated using hydrological data collected over a one-year period. The system was extended about 25 km along a river composed of three small reservoirs having the command area of 31, 55, and 55 ha, respectively. In this system, about 46% of seepage water from tanks entered the paddy fields of the command area. The crop consumed part of the water and the rest returned to the downstream tank through the drainage canals. Percolation loss in the command areas was low (3.6 mm/day) since a considerable portion of the percolation returned to the downstream tank. These results showed that return flows, which are generally disregarded in the water budget, contributed considerably to the water supply of the tank cascade system.

The papers in this issue are a selection of the presentations made at the second International Conference on Forests and Water in a Changing Environment. This special issue ‘Forest Ecohydrological Processes in a Changing Environment’ covers the topics regarding the effects of forest, land use and climate changes on ecohydrological processes across forest stand, watershed and regional spatial scales. Copyright © 2011 John Wiley & Sons, Ltd.