Measuring forest floor and canopy interception in a savannah ecosystem

Interception is one of the most underestimated processes of the hydrological cycle. However, it amounts to a substantial part of the terrestrial evaporation and forms a direct feedback of moisture to the atmosphere which is important to sustain continental rainfall. Most investigations on interception focus on canopy interception only, whereas the interception by the surface and forest floor may be of same order of magnitude. Moreover there is a regional bias. Most research has been carried out in Europe and America and little is known about interception in Africa. This paper presents a study on forest floor and canopy interception in a savannah ecosystem. The study deals with both interception storage capacity of different vegetation types and the related moisture fluxes. The interception storage capacity of Msasa leaf litter and of Thatching grass is 1.8 mm and 1.5 mm respectively. This water storage capacity is dependent on storm intensity, with high intensity storms resulting in smaller storage capacity than less intensive storms. Canopy interception for the study period averaged 25% of the total rainfall, which is comparable with other studies. More importantly, the study revealed that combining canopy and forest floor interception yields a total interception flux amounting to 37% of the rainfall, or close to 50% of the total evaporation. This is a significant amount which implies that interception of both canopy and forest floor should be included in hydrological modelling and that interception is relevant for water management.     

Spatial and temporal variability of canopy and forest floor interception in a beech forest

Depending on season, rainfall characteristics and tree species, interception amounts to 15–50% of total precipitation in a forest under temperate climates. Many studies have investigated the importance of interception of different tree species in all kinds of different climates. Often authors merely determine interception storage capacity of that specific species and the considered event, and only sometimes a distinction is made between foliated and non‐foliated trees. However, interception is highly variable in time and space. First, since potential evaporation is higher in summer, but secondly because the storage capacity has a seasonal pattern. Besides weather characteristics, such as wind and rain intensity, snow causes large variations in the maximum storage capacity. In an experimental beech plot in Luxembourg, we found storage capacity of canopy interception to show a clear seasonal pattern varying from 0·1 mm in winter to 1·2 mm in summer. The capacity of the forest floor appears to be rather constant over time at 1·8 mm. Both have a standard deviation as high as ± 100%. However, the process is not sensitive to this variability resulting only in 11% variation of evaporation estimates. Hence, the number of raindays and the potential evaporation are stronger driving factors on interception. Furthermore, the spatial correlation of the throughfall and infiltration has been investigated with semi‐variograms and time stability plots. Within 6–7 m distance, throughfall and infiltration are correlated and the general persistence is rather weak. Copyright © 2010 John Wiley & Sons, Ltd.     

Comparison of interception,forest floor evaporation and transpiration in Pinus radiata and Eucalyptus globulus plantations

Studies of evapotranspiration (ET) processes in forests often only measure one component of total ET, most commonly interception. This study examined all three components of annual ET (interception, evaporation from the forest floor and transpiration) and the correlations between them at 18 plantation forest sites in two species. All plantations had closed canopies, and sparse or no understorey. Single‐sided leaf area index averaged 3.5 (standard deviation ±0.5) in Eucalyptus globulus Labill. and 6.1 (±0.8) in Pinus radiata D.Don. Measurements included annual totals of rainfall in the open and under the canopy, stem flow (four sites only), evaporation from the forest floor and transpiration by the overstorey. Interception (I) averaged 19% (±4.9) of annual rainfall in E. globulus compared with 31% (±11.1) in P. radiata. However, higher annual interception in P. radiata did not result in higher total ET because annual evaporation from the forest floor (E) averaged 29% (±4.9) of rainfall in E. globulus but only 15% (±3.5) in P. radiata. Hence, the relative contribution of annual I plus E to ET did not differ significantly between the two species, averaging 48% (±7.3) of annual rainfall in E. globulus compared with 46% (±11.8) in P. radiata. As reported previously, transpiration did not differ significantly between the two species either, but was strongly related to depth‐to‐groundwater. In closed canopy plantations, mean annual ET did not differ between the two species. We conclude that when grown in plantations under similar soil and climatic conditions, conifer and broad‐leaved tree species can have similar annual ET, once the canopy of the plantation has closed. Lower average annual interception in broad‐leaved trees was offset by higher soil evaporation. These results highlight the importance of measuring all components of ET in studies of vegetation water use. Copyright © 2014 John Wiley & Sons, Ltd.     

Monitoring spatial variability of forest interception

Field instrumentation was designed and installed to quantify the influence of forest interception on the spatial and temporal distribution of water flux onto and into the forest soil at the plot scale. An application is presented which demonstrates that the instrumentation has the required resolution to monitor the spatial variability and dynamics of the flux processes. The observations show that spatial variability of interception may play an important role, not only in small scale soil moisture heterogeneity, but also in the hydrological response of a forested catchment at the hillslope scale. They also highlight the need of gathering more field information on the effects of vegetation on the spatial variability of soil surface water input.     

Problems of throughfall and interception assessment under tropical forest

An experiment set up in a highland area of northern Tanzania is described. The influence of skewness and great variability of throughfall data is illustrated. For heavy storms, the large standard errors make interception estimates unreliable. The techniques of analysis adopted have a considerable influence on throughfall and interception estimates.     

Partitioning of rainfall in a eucalypt forest and pine plantation in southeastern australia: IV the relationship of interception and canopy storage capacity,the interception of these forests,and the effect on interception of thinning the pine plantation

A study of partitioning of rainfall into throughfall, stemflow, and interception was conducted in a dry sclerophyll eucalypt forest and an adjacent pine plantation over a period of seven years, on a rainfall event basis. The following three issues are discussed: (1) the relationship between canopy storage capacity and interception of continuous events, (2) interception, throughfall, and stemflow, and (3) the effect on interception of thinning the pine plantation.

• 1 The canopy storage capacity/interception interaction for the eucalypt forest was assessed by comparing a gravimetric estimate of canopy storage capacity with interception. The maximum possible value for canopy storage capacity was found to be a small proportion of interception for events of all sizes. This suggests that evaporation of intercepted water during the continuous events was responsible for most of the interception. This ‘within event’ evaporation appears to be responsible also for the net rainfall/gross rainfall estimate of canopy storage capacity being four times the gravimetric value. For the pines the regression estimate was more closely related to interception.
• 2 Interception, throughfall, and stemflow of these forests were measured for four years. Data are presented for each year with overall average interception being 11-4 per cent of precipitation for the eucalypt forest and 18-3 per cent for the pine plantation. Topography and rainfall event type are considered in the comparison.

Species composition and tree type are considered when comparing these results with published studies from similar forest types in southeastern Australia. The periodic (annual) variations of interception in this and the other studies makes comparison difficult.

• 3 The effect of thinning on the throughfall, stemflow, and interception in a Pinus radiata plantation is examined. Throughfall increased, interception decreased but not in proportion to the removed biomass; stemflow decreased on an area basis, but increased on a per tree basis. A positive relationshiip is established between interception and stemflow on the thinned plantation but not in the unthinned. Reasons for this are suggested. The results are compared to those reported from similar experiments in other forests.
• 4 The periodic variations in interception and errors inherent in its estimation suggest that caution should be exercised when using average interception figures in water balance studies.

     

Factors affecting rainfall interception determined by a forest simulator and numerical model

To elucidate the factors involved in interception loss, we conducted experiments in which we measured environmental variables such as rainfall intensity, forest structure, and weather conditions. An artificial forest consisting of 24 vinyl trees was used to examine the influences of forest structure and rainfall conditions on interception loss. The interception rate was higher at higher plant area index (PAI) values and wind speeds and lower with greater rainfall intensity. We confirmed the factors affecting interception loss by using an interception model based on the tank model. The artificial forest simulations provide new evidence that interception loss is influenced by the PAI, rainfall intensity, saturation deficit, and wind speed. The effect of the saturation deficit on the interception rate was unclear from the experimental results, but the single‐tank model revealed that wind speed strongly influences the effects of the saturation deficit on interception loss. Thus, whereas interception loss was not significantly affected by the saturation deficit at low wind speeds, it increased significantly with the saturation deficit under windy conditions. The model simulation also showed the sensitivity of each factor with regard to the interception rate. The sensitivity of rainfall intensity decreased as the PAI increased, and the sensitivity of the saturation deficit increased as the wind speed increased. The experiments and model calculations clarified the main elements affecting interception loss and their sensitivities. Compared with previous studies on interception loss, this study revealed a positive relationship between the PAI and interception loss, a negative exponential relationship with rainfall intensity, and the effects of the saturation deficit on interception loss. Copyright © 2008 John Wiley & Sons, Ltd.     

Snowfall interception in a deciduous Nothofagus forest and implications for spatial snowpack distribution

Native Nothofagus forests in the midlatitude region of the Andes Cordillera are notorious biodiversity hot spots, uniquely situated in the Southern Hemisphere such that they develop in snow‐dominated reaches of this mountain range. Spanning a smaller surface area than similar ecosystems, where forests and snow coexist in the Northern Hemisphere, the interaction between vegetation and snow processes in this ecotone has received lesser attention. We present the first systematic study of snow–vegetation interactions in the Nothofagus forests of the Southern Andes, focusing on how the interplay between interception and climate determines patterns of snow water equivalent (SWE) variability. The Valle Hermoso experimental catchment, located in the Nevados de Chillán vicinity, was fitted with eight snow depth sensors that provided continuous measurements at varying elevations, aspect, and forest cover. Also, manual measurements of snow properties were obtained during snow surveys conducted during end of winter and spring seasons for 3 years, between 2015 and 2017. Each year was characterized by distinct climatological conditions, with 2016 representing one of the driest winters on record in this region. Distance to canopy, leaf area index, and total gap area were measured at each observational site. A regression model was built on the basis of statistical analysis of local parameters to model snow interception in this kind of forest. We find that interception implied a 23.2% reduction in snow accumulation in forested sites compared with clearings. The interception in these deciduous trees represents, on average, 23.6% of total annual snowfall, reaching a maximum measured interception value of 13.8‐mm SWE for all snowfall events analysed in this research.     

Rainfall interception in a lower montane forest in Ecuador: effects of canopy properties

Rainfall interception in forests is influenced by properties of the canopy that tend to vary over small distances. Our objectives were: (i) to determine the variables needed to model the interception loss of the canopy of a lower montane forest in south Ecuador, i.e. the storage capacity of the leaves S and of the trunks and branches S_t, and the fractions of direct throughfall p and stemflow p_t; (ii) to assess the influence of canopy density and epiphyte coverage of trees on the interception of rainfall and subsequent evaporation losses. The study site was located on the eastern slope of the eastern cordillera in the south Ecuadorian Andes at 1900–2000 m above sea level. We monitored incident rainfall, throughfall, and stemflow between April 1998 and April 2001. In 2001, the leaf area index (LAI), inferred from light transmission, and epiphyte coverage was determined. The mean annual incident rainfall at three gauging stations ranged between 2319 and 2561 mm. The mean annual interception loss at five study transects in the forest varied between 591 and 1321 mm, i.e. between 25 and 52% of the incident rainfall. Mean S was estimated at 1·91 mm for relatively dry weeks with a regression model and at 2·46 mm for all weeks with the analytical Gash model; the respective estimates of mean S_t were 0·04 mm and 0·09 mm, of mean p were 0·42 and 0·63, and of mean p_t were 0·003 and 0·012. The LAI ranged from 5·19 to 9·32. Epiphytes, mostly bryophytes, covered up to 80% of the trunk and branch surfaces. The fraction of direct throughfall p and the LAI correlated significantly with interception loss (Pearson's correlation coefficient r = −0·77 and 0·35 respectively, n = 40). Bryophyte and lichen coverage tended to decrease S_t and vascular epiphytes tended to increase it, although there was no significant correlation between epiphyte coverage and interception loss. Our results demonstrate that canopy density influences interception loss but only explains part of the total variation in interception loss. Copyright © 2004 John Wiley & Sons, Ltd.     

Effect of leaf distribution pattern on the interception storage capacity of leaf litter under simulated rainfall conditions

Rainwater interception by leaf litter is an important part of forest hydrological processes. The objective of this study was to investigate the interception storage capacity (ISC) of woodland leaf litter for three leaf distribution patterns, one flow path, two flow paths, and three flow paths, manually simulated via one-by-one leaf connection in the top leaf litter layer. A random pattern served as the control. Three different slopes (0°, 5° and 25°, representing flat, gentle and steep slopes, respectively) and two contrasting leaf litters (needle-leaf litter, represented by P. massoniana leaves, and broad-leaf litter, represented by C. camphora leaves) with a biomass of 0.5 kg/m² per unit area were applied, at a rainfall intensity of 50 mm/h. Results suggested that leaf distribution pattern greatly impacts litter drainage and hence affects leaf litter ISC. The delaying capacity of litter drainage initiation and ISC of broad-leaf litter were higher than those of needle-leaf litter under the same slope conditions. The maximum ISC (C_max) and minimum ISC (C_min) of leaf litter at flat and gentle slopes were higher than those at steep slope. C_min of the broad-leaf litter was two times higher than that of needle-leaf litter on average. When raindrops reached the litter layer, some were temporarily intercepted by the top litter layer while others infiltrated leaf litter sublayer along leaf edges, and in the process, some rainwater flowed through litter layer and contributed to lateral litter drainage along the potential flow path formed by leaves. The lateral litter drainage of broad-leaf litter was higher than that of needle-leaf litter, and the partitioning of rainwater into lateral litter drainage increased with increases in slope. The difference in leaf litter C_max among different slopes and leaf shapes decreased with flow path increasing. Therefore, leaf distribution pattern notably impact leaf litter ISC, which is similar to leaf shape and slope impacts. On inclined slopes, avoiding leaf accumulation to form flow path is helpful for improving ISC.     

Rates, timing, and mechanisms of rainfall interception loss in a coastal redwood forest

Rainfall, throughfall, and stemflow were monitored at 5-min intervals for 3 years in a 120-year-old forest dominated by redwood (Sequoia sempervirens) and Douglas-fir (Pseudotsuga menziesii) at the Caspar Creek Experimental Watersheds, located in northwest California, USA. About 2.5% of annual rainfall reaches the ground as stemflow at the site, while 22.4% is stored on foliage and stems and evaporates before reaching the ground. Comparison of the timing of rainfall and throughfall indicates that about 46% of the interception loss occurs through post-storm evaporation from foliage and 54% is either evaporated during the storm or enters long-term storage in bark. Until bark storage capacity is saturated, the proportion of rainfall diverted to bark storage would be relatively constant across the range of rainfall intensities encountered, reflecting primarily the proportional incidence of rainfall on surfaces contributing to bark storage. In any case, loss rates remain high—over 15%—even during the highest-intensity storms monitored. Clearcut logging in the area would increase effective annual rainfall by 20–30% due to reduction of interception loss, and most of the increase would occur during large storms, thus potentially influencing peakflows and hillslope pore-pressures during geomorphically significant events.     

Canopy interception by a spruce forest in the upper reach of Heihe River basin,Northwestern China

The aim of this study is to understand the canopy interception of Qinghai spruce forest under conditions of different precipitation characteristics and canopy structures in the upper reach of Heihe River basin, northwestern China. On the basis of a continuous record covering our investigating period by an automatic throughfall‐collecting system, we analysed the relationships between the canopy interception and the precipitation characteristics. Our results support the well‐established exponential decay relationship between the gross precipitation and the interception percentage after the canopy is saturated. But our results sufficiently illustrate a notable point that the variations in the interception percentage are almost independent from the variations in the gross precipitation before the canopy is saturated. Our examination into the relationship between the interception and the 10‐min average intensity of precipitation demonstrates a divergent relationship, and the divergent relationship is bracketed by an upper ‘dry line’ indicating that 100% of gross precipitation was intercepted before saturation and by a lower ‘wet line’ suggesting that the actual canopy storage capacity reached the maximum and evaporation was the only component of the interception. To search for the relationship between canopy structures and interception, we grouped the canopy covers over the 90 throughfall‐collecting tanks into ten categories ranging from 0 (no cover) to 0.9 (nearly completely covered), and the corresponding canopy interception was calculated by subtracting the averaged throughfall of each canopy‐cover category from the gross precipitation. The results show that the interception percentage increases faster with increasing canopy cover under intermediate rainfall conditions than that under heavy rainfall conditions. Unexpectedly, under light rainfall conditions the increasing rate of interception percentage with increasing canopy cover and also with increasing plant area index is not faster than that under the intermediate rainfall conditions simply because the tank‐measured percentage of interception was extremely high at near‐zero canopy cover conditions. Copyright © 2013 John Wiley & Sons, Ltd.     

Rainfall interception in three contrasting lowland rain forest types in Central Kalimantan, Indonesia

Rainfall interception was quantified and modeled for a Lowland Evergreen Rain Forest (LERF) and two Heath Forests (HF) of contrasting stature during a one-year period at a remote site in Central Kalimantan, Indonesia. Throughfall was measured using a roving gauge approach using 18–20 gauges per forest type. Throughfall was 82.8% of incident rainfall in the LERF vs. 89.1% in tall HF and 76.7% in stunted HF. Corresponding stemflow fractions were 0.8%, 1.3% and 2.0% of rainfall, respectively. Interception losses derived for the LERF (16.4%) and the tall HF (9.6%) were close to findings obtained for similar forest types elsewhere but the high interception loss for the stunted HF (21.3%) was unexpected. On the basis of canopy structural characteristics and wet canopy evaporation modeling it is concluded that throughfall in the stunted HF was underestimated and that the specific nature of the HF required a more intensive sampling arrangement. Throughfall sampling schemes in tropical forests, associated errors and hydrological implications are discussed.     

Measurement of evaporation from the forest floor in a deciduous forest throughout the year using microlysimeter and closed‐chamber systems

Evaporation from the forest floor (E_FF) in a deciduous broadleaf forest was measured using microlysimeter and closed‐chamber systems. The microlysimeter was used at six points in the experimental basin, and measurements gave different E_FF values at different points. This could be attributed to the local photoenvironment of each sampling point, rather than to litter conditions, if the spatial variation in air temperature (Ta) or vapour pressure deficit (VPD) at the forest floor was small within this basin. A detachable microlysimeter measured condensation in the litter layer during the night, indicating that the litter layer, as well as the mulch layer, played a role in preventing evaporation from the soil layer. The closed‐chamber system made it possible to continuously measure long‐term E_FF. E_FF was closely related to VPD; even during the night, when solar radiation was zero, E_FF amounted to 14·0% of the daily E_FF. The daily E_FF was 0·20 ± 0·13 mm day^?1 during the study period, with two seasonal peaks: in late spring (0·31 mm day^?1 in April) and early fall (0·22 mm day^?1 in September). The former peak has been reported from two deciduous forests in Japan and is strongly related to the solar radiation reaching the forest floor when the trees are dormant. Copyright © 2008 John Wiley & Sons, Ltd.     

Estimation of SCS Curve Number variation following forest fires

The Soil Conservation Service Curve Number (CN) method is routinely used to estimate the effects of forest fires on hydrological response. However, despite recent efforts, CN values are still not well known for burned conditions. A major forest fire in Attica, Greece, which affected the Lykorrema stream experimental watershed, provided an opportunity for the estimation of post-fire CN variation using detailed pre-fire and post-fire rainfall–runoff datasets. The CN values for both periods were estimated and compared using a wide range of available methods. Methods considering the spatial variability of soil-cover complexes were also used to investigate the effect of spatial heterogeneity. The post-fire watershed response changed from complacent to standard. Direct runoff depths and peak flows increased by a factor of more than 7.7 and 11.8, respectively. On average, the estimated post-fire CN values for the studied soil-cover complexes increased by about 25 units. This study may assist the improvement of existing post-fire hydrological assessment tools.     

A forest floor model for heat and moisture including a litter layer

Forest soils are often covered with a litter that influences the rate of mass and energy transfer between the soil and the air above, thereby modifying the temperature and moisture fields in the soil. The presence of a litter should therefore be accounted for in forest SVAT models, especially when long-term simulations are to be performed. A heat and moisture litter model has been developed by adding two dynamical equations to a force-restore type soil model. The experimental data used for the model validation was collected in a pine forest canopy in the South-West of France, that was part of the Euroflux network. The model is tested and validated over a two-year period. It is shown to provide a fairly good simulation of soil and litter moisture, soil and litter temperature and turbulent fluxes measured above the forest floor. It is also shown that simulations without the litter layer are unable to reproduce all these variables simultaneously. We then perform a sensitivity analysis to the parameters whose values are either uncertain or likely to be variable in time and space, such as the litter thickness, the rainfall fraction intercepted by the litter or the maximum value of the surface resistance. A threshold value of the litter moisture used in the surface resistance parameterisation turns out to be the most critical parameter. Further work is needed to investigate the possible relationships between the various parameters describing the litter, but the present litter model can already be used in combination with other forest SVAT models.     

A hillslope forest floor (duff) water budget and the transition to local control

The partially decomposed organic layer (duff: F and H layers) of the forest floor is an important boundary between the soil and atmospheric processes. Here we use both empirical data and a three‐dimensional coupled heat and water budget model to explain the duff hydrological hillslope shift between very brief wet periods when lateral flow in the duff and infiltration into the mineral soil occur and dry periods when evaporative flow dominates and both lateral and mineral soil flow are not important. The duff moisture transitions from wet to dry periods were the result of low lateral flow which moves liquid and water vapour only centimetres to metres, very rapidly and mostly in the H layer immediately after precipitation. During wet periods, the net lateral fluxes were negative on divergent areas and positive on convergent areas of the hillslope, leading to a net moisture loss in divergent areas and a net gain in convergent areas. The response to lateral flow in the H layer was more rapid than in the F layer. The transition from the lateral downwards flow to mineral soil to evaporative control was within approximately 48 h of precipitation. Canopy species and aspect were important with lodgepole pine, southwest aspect and 4‐cm deep duff controlled by evaporative processes while Engelmann spruce, northeast aspect and 30‐cm duff were more controlled by hillslope redistribution processes. Copyright © 2010 John Wiley & Sons, Ltd.