Stemflow Volume Research Articles

Interspecific variation of bark water storage capacity of three deciduous tree species in relation to stemflow yield and solute flux to forest soils

The purpose of the study was to: (1) test if normative bark water storage capacities differed significantly among three co-occurring deciduous tree species; and (2) examine the extent to which stemflow production and resulting solute inputs in temperate deciduous forests are affected by bark water storage capacity. Normative bark water storage capacities were determined for: Betula lenta L. (sweet birch), Carya glabra Mill. (pignut hickory), and Quercus rubra L. (northern red oak). Using the computed normative bark water storage capacities, previously published allometric equations relating tree diameter to aboveground woody surface area, and stemflow yield and chemistry data from ten precipitation events, the influence of bark water storage capacity on stemflow production and solute inputs was examined. Results demonstrated normative bark water storage capacities differed significantly among the tree species examined. Quercus rubra was found to hold the most water and B. lenta the least. For a 30 cm diameter tree, normative bark water storage capacities ranged from approximately 100 l for B. lenta to 250 l for Q. rubra. Despite a higher normative bark water storage capacity, Q. rubra was computed to have the largest stemflow solute inputs. Differences in stemflow quantities as well as solute inputs were attributable to interspecific variation in bark morphology and branching architecture, characteristics that affect amount of storage and detention times. Bark water storage capacity is linked with the geoecology of temperate deciduous forests because stemflow volume and solute inputs are partly determined by bark water storage capacity.

The stemflow of trees in a Bornean lowland tropical forest

AbstractStemflow volume generation in lowland tropical forests was measured over a 1‐year period in the Malaysian state of Sarawak. The stemflow volume generated by 66 free‐standing trees with a diameter at breast height (DBH) over 1 cm and a tree height over 1 m were measured daily in a representative 10 m × 10 m plot of the forest. Throughfall in the plot was also measured using 20 gauges in a fixed position. Of the 2292 mm of total rainfall observed during the year‐long period, stemflow accounted for 3·5%, throughfall for 82% and there was an interception loss of 14·5%. Understory trees (DBH < 10 cm) played an important role in stemflow generation, producing 77% of the overall stemflow volume and 90% during storms with less than 20 mm of rainfall. Also, owing to their efficiency at funneling rainfall or throughfall water received by their crowns, some understory trees noticeably reduced the catches of the throughfall gauges situated under the reach of their crown areas. During storms producing greater than 20 mm of rainfall, 80% of the total stemflow occurred; trees with a large DBH or height and for which the ratio between crown's diameter and depth is less than 1, tended to generate more stemflow volume in these storms. Mean areal stemflow as a fraction of rainfall in this lowland tropical forest was 3·4%, but may range from 1–10% depending upon the proportion of trees that are high or poor stemflow yielders. Trees with DBH greater than 10 cm were likely to contribute less than 1% of the 3·4% mean areal stemflow in the forest. Copyright © 2004 John Wiley & Sons, Ltd.

Differential winter stemflow generation under contrasting storm conditions in a southern New England broad‐leaved deciduous forest

AbstractDespite the importance of stemflow as a hydroecological process, differential winter stemflow generation among and within precipitation types (e.g. snow, rain‐to‐snow) is poorly understood. The purpose of the present study is to understand winter stemflow generation better under contrasting meteorological conditions in a leafless deciduous forest. It is hypothesized that stemflow volume and production, expressed as a funnelling ratio, will differ significantly among and within precipitation event types. Acceptance of this hypothesis would mean that intra‐event meteorological conditions have a significant and discernible effect on stemflow production in deciduous forests, differentially affecting soil moisture, groundwater recharge, soil solution chemistry, and nutrient cycling. Stemflow volumes generated from seven canopy trees were monitored in a leafless deciduous forest of southern New England on a precipitation event basis over two successive winter seasons. Stemflow volume was found to differ significantly among different and same‐type precipitation events. A rain event that occurred on 8 December 1998 and snow‐to‐rain event on 18 January 1999 were of similar duration, magnitude, and intensity, yet produced stemflow volumes that differed by a factor of 4. For two snow‐to‐rain events, stemflow volumes differed by a factor of 11. Stemflow production even varies widely within a discrete precipitation event as a function of meteorological conditions. Significant differential stemflow yield under contrasting storm conditions obviates generalizations concerning stemflow production as a function of precipitation event type and necessitates a deeper, process‐level understanding of winter stemflow generation before modelling exercises are undertaken. Copyright © 2004 John Wiley & Sons, Ltd.

Snowmelt induced stemflow in northern hardwood forests: a theoretical explanation on the causation of a neglected hydrological process

Snowmelt induced stemflow is an under-studied and poorly understood hydrological process. The aim of the present manuscript is to set forth several testable hypotheses and conceptual models concerning the production of snowmelt induced stemflow. The first testable hypothesis is that tree species with lower canopy albedos of snow-covered and snow-free surfaces (portions of the canopy are usually exposed and snow-free while the remainder is covered) will generate larger snowmelt induced stemflow volumes than trees with higher canopy albedos of snow-covered and snow-free surfaces. It is also hypothesized that snow-to-rain events produce larger volumes of snowmelt induced stemflow than snow events of the same magnitude because of differences in precipitation type (snow-to-rain vs. snow) and the additional melt energy from the latent heat of fusion associated with rain infiltrating the intercepted snow mass. For deciduous tree species with similar branching geometries, it is hypothesized that larger longwave radiative flux will generate larger snowmelt induced stemflow volumes. The conceptual models describe the components and the flux directions necessary to generate snowmelt induced stemflow. Through a test of the hypotheses and conceptual models in subsequent work, it is hoped that quantitative models are developed that will be able to estimate snowmelt induced stemflow volumes and yield a new understanding of winter water flux in forested ecosystems.

Stemflow estimation in a redwood forest using model‐based stratified random sampling

AbstractModel‐based stratified sampling is illustrated by a case study of stemflow volume in a redwood forest. The approach is actually a model‐assisted sampling design in which auxiliary information (tree diameter) is utilized in the design of stratum boundaries to optimize the efficiency of a regression or ratio estimator. The auxiliary information is utilized in both the design and estimation phases. Stemflow and its variance were modelled as powers of diameter and a generalized non‐linear least squares model was used to estimate the exponents and to impute values for missing storm events prior to application of the ratio estimator. The advantage of the ratio estimator over standard stratified sampling formulas is greatest for species in which stemflow is strongly dependent on diameter. With measurements on 24 trees in a 1‐hectare stand, annual stemflow was estimated with coefficients of variation of 9 per cent and 10 per cent in 2 years of study. Striking stemflow differences were found between species. Copyright © 2003 John Wiley & Sons, Ltd.

Winter stemflow leaching of nutrient-ions from deciduous canopy trees in relation to meteorological conditions

Leaching of nutrients from aboveground vegetative surfaces of canopy trees represents an important component of the intra-system nutrient cycle in forested ecosystems. The hypothesis tested is that there is no difference in winter stemflow leachate concentration or quantity among different storm types and meteorological conditions. Non-hierarchical cluster analysis demonstrated that there was an association between precipitation event type and chemical enrichment of stemflow drainage from the leafless crowns of deciduous canopy trees monitored over two successive winter seasons. Three clusters were derived on the basis of chemical enrichment. Stemflow was most enriched during snow-to-rain events. Rain and rain-to-snow events also enriched stemflow but to a lesser extent than snow-to-rain events. Snow, sleet, and rain shower events enriched stemflow the least. Stemflow from all precipitation events were chemically enriched compared to the incident bulk precipitation. The extent of chemical enrichment was inversely proportional to stemflow volume generated by canopy trees during discrete precipitation events. Other factors affecting the amount of winter stemflow leaching were precipitation type (influences interception efficiency) and intra-storm temperature regime. Stemflow from precipitation events of long duration and low intensity with air temperature oscillations around the freezing point was the most enriched because of the longer residence time of intercepted precipitation with the leafless crown. Because stemflow leachate quantities were found to differ among precipitation events as a function of meteorological conditions, it is possible that the winter biometeorology of deciduous forests could potentially impact tree vigor, forest stand productivity, and species composition of deciduous forests.

Winter stemflow nutrient inputs into a southern new england broadleaved deciduous forest

Stemflow leaching from the above‐ground vegetative surfaces of broadleaved deciduous canopy trees has been well documented during the growing season. Winter stemflow leaching from the leafless crowns of deciduous trees is less well understood, especially in the context of global climate change. Boreal and northern temperate forests are forecast to have a lower incidence of snow events and an increased frequency of rain and mixed precipitation events. A change in the seasonal precipitation regime may be significant due to linkages among global change, stemflow leaching, and biogeochemical processes. The aim of this paper is to (1) demonstrate the extent of winter stemflow nutrient leaching from deciduous trees; (2) explore how winter stemflow leachate quantity may vary as a function of the type of precipitation event; and (3) quantify the extent to which an increase in the incidence of snow‐to‐rain events would enhance stemflow leaching. Measuring meteorological conditions, stemflow volumes, and stemflow chemistry over two successive winter seasons in New England demonstrated that winter stemflow drainage was significantly enriched compared to the incident bulk precipitation: 162 times greater for K+, 44 times greater for Ca2+, and 29 times greater for Mg2+. Snow‐to‐rain events leached considerably greater quantities of base cations from the deciduous trees than all other types of precipitation events. An increased frequency of snow‐to‐rain events from 13.8% to 25% of winter precipitation events would substantially increase mean stemflow nutrient inputs to the bases of forest trees during winter. Implications for significantly increased winter stemflow leachate inputs, corresponding to an increased incidence of snow‐to‐rain events, include a shift in the biogeographic range of species, delayed leaf emergence, and increased soil respiration.

Generation of stemflow volume and chemistry in a mature Japanese cypress forest

AbstractLong‐term monitoring of rainfall, throughfall and stemflow was conducted in a mature Japanese cypress forest at the University Forest in Chiba, Boso Peninsula, Japan. Two stemflow collectors were installed in a tree at 19·1 m (just under the tree crown) and at 1·3 m above the ground to determine the processes controlling the generation, volume and quality of stemflow at the stem, leaves and branches. The canopy‐only stemflow and stem‐only stemflow for this tree, and the canopy‐and‐stem stemflow for three other trees, were measured and sampled for chemical analysis. Total stemflow increased after installation of the 19·1 m collector just under the tree crown, suggesting a large water storage capacity of the bark. All ion concentrations in stemflow except H+ were highest in the first sample collected during a rain event and declined thereafter. Hydrogen ion concentrations of canopy‐only stemflow and canopy‐and‐stem stemflow were nearly balanced with the charge balance deficit (CBD) concentrations for all samples. Following comparison of ion concentrations among canopy‐only stemflow, canopy‐and‐stem stemflow and throughfall, the ions were categorized into four groups (NO3−, SO42−, H+, CBD), (Na+, Cl−), (Mg2+, Ca2+) and (K+, NH4+). The groupings were based on differences in stemflow and throughfall chemistry generation processes between and among the groups. Copyright © 2001 John Wiley & Sons, Ltd.

Chemical Composition of Precipitation in Beijing Area, Northern China

Variations of anions (SO42-, NO3-,NO2-, Cl- and F-),cations (K+,Na+, Ca2+, Mg2+ and \({NH}_4^+\)+) and pH values in precipitation, througfall and stemflow samples collected overa four-year period (1995–1998) in Beijing (two sitesZhongguancun and Mangshan) are presented. The annualvolume-weighted range of pH values were 6.57–7.11 inprecipitation, 5.46–6.86 in thoughfall and 5.32–6.41 instemflow. The fominant anion was \({SO}_4^{2-}\), while Ca+and NH4+ were the main cations in precipitation,throughfall and stemflow. Most of ion concentrations with precipitation, throughfall and stemflow volume showed negative correlation, except for some ones. Significant correlationvalues were also found between ions (SO42-,NO3-, Cl-, F-, Ca2+,Mg2+ andNa+) in precipitation, throughfall and stemflow indicatedthe common sources of these ions such as coal combustion,automobile emission and fertilizers application. Compared toprecipitation, there was an increased ion concentration inthroughfall or in stemflow. Changes of ion concentrations werein Quercus liatungensis Koiz. and Pinus tabulaefornisCarr. throughfall (or stemflow) because of different crown andbark qualities of tree species.

Physical properties of water in relation to stemflow leachate dynamics: implications for nutrient cycling

Stemflow leachate chemistry from a deciduous canopy tree species monitored during late winter and early spring precipitation events demonstrated significant chemical enrichment. By considering stemflow volume and chemical concentration in relation to the quantity that would be expected in a rain gage occupying an area equivalent to the trunk basal area, manganese was found to be enriched by a mean factor of 1450 and potassium by a mean factor of 580. The most pronounced enrichment was documented during a late winter rain-on-snow event characterized by temperature oscillations near the freezing point. During this event, manganese was enriched by a factor of 4400 and potassium by 1715. We conclude that mixed precipitation events with multiple freeze-melt cycles can generate significantly more leachate than spring rainfall events because of lower air temperatures and increased kinematic viscosity and surface tension of stemflow drainage. These physical properties lengthen the residence time of intercepted precipitation on the woody frame of the tree and promote its funneling from inclined branches. Stemflow represents a spatially localized and enriched point input that may affect tree vigor in early spring. The influence of localized aqueous chemical fluxes to the forest floor on forest biogeochemistry and ecophysiological functioning are discussed.

Partitioning of rainfall into throughfall, stemflow and interception: effect of forest type, ground cover and climate

This paper reviews current knowledge of the effect of forest type, ground cover and climate on rainfall partitioning into throughfall, stemflow and interception. It considers the variety and reliability of measurement techniques that have been used and interprets the results of interception studies in relation to methodological, vegetational and climatic factors. The review illustrates that it is difficult to draw general conclusions about interception losses by particular forest types because they almost always depend on the type of rainfall and other meteorological conditions during the study period. Characteristics of a forest that affect interception are not always easy to identify and quantify. Characteristics such as trees/ha, branch angle, the uniformity or lack of uniformity in crown height, the nature and thickness of the bark layer, leaf shape and inclination, and leaf area index will all influence interception. The major difficulty is reliable estimation of throughfall. It is not unusual for measured throughfall to exceed the rainfall value, causing interception to be negative. The difficulty in estimation of throughfall is discussed, while also addressing the need for accurate measurement of rainfall. In forests where stemflow volumes are large enough to significantly influence the interception values, the methodological approach is even more difficult. Measurement of interception on a periodic basis presents fewer problems than estimation on an individual event basis because the variables that affect the values of throughfall and stemflow can change over very short periods, for example the intensity and angle of rainfall, and wind speed and direction. However, carefully conducted event-based studies can quantify the influence of a number of the variables. Copyright © 2000 John Wiley & Sons, Ltd.

The measurement and modelling of the variation of stemflow in a laurel forest in Tenerife, Canary Islands

The stemflow of 30 sample trees belonging to six different species was quantified in the laurel forest of Agua Garcı́a, Tenerife, Canary Islands, during a complete year. Different methods described in the literature, for scaling up the stemflow values from tree level to stand level were applied. The best for this type of forest is based on the application of annual regressions between basal area and stemflow in each species, giving a total annual stand stemflow of 42.86 l m −2. This represents 6.85% of annual gross precipitation. This is not a very high value but important to be taken into account in the determination of the water balance of the forest. An antecedent precipitation of 2 mm was necessary for the initiation of the stand stemflow. Stemflow attained 10% of gross precipitation during big rain events, being positively correlated with the mean hourly rain intensity. Differences among trees and species were also studied resulting that trees with the biggest crown area, not dominated by their neighbours, and with the smoothest bark collected the largest volumes of stemflow in the stand. Bark roughness, divergence angle of the branches with respect to the trunk and also the size of the leaves, were responsible for the differences between species. As a consequence of the stemflow the rain can be concentrated up to 12.8 times in the infiltration areas of the trees which could produce better soil conditions for tree growth. The results obtained in the Agua Garcı́a laurel forest stand are the first stemflow data obtained for the Canarian laurel forest and they are in the range found in other types of forests.

Stemflow determination in forest stands

A stemflow measurement technique is described, based on: (1) correlations between stemflow volume and tree diameters at breast height (DBH), determined on the basis of a relatively low number of tree samplings; (2) the distribution of trees in the stand relative to their DBH. Stemflow for trees in a stand (mm) can thus be determined from total rainfall (mm). The results were compared to those obtained with two other stemflow determination techniques. The proposed method was found to be more reliable, precise and easier to apply than the standard methods.

MID-WINTER STEMFLOW DRAINAGE FROM BIGTOOTH ASPEN (POPULUS GRANDIDENTATA MICHX.) IN CENTRAL MASSACHUSETTS

Under winter conditions, stemflow drainage in forested ecosystems is often assumed to be a negligible component of the hydrological cycle. This paper reports on mid-winter stemflow drainage from the broadleaved deciduous tree species Populus grandidentata. Stemflow volumes from this species at air temperatures of < 0°C were found to be comparable to rainfall-generated stemflow during summer. Over the three-month period January–March 1993, stemflow ranged from 5.4 to 9.9% of the incident gross precipitation. Expressed as depth equivalents per unit trunk basal area, these stemflow inputs ranged from 1.8 to 4.9 m. These concentrated mid-winter inputs of liquid water to the bases of canopy trees were attributable to: (1) snow interception by the leafless woody frame of each tree; (2) snow retention by glazed ice precipitation associated with the snowfall event; (3) increased temperature at the bark/snow interface caused by the low albedo of the bark tissue; and (4) convergence of snowmelt drainage from steeply inclined upthrust primary branches. The hydrological and ecological significance of liquid water inputs to the forest floor under sub-zero conditions are discussed. © 1997 by John Wiley & Sons Ltd.

Stemflow influences on the formation of solution pipes in Bermuda eolianite

Soil-filled pipes consisting of calcite casings were found in all of the eolianite formations of Bermuda except the Rocky Bay Formation. The pipes are cylindrical in shape, taper vertically downwards, and occur in aggregated clusters. They are most commonly exposed in areas of shoreline erosion in the form of soil-filled cylindrical rims of calcite projecting up from the surrounding host eolianite. In the early literature on the geology of Bermuda, it was suggested that these features were fossilized stumps of the Bermuda palmetto tree Sabal bermudiana. In the study presented here, this hypothesis is rejected on the basis of the morphology and stratigraphic position of these geomorphic features and their basal area and areal density compared with existing natural stands of S. bermudiana. The soil-filled pipes of Bermuda, however, cannot be simply explained as random solutional features that correspond to points where percolating soil water happened to converge and cause localized subsurface dissolution. In situ measurements of stemflow from indigenous tree species suggest that the soil-filled solution pipes of Bermuda may be the product of stemflow drainage from particular tree species. Stemflow from S. bermudiana was the most acidic, with pH values as low as 4.29 compared to the gross rainfall in the open, which had a pH of 5.58. The crowns of the Bermuda cedar ( Juniperus bermudiana) and the subtropical tree species Eugenia axillaris had the most significant funnelling effect. Stemflow volumes from individual trees representing these two species were 20 times greater than the volume of water that would have been collected in rain gauges with orifice areas equivalent to the trees' trunk basal areas. Tree species capable of acidifying intercepted rainwater and funnelling large quantities down their trunks as stemflow may be the cause of subsurface dissolution loci extending vertically downward through Bermuda's soluble eolianite formations. Stemflow from long-lived trees representing these species could account for the diameters, the cyclindrical shape, the vertical orientation, and the soil-filled interiors of the pipes. Successive generations of these trees in preferred habitats could explain their clustered distribution and high areal densities (>10 3 pipes/ha).

Throughfall below grassland canopies: A comparison of conventional and ion exchange methods

Sampling of canopy fluxes (throughfall and stemflow) below low structured vegetation with a small-scale, intricate canopy architecture is difficult, and representative sampling with most methods is questionable. In the present study, two sampling methods for canopy fluxes below grassland vegetation are compared. Method I sampled canopy fluxes of moisture inefficiently, because stemflow volumes were not quantitatively included. Canopy fluxes of ions calculated with method I necessitated assumptions on equal concentrations in actually sampled throughfall and non-sampled stemflow. Method II sampled canopy fluxes of ions quantitatively, because the total volume of throughfall and stemflow percolated through a mixed bed of ion exchange resins below the canopy. Ion-specific differences between the two methods were observed. For ions with foliar leaching, such as K + and Ca 2+, higher canopy fluxes were recorded with method II than with method I. In contrast, for ions with foliar uptake, such as NH 4 + and NO 3 −, canopy fluxes were found to be less with method II than with method I. Canopy fluxes of inorganic nitrogen below Mesobrometum grassland were 2·35 and 1·52 kmol cha −1 year −1 for methods I and II, respectively, and 2·85 and 7·90 kmol cha −1 year −1 for K +. It is argued that these differences result from under-estimated (foliar leaching) or over-estimated (foliar uptake) concentrations in stemflow by the first method. Canopy fluxes for SO 4 2− were not statistically different, indicating that canopy exchange of SO x was quantitatively unimportant, and that both methods estimated atmospheric input equally well.

Throughfall, stemflow and interception of rainwater in an evergreen broadleaved forest

AbstractIn order to analyze the dyanmics of heavy metals in a forest ecosystem, throughfall and stemflow were collected for individual rain showers in an evergreen broad‐leaved forest dominated by Castanopsis cuspidata. The relation between throughfall (or stemflow) (Px) and gross rainfall (P) was approximated by a linear regression equation Px=a(P‐b). The values of coefficient “a” were 0.32–2.02 for throughfall at each sampling point and the mean values for 1976 and 1977–1978 were 0.682 and 0.767, respectively. The stemflow volume differed widely among individual trees, depending mainly on the tree form of each species. In particular, the tree form of Pasania edulis was found to be especially suited to collecting stemflow. One P. edulis tree collected 64% of the rainwater that fell onto its crown as stemflow. The ratio of stemflow to gross rainfall decreased in summer resulting from an increase in leaf biomass and an increase in air temperature. In fact, the values of coefficient “a” for mean stemflow per unit area were 0.180 for summer and 0.229 for other seasons in 1976, and 0.145 for summer and 0.155 for other seasons in 1977–1978 for different sampled trees.

Infiltration‐excess caused by Stemflow in a cyclone‐prone tropical rainforest

AbstractTropical rainforest canopy trees that have large projected areas of upwardly inclined branches are capable of funnelling large volumes of rainwater down their trunks. During periods of prolonged heavy rainfall on Mount Bellenden Ker in northeast Queensland, Australia, stemflow volumes were found to be as much as two orders of magnitude greater than the volume of incident rainfall expected in a rain gauge occupying an area equal to the trunk basal area. Stemflow totals ranging from 6000 to 70000 litres were generated by individual trees from 7800 mm of rainfall over two successive wet seasons.The combination of high intensity rainfall and the funnelling effect results in significant quantities of infiltration‐excess at the ground surface. Stemflow fluxes as high as 31.4 cm3 min−1 per cm2 of basal area (i.e. the equivalent of 314 mm min−1) were recorded when rainfall intensity was only 2 mm min−1. The mean infiltration capacity of the topsoil was determined to be 6.2 mm min−1. The areas over which the stemflow would have had to spread in order to infiltrate were computed to be as much as 3 m2 around the bases of individual canopy trees. Approximations of the distances that the infiltration‐excess would have travelled away from the tree bases were calculated by assuming that the infiltration area either expands radially outward in the form of an annulus or extends straight downslope from the tree base.

Canopy interception of nitrogen in bulk precipitation by annually burned and unburned tallgrass prairie.

Nitrogen content of bulk precipitation and throughfall (canopy leachates) was measured on annually burned and unburned tallgrass prairie during a 20 month period. Throughfall amounts averaged 58% of precipitation on unburned prairie while throughfall on annually burned sites averaged 76% of precipitation inputs. Stemflow was measured in late summer and autumn. Volumes were correlated with stem density; maximum stemflow volumes measured in this study averaged about 50% of throughfall volumes.Bulk precipitation averaged 530, 456, and 420 μg/l of nitrate, ammonium and organic nitrogen, respectively. Throughfall on burned sites averaged 345, 344 and 980 μg/l of nitrate, ammonium and organic nitrogen, and throughfall on unburned sites averaged 258, 196 and 1701 μg/l of nitrate, ammonium and organic nitrogen. Microbes on standing dead vegetation and litter of the unburned sites were estimated to remove more inorganic nitrogen from bulk precipitation than did foliage on burned sites. Only a portion of the inorganic nitrogen in bulk precipitation is immediately available for plant use, and this availability is influenced by the amount of detritus present on the prairie.

The effect of crop covers on rainfall characteristics and splash detachment

A laboratory study is presented in which the effects of increasing canopy cover on rainfall characteristics and splash detachment under crops were investigated. Data were collected using a simulated rainfall of 61 mm/h for 5 min duration typical of an erosive storm in the Silsoe area, and a local erodible sandy soil of the Cottenham series. Experiments were carried out for three locally-grown crops; Brussels sprouts, sugar beet and potatoes. Measurements were made of the percentages of the storm rainfall reaching the ground surface as throughfall, leaf drips and stemflow. Soil detachment, rainfall energy and rainfall intensity were determined at three discrete positions under the plants, under the canopy, below the canopy margin and beneath the leaf drips. As the plant grows, it is shown that the percentage interception area increases, the volume of throughfall decreases and the volumes of stemflow and leaf drips increase. Soil detachment is reduced except at the leaf drip points under Brussels sprouts and sugar beet where it increases. Although the kinetic energy of the rainfall under the plants in all three positions is less than that of the control storm, it increases as canopy area increases. This results in a negative relationship between soil detachment and kinetic energy, a trend reported in previous work but still not fully explained.