Stemflow and soil water recharge during rainfall in a red pine chronosequence on the Oak Ridges Moraine, southern Ontario, Canada

Journal of Hydrology 517 (2014) 777–790

Contents lists available at ScienceDirect

Journal of Hydrology journal homepage: www.elsevier.com/locate/jhydrol

Stemflow and soil water recharge during rainfall in a red pine chronosequence on the Oak Ridges Moraine, southern Ontario, Canada J.M. Buttle a,⇑, H.J. Toye a, W.J. Greenwood b, R. Bialkowski b a b

Department of Geography, Trent University, Peterborough, ON K9J 7B8, Canada Environmental and Life Sciences Graduate Program, Trent University, Peterborough, ON K9J 7B8, Canada

a r t i c l e

i n f o

Article history: Received 4 October 2013 Received in revised form 10 April 2014 Accepted 11 June 2014 Available online 21 June 2014 This manuscript was handled by L. Charlet, Editor-in-Chief, with the assistance of Jiin-Shuh Jean, Associate Editor Keywords: Stemflow Throughfall Interception Soil water recharge Red pine Chronosequence

s u m m a r y Stemflow focusses water delivery to the forest floor in a relatively small area surrounding the tree bole, with the potential to enhance soil water contents and water recharge relative to more distal sites beneath the canopy. These stemflow fluxes may decrease as a given tree species ages due to changes in branch orientation and bark roughness, suggesting that the relative contribution of stemflow to water recharge near the bole will decline with time. The hypothesis that stemflow fluxes decline with tree age was tested in a chronosequence of red pine (Pinus resinosa Ait.) stands in a managed forest in southern Ontario, Canada, and stemflow contributions to soil water recharge below 1 m depth were quantified. Throughfall, stemflow and sub-canopy soil water contents (0.1 m and 1.5 m from the tree bole to 1 m depth) in stands ranging from 28 to 80 years in age were studied from late-Spring to Fall in 2012, supplemented by artificial irrigations of stemflow to examine short-term soil wetting at the two distances from the bole. The hypothesized decline in stemflow with increasing tree age was not supported, and canopy cover variations and forest management exerted a greater control on inter-stand differences in stemflow fluxes. Stemflow contributions generally resulted in greater soil water recharge below 1 m depth at 0.1 m from the tree bole compared to the 1.5 m distance. This enhanced recharge was greatest for the youngest stand and differences in recharge between the 0.1 m and 1.5 m distances from the bole were likely not significant for stands between 40 and 80 years of age. Nevertheless, the relative contribution of stemflow to soil water recharge may increase in this managed forest as red pine stands give way to a mixed hardwood-conifer forest, due to greater stemflow fluxes from hardwood species in this landscape. Ó 2014 Elsevier B.V. All rights reserved.

1. Introduction Forests partition above-canopy precipitation (Pg) between throughfall (TF), stemflow (SF) and canopy interception (Ic), the latter returned to the atmosphere via evaporation. Of these, SF is considered to be the smallest fraction of Pg (Helvey and Patric, 1965; Levia and Frost, 2003), although its ability to concentrate water delivery to and beneath the soil surface at the tree bole has been widely recognized (Voigt, 1960; Ford and Deans, 1978; Durocher, 1990; Chang and Matzner, 2000). This focussing of incident rainfall by SF to soil around the bole and its subsequent transport belowground along tree roots and other preferential flow paths has been characterized as a double-funneling process (Johnson and Lehmann, 2006; Schwärzel et al., 2012). Despite the potential of SF to make a disproportionate contribution to soil water and groundwater recharge (Johnson and ⇑ Corresponding author. Tel.: +1 705 7481011; fax: +1 705 7481205. E-mail address: [email protected] (J.M. Buttle). http://dx.doi.org/10.1016/j.jhydrol.2014.06.014 0022-1694/Ó 2014 Elsevier B.V. All rights reserved.

Lehmann, 2006; Nàvar, 2011), its role in soil water dynamics in forest landscapes has often been ignored due to its generally small fraction of Pg (Liang et al., 2009; Tanaka, 2011). Nevertheless, the influence of SF on soil water and groundwater recharge has received some attention. Empirical studies include Taniguchi et al.’s (1996) mass balance estimate that SF contributes up to 20% of total groundwater recharge in Japanese red pine, fluorescent dye tracing of SF channelization along root systems (MartinezMeza and Whitford, 1996; Schwärzel et al., 2012), Liang et al.’s (2011) study of SF, soil water content and pore water dynamics around a tree on a steep slope, and Germer’s (2013) investigation of perched water table development in response to SF from babassu palms. Modelling (e.g. Tanaka et al., 1996; Chang and Matzner, 2000; Liang et al., 2009) has also shown that SF can enhance soil and groundwater recharge significantly within a relatively small area surrounding the tree bole compared to more distal regions beneath the canopy. Stemflow volumes vary as a function of tree species, largely due to differences in bark roughness and branching geometry (Levia

778

J.M. Buttle et al. / Journal of Hydrology 517 (2014) 777–790

Nomenclature AET c CV DBH E Fg GF Ic LAI OM ORM

actual evapotranspiration (mm, mm d1) canopy cover coefficient of variation diameter at breast height (m) soil evaporation plus transpiration (mm) gap fraction Ganaraska Forest canopy interception (mm) leaf area index (m2 m2) organic matter (%) Oak Ridges Moraine

et al., 2010). Lower branch angles increase the amount of water flow along the branch lost to TF, thus reducing contributions to SF (Herwitz, 1987). The relative importance of SF in partitioning net water delivery to the soil surface in forests may also change with stand age for a given species. Individual tree branch angles decrease as branches age, while branches slope away from the trunk for much of their length as a forest ages (Ford and Deans, 1978). Stemflow is greater in younger relative to older hardwood stands in the eastern US due to greater stem density, smoother bark and the tendency of branches to grow up rather than out in younger stands (Helvey and Patric, 1965). Increasing bark roughness for older trees of the same species leads to greater water storage capacity and decreased SF (Levia and Frost, 2003). Stemflow as a% of net precipitation (TF + SF) appeared to increase from age 0 to 14 years and then declined as trees continued to age in upland catchments in the UK (Johnson, 1990). Stemflow contributions to net precipitation reaching the forest floor in red pine (RP, Pinus resinosa Ait.) stands in the Ganaraska Forest on the Oak Ridges Moraine (ORM) in southern Ontario were hypothesized to decrease as stands age (Buttle and Farnsworth, 2012), due to increasing branch length and consequent change in branch orientation relative to the bole from near horizontal to sub-horizontal (Levia and Frost, 2003) combined with greater water sorption by RP’s irregular and furrowed bark (Voigt and Zwolinski, 1964; Iida et al., 2005). Mature RP branches are fewer in number, much thicker, longer and flatter than for young RP and often curve downward towards the outer end (Voigt, 1960), reducing channelling of SF to the bole. This suggests that SF fluxes may decline, and by implication the influence of SF on soil water and groundwater recharge may lessen, as trees age. However, there is currently no evidence to support this assumption. The objectives of this paper were to: (1) characterise SF fluxes at the scale of an individual tree during rainfall in a chronosequence of RP stands in the Ganaraska Forest; (2) test the hypothesis that SF decreases with tree age in this managed forest; (3) estimate the area surrounding the tree bole that infiltrates this SF; and (4) quantify the role of SF in soil water recharge across this chronosequence. 2. Study area The study was conducted in the western portion of the Ganaraska Forest (GF, 44°50 N, 78°300 W) on the crest and flanks of the Oak Ridges Moraine (ORM, Fig. 1). The ORM is an interlobate kame moraine consisting of sand and gravel hills and high ridges comprised of interlayered gravels, sands, silts, clays and minor diamictons up to 150 m thick (Barnett et al., 1998). The upper several m of surficial deposits in the GF along the crest of the ORM are dominated by sand with low silt and clay content, and water well records indicate the water table is at least 30 m below the ground surface (Funk, 1977). Maximum elevation in the GF is 384 m asl,

PET Pg Pnet RP SD SE SF SWR SWC TF

potential evapotranspiration (mm, mm d1) above-canopy precipitation (mm) net precipitation (mm) red pine standard deviation standard error stemflow (L, mm) soil water recharge below 1 m depth (mm) soil water content (m3 m3) throughfall (mm)

309 m above Lake Ontario to the south. The region has a humid mid-latitude climate (Köppen DfB) and mean annual precipitation ranges from 950 mm on the western edge of the GF to 825 mm on its eastern edge (Ganaraska Region Conservation Authority, 2008) with no marked seasonality in precipitation (Buttle, 2011). About 20% of mean annual precipitation falls as snow. Mean daily air temperatures in January and July are 7.2 °C and 20.5 °C, respectively, while annual average regional evapotranspiration is 530 mm (Buttle and Farnsworth, 2012). Soils are brunisolic gray brown luvisols (Soil Classification Working Group, 1998; FAO equivalent: arenosol) belonging to either the Pontypool sand or gravelly sand series. Typical profiles have a 0–5 cm LFH layer, a dark greyish-brown loamy sand Ah horizon (0–5 cm), a light gray sandy Ae horizon (5–10 cm), a reddish brown sand Bfh horizon (10–28 cm), a light reddish brown sandy Bm horizon (28–58 cm), a dark brown sandy loam Btj horizon (58–64 cm), and an underlying gray medium sand Ck horizon (Hoffman and Acton, 1974). Soils have occasional layers of coarse and fine sand or gravel, and singlegrain structure with the exception of the Ah (weak, fine granular) and Btj (medium subangular blocky) horizons. Both the Bm and Btj horizons have lower boundaries with deep tongues that extend into the underlying horizon. Soils in the GF have large saturated hydraulic conductivities, with mean values to 1 m depth >266 mm h1 (Greenwood and Buttle, 2014). The ORM was deforested in the 19th century through logging and agricultural development. Large-scale reforestation of marginal and sub-marginal land in the GF began after 1945, ending in 1985 (described in more detail in Buttle, 2011). Red pine and white pine (Pinus strobus L.) were generally planted in combination at densities of 2100 trees ha1, with RP being the main species planted. A ridge-and-furrow system was not used during planting. Plantations serve as overstory for regenerating mixed hardwoods and conifers, and are selectively thinned and harvested to promote conversion to a mixed hardwood-conifer forest. Red pine plantations are thinned every 10–12 years with a basal area at the time of thinning of 28–32 m2 ha1. Thinning reduces the basal area to 20–22 m2 ha1 (R. Penwell, Ganaraska Region Conservation Authority, oral communication, 2010). Trees are also pruned (live and dead branches removed from standing trees) to within 0.1– 0.2 m of the bole up to 7.4 m from the ground to produce knot-free sawlogs. 3. Methods 3.1. Stand characterization Six stands consisting of at least 50% RP at time of planting were selected to span the range in RP plantation ages in the GF, varying in age from 28 to 80 years (Fig. 1, Table 1). The maximum distance between stands was <6 km. Species, diameter at breast height

779

J.M. Buttle et al. / Journal of Hydrology 517 (2014) 777–790

Fig. 1. The Oak Ridges Moraine, the central and western portions of the Ganaraska Forest, and locations of the study stands (see Table 1 for details), rain gauges used to measure above-canopy rainfall (Pg), tipping bucket rain gauge and eddy covariance tower.

(DBH, measured 1.5 m above ground surface), basal area, density of all species and of RP, height, distance from ground surface to lowest live branch for RP, and length of the RP bole consisting of live canopy was recorded for all trees within a 12 m radius of the mid-point of the sampled trees described in Section 3.2. The latter metric was used as a surrogate for tree biomass, found to be positively correlated with SF from European beech saplings (Levia et al., in press). It assumes that trees with a greater length of live canopy have a larger canopy volume that could intercept incoming rainfall and deliver it to the bole. These trees would produce more SF compared to trees with a smaller live canopy length where a greater fraction of SF would be generated by direct rainfall delivery to the relatively-small surface area of the bole. Upward digital hemispherical photographs were taken using a Pentax™ K10D SLR digital camera with a 10–17 mm fish-eye lens following the

method of Zhang et al. (2005) above the mid-point of the trough used to measure TF (Section 3.2) in each stand on 14 August 2012. Images were processed using the Can-Eye version 6.1 imaging software (Weiss and Baret, 2010), and gap fraction (Fg) and leaf area index (LAI, m2 m2) were determined for each image. Canopy cover (c) was estimated as 1  Fg. The approach of Zhang et al. (2005) was used to select images to represent Fg, c and LAI for a given stand. 3.2. Stemflow, throughfall, above-canopy rainfall, canopy interception and soil water content Four RP in a level area in each stand were selected (Table 2). Selection was based on similar DBH, proximity (within 10 m of one another), no overlap between projected canopy area of the tree

Table 1 Characteristics of study stands in the Ganaraska Forest.

* +

Stand

Area (ha)

Age (y)

Mean height ± SD, non-RP (m)

Mean DBH ± SD, non-RP (m)

Mean height ± SD, RP (m)

Mean DBH ± SD, RP (m)

Mean distance from ground surface to lowest live branch, RP (m)

Mean length of live canopy, RP (m); proportion of mean height

1

18

28

–

–

11.6 ± 0.9

0.16 ± 0.01

5.7 (2.0*)

5.9; 0.51 (9.6; 0.83*)

n= 98+

n = 98

n=4

Mean canopy projected area, RP (m2)

Density, all trees (N ha1)

Density, RP (N ha1)

c

LAI (m2 m2)

9.0

2166

2166

0.94

3.09

2

13

40

14.8 ± 10.2 n=3

0.13 ± 0.09 n=3

19.3 ± 3.4 n = 14

0.27 ± 0.03 n = 14

12.6 n=4

6.7; 0.34

19.4

376

309

0.51

0.77

3

9

54

16.7 ± 3.3 n = 10

0.24 ± 0.05 n = 10

20.5 ± 4.9 n = 11

0.33 ± 0.02 n = 11

13.2 n=4

7.3; 0.36

20.5

464

265

0.64

1.10

4

3

62

16.8 ± 3.2 n = 11

0.24 ± 0.05 n = 11

19.3 ± 3.3 n = 10

0.32 ± 0.02 n = 10

11.2 n=4

8.1; 0.42

21.5

464

221

0.86

2.14

5

5

70

11.2 ± 4.0 n=7

0.11 ± 0.08 n=7

16.2 ± 3.3 n = 12

0.33 ± 0.02 n = 12

11.0 n=4

5.2; 0.32

21.1

420

265

0.60

0.98

6

9

80

7.9 ± 3.7 n = 72

0.12 ± 0.09 n = 72

14.9 ± 3.8 n=5

0.34 ± 0.04 n=5

12.7 n=4

2.2; 0.15

19.4

1720

111

0.56

0.88

Includes dead branches 0.6 m long below the live canopy. Number of trees measured.

780

J.M. Buttle et al. / Journal of Hydrology 517 (2014) 777–790

and that of surrounding trees, and position at least 20 m from the stand boundary to minimize edge effects. Stemflow was measured at three trees using collars of clear tygon tubing (25.4 mm i.d.) slit longitudinally, stapled to the tree bole in a spiral beginning 1.5 m above the ground surface and sealed to the bole with silicone caulking. Collars drained to a collector, and no overtopping of collars or flow beneath collars via the tree bark was observed during events. Stemflow volumes for each tree were expressed as a depth relative to the tree’s projected canopy area, and mean SF fluxes in a stand were assigned to the 4th tree in that stand. One tree measured for SF in each stand was randomly selected, and TF beneath its canopy was measured using a 3.14 m long, 0.1 m wide trough draining to a collector. Trough length was selected to ensure it did not extend beyond the canopy of the youngest trees studied (stand 1), and the same length was used for all stands. Troughs were positioned 1 m above the ground surface above any understory vegetation at angles ranging from 0.9° to 6° from the horizontal. The trough mid-point was sited 0.25 m from the north side of the bole and the trough was oriented in an east–west direction parallel to the prevailing wind direction. Volume of TF was expressed as a depth relative to the trough plan area. Above-canopy rainfall was measured at four open sites (Fig. 1) using a standard rain gauge at each site (Buttle and Farnsworth, 2012). Sites were selected based on their proximity to the study stands as well as landowner permission to access their property. Each open site was >40 m in diameter, and Pg for each study stand was estimated using the inverse distance squared method (Smith, 1993). Rainfall intensity was measured with a tipping bucket rain gauge at the open site northeast of stand 4 (Fig. 1). Canopy interception (Ic) was the difference between Pg and TF plus SF (Buttle and Farnsworth, 2012). Soil water contents (SWC) were measured using a Delta T PR2/6 Profile Probe™ (http://www.delta-t.co.uk, last accessed April 4 2014) calibrated for GF soils. The calibration curve (Fig. 2) was developed by installing three vertical ATL-1 access tubes (1154 mm  28 mm i.d.) using a PR-AUG4 auger within 1 m of each other and adding differing quantities of water to the ground surface at each tube. Water was allowed to percolate through the soil column in order to produce a range of SWCs with

Table 2 Details for sampled trees in each study stand. SF = stemflow collar, TF = throughfall collector, SWC = soil water content access tubes at 0.1 and 1.5 m from tree bole. Standtree

Instrumentation

DBH (m)

Average canopy radius (m)

Height (m)

1-1 1-2 1-3 1-4 2-1 2-2 2-3 2-4 3-1 3-2 3-3 3-4 4-1 4-2 4-3 4-4 5-1 5-2 5-3 5-4 6-1 6-2 6-3 6-4

SF SF, TF SF SWC SF, TF SF SF SWC SF, TF SF SF SWC SF SF SF, TF SWC SF SF, TF SF SWC SF, TF SF SF SWC

0.16 0.14 0.16 0.16 0.25 0.24 0.26 0.25 0.30 0.34 0.30 0.33 0.33 0.33 0.33 0.31 0.31 0.32 0.33 0.32 0.32 0.38 0.33 0.31

1.8 1.6 1.7 1.8 2.7 2.2 2.6 2.3 2.8 2.4 2.5 2.6 2.9 2.4 2.5 2.5 2.9 2.4 2.9 2.6 2.5 2.3 2.7 2.5

10.9 12.2 11.6 13.0 21.1 17.8 17.7 14.0 19.6 24.0 18.9 17.5 19.6 24.0 18.9 18.0 12.2 19.7 15.9 18.5 12.8 21.1 11.0 14.7

Fig. 2. Field calibration relationship between SWC and Profile Probe output for soils in the Ganaraska Forest.

depth at each access tube. Profile Probe readings (mV) were made at 0.1, 0.2, 0.3, 0.4, 0.6 and 1 m depths, a 1 m3 soil pit was excavated adjacent to each access tube and soil cores (6 cm long, 4.8 cm i.d.) were extracted from the pit face within 6 cm of the access tube at each measurement depth. The process was repeated at a second location. The cores were dried at 105 °C, and volumetric water content was determined and regressed against corresponding mV readings. Measurements of SWC in each study stand were made in vertical access tubes installed in holes augered 0.1 m and 1.5 m away from the bole of the 4th tree. The 0.1 m distance was the closest to the bole that an access tube could be installed using the PR-AUG4 auger while the 1.5 m distance was the maximum distance from the smallest trees sampled (stand 1) that would ensure the access tube remained beneath the canopy. The same distances were used at all stands to ensure comparability of results. The access tube transect was on the west side of the tree, since prevailing westerly winds might maximize the potential to observe SF contributions to SWC on the tree’s windward side. Measurements of Pg, SF, TF and SWC were made 23 times at each stand between Day of Year (DOY) 171 (June 19) and DOY 290 (October 16), 2012.

3.3. Irrigation experiment A simple irrigation experiment was conducted in 2013 to examine the effects of SF on SWC at 0.1 m and 1.5 m distances from the bole in greater detail. Collars were fitted at 1.5 m height to the bole of the 4th tree instrumented with SWC access tubes in each study stand. Each collar consisted of aluminum foil shaped into an inverted cone around the bole and secured with a bungee cord, such that water applied inside the collar would percolate down the bole along the bark surface. The maximum volume of SF measured at any tree in that stand in 2012 was applied to the collar over a 5–10 min period (Table 4) to maximize the potential for SF infiltrating the soil around the bole to move away from the tree via macropores and soil surrounding tree roots and be detected at 1.5 m from the bole. Water was applied to the collar in an attempt to produce an even flow of SF down the bole. Soil water contents were measured immediately before application, and at 1, 2, 3, 4, 5, 6 and 24 h following application at 0.1 m and 1.5 m from the bole. We assumed the water infiltrated in a ring around the bole, similar to the cylindrical infiltration model of Tanaka et al. (1996) which was supported by dye tracing of SF infiltration around a beech tree (Schwärzel et al., 2012). The maximum

781

J.M. Buttle et al. / Journal of Hydrology 517 (2014) 777–790

Table 3 Total above-canopy precipitation (Pg), throughfall (TF), stemflow (SF) and interception (Ic) for the study stands, June 19–October 26, 2012. Average, standard deviation (SD) and coefficient of variation (CV) of total SF from the three instrumented trees in each stand is reported.

a

Stand

Pg (mm)

TF (mm)

SF ± SD (L)

CV

SF ± SD (mm)a

CV

Ic (mm)

SF:Pg

Ic:Pg

1 2 3 4 5 6

368 417 408 417 369 409

230 299 233 252 255 304

24.6 ± 15.4 4.4 ± 2.9 14.3 ± 3.0 46.3 ± 9.1 8.9 ± 3.8 23.8 ± 6.2

0.62 0.66 0.21 0.20 0.47 0.26

2.8 ± 1.8 0.2 ± 0.1 0.7 ± 0.1 2.2 ± 0.7 0.4 ± 0.2 1.2 ± 0.4

0.64 0.50 0.14 0.32 0.50 0.33

135 118 174 163 114 104

0.008 0.001 0.002 0.005 0.001 0.003

0.37 0.28 0.43 0.39 0.31 0.25

Based on canopy projected area.

3.5. Soil characterization Table 4 Irrigation experiment results: amount of water applied to collar, maximum change in soil water depth held in upper 1 m of soil at 0.1 m access tube, and width of ring around tree bole required to convert volume of applied water to an equivalent water depth equal to maximum change in soil water depth. Stand

DBH (m)

Volume of water applied (L)

Maximum change in soil water depth in upper 1 m of soil (mm)

Width of ring (m)

1 2 3 4 5 6

0.16 0.25 0.33 0.31 0.32 0.31

13.7 2.5 5.2 18.1 3 9.4

27.9 14.9 12.2 35.1 9.7 21.8

0.18 0.23 0.24 0.11 0.28 0.16

observed increase in water depth stored in the upper 1 m of soil based on SWCs at the 0.1 m access tube following water application was used to estimate the width of the ring of soil around the bole needed to convert the applied SF volume into an equivalent water depth.

3.4. Soil water recharge The large sand contents of the soils and underlying surficial geological deposits and the large depths to the water table along the crest of the ORM suggest minimal lateral subsurface water movement within the upper several m of the level areas monitored in the study stands, and that below-ground water fluxes were essentially vertical. Soil water recharge (SWR) below 1 m depth for the DOY 171–290 period was estimated at 0.1 m and 1.5 m distances from the tree using a 1-d water balance approach:

SWR ¼ Pnet  E  DSWC

Vertical profiles of soil texture, organic matter content and porosity were obtained from a 1 m3 soil pit dug 1 m away from a line joining the 0.1 m and 1.5 m access tubes midway between the tubes. Tree roots encountered during excavation were cut and removed. Horizontal cores (6 cm long, 4.8 cm i.d.) beginning at 0.05-m depth were taken at 0.1-m intervals to 0.95 m depth. Bulk density was determined for each core and used to estimate porosity assuming a particle density of 2650 kg m3 (Carter and Ball, 1993). Organic matter content was determined by loss-on-ignition (Tiessen and Moir, 1993). Gravel content (particles >2-mm diameter) of each core was determined by sieving, and sand, silt and clay portions of the fraction 62-mm diameter were determined using a Horiba LA-950 V2 Laser Scattering Particle Size Distribution Analyzer. 4. Results 4.1. Stand characteristics There was no progressive increase in mean height, projected canopy area, density, c or LAI with stand age (Table 1). Mean RP height and DBH increased from 28-year old to 54-year old stands, but whereas mean DBH remained relatively constant there was a decrease in mean RP height for stands older than 54 years. This likely reflects management practices in the GF, where RP stands are generally thinned every 10–12 years after reaching a basal area of 28–32 m2 ha1. All stands had been thinned at least once with the exception of the youngest (stand 1) which had never been thinned or pruned. As the stands age, they have changed from 100% RP at time of planting (e.g. stand 1) to a mixed hardwoodconifer composition at the other stands that includes eastern white pine, sugar maple (Acer saccharum Marsh.), white birch (Betula papyrifera Marsh.), red oak (Quercus rubra L.), hop hornbeam (Ostrya virginiana) and American beech (Fagus grandifolia Ehrh.).

ð1Þ 4.2. Soil characteristics

Pnet was equal to TF at the 1.5 m access tube and TF plus SF at the 0.1 m access tube. Stemflow volume was converted to an equivalent depth by assuming it infiltrated over the annular area of a ring extending around the tree bole, the width of which was estimated from the SF irrigation experiment described in Section 3.3. Direct soil evaporation plus water uptake by tree roots for transpiration (E) was actual evapotranspiration (AET) minus evaporation of intercepted water (study period Ic). Actual evapotranspiration was estimated by multiplying potential evapotranspiration (PET) for the study period from the Hamon equation (Dingman, 2002) by the ratio of total AET (measured via eddy covariance at the open site northeast of stand 4, Fig. 1) to PET for 2009 (Fig. 3a). The standard error (SE) in SWR was estimated following Davidson (1978) assuming relative errors of TF (0.15), SF (0.24–0.48, based on the observed SE of mean SF in each stand), E (0.2), Ic (0.13, based on the average ratio of the SE of Ic to mean Ic measured for RP stands in the GF by Buttle and Farnsworth, 2012) and DSWC (0.15).

Some stands (e.g. 6) showed a marked decline in porosity with depth, while others (e.g. 5) showed little change (Fig. 4a and c). Near-surface% organic matter content ranged from 2.5% (stand 2) to 9% (stand 6), and values generally declined with depth (Fig. 4b and d). Soil texture was dominated by sand while clay content was negligible at all stands (Fig. 5). Some stands (2, 3, 6) also had appreciable gravel contents. Silt content was relatively consistent with depth at between 15% and 40% in three stands (2, 3, 5), but showed greater variability in the other stands (1, 4, 6) with evidence of distinct silt-rich layers in the profile bounded by sandy layers. 4.3. Above-canopy rainfall, evapotranspiration, throughfall, interception and stemflow Total Pg for the 2012 study period varied from 368 mm to 417 mm across the stands (Table 3, Fig. 3b). Summer and early Fall

782

J.M. Buttle et al. / Journal of Hydrology 517 (2014) 777–790

Fig. 3. Cumulative potential evapotranspiration (PET) estimated using the Hamon equation in 2009 based on mean daily air temperature at the Peterborough Airport (40 km northeast of the Ganaraska Forest) and actual evapotranspiration (AET) measured using eddy covariance at the open site northeast of stand 4 (Fig. 1) (a); cumulative above-canopy precipitation (Pg) estimated for each stand, cumulative PET and estimated AET (PET  total AET:PET for 2009 in (a)) in 2012 (b).

2012 were slightly wetter than normal, and June–October rainfall at the Peterborough Airport 40 km northeast of the GF (447.6 mm) was 15% greater than the 1981–2010 normal. Monthly rainfall at the Peterborough Airport in 2012 ranged from as low as 62% below (July) to as much as 49% above (June) the respective monthly normal. Thus, SF fluxes reported here may be slightly larger than average values for this region. Maximum hourly rainfall intensity during the intervals between SF measurements ranged from 0.05 to 19.9 mm h1 (average = 5.4 mm h1, median = 4.1 m m h1), and rainfall duration varied from 0.5 h to 33.5 h (average = 10.1 h, median = 8.3 h). While maximum hourly intensity was not significantly correlated with rainfall duration (p = 0.05

level), both were significantly correlated with Pg for all stands. Potential evapotranspiration was 337 mm and AET was 283 mm for the study period, based on a total AET:PET ratio of 0.84 determined for 2009 (Fig. 3). Mean daily estimated AET was 2.4 mm (SD = 1.1 mm), ranging from 0.5 mm to 5.5 mm. All stands showed similar temporal patterns in TF (Fig. 6), although absolute amounts varied considerably between stands at times during the study period. Total Ic as a fraction of Pg was >25%, while total SF as fraction of Pg was <1% in all stands. A Friedman non-parametric analysis of variance test (Winkler and Hays, 1975) indicated significant and consistent differences in SF from the three trees in a stand across the 23 measurements

Fig. 4. Vertical profiles of soil porosity (a and c) and % organic matter (b and d) at the study stands.

J.M. Buttle et al. / Journal of Hydrology 517 (2014) 777–790

783

Fig. 5. Changes in soil texture with depth at the study stands: gravel: >2 mm dia. sand 2–0.05 mm; silt 0.05–0.002 mm; clay <0.002 mm.

at all but one stand (6). Nevertheless, all stands showed significant (p < 0.05) correlations between SF volumes from the three trees in a given stand across the 23 measurements. The relative variability about mean total SF in each stand (as indexed by the coefficient of variation, CV, Table 3) ranged from 0.20 to 0.66 for SF volume (mean of 0.40, standard deviation of 0.21) and from 0.14 to 0.64 for SF depth relative to projected canopy area (mean of 0.41, standard deviation of 0.18). The number of measurements of SF from at least one tree in a stand ranged from 16 (stands 1 and 3) to 19 (stand 4), while there were only 4 times that SF was not observed at all trees across all stands. The proportion of measurements with SF from all three trees in a given stand relative to the total number of measurements with SF from at least one tree in a stand ranged from 0.89 to 0.94, indicating a similar response to Pg from sampled trees in each stand in terms of whether or not they generated SF. Those four measurements that did not observe SF from all three trees in a stand tended to have small Pg (3.8–9.4 mm), low maximum intensity (0.8–4.3 mm h1) and relatively short rainfall duration (5 to 22 h, with three of the four having rainfall durations

<10 h). Stemflow volumes increased with Pg, maximum hourly intensity and rainfall duration, although significant correlations between all three rainfall metrics complicate assessing their relative influence on SF. Nevertheless, SF as a fraction of Pg was positively correlated with Pg for 16 of 18 trees (exceptions trees 2-2 and 5-1), indicating a greater fraction of Pg was translated to SF in larger events. Fig. 6 shows the mean SF volume for each stand expressed relative to a 0.2 m wide ring around the bole of tree 4 (used for SWC measurements). Total mean SF volume or depth across all stands was not correlated with DBH, tree height or projected canopy area. Stemflow depths and volumes did not support the hypothesized change with tree age; instead, mean stand SF increased with c (Fig. 7) and LAI. Stemflow also tended to be greater at stands with larger canopy depth; thus, the stand with the largest SF volume (4) had the greatest mean length of live canopy in addition to the second-largest c and LAI values. Mean RP height was significantly larger than that of other species at four of the five stands with a range of tree species (stand 1 was 100% RP), which might

784

J.M. Buttle et al. / Journal of Hydrology 517 (2014) 777–790

Fig. 6. Throughfall (TF), stemflow (SF, expressed relative to 0.2 m wide ring around bole) measured on each sampling day, and cumulative change in soil water (SW) storage relative to storage on Day of Year 171 (June 19): stand 1 (a), stand 2 (b), stand 3 (c), stand 4 (d), stand 5 (e) and stand 6 (f).

lead to greater interception of Pg and more SF from the RP compared to stands where the heights of RP and other tree species were similar. This was supported by results from stand 2, which had the smallest SF of any stand (Fig. 6), no significant difference in the mean heights of RP and other species, as well as the lowest

RP c and LAI values amongst all stands (Table 1). All stands exhibited threshold relationships between Pg and SF. The minimum Pg that generated measurable SF varied from 0.4 to 2.3 mm, with no obvious relationships with stand characteristics.

J.M. Buttle et al. / Journal of Hydrology 517 (2014) 777–790

Fig. 7. Total mean stemflow (SF) vs. canopy cover (c) for the study stands. Line of best fit is significant at p < 0.05.

4.4. Soil water content and storage during the study period Near-surface (0.1 m depth) SWC showed little response to TF and SF inputs, either at 0.1 m or 1.5 m from the bole (Figs. 8 and 9). Instead, maximum SWC was measured at depth in all stands. Pulses of SWC increases and declines of several days duration were observed at depth at many stands, implying vertical and lateral transfer of SF along tree roots and storage in the surrounding soil. These pulses could also be partly a function of changes in texture with depth. For example, stands 1, 4 and 6 had large changes in silt content with depth (Fig. 5), suggesting some of the increases in SWC reflect detention of infiltrating water above textural discontinuities in the profile. Total soil water stored in the upper 1 m of soil was similar at 0.1 m and 1.5 m distances from bole throughout the 2012 study period for the two youngest stands; however, soil water storage was consistently greater at 1.5 m for the remaining stands (data not shown). Cumulative soil water storage (Fig. 6) exhibited less temporal variability at 0.1 m compared to 1.5 m from the bole at all stands. This was particularly the case at stand 4, and suggests soil water storage at 0.1 m distance was sustained by combined SF and TF inputs to a greater extent than at 1.5 m where TF was the sole input to the soil surface. Soil water storage at the end of the study period (DOY 290, October 16) at 1.5 m showed the greatest decline relative to storage on DOY 171 (June 19) at stands with the largest Ic:Pg ratios (3:37 mm, 4:107 mm, 1:39 mm, Table 3). This may reflect lower net input of water relative to the other stands, where soil water storage at the end of the period was within ±20 mm of the value on DOY 171. All stands showed a marked decrease in soil water storage for DOY 171–201 (Fig. 6). Total TF and SF (expressed as a depth over a 0.2 m wide ring around the bole of tree 4 at each stand) for this period ranged from 19 mm and 26 mm (stand 4) to 47 mm and 52 mm (stand 6). All stands exhibited similar responses to large TF and SF inputs in excess of increases in soil water storage (e.g. DOY 201–208, DOY 249–253), suggesting SWR below 1 m depth. Stands 1, 2, 3, 4 and 6 saw their largest TF inputs during DOY 201–208 while stands 3, 4 and 6 also had their largest SF inputs at this time, much of this going into storage. 4.5. Irrigation experiment No overland flow was observed during the experiment, and all SF infiltrated the soil around the bole. Greatest increases in SWC at 0.1 m from the bole relative to SWC prior to the irrigation were in stands receiving the largest SF application (Fig. 10). Changes in

785

SWC at 0.2 and 0.3 m depths generally exceeded those at 0.1 m, suggesting SF bypassed the upper profile via the root network. Appreciable (P0.04 m3 m3) increases in SWC at 0.4 m depth were seen at stands with >10 L of SF application (1, 4). There was little change in SWC at 0.6 m and 1 m depths 24 h after application, indicating retention of applied water in the upper 1 m of the profile. The maximum increase in water depth in the upper 1 m of soil ranged from 10 mm (stand 5) to 35 mm (stand 4). Irrigation did not produce any changes in SWC at 1.5 m from the bole at any depth. Stands with the largest increases in SWC at depth at 0.1 m from the bole also saw substantial declines in SWC within 24 h of water application (e.g. Fig. 10a). Thus, Figs. 8 and 9 give an incomplete picture of short-term SWC dynamics at the stands, given that measurements in 2012 were often made more than 24 h following rainfall. Stands in Figs. 7 and 8 showing relatively persistent periods of increased SWC more than several days in duration (e.g. 1, 4, 6) likely experienced soil water retention above textural discontinuities, as noted above. The width of the annular ring needed to give a SF depth at the soil surface equal to the maximum increase in water depth stored in the upper 1 m of soil following SF application ranged from 0.11 m (stand 4) to 0.28 m (stand 5), and averaged 0.2 m (Table 4). This average width was used to estimate SWR around the bole and compared to SWR at 1.5 m where TF is the only water input. Total mean SF volume in the 2012 study period as an equivalent depth over this 0.2 m wide ring averaged 69 mm (±50 mm) across the stands, ranging from 15 mm (2) to 145 mm (4). These depths expressed relative to total Pg and TF at each stand averaged 17% (SD = 13%) and 20% (SD = 12%), respectively. 4.6. Soil water recharge Soil water recharge (Eq. (2)) differed considerably between stands for a given distance from the bole (Table 5). Recharge at 0.1 m from the bole ranged from 126 mm (stand 5) to 285 mm (stand 4) and averaged 189 mm (±57 mm), while SWR at 1.5 m ranged from 84 mm (stand 5) to 239 mm (stand 4) and averaged 148 mm (±52 mm). Average total SWR 1.5 m from the bole across all stands was 37% of Pg averaged across all stands (398 mm ± 23 mm), while average SWR 0.1 m from the bole across all stands was 47% of Pg. The proportion of total SWR 0.1 m from the bole supplied by SF ranged from 10% (stand 2) to 51% (stands 1 and 4) and averaged 33% (±16%). All but one stand showed greater SWR near the tree bole than at 1.5 m, and differences in total SWR between the 0.1 m and 1.5 m locations ranged from 9 mm (stand 3) to 94 mm (stand 1). We used a modified Z-score (Winkler and Hays, 1975) to compare SWR at 0.1 m (SWR0.1m) and 1.5 m (SWR1.5m):

Z¼

SWR0:1m  SWR1:5m SESWR1:5m

ð2Þ

where SESWR1:5m is the SE of the estimated SWR 1.5 m from the bole. Only the SWR values at the youngest stand (1) differ significantly at 0.1 and 1.5 m (p < 0.1, one-tailed test). 5. Discussion 5.1. Interception and stemflow in the GF stands Ratios of Ic to Pg exceeded values reported for RP in the GF (Buttle and Farnsworth, 2012) and elsewhere (e.g. Voigt, 1960; Helvey, 1971; Mahendrappa, 1990). The former result likely reflects Buttle and Farnsworth’s (2012) greater estimate of TF at the stand level (including canopy gaps between tree rows) using rain gauges, leading to smaller Ic compared with this study where

786

J.M. Buttle et al. / Journal of Hydrology 517 (2014) 777–790

Fig. 8. Interpolated vertical profiles of soil water content (m3 m3) 0.1 m from the bole of tree 4 in each of the stands during the 2012 study period.

TF was measured by a trough beneath a single tree. Troughs may be more effective than groups of gauges in overcoming microscale variations in TF delivery to the forest floor (e.g. Crockford and Richardson, 1990a); thus, we believe our trough estimates of TF under a single tree in each stand are realistic. Few studies report the standard deviation associated with mean SF depth or volume (Levia and Frost, 2003); nevertheless, previous work (Carlyle-Moses and Price 1999, 2006; Levia et al., 2010, in press; Van Stan and Levia, 2010) indicates a range in the CV of total SF from 0.35 (Levia et al., 2010, three trees sampled) to 1.82 (Van Stan and Levia, 2010, 15 trees sampled) with a mean of 0.93 (±0.48). The relatively small CVs for total SF in the GF stands (Table 3) highlight the similar ability of trees in a given stand to generate SF, which is perhaps unsurprising given the small range of DBH for monitored trees in a given stand (Table 2). It also supports our assumption that the mean SF depths for the three trees

monitored for SF are a reasonable estimate of SF for the fourth tree in each stand that was monitored for SWC. Total SF for the study period was a small proportion of Pg, similar to SF:Pg for RP in the GF (Buttle and Farnsworth, 2012) and central New Brunswick (Mahendrappa, 1990) but an order of magnitude less than SF:Pg for RP in the eastern US (Voigt, 1960; Rogerson and Byrnes 1968; Helvey, 1971). The hypothesized decrease in SF:Pg with increasing stand age was not supported. Stemflow as a fraction of Pg was greatest for the youngest stand (1); however, there was no consistent change in SF:Pg with increasing stand age. Stemflow increased with tree size in American beech and yellow poplar (Liriodendron tulipifera L.) in the eastern USA (Levia et al., 2010) and with DBH in forest communities in northeastern Mexico (Nàvar, 2011); however, tree size characteristics were inadequate in explaining differences in SF between trees of the same species at the stand scale in other studies (e.g. Ford and

J.M. Buttle et al. / Journal of Hydrology 517 (2014) 777–790

787

Fig. 9. Interpolated vertical profiles of soil water content (m3 m3) 1.5 m from the bole of tree 4 in each of the stands during the 2012 study period.

Deans, 1978; Lostau et al., 1992). Instead, the relationship between SF and c (Fig. 6) suggests the latter exerts a greater control on the partitioning of Pg to SF in RP stands in the GF than changes in branch angle and bark roughness with increasing tree age. This is supported by explanations of inter-tree differences in SF in terms of canopy volume or area (Martinez-Meza and Whitford, 1996) or total dry biomass (Levia et al., in press). Management practices such as thinning and pruning can also significantly affect water partitioning between TF, SF and Ic, and must be considered when predicting how this partitioning changes as RP stands age and shift to mixed hardwood-conifer stands. Thinning would increase both the height of residual RP relative to regenerating hardwoods and the open areas around these residual RP, which may increase rainfall reaching the tree and contributing to SF (cf. Crockford and Richardson, 1990b). Thus, thinning may counteract the possible effects of tree age on SF generation in the GF.

5.2. Stemflow infiltration near the tree bole The irrigation experiment showed that SF influenced a relatively small area of soil adjacent to the bole. Despite rapid application of the maximum SF volume measured in a given stand in 2012, SWC at 1.5 m from the bole and at depths P0.6 m at 0.1 m from the bole did not increase in any stand. Our results support the cylindrical infiltration model near the tree bole of Tanaka et al. (1996), and are consistent with dye tracing results (e.g. Martinez-Meza and Whitford, 1996; Schwärzel et al., 2012) indicating infiltration of SF in a circular area around the bole. The estimated radius of the annular area of SF infiltration (average of 0.2 m) also agrees with values from empirical (e.g. Voigt, 1960; Pressland, 1976; Tanaka et al., 1996; Nàvar, 2011) and modelling studies (Chang and Matzner, 2000). The equivalent depth of SF applied to this annular area of infiltration can represent a significant fraction of incident

788

J.M. Buttle et al. / Journal of Hydrology 517 (2014) 777–790

Fig. 10. Soil water profiles prior to and following water application at stands 4 (a – 18.1 L of applied stemflow) and 2 (b – 2.5 L of applied stemflow).

Table 5 Summary of water inputs, changes in storage in upper 1 m of soil profile and estimated soil water recharge (SWR) for the DOY 171–DOY 290 period in 2012.

a

Stand

Pg (mm)

TF (mm)

SF (mm)a

Total change in soil water storage at 0.1 m access tube (mm)

Total change in soil water storage at 1.5 m access tube (mm)

SWR (± estimated error) at 0.1 m access tube (mm)

SWR (± estimated error) at 1.5 m access tube (mm)

1 2 3 4 5 6

368 417 408 417 369 409

230 299 233 252 255 304

109 15 43 145 27 76

24 6 16 8 13 2

39 2 37 107 2 19

215 ± 103 155 ± 91 152 ± 63 285 ± 143 126 ± 59 199 ± 90

121 ± 58 136 ± 80 161 ± 67 239 ± 120 84 ± 39 144 ± 65

Obtained by dividing total mean SF volume by area of 0.2 m wide ring around tree bole.

precipitation, and averaged 17% (SD = 13%) in the GF. This agrees with SF:Pg for similar annular areas around RP (12.1%; Voigt, 1960) and Japanese red pine (7.8–20.5%; Tanaka et al., 1996). 5.3. Uncertainty in soil water recharge estimates Assumptions underlying use of Eq. (2) to estimate SWR introduce considerable uncertainty in both absolute SWR values in a given stand and relative differences with distance from the bole. Previous studies report increasing (Voigt, 1960; Ford and Deans, 1978; Herwitz, 1987), decreasing (Johnson, 1990; Beier et al., 1993; Whelan et al., 1998), and inconsistent relationships between TF and proximity to the tree bole (Helvey and Patric, 1965; Lostau et al., 1992; Keim et al., 2005). Given the lack of consensus on this issue, we feel our assumption of uniform TF with distance from the bole in estimating SWR is appropriate. A second source of uncertainty is the E term, derived as a fraction of estimated PET (Hamon model) and applied to all locations at all stands. While the Hamon model can perform well relative to other temperature- and radiation-based PET models (e.g. Lu et al., 2005) and the estimated AET:PET ratio in 2012 in the GF (0.84) is in the range reported for forest landscapes (e.g. Ladekarl, 1998), the latter was based on measurements in 2009. The AET:PET ratio in forests varies between years (e.g. Gholz and Clark, 2002; Stoy et al., 2006), and the actual ratio in the GF in 2012 is unknown. Nevertheless, while differences from the 2009 value would affect absolute SWR values, they would not affect relative differences in SWR either between or within stands. The assumption of uniform E across stands is more problematic. Evapotranspiration may change with tree age, in response to changes in LAI (Gholz and Clark, 2002) and possibly greater utilization of radiant energy and advective heat and enhanced air turbulence with increasing

stand height (Douglass, 1966). However, there was no systematic change in tree height or LAI with stand age in the GF (Table 1), and work elsewhere (e.g. Gholz and Clark, 2002; Delzon and Lostau, 2005) indicates AET may be similar across the RP stands studied here. This suggests our assumption of uniform AET across stands is reasonable, particularly given the relatively short distances between study stands. Inter-stand differences in soil evaporation and transpiration were partly captured by our reduction of AET by stand Ic; nevertheless, future efforts to estimate SWR would benefit from direct measurement of these fluxes. We also assume E is the same at 0.1 m and 1.5 m from the bole in a given stand. Horizontal roots of RP can extend more than 6 m from the bole, reaching their maximum extent within the first 20 years of growth (Fayle, 1975). This implies that water uptake from the soil by RP roots is similar at the two distances from the bole, and is consistent across the range of stand ages studied here. Assessment of SF contributions to SWR at the 0.1 and 1.5 m from the bole is complicated by the relatively large error associated with SWR estimates. The SE values for the terms in Eq. (2) were conservative estimates based on a combination of observed and assumed variability. Further effort is needed to reduce the magnitude of these errors, such as by sampling more trees to reduce the SE of SF or through direct measurement of AET in stands of different ages. Alternative approaches to estimating SWR, such as using tracers (e.g. Redding and Devito, 2008), should also be applied at proximal and distal locations relative to the bole to assess the role of SF in soil water and groundwater recharge. 5.4. Stemflow contributions to soil water recharge As with SF, there was no consistent change in either absolute or relative SF contributions to SWR with increasing stand age. The

J.M. Buttle et al. / Journal of Hydrology 517 (2014) 777–790

only significant (p < 0.1) difference between SWR at 0.1 m and 1.5 m distances was at the youngest stand. Nevertheless, SF fluxes tend to increase SWR below 1 m depth near the bole relative to distal locations, and there was a strong correlation between total SF and total SWR 0.1 m from the bole, whether SF was expressed as a volume (r = 0.96, p = 0.003) or as a depth over the annular area around the bole (r = 0.95, p = 0.003). Thus, SF plays a significant role in recharge in forest stands in the GF that is increased where tree characteristics such as canopy cover, bark roughness, branching geometry and relative exposure to rainfall enhance SF generation. Greater recharge near the bole relative to distal areas beneath the canopy has been noted previously. Taniguchi et al. (1996) estimated SF proportions of total recharge of 10.9–19.1% using a chloride mass balance in Japanese red pine, while Tanaka et al.’s (1996) modelling of SF contributions to recharge in the same stands gave values of 9.1– 22.9%. These are similar to the SF contributions to SWR at 0.1 m from the bole estimated here. However, Chang and Matzer’s (2000) simulated SF contributions to soil water fluxes in a mixed beech/oak stand in Germany estimated 1470 mm y1 of recharge below 0.6 m depth within 0.2 m of the bole and 278 mm y1 beyond that distance. Their relatively greater recharge near the bole compared to our findings may reflect larger SF contributions in the beech/oak forest (5% of Pg) compared to RP in the GF. This is an important issue in the GF, where RP stands are gradually transitioning to a mixture of hardwoods and conifers. Buttle and Farnsworth (2012) found that hardwood stands in the GF had SF fluxes up to an order-of-magnitude larger than those from 57 to 62 year-old RP stands. This suggests the following temporal sequence for SF contributions to SWR in the GF: relatively large SF fluxes in young (<40 y old) RP stands leading to significantly greater SWR near the bole relative to distal areas beneath the canopy ? no clear differences in SF fluxes from RP stands ranging in age from 40 to 80 years but slightly greater SWR near the bole relative to distal areas ? partitioning of Pg to SF in mixed hardwood-conifer stands equal to or exceeding that in young RP stands, with potentially greater SWR near the bole relative to distal areas. The hypothesized role of enhanced SF fluxes on SWR beneath the canopy in this last stage remains to be tested; nevertheless, Schwärzel et al. (2012) suggest a similar temporal trajectory of SF generation as German spruce plantations with minor SF fluxes are replaced by mixed spruce-beech forests which partition a much larger fraction of Pg to SF. 6. Conclusions Monitoring of rainfall partitioning in a chronosequence of red pine stands in a managed forest did not support the hypothesized reduction in SF fluxes with increasing tree age. Instead, canopy cover as well as stand thinning and tree pruning exerted a greater control on differences in SF delivery to the forest floor between trees of increasing age. Nevertheless, SF contributions to soil surrounding the tree bole generally led to greater recharge of soil water below 1 m depth compared to recharge at 1.5 m from the bole, although this difference was most pronounced in the youngest stand examined. Greater SF fluxes for hardwoods in the GF imply that the role of SF in enhancing soil water and groundwater recharge near the tree bole will increase as red pine stands transition to a mixed hardwood-conifer forest. This has important implications for water availability to trees and for biogeochemical processes in the root zone, as well as for soil water and groundwater recharge in the GF and on this part of the ORM. Acknowledgements This study was supported by the Natural Sciences and Engineering Research Council of Canada. Thanks to the Ganaraska Region

789

Conservation Authority (GRCA) for permission to conduct research within the GF, to R. Penwell of the GRCA for providing forest inventory data, and to G. Buttle for assistance with data collection. We thank three anonymous reviewers for their constructive comments on an earlier version of this paper. References Barnett, P.J., Sharpe, D.R., Russell, H.A.J., Brennand, T.A., Gorrell, G., Kenny, F., Pugin, A., 1998. On the origin of the Oak Ridges Moraine. Can. J. Earth Sci. 35, 1152– 1167. Beier, C., Hanson, K., Gunderson, P., 1993. Spatial variability of throughfall fluxes in a spruce forest. Environ. Poll. 81, 257–267. Buttle, J.M., 2011. Streamflow response to headwater reforestation in the Ganaraska River basin, southern Ontario, Canada. Hydrol. Proc. 25, 3030–3041. Buttle, J.M., Farnsworth, A.G., 2012. Measurement and modeling of canopy water partitioning in a reforested landscape: the Ganaraska Forest, southern Ontario, Canada. J. Hydrol. 466–467, 103–114. Carlyle-Moses, D.E., Price, A.G., 1999. An evaluation of the Gash interception model in a northern hardwood stand. J. Hydrol. 214, 103–110. Carlyle-Moses, D.E., Price, A.G., 2006. Growing season stemflow production within a deciduous forest of southern Ontario. Hydrol. Proc. 20, 3651–3663. Carter, M.R., Ball, B.C., 1993. Soil porosity. In: Carter, M.R. (Ed.), Soil Sampling and Methods of Analysis. Lewis Publishers, Boca Raton, FL, pp. 581–588. Chang, S.-C., Matzner, E., 2000. The effect of beech stemflow on spatial patterns of soil solution chemistry and seepage fluxes in a mixed beech/oak stand. Hydrol. Proc. 14, 135–144. Crockford, R.H., Richardson, D.P., 1990a. Partitioning of rainfall in a eucalypt forest and pine plantation in southeastern Australia: I throughfall measurement in a eucalypt forest: effect of method and species composition. Hydrol. Proc. 4, 131– 144. Crockford, R.H., Richardson, D.P., 1990b. 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. Hydrol. Proc. 4, 169–188. Davidson, D.A., 1978. Science for Physical Geographers. Arnold, London, 187p. Delzon, S., Lostau, D., 2005. Age-related decline in stand water use: sap flow and transpiration in a pine chronosequence. Agric. For. Meteor. 129, 105–119. Dingman, S.L., 2002. Physical Hydrology, second ed. Prentice-Hall, Upper Saddle River, NJ, 646p. Douglass, J.E., 1966. Effects of species and arrangement of forests on evapotranspiration. In: Sopper, W.E., Lull, H.W. (Eds.), Forest Hydrology. Pergamon, New York, NY, pp. 451–461. Durocher, M.G., 1990. Monitoring spatial variability of forest interception. Hydrol. Proc. 4, 215–229. Fayle, D.C.F., 1975. Extension and longitudinal growth during the development of red pine root systems. Can. J. For. Res. 5, 109–121. Ford, E.D., Deans, J.D., 1978. The effects of canopy structure on stemflow, throughfall and interception loss in a young Sitka spruce plantation. J. Appl. Ecol. 15, 905–917. Funk, G., 1977. Geology and Water Resources of the Bowmanville, Soper and Wilmot Creeks IHD Representative Drainage Basin. Water Resources Report 9a, Water Resources Branch, Ontario Ministry of the Environment, Toronto, ON, 113p. Ganaraska Region Conservation Authority. 2008. Tier 1 Water Balance and Stress Assessment. Ganaraska Region Conservation Authority, Port Hope, ON, 146p. Germer, S., 2013. Development of near-surface perched water tables during natural and artificial stemflow generation by babassu palms. J. Hydrol. 507, 262–272. Gholz, H.L., Clark, K.L., 2002. Energy exchange across a chronosequence of slash pine forest in Florida. Agric. For. Meteor. 112, 87–102. Greenwood, W.J., Buttle, J.M., 2014. Effects of reforestation on near-surface saturated hydraulic conductivity in a managed forest landscape, southern Ontario. Canada. Ecohydrol. 7, 45–55. Helvey, J.D., 1971. A summary of rainfall interception by certain conifers of North America. In: Biological Effects in the Hydrological Cycle. Proceedings of the 3rd International Seminar for Hydrology Professors, Lafayette, IN, pp. 103–113. Helvey, J.D., Patric, J.H., 1965. Canopy and litter interception of rainfall by hardwoods of eastern United States. Water Resour. Res. 1, 193–206. Herwitz, S.R., 1987. Raindrop impact and water flow on the vegetative surfaces of trees and the effects on stemflow and throughfall generation. Earth Surf. Proc. Landforms 12, 425–432. Hoffman, D.W., Acton, C.J., 1974. The Soils of Northumberland County. Report No. 42, Ontario Soil Survey Research Branch, Agriculture Canada and the Ontario Agricultural College, Ottawa, ON, 117p. Iida, S., Tanaka, T., Sugita, M., 2005. Change of interception process due to the succession from Japanese red pine to evergreen oak. J. Hydrol. 326, 166–180. Johnson, M.S., Lehmann, J., 2006. Double-funneling of trees: stemflow and rootinduced preferential flow. Ecoscience 13, 324–333. Johnson, R.C., 1990. The interception, throughfall and stemflow in a forest in highland Scotland and the comparison with other upland forests in the U.K. J. Hydrol. 118, 281–287. Keim, R.F., Skaugset, A.E., Weiler, M., 2005. Temporal persistence of spatial patterns in throughfall. J. Hydrol. 314, 263–274.

790

J.M. Buttle et al. / Journal of Hydrology 517 (2014) 777–790

Ladekarl, U.L., 1998. Estimation of the components of soil water balance in a Danish oak stand from measurements of soil moisture using TDR. For. Ecol. Manage. 104, 227–238. Levia, D.F., Frost, E.E., 2003. A review and evaluation of stemflow literature in the hydrologic and biogeochemical cycles of forested and agricultural ecosystems. J. Hydrol. 274, 1–29. Levia, D.F., Van Stan II, J.T., Mage, S.M., Kelley-Hauske, P.W., 2010. Temporal variability of stemflow volume in a beech-yellow poplar forest in relation to tree species and size. J. Hydrol. 380, 112–120. Levia, D.F., Michalzik, B., Näthe, K., Bischoff, S., Richter, S., Legates, D.R., in press. Differential stemflow yield from European beech saplings: the role of individual canopy structure metrics. Hydrol. Proc. Liang, W.-L., Kosugi, K., Mizuyama, T., 2009. A three-dimensional model of the effect of stemflow on soil water dynamics around a tree on a hillslope. J. Hydrol. 366, 62–75. Liang, W.-L., Kosugi, K., Mizuyama, T., 2011. Soil water dynamics around a tree on a hillslope with or without rainwater supplied by stemflow. Water Resour. Res. 47. http://dx.doi.org/10.1029/2010WR009856. Lostau, D., Berbigier, P., Granier, A., Hadj Moussa, F., 1992. Interception loss, throughfall and steamflow in a maritime pine stand. I. Variability of throughfall and stemflow beneath the pine canopy. J. Hydrol. 138, 449–467. Lu, J., Sun, G., McNulty, S.G., Amatya, D.M., 2005. A comparison of six potential evapotranspiration methods for regional use in the southeastern United States. J. Am. Water Resour. Assoc. 41, 621–633. Mahendrappa, M.K., 1990. Partitioning of rainwater and chemicals into throughfall and stemflow in different forest stands. For. Ecol. Manage. 30, 65–72. Martinez-Meza, E., Whitford, W.G., 1996. Stemflow, throughfall and channelization of stemflow by roots in three Chihuahuan desert shrubs. J. Arid Environ. 32, 271–287. Nàvar, J., 2011. Stemflow variation in Mexico’s northeaster forest communities: its contribution to soil moisture content and aquifer recharge. J. Hydrol. 408, 35– 42. Pressland, A.J., 1976. Soil moisture redistribution as affected by throughfall and stemflow in an arid zone shrub community. Aust. J. Bot. 24, 641–649. Redding, T.E., Devito, K.J., 2008. Lateral flow thresholds for aspen forested hillslopes on the Western Boreal Plain, Alberta, Canada. Hydrol. Proc. 22, 4287–4300. Rogerson, T.L., Byrnes, W.R., 1968. Net rainfall under hardwoods and red pine in central Pennsylvania. Water Resour. Res. 4, 55–57. Schwärzel, K., Ebermann, S., Schalling, N., 2012. Evidence of double-funneling effect of beech trees by visualization of flow pathways using dye tracer. J. Hydrol. 470–471, 184–192.

Smith, J.A., 1993. Precipitation. In: Maidment, D.R. (Ed.), Handbook of Hydrology. McGraw-Hill, New York, NY, pp. 3.1–3.47. Soil Classification Working Group, 1998. The Canadian System of Soil Classification. Agriculture and Agri-Food Canada, Ottawa, ON. Publ. 1646. (Revised). Stoy, P.C., Katul, G.G., Siqueira, M.B.S., Juang, J.-Y., Novick, K.A., McCarthy, H.R., Oishi, A.C., Uebelherr, J.M., Kim, H.-S., Oren, R., 2006. Separating the effects of climate and vegetation on evapotranspiration along a successional chronosequence in the southeastern US. Global Change Biol. 12, 2115–2135. Tanaka, T., 2011. Effects of the canopy hydrologic flux on groundwater. In: Levia, D.F., Carlyle-Moses, D., Tanaka, T. (Eds.), Forest Hydrology and Biogeochemistry: Synthesis of Past Research and Future Directions. Springer, Dordrecht, Germany, pp. 499–518. Tanaka, T., Taniguchi, M., Tsujimura, M., 1996. Significance of stemflow in groundwater recharge. 2: a cylindrical infiltration model for evaluating the stemflow contribution to groundwater recharge. Hydrol. Proc. 10, 81–88. Taniguchi, M., Tsujimura, M., Tanaka, T., 1996. Significance of stemflow in groundwater recharge. 1: evaluation of the stemflow contribution to recharge using a mass balance approach. Hydrol. Proc. 10, 71–80. Tiessen, H., Moir, H.O., 1993. Total and organic carbon. In: Carter, M.R. (Ed.), Soil Sampling and Methods of Analysis. Lewis Publishers, Boca Raton, FL, pp. 187– 199. Van Stan II, J.T., Levia, D.F., 2010. Inter- and intraspecific variation of stemflow production from Fagus grandifolia Ehrh. (American beech) and Liriodendron tulipifera L. (yellow poplar) in relation to bark microrelief in the eastern United States. Ecohydrology 3, 11–19. Voigt, G.K., 1960. Distribution of rainfall under forest stands. For. Sci. 6, 2–10. Voigt, G.K., Zwolinski, M.J., 1964. Absorption of stemflow by bark of young red and white pines. For. Sci. 10, 277–282. Weiss, M., Baret, F., 2010. Can-Eye Version 6.1 User Manual. EMMAH (Mediterranean environment and agro-hydro system modelisation) laboratory, French National Institute of Agricultural Research (INRA, Avignon, 47p. Whelan, M.J., Sanger, L.J., Baker, M., Anderson, J.M., 1998. Spatial patterns of throughfall and mineral ion deposition in a lowland Norway spruce (Picea abies) plantation at the plot scale. Atmos. Environ. 32, 3493–3501. Winkler, R.L., Hays, W.L., 1975. Statistics: Probability, Inference, and Decision, second ed. Holt, Rinehart and Winston, New York, NW, 889p+appendices. Zhang, Y., Chen, J.M., Miller, J.R., 2005. Determining digital hemispherical photograph exposure for leaf area index estimation. Agric. For. Meteorol. 133, 166–181.