Relation of soil-, surface-, and ground-waterdistributions of inorganic nitrogen withtopographic position in harvested andunharvested portions of an aspen-dominatedcatchment in the Boreal Plain

M.L. Macrae, K.J. Devito, I.F. Creed, and S.E. Macdonald

Abstract: Spatial distributions of soil extractable nitrate (NO3−) and ammonium (NH4

+) concentrations were related tosurface- and ground-water NO3

− and NH4+ concentrations in harvested and forested sections of a catchment dominated

by trembling aspen (Populus tremuloides Michx.) in the subhumid boreal forest of Alberta, Canada. NO3− and NH4

+ con-centrations in soils varied spatially throughout the catchment and were larger in surface soils than in subsurface soils.Spatial distributions of soil inorganic nitrogen (N) concentrations were not explained by the harvested versus the unhar-vested condition; heterogeneity was instead related to topographic position. NO3

− concentrations in both surface andsubsurface soils were largest in ephemeral draws and wetlands. NH4

+ concentrations in subsurface soils were largest inephemeral draws and wetlands, but this pattern was not apparent for surface soils. Soil NO3

− and NH4+ availability and

surface- and ground-water NO3− and NH4

+ concentrations reflected soil NO3− and NH4

+ concentrations. N-rich surfacesoils in both forested and harvested areas have a large potential for releasing N to surface waters. This study indicatesthat even though topography is subtle in this catchment, topographic position and its soil moisture relations, along withvegetation demand, can influence N transformation and transport in both forested and harvested portions of the BorealPlain landscape.

Résumé : La distribution spatiale des concentrations de nitrate (NO3−) et d’ammonium (NH4

+) extractibles du sol a étéreliée aux concentrations de NO3

− et NH4+ dans les eaux de surface et souterraine dans les sections boisées et coupées

d’un bassin dominé par le peuplier faux-tremble (Populus tremuloides Michx.) dans la forêt boréale sub-humide del’Alberta, au Canada. La concentration de NO3

− et NH4+ dans le sol variait spatialement partout dans le bassin et elle

était plus élevée dans le sol de surface que sous la surface. La distribution spatiale de la concentration d’azote (N)inorganique n’était pas expliquée par les conditions associées à la coupe ou à l’absence de coupe et l’hétérogénéitéétait plutôt reliée à la position topographique. La concentration de NO3

− dans le sol, tant en surface que sous la surface,était la plus élevée dans les ravines temporaires et les zones humides. La concentration de NH4

+ dans le sol sous la sur-face était la plus élevée dans les ravines temporaires et les zones humides mais ce n’était pas le cas dans le sol en sur-face. La disponibilité de NO3

− et NH4+ dans le sol et la concentration de NO3

− et NH4+ dans les eaux de surface et

souterraine reflétaient la concentration de NO3− et NH4

+ dans le sol. Les sols de surface riches en N, tant dans les zonesboisées que coupées, ont une grande capacité de libérer du N dans l’eau de surface. Cette étude indique que, même sila topographie est peu contrastée dans ce bassin, la position topographique et sa relation avec l’humidité du sol, combi-nées à la demande créée par la végétation, peuvent influencer la transformation et le transport de N, tant dans les por-tions boisées que coupées du paysage de la plaine boréale.

[Traduit par la Rédaction] Macrae et al. 2103

Introduction

There has been a dramatic increase in industrial exploita-tion of forested lands in the mixedwood Boreal Plain of the

western boreal forest in Canada (Alberta EnvironmentalProtection 1998). Harvesting of trembling aspen (Populustremuloides Michx.), which has become an important com-mercial source of fibre in recent decades, is extensive and

Can. J. For. Res. 36: 2090–2103 (2006) doi:10.1139/X06-101 © 2006 NRC Canada

2090

Received 1 October 2005. Accepted 5 April 2006. Published on the NRC Research Press Web site at http://cjfr.nrc.ca on 17 August2006.

M.L. Macrae1,2 and K.J. Devito. Department of Biological Sciences, University of Alberta, Edmonton, AB T6G 2G8, Canada.I.F. Creed. Department of Biology, University of Western Ontario, London, ON N6A 5B7, Canada.S.E. Macdonald. Department of Renewable Resources, University of Alberta, Edmonton, AB T6G 2H1, Canada.

1Corresponding author (e-mail: mmacrae@fes.uwaterloo.ca).2Present address: Department of Geography, University of Waterloo, Waterloo, ON N2L 3G1, Canada.

may impact soil properties and processes in the boreal for-ests of Alberta (Schmidt et al. 1996; Westbrook and Devito2002). To predict, and potentially mitigate, the impact ofaspen harvesting on surface water quality in this subhumidregion we need to understand the natural variability of nutrientdynamics in aspen forests and the effects of forest harvestingupon this.

Little is known about nitrogen (N) dynamics in aspen forestsin subhumid climates. Our understanding of these dynamicscomes largely from humid climates on the Boreal Shield,where precipitation exceeds potential evapotranspiration onan annual basis (Buttle et al. 2000, 2005). Studies from for-ests in humid climates have shown that timber harvestingmay increase soil moisture and temperature, owing to reducedplant uptake and shading; this in turn may lead to increasedmineralization and nitrification of organic N (Covington1981; Hornbeck et al. 1986; Martin et al. 2000). Severalstudies have reported large NO3

− concentrations in runofffollowing a harvest (e.g., Likens et al. 1970; Wiklander1981), while in other studies, only small increases in NO3

−

export have been observed (Vitousek and Melillo 1979;Krause 1982; Miller and Newton 1983). When NO3

− concen-trations in stream flow have been elevated following harvest,they have generally returned to pre-harvest conditions withinseveral years of the harvest as the vegetation regenerates(Pierce et al. 1972; Bormann et al. 1974; Martin et al. 2000).

The influence of harvesting aspen in the subhumid BorealPlain, where precipitation equals or is less than potentialevapotranspiration on an annual basis, may be less than isobserved in more humid regions (Swank and Johnson 1994).In subhumid climates, soil moisture storage capacity maynot be exceeded in many years, resulting in infrequenthydrologic flushing, with significant surface runoff occur-ring once every 10–20 years (Devito et al. 2005a). Infre-quent flushing of near-surface soils may result in a buildupof nutrients and thus little difference between harvested andforest portions of a catchment, as has been observed forphosphorus (Macrae et al. 2005). Furthermore, NH4

+ and (or)NO3

− may be quickly taken up by aspen, which regrows veryrapidly by vegetative suckering following harvesting (West-brook and Devito 2002). This could actually lead to a reduc-tion in soil moisture and temperature due to shading andtherefore reduced mineralization of organic N and subse-quent nitrification. Although Carmosini et al. (2003) mea-sured higher gross N mineralization rates in harvested aspenstands the first year following harvesting, those high gross Nmineralization rates were balanced by high rates of N immo-bilization by microbes and plants, leading to a lack of differencein rates of net N mineralization and nitrification betweenharvested and forested aspen stands. In contrast, other stud-ies that have been conducted in the subhumid regions of theBoreal Plain have shown higher rates of net N mineraliza-tion and larger concentrations of inorganic N in soils of har-vested forests (Walley et al. 1996).

The potential effects of harvesting on N pools and fluxesin soils can be masked by the natural heterogeneity of Ndynamics in a catchment. Heterogeneity in soil moisture lev-els, N pools, and rates of N transformations, including min-eralization, nitrification, and denitrification, have all beenfound to vary with topography (Zak et al. 1986; Groffman

and Tiedje 1989; Creed et al. 2002). Furthermore, differenttopographic features (e.g., uplands versus wetlands) havebeen found to have distinct roles as either sources or sinksof N as a result of differences in soil moisture level or vege-tation or soil characteristics (e.g., Hill 1996; Devito et al.1999; Price et al. 2005). Differences in transformation anddistribution of N with topographic position within harvestedand forested (uncut) areas within a catchment may confoundinterpretation of potential effects of harvesting on soil Ndynamics. It is therefore important to describe the influenceof topographic position and the corresponding soil and mois-ture characteristics on N dynamics and to discriminatebetween these and anthropogenic sources of variation in Ndynamics in harvested catchments.

For both humid and subhumid regions, what is knownabout the natural variability of N dynamics in runoff and thepotential effects of timber harvesting on these dynamics hasbeen derived from studies that have focused on catchmentrunoff rather than the distribution of surface and groundwaters that may (or may not) contribute to catchment runoff(e.g., Schiff et al. 2002). Recently, Devito et al. (2005b)highlighted the dominance of soil storage and verticalgroundwater fluxes that regulate potential evapotranspirationin hydrologic budgets of the subhumid climates typical oflarge portions of the Boreal Plain. When horizontal ground-water fluxes occur along the land-to-lake continuum,groundwater (rather than surface water) fluxes may domi-nate hydrologic and nutrient export (Devito et al. 2000;Evans et al. 2000). The effects of harvesting on surface-water or groundwater N dynamics on the Boreal Plain arenot well documented(Buttle et al. 2005).

We examined the influence of timber harvesting on soilNO3

− and NH4+ concentrations and groundwater and surface-

water NO3− and NH4

+ concentrations in a topographicallysubtle aspen-dominated catchment in the subhumid BorealForest of northern Alberta. The hydrology and phosphorusdynamics of the catchment have been reported previously(Devito et al. 2005a; Macrae et al. 2005). Significant naturalvariability in precipitation and runoff regimes occurredthroughout the study period (Devito et al. 2005a), providingthe opportunity to contrast hydrologic and biogeochemicalcycling in relatively dry, wet, and average hydrologic years.The objectives of this study were to answer the followingquestions: (i) Does topographic position control the distribu-tion of soil NO3

− or NH4+ concentrations? (ii) Are differences

in soil NO3− and NH4

+ concentrations between forested andharvested portions of the catchment larger than among topo-graphic units? and (iii) Is there a relation between the distri-bution of soil extractable inorganic N concentrations andsoil-, surface-, and ground-water and runoff NO3

− and (or)NH4

+ concentrations in forested or harvested portions of thecatchment?

Materials and methods

Study siteThis study was conducted in a 53 ha catchment (LLB20)

draining into Moose Lake, which is about 250 km northeastof Edmonton, Alberta, in the subhumid climatic region ofnorth-central Alberta (55.1°N, 113.8°W) (Devito et al. 2005a;

© 2006 NRC Canada

Macrae et al. 2091

Macrae et al. 2005). The regional climate is characterized bywarm summers and cold winters, with mean January andJuly temperatures of –16.7 and 16.3 °C, respectively. Long-term mean total precipitation is 462 mm·year–1, with >60%occurring between June and August as short, localized con-vective storms (Devito et al. 2005c). Long-term mean annualpotential evapotranspiration (538 mm·year–1) exceeds pre-cipitation in most years (Environment Canada 2002). Totalprecipitation from 1997 to 2001 was 486, 329, 318, 529, and444 mm·year–1 (Devito et al. 2005a). In the first year postharvest (1997), soils were saturated following a very wetyear and there was significant annual runoff (251 mm·year–1).In contrast, runoff was 12, 2, 2, and 1 mm in 1998, 1999,2000, and 2001, respectively.

LLB20 is a first-order catchment with three subcatchmentscontaining ephemeral streams that drain into a minerotrophicmixedwood wetland prior to draining into Moose Lake (Fig. 1).Well-drained uplands are dominated by Gray Luvisols andEutric Brunisols, while poorly drained low-lying areas aredominated by Gleysols and organic soils (Carmosini et al.2003; Devito et al. 2005a). Soil profiles in uplands consistof forest floor (litter–fibric–hemic layers; ~10 cm) and Ahorizons. The A horizons contain a mixture of organic mate-rial and fine sand and silt. Below 20 cm, the B horizons con-

sist of sandy clays that extended to at least 130 cm depth inmost areas of the catchment. Soil profiles in low-lying areas,ephemeral draws, and wetlands contain an organic (peat)horizon (~30 cm) underlain by mineral horizons containinga mixture of sandy clays and organic material. Tremblingaspen is the dominant tree species, with small amounts ofwhite spruce (Picea glauca (Moench) Voss) in the southeastportion of the catchment. Balsam poplar (Populusbalsamifera L.) and paper birch (Betula papyrifera Marsh.),with some black spruce (Picea mariana (Mill.) BSP) andtamarack (Larix laricina (Du Roi) K. Koch), occur inephemeral draws and wetlands.

Approximately 50% of the study catchment (Fig. 1) washarvested in winter, when the ground was frozen, as part ofthe Terrestrial and Riparian Organisms, Lakes and Streamsresearch project (Prepas et al. 2001). Timber harvesting oc-curred in January 1997 in portions of the three subcatch-ments within the study catchment. Two subcatchments were>70% harvested and a small portion (<35%) of the upperreaches of the third subcatchment was harvested. The forestin the extreme upper reaches and the lower portion of thecatchment, including the minerotrophic mixedwood wetlandin its lower reaches, was not harvested (Fig. 1). Trees werecut by feller buncher and logs were skidded to nearby roads

© 2006 NRC Canada

2092 Can. J. For. Res. Vol. 36, 2006

Harvesting Haul Road

Harvested Section

Soil-sampling Transect (Upland)

Soil-sampling Transect (Low-lying)

Soil-sampling Transect (Wetlands and Draws)

Weir

Ephemeral Stream

Forested

Forested

Harvested

N

0 m 100 m 200 m 300 m 400 m

Subcatchment C

Subcatchment B

Subcatchment A

Subcatchment D

Fig. 1. Map of subcatchment LLB20 of the Moose Lake watershed, showing the harvested section, harvesting haul roads, ephemeralstreamsline), and soil-sampling transects.

© 2006 NRC Canada

Macrae et al. 2093

and landings. Only one main harvesting haul road was con-structed, which ran roughly midway through the harvestedblock perpendicular to the slope (Fig. 1). Logging slash andherbaceous growth were not removed from harvested areas.Rapid forest regeneration occurred in the harvested blocks,with suckers of trembling aspen exceeding 30 cm in heightby the end of the 1997 growing season and 1 m in height bythe end of the 1998 growing season (Devito et al. 2005a).

Soil samplingTwenty sampling transects of 100 m length were estab-

lished along topographic positions with a range of slopes,including uplands (upper- and mid-slope positions), low-lying areas (low-slope positions and flats), and ephemeraldraws and wetlands in both forested and harvested por-tions of the catchment (Fig. 1). Along each transect, fivelocations (20 m apart) were sampled at two depths. In up-lands, low-lying areas, and ephemeral draws, the surfacesample consisted of the litter–fibric–hemic layer, whilethe subsurface sample consisted of the upper mineral soil(0–10 cm depth of the A and B horizons). In wetlands, thesoils were predominantly organic and samples were col-lected at 0–10 and 10–20 cm depths. Soil samples werecollected on all transects in July 1998 and June and Au-gust in 1999. Of the original 20 transects, one was sam-pled in 1998 only and two were relocated during thestudy. Soil cores were collected on a subset of 10transects in June 2000 and seven transects in June 2001.

Soil samples were analyzed for moisture content, bulkdensity, and KCl-extractable NO3

− and NH4+ concentrations.

Soil-core samples were placed in bags and hand-mixed untilvisually homogenized, and a subsample of approximately5 g (dry mass) was removed. Subsamples were extracted in50 mL of 2M KCl for colorimetric analysis of NO3

− andNH4

+ using a Technicon Autoanalyser (Technicon IndustrialSystems 1973a, 1973b). The remaining core sample wasdried to determine gravimetric moisture content and deter-mine bulk density using the dimensions of the core sampler.All moisture contents are reported as volumetric moisturecontent (m3·m–3).

To determine NO3− and NH4

+ availability, ion exchangeresin bags were placed at the interface of the litter–fibric–hemic and mineral soil, or at 10 cm depth in organic soils,within 0.30 m of soil-extraction cores at five locations alongeight of the 27 transects in 1998 (after Binkley and Matson1983). Resin bags contained 45 mL of Amberlite IRN-150(Sigma Chemicals) mixed-bed resin (25 mmol/L) enclosedin nylon-stocking bags and were charged by soaking in50 mL of 1 mol/L NaCl three times. After 4 weeks’ incuba-tion, the resins were extracted by shaking with 100 mL of2 mol/L KCl for 90 min and filtered through a WhatmanNo. 42 filter. Extractants were subsequently analyzed forNH4

+ and NO3− using a Technicon Autoanalyser (Technicon

Industrial Systems 1973a, 1973b).

Surface water and groundwaterThe monitoring of surface runoff, saturated areas, and

groundwater is detailed in Devito et al. (2005a). Runoff wassampled at weirs that were installed in the ephemeral streamsin the catchment. Surface water was sampled from a network

of standing pools that formed throughout the study. Lysimetersto capture shallow subsurface flow were installed in the top10 cm of the soil at the base of hill slopes and near the out-lets of subcatchments A (forested) and B and C (harvested).Lysimeters were constructed from 50 mm i.d. ABS pipesscreened between 0 and 10 cm beneath the ground surfaceand sealed at the bottom with PVC caps. Groundwater wassampled from piezometers and wells. Piezometers were con-structed from PVC pipe (19 mm i.d.) coupled to 0.2 m diameterslotted and shielded PVC heads. Piezometers were insertedinto prebored holes at 0.5 m depth intervals to 4 m depth, oruntil the confining material was intercepted, and then sealedwith bentonite clay over the interval from the ground surfaceto the top of the screened portion to ensure that no surfacewater entered the piezometer. Wells were inserted into pre-bored holes extending into the underlying confining materialat approximately 2–4 m depth and backfilled with aggregate(Devito et al. 2005a).

Significant surface saturation and surface runoff occurredonly in 1997 and early 1998 (Devito et al. 2005a). Standingwater that collected in shallow depressions and drawsthroughout the catchment was sampled from April to June in1998 and 1999. Surface water was collected using a coarsefilter, and care was taken not to disturb the sediments. Sur-face samples from near wells and groundwater samples werecollected approximately every 2 weeks from May to Octoberin 1997 and 1998. Samples were collected between Apriland September six times in 1999 and three times in 2000and 2001. Wells and piezometers were purged the day beforesampling. Piezometers were sampled with a tube attached toa Nalgene Erlenmeyer flask and hand vacuum pump or to asyringe with a three-way valve chain, and wells were sam-pled with a Nalgene bailer attached to a stainless-steel chain(Evans et al. 2000). All containers were prewashed with10% HCl and prerinsed with distilled water and sample. Allsamples were kept on ice for 12–48 h until prepared foranalysis. Samples were filtered through 0.45 µm Milliporemembrane filters and stored at 4 °C. NO3

− and NH4+ concen-

trations were determined using a Technicon Autoanalyser(Technicon Industrial Systems 1973a, 1973b).

StatisticsBecause of the non-parametric nature of the data, tempo-

ral and spatial variability in soil NO3− and NH4

+ concentra-tions from 1998 to 2001 was examined using box-and-whisker plots. The spatial analyses of interactions betweenharvested and forested treatments and topographic featureswere conducted using the non-parametric equivalent of thetwo-way ANOVA (Sheirer–Ray–Hare test; Sokal and Rohlf1995). For the Sheirer–Ray–Hare test, the five sampled siteson each transect were classified into one of three slopeclasses or topographic positions: uplands (slope >5°), low-lying flats (slope 2–5°), and ephemeral draws and wetlands(slope <2°). Soil moisture content was found to vary amongthese topographic classes in a previous study (Macrae et al.2005). If there was a significant difference (p < 0.05) in soilNO3

− or NH4+ concentrations among the three topographic

classes in the harvested versus forested areas, then a Mann–Whitney U test was conducted to identify the source of thissignificant difference.

Results and discussion

Distribution of soil extractable NO3− and NH4

+

Soil NO3− and NH4

+ concentrations were highly variablewithin the catchment. For all sampling dates and locations,the median NO3

− concentration in the surface soil layer(0.34 µg N·g–1, ranging from 0 to 92.00 µg N·g–1, with oneoutlier reaching 410.70 µg N·g–1) was double that of thesubsurface soil layer (0.11 µg N·g–1, ranging from 0 to92.82 µg N·g–1). In contrast, the median soil NH4

+ concentra-tion in the surface soil layer (21.19 µg N·g–1, ranging from 0to 319.18 µg N·g–1) was about 15 times higher than that inthe subsurface soil layer (1.96 µg N·g–1 (ranging from 0 to204.24 µg N·g–1). The observed differences in soil NO3

− andNH4

+ concentrations between mineral and organic soils andwith depth were consistent with those found in other studiesof boreal aspen forest soils covering subhumid to humid cli-mates (e.g., Huang and Shoenau 1996; Carmosini et al. 2003).

Measurements made in July 1998 and June and August of1999 showed no significant differences in surface and sub-surface soil NO3

− and NH4+ concentrations between harvested

and forested portions of the subcatchments (Table 1, Fig. 2).This trend held for the 5-year period covering a wet and drycycle (Fig. 2).

Soil NO3− concentrations were related to topographic posi-

tion (Fig. 2), and were smallest in uplands and largest inephemeral draws and wetlands. This is likely an effect oftopographically driven soil moisture, given that soil moistureis related to the slope of the topographic position in thiscatchment (Macrae et al. 2005). The Sheirer–Ray–Hare testshowed significant differences in both surface and sub-surface soil NO3

− concentrations among topographic classes(Table 1). Mann–Whitney U tests (Table 2) indicated signifi-cant differences between the three topographic classes (p <0.05), with the smallest concentrations in the upland areas,intermediate concentrations in the low-lying areas, and larg-est concentrations in the ephemeral draws and wetlands(Fig. 2).

The observed pattern of soil moisture levels and topogra-phy and soil NO3

− concentrations (Fig. 2) can be expected inheadwater catchments in subhumid or semi-arid climates, asannual water deficits can result in wetlands and draws thatdrain regularly (Devito et al. 2005a, 2005b). Wetlands anddraws may represent areas where moisture is sufficient fornitrification and buildup of soil NO3

− but over extended peri-

ods may not be enough to facilitate denitrification. This con-trasts with findings from the eastern boreal forest, where soilNO3

− concentrations were small throughout a catchment andsmallest in wetlands because of waterlogged conditions,which inhibit nitrification and promote denitrification (Hilland Shackleton 1989; Devito et al. 1999). However, wettingand drying periods can shift N transformation from denitri-fication to nitrification and result in large NO3

− concentra-tions in soils and streams with rewetting in humid climates(Devito and Dillon 1993; Creed et al. 1996; Hill and Devito1997; Ohrui and Mitchell 1998).

Surface-soil NH4+ concentrations tended to be similar

among the topographic classes (Tables 1 and 2), whereassubsurface-soil NH4

+ concentrations varied with topographicposition (Tables 1 and 2). In subsurface soils there was nosignificant difference between upland and low-lying areasbut there was a significant difference (p < 0.000) betweenthese two topographic classes and ephemeral draws and wet-lands (Table 2, Fig. 2).

In more humid eastern areas, soil NH4+ concentrations are

expected to be smaller in ephemeral draws and wetlandsbecause of the higher immobilization rates, with high C:Nratios and waterlogged soils causing the lower mineraliza-tion rates that characterize these topographic features (Pat-rick 1982; Devito et al. 1999; Ohrui et al. 1999; Westbrookand Devito 2002). However, in this subhumid region, flush-ing does not occur and the wet–dry cycles that occur inephemeral draws and wetlands can raise gross NH4

+ mineral-ization rates and lower NH4

+ immobilization rates duringrewetting of soils, thereby increasing soil NH4

+ concentra-tions in these locations following a hydrologic event (Devitoet al. 1999). In general, soil moisture levels are below ornear the critical threshold (i.e., 25%) above which mineral-ization can occur. NH4

+ concentrations are larger in sub-surface soils of wetlands and ephemeral draws than inupland subsoils because deep organic deposits are present inwetlands but not in mineral subsoils in upland areas (Devitoand Dillon 1993).

In general, longer term temporal trends in soil NO3− and

NH4+ concentrations corresponded to year-to-year weather

variability rather than to time following harvest (Fig. 2), asindicated by the lack of a significant interaction betweentreatment and topographic position (Table 1). The absence ofa change in soil NO3

− and NH4+ concentrations following har-

vest suggests either that there was no change in N mineral-

© 2006 NRC Canada

2094 Can. J. For. Res. Vol. 36, 2006

NO3− concentration NH4

+ concentration

Depth Source Sum of squares F p Sum of squares F p

Surface horizons Treatment 342 0.02 0.894 3 349 0.2 0.677Position 2 12 224 11.0 0.004 11 709 0.6 0.739Treatment × position 48 131 2.5 0.286 82 256 4.3 0.119

Deeper soil Treatment 4 995 0.2 0.631 1 078 0.05 0.824Position 2 62 593 12.1 0.002 3 49 435 16.0 <0.001Treatment × position 28 994 1.3 0.513 36 592 1.7 0.432

Note: The analysis was based on soil NO3− and NH4

+ concentrations measured at five sampling sites on 17 transects established and sampled on eachsampling date in 1998 (July) and 1999 (June, August).

Table 1. Results of a Sheirer–Ray–Hare test (non-parametric equivalent of two-way ANOVA) examining the effects of treatment (har-vested vs. forested), topographic position (upland vs. lowland vs. ephemeral draw and wetland), and their interaction on soil NO3

− andNH4

+ concentrations at two depths (surface and subsurface layers).

ization with harvest, or that differences in N mineralizationwere offset by rapid microbial immobilization and (or) as-pen regrowth. Previous studies have shown that at least onegrowing season must pass before appreciable increases in Nmineralization occur following a harvest (Kimmins 2004),but that increases in N mineralization are limited to the first

few years following a harvest because plant regrowth causesthe microclimate at the soil surface to quickly return topreharvest conditions (Marks 1974; Covington 1981). Ourresults suggest that any increase in soil NO3

− and NH4+ con-

centrations which occurs after a harvest is followed by adecline within a year, likely because of rapid regeneration of

© 2006 NRC Canada

Macrae et al. 2095

1.0

0.8

0.6

0.4

0.2

0.0

150

100

50

0

150

100

50

0

Upland

Areas

Low-lying

Areas

Mois

ture

Ephemeral Draws

and WetlandsJuly

1998

20

10

0

20

10

0

(a)

(b)

(c)

(d)

(e)

June

1998

Aug. 1

998

July

2000

July

2001

July

1998

June

1998

Aug. 1

998

July

2000

July

2001

July

1998

June

1998

Aug.1998

July

2000

July

2001

nd

nd

nd

nd

nd nd

nd

32

NH

(µg

N·g

)4+

–1

NH

(µg

N·g

)4+

–1

NO

(µg

N·g

)3–

–1

NO

(µg

N·g

)3–

–1

(m·m

)3

–3

Fig. 2. Box-and-whisker plots of soil moisture level (a) and soil extractable NO3− (b and c) and soil extractable NH4

+ (d and e) concen-trations in soils in upland areas, low-lying areas, and ephemeral draws and wetlands from 1998 to 2001 in the forested (shaded) andclear-cut (open) sections of the catchment. Soil NO3

− and NH4+ concentrations are given for the forest floor (b and d, respectively) and

subsurface soil layer (c and e, respectively). The 25th, 50th, and 75th percentiles are indicated by boxes and the 10th and 90th percen-tiles by whiskers. Soil NO3

− and NH4+ concentrations are based on all available data (i.e., five sampling sites on 20 transects in 1998

(n = 100), 19 transects in 1999 (n = 95 in June and August), 10 transects in 2000 (n = 50), and 7 transects in 2001 (n = 35)); “nd”denotes sites and sampling dates where samples were not collected.

aspen and (or) microbial immobilization (Carmosini et al.2003).

Spatial variation in NO3− availability, as indicated by ion

exchange resin bags (Binkley and Matson 1983), showed nodifference between the harvested and forested portions of thecatchment, but there were substantial differences along thetopographically driven moisture gradient between organi-cally rich wetlands and ephemeral draws and mineral soilsof upland and low-lying slopes (Fig. 3). Spatial trends inNO3

− supply rates suggest low availability in uplands andlow-lying areas but high availability in ephemeral draws andwetlands. This pattern is consistent with the higher soil NO3

−

pools observed in ephemeral draws and wetlands (Fig. 2),suggesting that the observed spatial differences in soil NO3

−

concentration are due to differences in NO3− production in

repeatedly exposed deeper organic soils (Hill and Devito1997) rather than to NO3

− immobilization or vegetation up-take.

Spatial trends in NH4+ availability, as indicated by ion ex-

change resin bags, showed substantial differences betweenthe harvested and forested sections in upland portions of thecatchment, but no differences along the topographicallydriven moisture gradient (Fig. 3). In harvested areas, higherNH4

+ availability (Fig. 3) combined with lower soil NH4+

concentrations in upland locations relative to other topo-graphic positions within the catchment suggests greater veg-etation uptake or immobilization. Although this patterncould also be explained by leaching of NH4

+ from uplandareas, this is unlikely given the large soil moisture storagepotential and low runoff rates observed in this region(Devito et al. 2005a) and the fact that there is no differencein soil moisture in surface soils between the forested andharvested sections of the catchment (Macrae et al. 2005).The work of Carmosini et al. (2002) in a nearby catchmentshowed that gross mineralization rates were higher in har-vested areas than in forested areas, but these were matchedby high immobilization rates, resulting in lower net N min-eralization rates in harvested aspen stands. Although immo-bilization rates are high, uptake of available NH4

+ by aspensuckers may also explain the lack of difference in soil NH4

+

concentrations between forested and harvested areas, whichsuggests that in the absence of aspen suckers, the increased

© 2006 NRC Canada

2096 Can. J. For. Res. Vol. 36, 2006

Upland vs. low-lying Upland vs. wetlandsLow-lyingvs. wetlands

Upland + low-lyingvs. wetlands

NO3− NH4

+ NO3− NH4

+ NO3− NH4

+ NO3− NH4

+

Surface horizons Mann–Whitney U 2830 3121 1281 2698 1830 2636 3111 5334Wilcoxon’s W 6070 6862 4521 5938 5571 6377 16972 19195Z –2.00 –1.03 –6.12 –0.95 –4.64 –1.87 –6.15 –1.63Asymtotic significance

(two-tailed)0.046 0.303 0.000 0.343 0.000 0.062 0.000 0.104

Deeper soil Mann–Whitney U 3550 3714 1680 1069 1508 1130 3187 2198Wilcoxon’s W 7645 7809 5775 5164 5603 5225 19477 18488Z –1.46 –0.96 –5.58 –7.55 –6.17 –7.35 –6.71 –8.48Asymtotic significance

(two-tailed)0.145 0.336 0.000 0.000 0.000 0.000 0.000 0.000

Table 2. Results of a Mann–Whitney U test of variability in soil NO3− and NH4

+ concentrations (µg N·g–1) among topographic positionswithin the catchment.

(a)

(b)

Draws

and Wetlands

Low-Lying

Areas

Upland

Areas

µg

NO

3-N

·mL

resin

µg

NH

4-N

mL

resin

10

0

135

125

100

75

50

25

0Draws

and Wetlands

Low-Lying

Areas

Upland

Areas

Na

va

ilab

ility

Na

va

ilab

ility

((

))

·

Fig. 3. Box-and-whisker plots of soil N availability (from ionexchange resin bag incubations) in upland areas, low-lying areas,and ephemeral draws and wetlands in the forested (shaded) andclear-cut (open) sections of the catchment. The 25th, 50th, and75th percentiles are indicated by boxes and the 10th and 90thpercentiles by whiskers.

NH4+ in soils following harvesting could be released to run-

off during wet periods. A recent review by Schimel andBennett (2004) also emphasizes the importance of soil organicN as a source of N for both vegetation and microbes. A keyquestion that emerges from this is whether soil organic Npools change in harvested areas. Carmosini (2000) did notobserve differences in soil total organic N between the har-vested and forested sections of a nearby catchment. In thepresent study, levels of dissolved organic N in surface waterand groundwater are high in both forested and harvested ar-eas (Fig. 4). However, in light of changing views of N cy-cling (Schimel and Bennett 2004), distributions of organic Nbetween forested and harvested areas should be consideredin future studies.

N patterns in surface water and runoffOrganic N concentrations in surface water and groundwa-

ter are large (528–4570 µg N·L–1) and are often the domi-nant form of total dissolved N in this catchment (>85% insurface soils, 27%–75% in mineral subsoils). Althoughorganic N has been shown to increase (e.g., Qualls et al.2000; Nieminen 2004) or decrease (Hannam and Prescott2003) following clear-cutting, the influence of harvesting ondissolved organic N in surface runoff and groundwater incatchment LLB20 are difficult to determine because of thehigh variability and generally high values throughout (Fig. 4).

NO3− and NH4

+ concentrations in surface water and runoffclosely matched soil distributions. Similar to patterns ob-served in soils, NO3

− and NH4+ concentrations in lysimeters

capturing runoff at the forest floor – mineral soil interfaceon low-lying slopes were highly variable but did not differbetween the forested and harvested sections of the catchment(Fig. 5). One lysimeter in the harvested section of the catch-ment had much larger NH4

+ concentrations than other lysi-meters in both the harvested and forested sections of thecatchment. This lysimeter was likely located in an area with

elevated production of NH4+ (i.e., a hot spot) and the high

values observed are a result of natural spatial variability.The occurrence of such areas (hot spots) is apparent frompatterns in surface runoff throughout the catchment in boththe harvested and the forested areas (Figs. 6 and 7).

Surface runoff occurred in 1997 and 1998 but little wasobserved in 1999–2001. As the catchment dried out in 1998,NO3

− (Fig. 6) and NH4+ (Fig. 7) concentrations in standing

water in ephemeral draws and wetlands during 1997and 1998 were variable and high, often exceeding 500 and1000 µg·L–1, respectively, but did not differ noticeably be-tween the harvested and forested areas. This is in contrastto the results of work in other forest regions, where severalstudies have shown increases in NO3

− in runoff following aharvest (Likens et al. 1970; Vitousek and Melillo 1979;Krause 1982; Miller and Newton 1983).

NO3− and NH4

+ concentrations at the catchment outflow(Fig. 8) were similar to those observed in the lysimeters(Fig. 5) and most surface-saturated areas (Figs. 6 and 7),although concentrations at the outflow were not as large asthose observed in some hot spots throughout the catchment.

© 2006 NRC Canada

Macrae et al. 2097

0

1000

2000

3000

4000

5000

HarvestedForested HarvestedForested

Forest Floor andShallow Soils (0-10 cm)

Deeper Soils (10-300 cm)

Dis

solv

ed

Org

anic

Nitro

gen

(µg·L

)–

1

Fig. 4. Box-and-whisker plots of dissolved organic N concentra-tions in surface water (forest floor and 0–10 cm depth in soil)and groundwater (10–300 cm depth in soil) in the harvested andforested sections of the catchment. The 25th, 50th, and 75th per-centiles are indicated by boxes and the 10th and 90th percentilesby whiskers. Outliers are indicated by circles.

0

20

40

60

80

100

120

Forested Harvested

0

2000

4000

6000

8000

10 000

Forested Harvested

(a)

(b)

NH

co

nce

ntr

atio

n4+ (µ

gN

·L)

–1

NO

co

nce

ntr

atio

n3– (µ

gN

·L)

–1

Fig. 5. Box-and-whisker plots of NO3− and NH4

+ concentrations insoil water collected from lysimeters in the forested and harvestedsections of the catchment, based on sampling between August1997 and June 1998. No water was observed in lysimeters afterJune 1998. The 25th, 50th, and 75th percentiles are indicated byboxes and the 10th and 90th percentiles by whiskers. Outliersare indicated by circles.

© 2006 NRC Canada

2098 Can. J. For. Res. Vol. 36, 2006

It appears that in 1998 there may have been more NH4+ mov-

ing from forested (versus harvested) areas, although soilconcentrations do not indicate a difference between the har-vested and forested sections of the catchment. This may beattributed to the organization of soil-NH4

+ -rich sections ofthe catchment with respect to the connectivity of surfaceflow in forested areas compared with harvested areas duringthis time period (Devito et al. 2005a).

Overall, our results indicate that the potential for N re-lease to receiving waters is similar in harvested and forestedareas during wet conditions and this is attributable to N-richsurface soils in the study catchment. However, in most of thestudy years, drainage networks were not well connected andsubsurface flow paths were predominant, thereby limiting Nexport to receiving streams and lakes in most years (Devitoet al. 2005a).

N patterns in groundwaterLike those in soil extracts, NO3

− and NH4+ concentrations

in groundwater influenced by soils over a range of depthswere variable but showed no consistent differences betweenharvested and forested areas (Tables 3 and 4). There washigh temporal variability in NO3

− concentration in groundwa-ter, with consistently higher NO3

− concentration pulses(>500 µg N·L–1) observed in ephemeral draws and wetlandswith organically rich soils (Table 3). The elevated concen-trations were likely the result of water-table drawdown,where conditions became oxic and higher rates of nitrifica-tion occurred (Kalef 2002).

There was large temporal variability in NH4+ concentration

in shallow and deep groundwater among topographic posi-tions and in emerging springs, with elevated concentrationsobserved at all depths in both harvested and tacked forests(Table 4). Consistently large minimum NH4

+ concentrations(>300 µg N·L–1) were observed at depth in wetland organicsoils and below springs, which occurred largely in the for-ested rather than the harvested portion of the catchment.Thus, natural spatial variability in soil characteristics within

N

0 m 100 m 200 m 300 m 400 m

Forested

Harvested

600 g N·Lµ –1

100 µg N·L–1

50 µg N·L–1

300 g N·Lµ –1

Fig. 6. Range of NO3− concentrations in surface water in the harvested and forested sections of the catchment. Minimum (solid circles)

and maximum (open circles) concentrations are given for each location; the results are based on data collected in 1998.

© 2006 NRC Canada

Macrae et al. 2099

the catchment may confound the comparison of harvestedand forested areas because it appears to override the effectsof harvesting.

Although soil N decreased with depth, groundwater Ngenerally did not. This contrasts with observations in easternboreal areas, where there is a strong decrease in concentra-tion with depth, owing to flushing of groundwater NO3

− andNH4

+ at deeper layers by lateral flow and resulting in largerNO3

− and NH4+ concentrations in organic soils at the surface

because of infrequent flushing (e.g., Creed et al. 1996). Thefact that groundwater N did not decrease with depth may re-flect low water movement and (or) that the study catchmentis an area of groundwater recharge (Kalef 2002; Devito et al.2005a), where groundwater flow is vertical to a depth of 2 mand lateral flushing is reduced.

The consistent high concentrations observed in groundwa-ter NH4

+ concentrations with depth have been observed in or-ganic deposits of peatlands on the Boreal Plain (Vitt andBayley 1984) as well as in other wetlands (Patrick 1982;

Devito and Dillon 1993). In the Boreal Plain of Albertathere are accumulations of buried organic compounds andcoal seams together with reworked glacial material. Conse-quently, larger NH4

+ concentrations are observed in deepergroundwater in some areas. Also, some surficial geologicalfeatures in glaciated Alberta have very large NO3

− concentra-tions that occur naturally (Hendry et al. 1984; Fujikawa andHendry 1991). These characteristics, as well as the influenceof groundwater recharge on groundwater N concentrations,may override the relations of groundwater N concentrationwith depth that has been shown in eastern boreal areas. Pre-liminary studies at catchment LLB20 indicate that deepgroundwater may provide an important pathway for higherN concentrations from recharge or buried substrates to betransported within and exported from the catchment (Devitoet al. 2000, 2005a; Kalef 2002).

ConclusionsIn our study, timber harvesting had little net effect on soil

N

0 m 100 m 200 m 300 m 400 m

Forested

Harvested

3000 g N·Lµ –1

750 µg N·L–1

1500 g N·Lµ –1

Fig. 7. Range of NH4+ concentrations in surface water in the harvested and forested sections of the catchment. Minimum (solid circles)

and maximum (open circles) concentrations are given for each location; the results are based on data collected in 1998.

NO3− and NH4

+ concentrations in this aspen-dominated sub-humid catchment. No difference in soil NO3

− and NH4+ con-

centrations was observed between the harvested and forestedsections of the catchment over a 4-year period followingharvesting. Spatial variation in soil NO3

− and NH4+ concen-

trations among topographic features representing a moisturegradient from upland to wetland was much larger than thedifferences between harvested and forested sections. Thelargest soil NO3

− and NH4+ concentrations occurred in

ephemeral draws and wetlands, most likely because of highmobilization rates during wetting and drying of organicallyrich soils (Kalef 2002). The spatial patterns of soil NO3

− andNH4

+ concentrations and soil available NO3− and NH4

+ indi-cate that even though relief is subtle in this catchment, andin mixedwood forests of the Boreal Plain in general, slope ortopographic position and differences in soil organic com-pounds and moisture level influence N pools and their trans-formations and should be considered when sampling to

© 2006 NRC Canada

2100 Can. J. For. Res. Vol. 36, 2006

Wet

land

s

Upl

and

and

low

-lyi

ngar

eas

Eph

emer

aldr

aws

Org

anic

soil

sM

iner

also

ils

Spr

ings

Dep

thin

soil

aT

reat

men

tM

edia

nR

ange

Med

ian

Ran

geM

edia

nR

ange

Med

ian

Ran

geM

edia

nR

ange

Fore

stfl

oor

and

surf

ace

orga

nic

laye

r0–

10cm

H39

1–38

939

2–54

912

2–80

2—

1010

–10

F41

4–95

351–

229

308–

235

——

712–

438

10–1

00cm

H15

5–25

663–

421

——

203–

944

182–

68F

353–

185

603–

585

168–

41—

—34

3–86

100–

200

cmH

103

1–37

414

23–

1776

——

133–

282

125

20–3

14F

401–

427

562–

728

143–

150

71–

56—

—>

200

cmH

266–

6140

2–12

0—

—27

6–77

——

F11

61–

851

105

2–73

6—

——

—11

2–33

a Rep

rese

nts

the

shal

low

est

soil

that

isin

ters

ecte

dan

din

flue

nced

byth

ew

ell

scre

en.

Tab

le3.

NO

3−co

ncen

trat

ions

( µg

N·L

–1)

ingr

ound

wat

ersa

mpl

esco

llec

ted

from

upla

nds,

ephe

mer

aldr

aws,

wet

land

s,an

dsp

ring

sin

harv

este

d(H

)an

dfo

rest

ed(F

)ar

eas

wit

hin

the

catc

hmen

t.

Date

2000

1500

1000

500

0

Date

(a)

(b)

200

150

100

50

0

MA

Y1

99

7

JU

LY

19

97

AU

G.

19

97

SE

PT.

19

97

MA

Y1

99

7

MA

Y1

99

8

JU

NE

19

98

JU

LY

19

98

AU

G.

19

98

SE

PT.

19

98

OC

T.

19

98

MA

Y1

99

9

JU

NE

19

99

AU

G.

19

99

JU

LY

19

97

AU

G.

19

97

SE

PT.

19

97

MA

Y1

99

7

MA

Y1

99

8

JU

NE

19

98

JU

LY

19

98

AU

G.

19

98

SE

PT.

19

98

OC

T.

19

98

MA

Y1

99

9

JU

NE

19

99

AU

G.

19

99

JU

LY

19

99

NH

concentr

ation

4+ (µg

N·L

)–1

(µg

N·L

)–1

NO

concentr

ation

3– (µg

N·L

)–1

(µg

N·L

)–1

Fig. 8. Box-and-whisker plots of NO3− and NH4

+ concentrations inrunoff from subcatchment outflows in the forested (shaded) andharvested (open) sections of the catchment between spring 1997and summer 1999. The 25th, 50th, and 75th percentiles areindicated by boxes and the 10th and 90th percentiles by whiskers.

assess the impacts of timber harvesting on N dynamics inthe Boreal Plain.

N concentrations in soil water, surface water, groundwa-ter, and runoff reflected the extractable inorganic N concen-trations in surface soils, suggesting that there is a largepotential for the release of N to surface runoff should wetconditions occur. Large N concentrations observed in soilsand surface water and groundwater indicate that the potentialfor N export during wet periods is high in both harvestedand tacked forest. However, subhumid climatic conditionsand rapid aspen regeneration limit the potential for surfacerunoff. In dry years, groundwater fluxes may be an impor-tant pathway by which N is exported. Further study of soil Ndynamics and surface water – groundwater interactions onthe Boreal Plain of Alberta is necessary, as hydrological andbiogeochemical cycling in this landscape is not consistentwith patterns observed in more easterly boreal regions.

Acknowledgements

We thank C. Applewaite, D. Bryant, N. Carmosini, G.Eerkes, A. House, N. Kalef, T. Redding, S. Reedyk, G.Tondeleir, and C. Westbrook for technical assistance andfieldwork. This research was funded by an operating grantfrom the Natural Sciences and Engineering Research Councilof Canada to K. Devito, by two grants from Alberta PacificForestry Industries Inc. to K. Devito and I. Creed, a NationalCentre for Excellence – Sustainable Forest ManagementNetwork grant to I. Creed and K. Devito, a National Centrefor Excellence – Sustainable Forest Management Networkgrant to K. Devito, and the Terrestrial and Riparian Organ-isms, Lakes and Streams research project.

References

Alberta Environmental Protection. 1998. The boreal forest naturalregion of Alberta: Edmonton, Alberta, Natural Resources Ser-vice, Recreation and Protected Areas Division, Natural HeritagePlanning and Evaluation Branch, Alberta Environmental Protec-tion, Edmonton, Alta. Binkley, D., and Matson, P.A. 1983. Ionexchange resin bag method for assessing forest soil nitrogenavailability. Soil Sci. Soc. Am. J. 47: 1050–1052.

Bormann, F.H., Likens, G.E., Siccama, T.G., Pierce, R.S., andEaton, J.S. 1974. The export of nutrients and recovery of stableconditions following deforestation at Hubbard Brook. Ecol.Monogr. 44: 255–277.

Buttle, J.M., Creed, I.F., and Pomeroy, J.W. 2000. Advances in Ca-nadian forest hydrology, 1995–1998. Hydrol. Process. 14: 1551–1578.

Buttle, J.M., Creed, I.F., and Moore, R.D. 2005. Advances in Ca-nadian forest hydrology, 1999–2003. Hydrol. Process. 19: 169–200.

Carmosini, N. 2000. Net and gross nitrogen mineralization and ni-trification in upland stands of the mixed-wood Boreal Forest fol-lowing harvesting. M.Sc. Thesis, University of Alberta,Edmonton, Alta.

Carmosini, N., Devito, K.J., and Prepas, E.E. 2002. Gross nitrogentransformations in harvested and mature aspen–conifer mixedforest soils from the Boreal Plain. Soil Biol. Biochem. 34:2949–1951.

© 2006 NRC Canada

Macrae et al. 2101

Wet

land

s

Upl

and

and

low

-lyi

ngar

eas

Eph

emer

aldr

aws

Org

anic

soil

sM

iner

also

ils

Spr

ings

Dep

thin

soil

aT

reat

men

tM

edia

nR

ange

Med

ian

Ran

geM

edia

nR

ange

Med

ian

Ran

geM

edia

nR

ange

Fore

stfl

oor

and

surf

ace

orga

nic

laye

r0–

10cm

H16

05–

1333

576

0–78

0084

1–69

7—

2312

–34

F61

66–

3673

121

3–71

319

94–

975

——

566

2–19

3310

–100

cmH

120

15–2

2629

03–

2360

——

999–

1833

175

11–6

16F

173

9–68

538

828

–169

873

762

3–80

9—

—13

0610

9–20

0110

0–20

0cm

H20

53–

1408

173

3–18

85—

—55

2–13

4925

147

–462

F21

60–

2454

564

2–89

6672

932

3–26

4185

59–

3924

——

>20

0cm

H27

914

–805

274

1–13

73—

—10

125

–523

——

F47

63–

3231

598

39–2

163

——

——

1467

337–

2232

a Rep

rese

nts

the

shal

low

est

soil

that

isin

ters

ecte

dan

din

flue

nced

byth

ew

ell

scre

en.

Tab

le4.

NH

4+co

ncen

trat

ions

( µg

N·L

–1)

ingr

ound

wat

ersa

mpl

esco

llec

ted

from

upla

nds,

ephe

mer

aldr

aws,

wet

land

s,an

dsp

ring

sin

harv

este

d(H

)an

dfo

rest

ed(F

)ar

eas

wit

hin

the

catc

hmen

t.

© 2006 NRC Canada

2102 Can. J. For. Res. Vol. 36, 2006

Carmosini, N., Devito, K.J., and Prepas, E.E. 2003. Net nitrogenmineralization and nitrification in trembling aspen forest soilson the Boreal Plain. Can. J. For. Res. 33: 2262–2268.

Covington, W.W. 1981. Changes in forest floor organic matter andnutrient content following clear cutting in northern hardwoods.Ecology, 62: 41–48.

Creed, I.F., Band, L.E., Foster, N.W., Morrison, I.K., Nicolson,J.A., Semkin, R.S., and Jeffries, D.S. 1996. Relation of nitrate-Nrelease from temperate forests: a test of the N flushing hypothe-sis. Water Resour. Res. 32: 3337–3354.

Creed, I.F., Trick, C.G., Band, L.E., and Morrison, I.K. 2002.Characterizing the spatial pattern of soil carbon and nitrogenpools in the Turkey Lakes watershed: a comparison of regres-sion techniques. Water Air Soil Pollut. (Focus) 2: 81–102.

Devito, K.J., and Dillon, P.J. 1993. The influence of hydrologicconditions and peat oxia on the phosphorus and nitrogen dynam-ics of a conifer swamp. Water Resour. Res. 29: 2675–2685.

Devito, K.J., Westbrook, C.J., and Schiff, S.L. 1999. Nitrogen min-eralization and nitrification in upland and peatland forest soils intwo Canadian Shield catchments. Can. J. For. Res. 29: 1793–1804.

Devito, K.J., Creed, I.F., Rothwell, R.L., and Prepas, E.E. 2000.Landscape controls on phosphorus loading to boreal lakes: im-plications for the potential impacts of forest harvesting. Can. J.Fish. Aquat. Sci. 57: 1977–1984.

Devito, K.J., Creed, I.F., and Fraser, C.J.D. 2005a. Controls onrunoff from an aspen dominated headwater catchment on theBoreal Plain: implications for forest management. Hydrol. Pro-cess. 19: 3–25.

Devito, K.J., Creed, I., Gan, T., Mendoza, C., Petrone, R., Silins,U., and Smerdon, B. 2005b. A framework for broad scale classi-fication of hydrologic response units on the Boreal Plain: Is to-pography the last thing to consider? Invited commentaries. HPToday, Hydrol. Process. 19: 1705–1714.

Devito, K.J., Fraser, C.J.D., and Creed, I.F. 2005c. Climate andrunoff characteristics of a Boreal Plain headwater catchment,Lac La Biche, AB, Canada: implication for forest management.Sustainable Forest Management–NCE report [online]. Availablefrom http://129.128.45.107/english/home/index.htm and http://www.sfm-1.biology.ualberta.ca/english/home/ [accessed 1 August 2005].

Environment Canada. 2002. Canadian climate normals, Meteoro-logical Service of Canada [online]. Available from http://www.climate.weatheroffice.ec.gc.ca/climateData/canada_e.html [accessed20 July 2004].

Evans, J.E., Prepas, E.E., Devito, K.J., and Kotak, B.G. 2000.Phosphorus dynamics in shallow subsurface waters in an uncutand cut sub-catchment of a lake on the Boreal Plain. Can. J.Fish. Aquat. Sci. 57(Suppl. 2): 60–72.

Fujikawa, J.I., and Hendry, M.J. 1991. Denitrification in clayey till.J. Hydrol. 127: 337–348.

Groffman, P.M., and Tiedje, J.M. 1989. Denitrification in northtemperate forest soils: spatial and temporal patterns and thelandscape and seasonal scales. Soil Biol. Biochem. 21: 613–620.

Hendry, M.J., McCready, R.G.L., and Gould, W.D.1984. Distribu-tion, source and evolution of nitrate in a glacial till of southernAlberta, Canada. J. Hydrol. 70: 177–198.

Hill, A.R. 1996. Nitrate removal in stream riparian zones. J. Envi-ron. Qual. 25: 743–755.

Hill, A.R., and Devito, K.J. 1997. Hydrologic–chemical interac-tions in headwater forest wetlands. In Northern forestedwetlands: ecology and management. Edited by C.C. Trettin,M.F. Jurgensen, D.F. Grigal, M.R. Gale, and J. Jeglum.CRC/Lewis Publisher, Boca Raton, Fla. pp. 217–233.

Hill, A.R., and Shackleton, M. 1989. Soil N mineralization and ni-trification in relation to nitrogen in a small forested watershed.Biogeochemistry, 8: 167–184.

Hornbeck, J.W., Martin, C.W., Pierce, R.S., Bormann, F.H., Likens,G.E., and Eaton, J.S. 1986. Clearcutting northern hardwoods: ef-fects on hydrologic and nutrient ion budgets. For. Sci. 32: 667–686.

Huang, W.Z., and Schoenau, J.J. 1996. Forms, amounts and distri-bution of carbon, nitrogen, phosphorus and sulfur in a boreal as-pen forest soil. Can. J. Soil Sci. 76: 373–385.

Kalef, N. 2002. Interlinking hydrological behaviour and inorganicnitrogen cycling in a forested boreal wetland. M.Sc. Thesis,University of Alberta, Edmonton, Alta.

Hannam, K.D., and Prescott, C.E. 2003. Soluble organic nitrogenin forests and adjacent clearcuts in British Columbia, Canada.Can. J. For. Res. 33: 1709–1718.

Kimmins, J.P. 2004. Forest ecology: a foundation for sustainableforest management and environmental ethics in forestry. 3rd ed.Pearson Prentice Hall, Saddle River, N.J.

Krause, H.H. 1982. Nitrate formation before and after clearcuttingof a monitored watershed in central New Brunswick, Canada.Can. J. For. Res. 12: 922–930.

Likens, G.E., Bormann, F.H., Johnson, N.M., Fisher, D.W., andPierce, R.S. 1970. Effects of forest cutting and herbicide treat-ment on nutrient budgets in the Hubbard Brook watershed–ecosystem. Ecol. Monogr. 40: 23–47.

Macrae, M.L., Redding, T.E., Creed, I.F., Bell, W.R., and Devito,K.J. 2005. Soil, surface water and ground water phosphorus re-lationships in a partially harvested Boreal Plain aspen catch-ment. For. Ecol. Manage. 206: 315–329.

Marks, P.L. 1974. The role of pin cherry (Prunus pensylvanica L.)in the maintenance of stability in northern hardwood ecosys-tems. Ecol. Monogr. 44: 73–88.

Martin, C.W., Hornbeck, J.W., Likens, G.E., and Buso, D.C. 2000.Impacts of intensive harvesting on hydrology and nutrient dy-namics of northern hardwood forests. Can. J. Fish. Aquat. Sci.57(Suppl. 2): 19–29.

Miller, J.H., and Newton, M. 1983. Nutrient loss from disturbedforest watersheds in Oregon’s Coast Range. Agroecosystems, 8:153–167.

Nieminen, M. 2004. Export of dissolved organic carbon, nitrogenand phosphorus following clearcutting of three Norway spruceforests growing on drained peatlands in southern Finland. SilvaFenn. 38: 123–132.

Ohrui, K., and Mitchell, M.J. 1998. Spatial patterns of soil nitratein Japanese forested watersheds: importance of the near-streamzone as a source of nitrate in stream water. Hydrol. Process. 12:1433–1445.

Ohrui, K., Mitchell, M.J., and Bischoll, J.M. 1999. Effect of land-scape position on nitrogen mineralization and nitrification in aforested watershed in the Adirondack Mountains of New York.Can. J. For. Res. 29: 497–508.

Patrick, W.H., Jr. 1982. Nitrogen transformations in submergedsoils. In Nitrogen in agricultural soils. Am. Soc. Agron. Monogr.No. 22. pp. 449–465.

Pierce, R.S., Martin, C.W., Reeves, C.C., Likens, G.E., andBormann, F.H. 1972. Nutrient loss from clearcutting in NewHampshire. In Watersheds in Transition: Symposium of theAmerican Water Resources Association, Colorado State Univer-sity, Fort Collins, Co. Edited by S.C. Csallany, T.G. McLaughlin,and W.D. Striffler. American Water Resources Association, Ur-bana, Ill. pp. 285–295.

Prepas, E.E., Pinel-Alloul, B., Planas, D., Methot, G., Paquet, S.,and Reedyk, S. 2001. Forest harvest impacts on water quality

and aquatic biota on the Boreal Plain: introduction to theTROLS lake program. Can. J. Fish. Aquat. Sci. 58: 421–436.

Price, J.S., Branfireun, B.A., Waddington, J.M., and Devito, K.J.2005. Advances in Canadian wetland hydrology, 1999–2003.Hydrol. Process. 19: 201–214.

Qualls, R.G., Haines, B.L., Swank, W.T., and Tyler, S.W. 2000.Soluble organic and inorganic nutrient fluxes in clearcut andmature deciduous forests. Soil Sci. Soc. Am. J. 64: 1068–1077.

Schiff, S.L., Devito, K.J., Elgood, R.J. McCrindle, P.M., Spoelstra,J., and Dillon, P. 2002. Two adjacent catchments: dramaticallydifferent nitrate export. Water Resour. Res. 38: 1278–1292.

Schimel, J.P., and Bennett, J. 2004. Nitrogen mineralization: chal-lenges of a changing paradigm. Ecology, 85: 591–602.

Schmidt, M.G., Macdonald, S.E., and Rothwell, R.L. 1996. Im-pacts of harvesting and mechanical site preparation on soilchemical properties of mixed wood boreal forest sites in Al-berta. Can. J. Soil Sci. 76: 531–540.

Sokal, R.R., and Rohlf, F.J. 1995. Biometry. 3rd ed. W.H. Freemanand Co., New York.

Swank, W.T., and Johnson, C.E. 1994. Small catchment research inthe evaluation and development of forest management practices.In Biogeochemistry of small catchments: a tool for environmen-tal research. Edited by B. Molden and J. Cerny. Rep. No. 51,Scientific Committee on Problems of the Environment, JohnWiley & Sons, Chichester, UK. pp. 382–408.

Technicon Industrial Systems. 1973a. Ammonia in water andwastewater. Industrial Method No. 696-82W, Technicon Indus-trial Systems, Tarrytown, N.Y.

Technicon Industrial Systems. 1973b. Nitrate and nitrite in waterand seawater. Industrial Method No. 100-70W, Technicon Indus-trial Systems, Tarrytown, N.Y.

Vitousek, P.M., and Mellilo, J.M. 1979. Nitrate losses from dis-turbed ecosystems: patterns and mechanisms. For. Sci. 25: 605–619.

Vitt, D.H., and Bayley, S.E. 1984. The vegetation and water chem-istry of four oligotrophic basin mires of northwestern Ontario.Can. J. Bot. 62: 1485–1500.

Walley, F.L., van Kessel, C., and Pennock, D.J. 1996. Landscape-scale variability of N mineralization in forest soils. Soil Biol.Biochem. 28: 383–391.

Westbrook, C.J., and Devito, K.J. 2002. Comparative analysis ofthe effects of clearcut harvesting and wildfire on physical andchemical properties of upland forest soil. In Ecological basis forstand management: a synthesis of ecological responses to wild-fire and harvesting. Edited by S.J. Song. Alberta ResearchCouncil Inc., Vegreville, Alta. pp. 4-1to 4-36.

Wiklander, G. 1981. Rapporteur’s comment on clearcutting. In Ter-restrial nitrogen cycles: processes, ecosystem strategies, andmanagement impacts. Edited by F.E. Clark and T. Rosswall.Ecol. Bull. NFR (Naturvetensk. Forstningsradet.), 33: 642–647.

Zak, D.R., Pregitzer, K.S., and Host, G.E. 1986. Landscape varia-tion in nitrogen mineralization and nitrification. Can. J. For.Res. 16: 1258–1263.

© 2006 NRC Canada

Macrae et al. 2103