A New Forest Floor Corer for Rapid Sampling, Minimal … › 4959 › 1332e543390c0ad7... ·...

Nalder and Wein A New Forest Floor Corer for Rapid Sampling, ...

373

A New Forest Floor Corer for RapidSampling, Minimal Disturbance andAdequate Precision

Ian A. Nalder and Ross W. Wein

Nalder, I.A. & Wein, R.W. 1998. A new forest floor corer for rapid sampling, minimaldisturbance and adequate precision. Silva Fennica 32(4): 373–382.

We describe an effective and inexpensive device for sampling forest floors. It is basedon a rechargeable, battery-powered drill that drives a sharpened steel coring tube. Thecorer is simple to fabricate, is lightweight (3.5 kg) and can be used easily by one personto obtain intact, natural volume cores of the forest floor. It has been used extensively toobtain samples in 114 boreal forest stands of western Canada. We found that coefficientsof variation were typically 30 % for forest floor organic matter and bulk density, andtended to be higher in Pinus banksiana stands than in Picea glauca and Populustremuloides. Ten samples per stand gave adequate precision for a study of forest floordynamics and autocorrelation did not appear to be a problem with five-metre samplingintervals. In addition to sampling forest floors, the corer has proven suitable for sam-pling moss and lichen layers and mineral soil down to about 20 cm. A similar poweredsystem can also be used for increment boring of trees.

Keywords corer, forest floor, moss, variability, bulk density, western Canada, borealAuthors' address University of Alberta, Department of Renewable Resources, 442 EarthSciences Building, Edmonton, Alberta, Canada T6G 2E3Fax +1 403 492 1767 E-mail [email protected] 14 May 1998 Accepted 4 August 1998

Silva Fennica 32(4) research notes

1 Introduction

In upland boreal forests, there is little faunalmixing of decomposing litter with the underly-ing mineral soil, and consequently a distinctiveorganic layer soon develops. This comprises L,F and H layers (Agriculture Canada Expert Com-mittee on Soil Survey 1987), which may or maynot be distinctly delineated. For many purposes

it is convenient to treat the three layers as oneand refer to them as the forest floor (Kimmins1987). The depth of the forest floor depends onmany factors, including forest type, climate andstand history, and commonly ranges from 2 to20 cm. The forest floor is important for nutrientcycling (van Cleve et al. 1983, Johnson 1995),for forest fire fuel loads (Van Wagner 1983,Forestry Canada Fire Danger Group 1992), and


374

for its role in regulating soil temperature andmoisture (Bonan and Shugart 1989). It is alsoimportant as a carbon pool, accounting for 28 %of soil carbon storage in non-peatland forests ofFinland (Liski and Westman 1997) and up to onehalf of soil carbon in upland stands in westernCanada (Huang and Schoenau 1996). Yet it isdifficult to characterise because spatial variabili-ty is high (Hokkanen et al. 1995, Liski 1995).Consequently, a high number of samples is need-ed to achieve reasonably precise estimates forvariables such as carbon density or bulk density.

The traditional method of forest floor sam-pling has been to extract a block, typically 30 by30 cm square, by cutting around the perimeter ofa template with a sharp knife, then lifting thisblock away from the mineral soil (e.g., van Cleveand Noonan 1971, Alban 1982, Quintilio et al.1991, Paré and Bergeron 1996, Gower et al.1997). This is time consuming and gives largequantities of forest floor material to process; as aresult, small sample numbers may not capturevariability efficiently. Cylindrical push-corers ofsmaller cross sectional area are not uncommon(e.g., Westman 1995, Zackrisson et al. 1995,Finer et al. 1997, Liski and Westman 1997), buthave two important limitations for forest floorsampling. First, they compress the layer and re-

duce its natural volume making it difficult toobtain reasonable bulk density measurements.Second, the fibric nature of forest floor materialscan cause some material to be pushed away fromthe cutting edge, either inward or outward, whichleads to overestimates or underestimates, respec-tively. This problem is particularly acute withsmall corers, where the ratio of perimeter tocross-sectional area is high. Push-corers also tendto deflect from particularly hard material, suchas roots and pieces of dead wood, resulting in acrooked core and a biased estimate of forestfloor mass. Ideally, a small diameter corer shouldmake a clean, vertical cut through the forestfloor with virtually no disturbance. We have de-veloped such a corer, which is lightweight, inex-pensive to fabricate, and allows cores to be takenmuch faster than traditional template methods.We have used the corer over three field seasonsfor sampling the forest floor, as well as the moss-plus-lichen layer and shallow mineral soil, instands dominated by Picea glauca (Moench)Voss, Populus tremuloides Michx. and Pinusbanksiana Lamb. in western Canada. Here wedescribe the corer and its operation, present dataon sampling variability and discuss other appli-cations.

Fig. 1. Use of corer system to take a forest floor sample:a) coring, b) extracting core onto examining tray.


375

2 Description and Operation

2.1 Design and Construction

The corer system consists of a sharpened lengthof steel water pipe driven by a battery-poweredrotary drill through an adaptor unit (Fig. 1). Thissystem has become feasible with the recent wideavailability of powerful drills driven by recharge-able batteries. Details of the corer and adaptorare shown in Fig. 2. The tubular corer is madefrom standard, mild steel water pipe. We usedtwo-inch pipe (55.9 mm OD, 52.6 mm ID). Theexact diameter is not critical, but larger diameterpipe would require more torque, and battery lifemay be limiting. A pipe length of 30 cm was suit-able for our purposes but it could be shorter orlonger depending on the application. Two holesat the upper end take a quick-connect pin thatcouples the corer to the adaptor. The machinedadaptor unit (Fig. 2) comprises a 9.6 mm hexag-onal shaft, which is clamped in the drill chuck, acylinder whose outer diameter is such that it fitsflush within the corer pipe, and a quick-connectpin that goes transversely though the cylinder.The motive force is provided by a standard, in-dustrial-quality drill, which should have a 12 mm

(1/2-inch) chuck. Because torque and battery lifeare critical, we would recommend that the drillhave a torque rating of at least 34 N-m (300 inch-pounds), low-reduction gearing providing speedsof 350 rpm or less, and a battery of at least 12volts. For most field applications, one or twospare rechargeable batteries will be required.

The last two components of the system are alight-weight core extractor and an examining tray(Fig. 1b). The core extractor is a piece of ABSwater pipe cut slightly longer than the corer andsealed at each end with plastic end caps. Thepipe diameter should be chosen so that the endcap fits snugly into the steel corer pipe. In theNorth American system, a 4.4 cm (1 3/4-inch)ABS pipe works well with the two-inch corer.An examining tray is made from white polyvinylchloride pipe of the same internal diameter andlength as the corer. It is cut in half vertically,making one “half” slightly larger than the other.The longitudinal edges of the larger half are thenflared out by heating and bending the plastic sothat the soil core is securely held in the examin-ing tray while the high sides prevent accidentalloss of material. To aid in measuring layer depths,we suggest drawing lines around the inner cir-cumference at 1-cm intervals.

A critical aspect of the system is a sharp corer.Sharpening can be done with the drill and arotary sanding disc, which is how the edge iskept sharp in the field, but a belt sander (orgrinder) is quicker and easier for the initial shap-ing of the edge. Sand first with coarse grit (#50)and work up to fine grit (#120). The angle of theedge should be 25–30 degrees and the edge mustbe scalloped. A peak-trough depth of 1–2 mmwith a cycle interval (peak-peak distance) of 2cm worked well. At this stage, it is also useful toscribe horizontal lines on the outside of the pipeto indicate the depth of coring (Fig. 2).

Apart from the adaptor, which must be ma-chined, the system can be put together by any-one with a modicum of mechanical skill. Oursystem, including the drill, two rechargeable bat-teries, construction materials and machine shopcharges, cost less than US$500.

Fig. 2. Corer and adaptor unit.

Adaptor Unit

Corer

Quick connect (QC) pin

Securing clip Side ViewEnd View

Slide over adaptor unitand secure with QC pin

Cross section of cutting edge (magnified)

1 cm graduation

5 cm graduation

Hole for QC pin

Scalloped cutting edge

Securing cord

Insert into drill chuck

0 5 10 15 20

Scale (cm)


376

2.2 Operation

To take soil cores, clamp the adaptor firmly inthe drill chuck, clip the corer onto the adaptor,and hold the corer vertically over the samplingpoint (a 2-D level glued to the top of the drill canbe helpful for keeping the corer vertical). Selectthe low-speed gear, engage the motor and slowlylower the corer, letting it cut through the mossand lichen layer (if present), through the L, Fand H layers and into mineral soil. Only a slightweight is required on the cutting edge, thereforesupport the weight of the drill while coring; toomuch pressure may compress the sample, partic-ularly moss, which will prevent proper bulk den-sity measurements. Core a few centimetres intothe mineral soil to create a base for the forestfloor sample. Stop the drill and lift the corer out.Keeping the corer upright, disconnect it from theadaptor, put the top of the corer onto the base ofthe examining tray, insert the core extractor intothe bottom of the corer and gently push the coreout of the corer and onto the examining tray(Fig. 1b). At this point, the layers can be identi-fied, any extraneous material removed, depths oflayers measured, and layers described as neces-sary. The required layers are then separated, usu-ally with a sharp knife, and transferred to plasticbags for later laboratory analysis.

2.3 Practical Aspects

Weighing only 3.5 kg, the complete corer sys-tem is easy to carry and can be operated by oneperson. It takes only a few seconds to drill a coreand a few more seconds to extract it into theexamining tray. The longest process is examina-tion, separation of layers, description, and bag-ging of samples. Field time can be reduced byreturning the cores to the laboratory intact, butfor our purposes, we found that field examina-tion of the cores was preferable. It avoided anychance of samples being disturbed in transport,examination was easier in daylight, and we couldrelate the core to site conditions and resolve anyqueries that arose.

Battery life, and hence the number of recharge-able batteries required, will depend on the depthof forest floor, depth cored into mineral soil and

texture of the mineral soil. In particular, penetra-tion of clay mineral soil requires full power,which can rapidly drain the battery. At one ex-treme, we found that coring through deep Popu-lus tremuloides forest floors and 15 cm into un-derlying clay could drain a battery in as few as10 cores. At the other extreme, one battery wassufficient for 30–40 cores in Pinus banksianaforests where there were thin forest floors oversandy mineral soil.

It is important that the corer be kept sharp. Innormal use, we found it necessary to sharpen theedge every 20–30 cores. As noted above, thiscan be done with the drill and a rotary sandingdisc, or with a sharpening stone for a light touch-up. When sharp, the corer slices cleanly throughvirtually all forest floor materials, including rootsup to 2 cm thick. It is possible to cut throughthicker roots, but the battery drain becomes ex-cessive. The corer is not suitable for stony soilsor for soils where bedrock is close to the surfaceas the cutting edge is easily dulled; if in doubt, itis wise to probe the soil with a thin metal rod todetermine stoniness or depth to bedrock.

2.4 Other Applications

We have also used the corer system for threeother applications. First, the corer has been ef-fective in obtaining shallow mineral soil sam-ples (down to about 20 cm) for determination ofbulk density, organic matter content and waterholding capacity. In the western Canadian borealforest stands that we have sampled, we havenoticed that fine roots are concentrated in theforest floor and shallow mineral soil, an obser-vation confirmed by Steele et al. (1997). Thissuggests that the corer would be ideal for inves-tigations of nutrient cycling in these areas, be-cause nutrient release from decomposition andnutrient uptake by roots is occurring in the soillayers that are easily sampled by the corer. Thesecond application was for preliminary evalua-tions of soil characteristics. The corer is muchfaster than digging a soil pit and provides a larg-er and cleaner core than the common 22 mm soilprobe. When the core is laid in the examiningtray, layers and horizons are clearly visible asare other characteristics such as soil colour, mot-


377

tles, fire scars, and fine roots. The third applica-tion is taking increment cores from trees, albeitwith a standard increment borer tube instead ofthe two-inch water pipe and an appropriate adap-tor. The concept is similar to a previously de-scribed borer powered by a chainsaw motor (Halland Bloomberg 1984), but the battery-powereddrill is much lighter, is easier to operate, has noexhaust fumes and virtually no fire risk. Wefound it provided adequate power for trees in ourregion where increment cores are seldom longerthan 25 cm. Apart from the advantage of speed,this “powered borer” enables cores to be taken atground level, which is important for reducinguncertainty when trees are being aged.

3 Sampling Considerations

3.1 Methods

We used the corer to determine stand level val-ues of forest floor organic matter and bulk densi-ty, as well as moss-plus-lichen layer biomass andbulk density. These means were used for a forestfloor dynamics study (Nalder and Wein Submit-ted). Here our objective is to present stand levelvariability data. We sampled 114 upland standsacross the western Canadian boreal forest. Standswere dominated by Picea glauca, Populus trem-uloides or Pinus banksiana. In each stand, wetook 10 cores at five-metre intervals along a ran-domly-oriented sampling transect. Cores weretaken vertically through moss/lichen and the for-est floor and 15–20 cm into mineral soil. Whencores could not be taken at the designated spotbecause of logs, large roots or rocks, the corerwas moved 50 cm further along the transect. Thedepth of each layer was recorded. We routinelychecked for compression of the sample by com-paring the length of the core in the examining trayagainst the depth cored. The moss-plus-lichenlayer was removed at the top of the litter layer,which was essentially at the base of the photo-synthesising tissue. As we were interested in for-est floor accumulations since the last stand-re-placing fire, the base of the forest floor layer wasdefined by the midpoint of the fire scar or at themineral soil surface if no fire scar was discerni-

ble. Forest floor samples were airdried, woodymatter was chopped into small fragments, thesample was ground until suitably homogenous,and a subsample of approximately 5 g was tak-en. This subsample was oven-dried to constantweight at 75 °C, then ashed at 450 °C for 16 hoursto determine organic matter content and percentorganic matter. Bulk density was calculated asoven-dry mass divided by volume (cross sectionalarea of corer by depth of forest floor). Cores con-taining unusually large accumulations of organicmatter, such as rotten logs or squirrel middens,were flagged for separate analysis. Biomass ofmoss-plus-lichen layer samples was determinedby oven-drying to constant weight at 75 °C andbulk density was calculated as for the forest floor.

3.2 Results and Discussion

3.2.1 Precision

Coefficients of variation (CV) were high for allvariables (Table 1). This was particularly so formoss-plus-lichen biomass, where CV’s wereclose to 80 %, mainly because many cores hadno moss or lichen, i.e., moss or lichen covertended to be patchy. For forest floor variables,CV’s were on average about 30 % (Table 1).Values for Pinus banksiana were generally muchhigher than the other two species, and values fororganic matter and bulk density were higher thanthose for depth and percent organic matter. Thevariability in percent organic matter was surpris-ingly high, and is likely due to the unavoidableinclusion of small amounts of mineral soil witheach forest floor sample. Similar CV’s for forestfloor variables have been reported elsewhere. InPinus sylvestris, there is a CV of 27 % for Cdensity (kg m–2) of the FH layer (Liski 1995)and 32 % for H layer thickness (Finer et al.1997). Across a range of forest types in the north-ern United States, Grigal et al. (1991) report aCV of 30 % for percent organic matter. Thelatter study noted that variability was correlatedwith means, which is also evident in our data(Fig. 3). The linear relationship between stand-ard deviations and means for each species indi-cates that CV’s remain fairly constant over awide range of means.


378

Data in Table 1 will tend to be conservativebecause unusually deep forest floor cores, par-ticularly those through rotten logs, were exclud-ed from this analysis. Considering forest floororganic matter as an example, inclusion of theexcluded cores increases CV’s to 37, 47 and 34% for Picea glauca, Pinus banksiana and Popu-lus tremuloides, respectively. Depending on thepurpose of the study, it may or may not be desir-able to treat these cores separately; for our dy-namics study, it enabled separation of compo-nents whose dynamics were quite different (Nal-der and Wein Submitted).

What are the implications of these high CV’sfor sampling intensity? For our dynamics study,10 samples per stand proved adequate to detectthe hypothesised effects of age, species and cli-mate on forest floor accumulation (Nalder andWein Submitted). In Finland, Liski (1995) rec-ommends 8–10 samples per stand for studyingorganic layer and mineral soil dynamics. As il-lustrated by Fig. 4, however, the number of sam-ples is very sensitive to the desired relative stand-ard error (RSE), defined as the standard error ofthe mean (Sx̄) as a percentage of the mean. Thismay vary considerably depending on the objec-tives of the study. At one extreme, 100 samples(RSE = 3 % with a CV of 30 %), would probablybe insufficient for detecting year-to-year chang-es in forest floor C pools because long-term ac-cumulation rates are generally less than 1 % peryear (Nalder and Wein Submitted). At the otherextreme, RSE’s of 20 % have been suggested asadequate for wildfire fuel load studies (Brown

Table 1. Mean coefficients of variation for within-stand variation for several variables (meanCV ± standard deviation of CV, number of stands in parentheses).

Variable Picea glauca Pinus banksiana Populus tremuloides

Forest floorOrganic matter (kg m–2) 34%±13% (23) 38%±14% (42) 29%±09% (49)Bulk density (Mg m–3) 32%±11% (17) 53%±29% (36) 32%±12% (47)Percent organic matter 21%±07% (23) 37%±17% (42) 21%±08% (49)Depth (cm) 30%±15% (17) 44%±14% (36) 22%±07% (48)

Moss-plus-lichenBiomass (kg m–2) 75%±51% (23) 81%±40% (37) NA1)

Bulk density (Mg m–3) 42%±17% (17) 45%±15% (33) NA1)

1) No moss-plus-lichen data are given for P. tremuloides because occurrence of moss or lichen was rare in these stands

Fig. 3. Linear relationships between means and stand-ard deviations: a) forest floor organic matter, b)forest floor bulk density.


379

1974, Van Wagner 1982) so that three samplesper stand may suffice. This assumes normally-distributed data, which was not true for any vari-able based on the Lilliefors test at 5 % signifi-cance level (SPSS Inc. 1993). Rather, all varia-bles except percent organic matter were log-nor-mally distributed. As shown by Grigal et al.(1991), fewer samples are required for the sameprecision under the log-normal assumption.

It may be possible to reduce variability bysampling with a larger cross-sectional area, suchas the traditional 30-cm square template. Cer-tainly the integrating effect of 900 cm2 has anintuitive appeal compared with the 22 cm2 of thecorer presented here. From our observations,however, forest floor variability is small over ascale of centimetres compared with a scale ofmeters. In support of this, organic layer accumu-lation is greater under canopies or close to treestems (Hokkanen et al. 1995, Liski 1995) whichin most stands provides a patterning on the scaleof meters. Consequently, we doubt that a largercross-sectional area would have a significant im-pact on sampling variability. If so, then the sam-pling efficiency (defined as 1/(n * t), where n isthe number of samples required and t is the timetaken for each sample) will be much lower forlarge template methods.

3.2.2 Spatial Autocorrelation

Spatial autocorrelation may lead to bias in esti-mates, particularly if sampling intervals are smallcompared with the scale of variation. We testedfor spatial autocorrelation with two methods, us-ing forest floor organic matter as an indicatorvariable. First, we calculated autocorrelation co-efficients for lags ranging from 1 to 8 and testedfor significance at the 5 % level using the Box-Ljung statistic (SPSS Inc. 1993). Of the 114stands sampled, only seven showed evidence ofautocorrelation, defined as having more than onesignificant lag interval. Three other stands hadone significant lag interval, but given the largenumber of tests performed, this is to be expectedby chance alone. Among the seven stands, therewas no apparent pattern. Two were Picea glau-ca, two were Pinus banksiana, three were Popu-lus tremuloides, and they spanned a wide rangeof ages and stem density, suggesting that theapparent autocorrelation was due to chance. Thesecond method was a semivariogram analysis, atechnique commonly used in geostatistics (Kita-nidis 1997). For each stand, we calculated semi-variances for lags from 5 to 35 m at five-metreintervals. Because there were insufficient pointsin each stand to develop a stable variogram, wenormalised semi-variances by expressing themas a fraction of the stand mean, then averagedthe normalised semi-variances across all stands.

Fig. 4. Relationship between relative standard error(RSE) and number of samples (N) for a range ofcoefficients of variation (CV) assuming a normaldistribution (RSE = CV / √N * 100 %).

Fig. 5. Semi-variogram showing spatial variability alongsampling transect, averaged across 114 stands.


380

Again there was little evidence of spatial auto-correlation: the composite semi-variogram indi-cates variability is fairly constant with distance(Fig. 5). This is consistent with two studies inPinus sylvestris stands. In one, there was noclear spatial dependence over intervals of 5–50m (Hokkanen et al. 1995), and in the other, spa-tial dependence occurred mainly at distances lessthan 5 meters (Liski 1995). We conclude thatspatial autocorrelation was not a concern withthe 5 m sampling intervals used in this study.

3.2.3 Other Variability

Spatial variability is just one component of vari-ation leading to the high CV’s in Table 1. Herewe consider three other sources, corer bias, layerseparation error and core compression. Corerbias may occur for two reasons. First the corernever makes a perfect vertical cross sectionthrough sampled layers; as noted previously, thecutting edge may push some material aside, orcapture some material that is outside the crosssectional area, or be diverted by hard objects.We cannot quantify this effect, but qualitativelywe judge it to be insignificant given the clean cutmade by our corer. Second, not all chosen pointsalong the transect can be cored because of ob-structions, particularly large logs representingthe boles of fallen trees. This may bias estimatesif forest floor characteristics are different at thesepoints. To illustrate, large logs obstructed 112 ofthe 1252 sampling points in this study. Becauselogs tend to screen the underlying ground fromlitterfall, it is likely that the forest floor is thinnerunder logs. Assuming a 50 % reduction, ourestimates of mean forest floor organic matterwould overestimate actual values by about 5 %.Clearly, there is a potential for significant bias.

The second potential source of error occursbecause identification and separation of layers isnever precise; layers often inter-grade and esti-mation of a layer boundary has an element ofsubjectivity. We found that assigning a layerboundary to the nearest 0.5 cm was as precise aswas warranted, even with the good quality coresproduced by our corer. This imprecision can rep-resent a significant source of error, particularlyin thin layers. Next to spatial variability, we

would judge this to be the next largest compo-nent of variability, and it probably accounts forthe higher CV’s in Pinus banksiana (Table 1).

The third potential source of error is core com-pression which affects bulk density. Significantcompression can occur, particularly with themoss-lichen layer, if the corer is not sharp or ifthe drill is pushed down while coring. However,with a sharp corer and the proper coring tech-nique, we were not able to measure any corecompression, so we doubt that this would be asignificant source of variation.

4 Conclusion

We have described an effective device for sam-pling forest floors, which is also useful for otherlayers. It is easy to fabricate, inexpensive, allowscores to be taken rapidly by one person, andgives undisturbed, natural-volume cores whencoring through materials as varied as moss, li-chen, litter, ferment, humus and mineral soil.Sampling CV’s for forest floor organic matterand bulk density were typically 30 % or higher.We attribute most of this variation to spatialvariability although the lack of precision in iden-tifying layer boundaries undoubtedly forms asignificant component, particularly for thin lay-ers. The impact of logs is not insignificant andneeds to be considered. It is clear that high preci-sion is not feasible in forest floor sampling. For astudy of forest floor dynamics, 10 samples perstand gave adequate precision and we think thiswould be sufficient for many studies. Obtainingmore precise estimates would soon become veryexpensive in sampling effort. Our data supportour choice of five-metre sampling intervals asadequate to avoid spatial autocorrelation.

Acknowledgements

Funding came from a Natural Sciences and En-gineering Research Council of Canada (NSERC)Research Grant to R.W. Wein and a NSERC67Scholarship and Boreal Alberta Research Grantto I. A. Nalder. We thank Dr. Jari Liski and Dr.


381

Ann Dolling for constructive reviews which im-proved this manuscript.

References

Agriculture Canada Expert Committee on Soil Survey1987. The Canadian System of Soil Classifica-tion. Publ. 1646. Agric. Can., Ottawa, Ontario.

Alban, D.H. 1982. Effects of nutrient accumulationby aspen, spruce, and pine on soil properties. SoilScience Society of America Journal 46: 853–861.

Bonan, G.B. & Shugart, H.H. 1989. Environmentalfactors and ecological processes in boreal forests.Annual Review of Ecological Systems 20: 1–28.

Brown, J.K. 1974. Handbook for inventorying downedwoody material. Gen. Tech. Rep. INT-16. Inter-mountain Forest & Range Experimental Station,USDA For. Serv., Ogden, Utah.

Finer, L., Messier, C. & De Grandpre, L. 1997. Fine-root dynamics in mixed boreal conifer–broad-leafed forest stands at different successional stag-es after fire. Canadian Journal of Forest Research27(3): 304–314.

Forestry Canada Fire Danger Group 1992. Develop-ment and structure of the Canadian Forest FireBehaviour Prediction System. Inf. Rep. ST-X-3.Forestry Canada, Ottawa, Ontario.

Gower, S.T., Vogel, J.G., Norman, J.M., Kucharik,C.J., Steele, S.J. & Stow, T.K. 1997. Carbon dis-tribution and aboveground net primary productionin aspen, jack pine, and black spruce stands inSaskatchewan and Manitoba, Canada. Journal ofGeophysical Research 102(D24): 29029–29041.

Grigal, D.F., McRoberts, R.E. & Ohmann, F.F. 1991.Spatial variation in chemical properties of forestfloor and surface mineral soil in the north centralUnited States. Soil Science 151: 282–290.

Hall, A.A. & Bloomberg, W.J. 1984. A power-drivenincrement borer. Forest Chronicle 60(6): 356–357.

Hokkanen, T.J., Järvinen, E. & Kuuluvainen, T. 1995.Properties of top soil and the relationship betweensoil and trees in a boreal Scots pine stand. SilvaFennica 29(3): 189–203.

Huang, W.Z. & Schoenau, J.J. 1996. Forms, amountsand distribution of carbon, nitrogen, phosphorusand sulfur in a boreal aspen forest soil. CanadianJournal of Soil Science 76: 373–385.

Johnson, D.W. 1995. Role of carbon in the cycling of

other nutrients in forested ecosystems. In: McFee,W.W. & Kelly, J.M. (Eds.). Carbon Forms andFunctions in Forest Soils. Soil Science Society ofAmerica, Inc., Madison, Wisconsin. p. 299–328.

Kimmins, J.P. 1987. Forest Ecology. Macmillan Pub-lishing Company, New York. 531 p.

Kitanidis, P.K. 1997. Introduction to Geostatistics:Applications in Hydrogeology. Cambridge Uni-versity Press, Cambridge. 249 p.

Liski, J. 1995. Variation in soil organic carbon andthickness of soil horizons within a boreal foreststand – effect of trees and implications for sam-pling. Silva Fennica 29(4): 255–266.

— & Westman, C.J. 1997. Carbon storage in forestsoil of Finland: 2. Size and regional patterns. Bio-geochemistry 36: 261–274.

Nalder, I.A. & Wein, R.W. (Submitted). Long-termforest floor carbon dynamics after fire in uplandboreal forests of western Canada. Global Biogeo-chemical Cycles.

Paré, D. & Bergeron, Y. 1996. Effect of colonizationtree species on soil nutrient availability in a claysoil of the boreal mixedwood. Canadian Journalof Forest Research 26(6): 1022–1031.

Quintilio, E., Alexander, M.E. & Ponto, R.L. 1991.Spring fires in a semimature trembling aspen standin central Alberta. Inf. Rep. NOR-X-323. Can.For. Serv., Edmonton, Alberta.

SPSS Inc. 1993. SPSS for Windows: Base SystemUser’s Guide, Release 6.0. SPSS Inc., Chicago,Illinois, 828 p.

Steele, S.S., Gower, S.T., Vogel, J.G. & Norman,J.M. 1997. Root mass, net primary production andturnover in aspen, jack pine and black spruce for-ests in Saskatchewan and Manitoba, Canada. TreePhysiology 17: 577–587.

van Cleve, K. & Noonan, L.L. 1971. Physical andchemical properties of the forest floor in birch andaspen stands in interior Alaska. Soil Science Soci-ety of America Proceedings 35: 356–360.

— , Olivier, L., Schlentner, R., Viereck, L.A. & Dyr-ness, C.T. 1983. Productivity and nutrient cyclingin taiga forest ecosystems. Canadian Journal ofForest Research 13: 747–766.

Van Wagner, C.E. 1982. Practical aspects of the lineintersect method. Inf. Rep. PI-X-12. Can. For.Serv., Edmonton, Alberta.

— 1983. Fire behaviour in northern conifer forestsand shrublands. In: Wein, R.W. & McLean, D.A.(Eds.). The Role of Fire in Northern Circumpolar


382

Ecosystems. John Wiley & Sons, Chichester. p.65–80.

Westman, C.J. 1995. A simple device for sampling ofvolumetric forest soil cores. Silva Fennica 29(3):247–251.

Zackrisson, O., Nilsson, M.-C., Steijlen, I. & Hörn-berg, G. 1995. Regeneration pulses and climate-vegetation interactions in nonpyrogenic borealScots pine stands. Journal of Ecology 83: 469–483.

Total of 27 references

A New Forest Floor Corer for Rapid Sampling, Minimal … › 4959 › 1332e543390c0ad7... ·...

Documents

Transcript of A New Forest Floor Corer for Rapid Sampling, Minimal … › 4959 › 1332e543390c0ad7... ·...