Nursery and site preparation interaction research in

New Forests 22: 43–58, 2001. © 2001 Kluwer Academic Publishers. Printed in the Netherlands.

Nursery and site preparation interaction research in the United States DAVID B. SOUTH1, ROBIN W. ROSE2 and KEN L. MCNABB1

1 School of Forestry and Wildlife Sciences and Alabama Agricultural Experiment Station, Auburn University, Auburn, AL 36849-5418, USA; 2 Department of Forest Science, Oregon State University, Corvallis, OR

Accepted 3 January 2001

Key words: growth and yield, interactions, regeneration, research designs, seedling quality, silviculture Abstract. Researchers in the United States have generally taken a “segregated” approach to regeneration research. As a result, trials involving nursery treatments are usually separated from trials involving site preparation treatments. Researchers have concentrated on just the main treatment effects. Little effort has been placed on examining potential interactions between nursery and site preparation treatments. Only a few nursery by site preparation (NxSP) studies have been established. The profession has generally assumed that gains from nursery and site preparation treatments are simply additive, or that the relative gains from nursery treatments would be insignificant when compared to intensive site preparation treatments. Both of these assumptions may be incorrect. This review examines some of the NxSP interaction trials that have been established in the United States during the last half of the 20th century. We predict that as rotation ages for plantations decrease, the need for NxSP interaction trials will increase. NxSP interaction studies will be required if researchers wish to include both nursery and site preparation treatments in their early establishment models. To date, only a few establishment models have been developed in the United States.

Introduction Researchers have been conducting artificial regeneration trials in the United States (US) for more than a century. Some of the early tree planting trials during the 20th century were conducted by James Toumey, Carl Schenck, Philip Wakeley, Paul Rudolf and others. Over time, findings by these researchers led to improvements in nursery and site preparation treatments. However, the way in which they segregated regeneration research has been carried forward to the present. Current artificial regeneration research tends to segregate around three groups: genetics, nursery and, for lack of a better word, site preparation.

44 For the purpose of this paper, the term “site preparation” includes not only mechanical cultivation but also the use of herbicides, fertilizers, insecticides and other treatments that are applied during plantation establishment. As a generalization, nursery researchers tend to outplant nursery treatments, geneticists tend to outplant progeny tests, and silviculturalists tend to evaluate site preparation methods. Due to this fragmented approach, trials that combine nursery treatments with site preparation treatments are rare. As a result, much is known about the main effects but little is known about their interactions. For purposes of this paper, “NxSP interaction” research will refer mainly to field trials where both nursery treatments and site preparation treatments are tested in the same experimental design. For example, a study in Texas (Barber et al. 1991) would qualify since both nursery treatments (fertilization rates) and site preparation treatments (herbicides) were compared in the same experimental design. Although nursery by genetic trials can also be of importance (Snyder and Allen 1963; Wakeley 1963; Park et al. 1973; Land 1983), they will not be discussed here. Genotype by silvicultural interactions have been discussed by others elsewhere (McKeand et al. 1997; McDonald et al. 1999). This review examines some of the NxSP interaction research in the Northern, Western, and Southern US. We will briefly discuss what we have learned from these studies and ponder what could be learned from future interaction studies.

Northern region A NxSP study was conducted on northern red oak (Quercus rubra L.) in Missouri (Johnson 1984). The study involved top-pruning, 4 stock types, and 3 overstory treatments. Although a factorial analysis was possible, no tests for interactions were conducted. Apparently, average height after 5 years of growth was the same in plots that were clearcut but growth differed among treatments when seedlings were slower growing due to shade. Top-pruning seedlings tended to increase survival regardless of overstory treatment. In Connecticut, Ward (1997) compared seedling size with three protective devices on early survival and growth of two hardwood and three conifer species. He found that for some species, tall seedlings planted with no protection grew as much as average seedlings with browsing protection. Only two NxSP studies have appeared in the Northern Journal of Applied Forestry (Table 1) and both were in Ontario. Regeneration research is more common in Ontario where about 120 million seedlings are planted annually (as compared to the northern US where 77 million trees are planted in 20

45 Table 1. The number of papers in Applied Journals of Forestry that relate to nursery research, site preparation research, or the interactions where seedlings were outplanted in the field. Journal

Volumes

Nursery

Site preparation

Interaction

Total

Northern Southern Western

1–14 1–22 1–10

21 36 6

27 80 11

2 2 0

528 917 417

northern states). A majority (85%) of the nursery stock for the northern US is produced in Michigan, Minnesota, Ohio and Indiana (Moulton 1999). Sutherland and Day (1988) summarized the Canadian literature on the effect of container volume on outplanting survival and growth of white spruce (Picea glauca [Moench] Voss), black spruce (Picea mariana [Mill] B.S.P), and jack pine (Pinus banksiana Lamb.). There was ample evidence that larger container cell volumes produced larger seedlings that in turn exhibited increased growth. Although such approaches are limited when looking for interactions, they do indicate the effect seedling quality has on long-term plantation performance. Wood (1990) compared three stock types and two planting dates for black spruce and jack pine. He found initial stock type and planting date effects “were still evident 10 years after planting.” Although there appeared to be an interaction between planting date and stock type for black spruce (bareroot transplants were by far the best when planted in May), tests for significant interactions were not reported. Pitt et al. (1999) planted two black spruce stock types at different times with four herbicide application rates. The authors did an exhaustive and detailed analysis of herbicide, planting date, and planting by herbicide interactions, concluding that early planting and herbaceous weed control were advantageous. Although they did not look for any potential interactions with stock type (separate analyses were conducted by stock type), the effects of stock type and herbicide rate appear to be additive). Initial seedling stem volumes at the time of planting were recorded but they were not looking for this variable to impact the interpretation of their results. Further analysis of the relationship of individual seedling stem volumes at the time of planting with herbicide treatment level might provide valuable insight into the possible interaction of these two variables. Payandeh (1996) modelled the growth of three planted species in northern Ontario based on operational outplanting and plantation assessment data. Stock type, planting date, and site preparation variables were found to signifi-

46 cantly affect prediction of plantation survival and height. While comparing only stock type (bareroot versus container) and not looking at stock size, it is noteworthy that the model incorporated a seedling quality factor to predict plantation height. These results led to one of the first establishment models in North America. Southern region Publications involving data from artificial regeneration trials were rare prior to 1920 (Dorman and Sims 1949; Toumey 1916). As far as we know, there were no NxSP studies conducted during the first half of the 20th century. Possibly the first interaction study in the South was installed by T.E. Maki and his student (Johansen 1955) using longleaf pine (P. palustris Mill.). The study involved various thinning densities in the nursery, root pruning in July in the seedbed, and three types of field sites (including a furrowed site and a herbicide-treated area). The study showed that both root pruning and low seedbed density increased survival, and these effects appear to be additive with site preparation. Another early interaction study (at the Savannah River Project in South Carolina) involved sowing date and furrowing for longleaf pine (Shipman 1958). In this study, it was learned that furrowing had a dramatic effect on old-field sites but did not improve survival when scrub oak sites were cleared before planting (Table 2). From these trials, Shipman concluded that large nursery stock was desirable “not only because such stock survives better, but also because large seedlings may reach merchantable size sooner.” Since the 1950’s, only a few interaction studies have been carried out in the southern US. Thus far, only 2 NxSP interaction studies have been published in the Southern Journal of Applied Forestry (Table 1). The study by Powers and Rowan (1983) may be the first to include not only nursery and fertilization treatments, but also different genotypes. NxSP studies have involved seedling grade and herbaceous weed control (Mitchell et al. 1988; South and Mitchell 1999), seedling grade, fertilization, soil cultivation and herbaceous weed control (South et al. 1995), and seedling nutrition and herbaceous weed control (Barber et al. 1991). Hitch et al. (1996) used survey data to develop a model for loblolly pine (P. taeda L.) that predicts survival from seedling quality attributes and silvicultural treatments. Users of this model can compare the relative effects of seedling size (height and number of lateral roots) and site preparation (disking and herbicides) on initial survival (Table 3). As far as we know, this was the first survival model for loblolly pine that allows the user to model potential interactions.

47 Table 2. The effect of sowing date and furrowing on the first-year suvival of longleaf pine on two planting sites (from Shipman 1958). This is possibly the second nursery by silvicultural interaction study in the southern United States. Site

Sowing date

Furrowing

Survival (%)

Old field Old field Old field Old field Scrub oak Scrub oak Scrub oak Scrub oak

November November March March November November March March

Yes No Yes No Yes No Yes No

96 62 87 55 81 78 76 77

Table 3. The effect of herbaceous weed competition (during the first growing season), disking, seedling height, and number of lateral roots on predicated 5th year survival of loblolly pine in Georgia (adapted from Hitch et al. 1996). Weeds

Disking

Seedling height (cm)

Lateral roots

Survival (%)

60% cover 60% cover 60% cover 60% cover 60% cover 60% cover 60% cover 60% cover 10% cover 10% cover 10% cover 10% cover 10% cover 10% cover 10% cover 10% cover

No No No No Yes Yes Yes Yes No No No No Yes Yes Yes Yes

25 25 25 25 20 20 20 20 25 25 25 25 20 20 20 20

7 9 7 9 7 9 7 9 7 9 7 9 7 9 7 9

77 79 83 85 86 88 90 91 82 85 88 90 90 92 94 94

48 Ptaeda2V
is the first model to allow the user to examine potential NxSP interactions for growth of loblolly pine. At the request of James Vardaman, this model was modified by Harold Burkhart to allow the user to predict the effects of combining various establishment treatments. The model allows the user to model the effects of planting first- or second-generation seedlings, subsoiling, use of seedlings grown at low seedbed densities, use of a herbicides at time of planting, use of a release herbicide, and mid-rotation fertilization (Vardaman 1998). Establishment models exist from other regions of the world that allow the user to model growth from various nursery and site manipulations. For example, one model predicts that planting larger pine seedlings will result in great early growth (Payandeh et al. 1992). Western region In 1975, the US Forest Service installed a large interaction study involving four site preparation treatments, a browse protection treatment, seven nursery stocks, and a herbicide treatment (Stein 1984). This may have been the first NxSP interaction study in Oregon. Unfortunately, this preliminary report listed only the main effects of stock type and weed control. Therefore, we do not know if there were any significant NxSP interactions. Variation due to the interaction term likely went into the experimental error term. However, in cases where the NxSP interaction is significant, this will be a mistake since the conclusions about main effects could be erroneous. For example, the inclusion of a NxSP term could make a big difference in the analysis of variance (Table 4). A fertilization by species by mycorrhizal inoculation study was established on a high elevation mine site in Colorado (Grossnickle and Reid 1982). The study involved 3 species, 4 mycorrhizal treatments, and 4 fertilization treatments. After four years of growth, seedling growth was greatest in plots amended with sewage sludge and wood chips. There were no significant interactions among tree species, mycorrhizal treatments and fertilization treatments. Since 1996, seven NxSP interaction studies have been installed by the Vegetation Management Research Cooperative at Oregon State University. Treatments include two stock sizes, a fertilizer treatment and a herbicide treatment. One goal of this research is to achieve 2-m tall seedlings two years after planting (Rose and Ketchum 1999). To date, NxSP studies have not been published in the Western Journal of Applied Research. Thus far, interest has concentrated on the main effects from nursery practices and silvicultural practices (Table 1). Without the

49 Table 4. An example illustrating the effect of leaving out the interaction term in an analysis of variance. The correct analysis results in a significant interaction effect while the incorrect analysis shows no significant nursery effect. Source

df

Sum of squares

F value

P>F

- - - - - - - - - - - - - - - - - - - - - - Correct analysis - - - - - - - - - - - - - - - - - - - - - Replication Insecticide (SP) Stock size (N) NxSP interaction Error

2 1 1 1 6

0.1888 2.0227 1.0817 2.0354 0.4147

1.36 29.26 15.65 29.45

0.325 0.002 0.007 0.002

- - - - - - - - - - - - - - - - - - - - - Incorrect analysis - - - - - - - - - - - - - - - - - - - - Replication Insecticide Stock size Error

2 1 1 7

0.1888 2.0227 1.0817 2.4501

0.27 5.78 3.09

0.771 0.047 0.122

This trial tests an insecticide treatment and two nursery stocks (small seedlings vs. large seedlings). The randomized block design has three replications.

benefit of research data, forest managers usually assume the gains from the main effects are additive.

Types of NxSP interactions Some of the various NxSP interactions are illustrated in Figure 1. In Figure 1a, high intensity site preparation treatments improved survival only when low performing nursery stock was planted. Figure 1b illustrates a hypothetical case where use of intensive site preparation increases the survival of small nursery stock but decreases survival of large-diameter stock. An example of no NxSP interaction for growth is illustrated in Figure 1c where gains from intensive site preparation and improvements in nursery stock performance are strictly additive. Figure 1d shows synergism where more growth gains result when combining high performing nursery stock with high site preparation. Figure 1e shows a hypothetical case where the effects of nursery treatments are only expressed when intensive site preparation is used. The opposite is illustrated in Figure 1f where nursery treatment made no difference when high site preparation treatments are applied.

50

Figure 1. Hypothetical examples of interactions between seedling quality and site preparation intensity.

A need to identify the “optimum” seedling The “optimum” seedling is defined as the ideotype that will minimize overall reforestation costs while achieving established goals for initial survival and growth (South and Mitchell 1999). However, the optimum seedling can not be properly identified by using main effect studies. Since economics is greatly affected by site preparation costs, NxSP interaction studies are necessary in order to determine the “optimum” seedling. Due to a national tendency to keep nursery and silvicultural trials separate, some organizations are attempting to minimize seedling costs while spending an unnecessary amount of money on intensive treatments in order to get the small seedlings to grow

51 well after planting. Due to a lack of an integrated approach to regeneration, inefficient systems result. Plantation establishment should be viewed as an entire process rather than a series of steps in isolation (Kimmins 1989). This means effort should be made to determine the “optimum” seedling. Cost of nursery and silviculture practices When nursery practices are fully integrated into silvicultural prescriptions, overall establishment costs might be reduced (Todd 1989, South et al. 1993; South and Mitchell 1999). Therefore, some industry leaders see the need for empirical data and have installed their own NxSP studies. After conducting several NxSP trials, Albert et al. (1980) said that “the nursery offers an area of yield improvement that is competitive, on a cost-benefit basis, with many other establishment practices.” Since costs have practical significance, a sampling of various establishment costs is provided in Table 5. Regeneration costs vary with region, site and organization. Therefore, the costs of treatments in Table 5 are only examples. Some of the nursery treatments are relatively inexpensive and can be easily justified economically. Nursery practices such as top-pruning and extra fertilization cost little, but for some species they have the potential to greatly improve survival after outplanting. In some cases, top-pruning can improve the balance between roots and shoots and can increase field survival by 12 percent or more (South 1996, 1998). Extra fertilization in the nursery can boost foliar concentrations of nutrients and this increased survival of longleaf pine and slash pine (P. elliottii Engelm. var elliottii) by 20% (Allen and Maki 1955; Irwin et al. 1998). The benefit/cost ratios of these practices can be high. Although toppruning might cost about $30/nursery ha, the additional cost per outplanted ha will be only a few cents! This will not be true for some other nursery treatments or for most intensive site preparation treatments. Combining some stock types with some site preparation techniques can cost more than $1,000 per ha (Brand 1991). In order to achieve the same return on investment, four times as much value (or wood) needs to be produced when investing $1,200 per ha as compared to $300 per ha. In some cases, investing more capital in intensive management will increase volume production but will also increase the average cost of producing a cubic meter of wood (Figure 2a). This will occur when the percent increase in volume production at harvest is less than the percent increase in the cost of establishment. The ideal situation would be to decrease the cost of wood production (per m3 ) while increasing wood production per ha. This will occur when the percentage gain in wood production is greater than the percentage gain in establishment costs (see Figure 2b). This result is more likely to occur when

52 Table 5. Break-even yield analysis for various regeneration practices. Practice

Region

Additional cost/ha1

Purportedly increases Survival2 growth3

$

Break-even volume/ha4 Cubic meters

Nursery treatments Top pruning Antitranspirant Extra nitrogen fertilization Vegetative mycorrhizal inoculum 1+0 bare-root seedlings 6–8 mm RCD Container grown stock 2–3 mm RCD 1+1 bare-root stock 5–8.5 mm RCD 1+1 bare-root stock 5.5–10 mm RCD 1+1 bare-root stock 6–10.5 mm RCD P+2 bare-root stock 4–7 mm RCD Tissue culture plantlets in containers5

South South South South South South West West West North South

0.03 0.06 0.15 14 27 95 190 234 265 670 780

Yes Yes No Yes Yes Yes Yes Yes Yes Yes Yes

No No Yes Yes Yes No Yes Yes Yes Yes Yes

Yes Yes Yes Yes No No No No Yes No Yes No No Yes Yes No Yes

No No No No Yes No Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes

0.003 0.006 0.016 0.016 2.9 10.3 20.5 25.3 28.6 72.4 84.2

Site preparation treatments Machine planting Shovel planting Additional planting supervision Shovel planting Burning (after chemical site preparation) Ectomycorrhizal spores in planting hole Insecticides for tip-moth P fertilization Herbicide – post planting Animal protection Ripping Single chop Aerial herbicide 3-in-1 plow Shear-rake-pile-bed Pile and burn Tree shelters

South South South West South South South South West West South South South South South West North

17 24 25 26 45 48 74 83 100 110 150 175 235 306 350 435 1800

1.8 2.6 2.7 2.8 4.8 5.2 7.9 8.9 10.8 11.9 16.2 18.9 25.4 33.0 37.8 46.9 194.4

1 Baseline treatment costs per ha:

North = dibble plant ($95) 2+0 bare-root seedlings ($120). South = hoedad plant ($72) 1+0 bare-root seedlings 3–4 mm RCD ($45). West = hoedad plant ($120) 2+0 bare-root seedlings 4–5.5 mm RCD ($230). 2 First-year survival. 3 Growth per seedling. 4 Economic assumptions: 1200 trees per hectare; pulpwood stumpage value $50/m3 ; 15 year rotation; a 6% interest rate; and a 26% tax bracket. 3 3 5 Example for tissue culture-plants in the South: ($825 − $45) = $780 = 84.2 m ×$30/m ×(0.74) . (1.06)15

53

Figure 2. Hypothetical examples of the effects of increasing establishment costs on yield and the cost of wood production. Squares = wood cost: Circles = wood yield.

a NxSP interaction like Figure 1d occurs. Therefore, we need to identify those treatment combinations that will produce responses like Figure 1d. Initially, empirical trials will be used to identify these combinations and later establishment models will be developed and used.

A need for establishment models Although various growth and yield models exist in the US, establishment models are rare. Establishment models (a subset of growth and yield models) predict survival and growth for various establishment practices for five or more years after planting. Few establishment models exist because of: (1) a lack of NxSP interactions studies; (2) a large amount of experimental variation during the first ten years of growth; and (3) an ability of critics to quickly verify errors in the model. Despite these limitations, we believe there is a real need to develop establishment models for the US. Such models could go a long way to improving the efficiency of stand establishment. We need models

54

Figure 3. The interaction of stock size, planting time and weed competition on initial seedling growth (adapted with permission from Kimmins 1989).

that would allow foresters to predict outcomes from factors such as planting date, seedling size and weed competition (Figure 3). Establishment models could be developed using our best guesses, but they will be more reliable if they are based on data from NxSP studies. An early model (in the form of a table) was prepared by Hatcher (1957) for survival of pine seedlings planted in the sandhill region of South Carolina. A version of this model is given in Figure 4. Hatcher claimed that instead of prescribing one planting density to all sandhill sites, spacing should be varied by seedling quality and method of site preparation. He said “the combination of planting in furrows and good quality graded seedlings reduced planting costs 23.4 percent in spite of supervision and service costs running above our three-year average.” Therefore, money would be wasted if the field forester assumed that seedling size and furrowing had no effect on seedling survival.

55

Figure 4. The predicted survival of P. elliottii on sandhill sites as affected by seedling grade (Wakeley 1954), furrowing and soil texture (from Hatcher 1957).

An equation-based survival model was developed by Hitch et al. (1996). This model predicts that survival can be increased by planting shorter trees and trees with more lateral roots. The model suggests that survival grains from disking and weed control are additive with that from planting “short and fat” seedlings. However, the gains are not completely additive since survival can not exceed 100 percent. Currently, there are at least three establishment models that are in the “public” domain in North America. The model developed by the Great Lakes Forestry Centre accounts for species, seed source, stock type, nursery storage, site preparation, planting date, planting method, weed control and fertilization (Payandeh et al. 1992). There are no models developed in the US that allow the user this many treatment options. In Georgia, an establishment model was developed through funding by American Cyanamid. Acorm
allows the user to input different chemical and mechanical site preparation methods, as well as different levels of weed competition (Anonymous 1995). Ptaeda2V
allows the user to model gains from genetics, fertilizer, ripping, seedlings size, and herbaceous weed control. Gains from planting high quality planting stock (seedlings grown at low seedbed densities) are modeled with a 1-year “establishment quality boost” (EQB) while gains from herbaceous weed control are modeled with either a 1-year or 2-year EQB. In theory, a 1-year EQB will predict that a stand at age 23 will yield the same volume per hectare as a 24-year-old stand with no EQB. Although the current Ptaeda2V
model needs improvement (output results depend on the order of input), this

56 computer model is a good tool for testing assumptions about best how to model NxSP interactions. Thus far, it has been assumed that all main effects that can be modeled with Ptaeda2V
are simply additive.

Conclusions To date, data from studies involving only nursery or site preparation treatments have guided decisions on preferred establishment practices. Due mainly to tradition, it is rare for researchers to include both nursery and site preparation treatments in the same experimental design. Without results from NxSP studies, researchers usually accept the hypothesis that main effects are always additive. This hypothesis should be tested with properly designed NxSP trials. In cases where establishment modes are based on empirical trials, NxSP studies will be useful in verifying the model’s predictive ability. We expect that in the future, results from NxSP studies will help guide the optimal combination of nursery and site preparation treatments.

Acknowledgement We thank the companies for providing seedling prices and cost for silvicultural treatments.

References Albert, D.J., Fry, G. and Pool, B.R. 1980. An industrial company’s view of nursery stock quality. New Zealand Journal of Forestry Science 10: 2–11. Allen, R.M. and Maki, T.E. 1955. Response of longleaf pine seedlings to soils and fertilizers. Soil Science 79: 359–362. Anonymous. 1995. Acorm User’s Guide. American Cyanamid Company, Princeton, New Jersey, 16 pp. Barber, B.L., Messina, J.S., van Buijtenen, J.P. and Wall, M.M. 1991. Influence of nursery fertilization, site quality, and weed control on first-year performance of outplanted loblolly pine. pp. 27–37. In: Coleman, Sandra S. and Neary, Daniel G. (Eds.) Proceedings of the Sixth Biennial Southern Silvicultural Research Conference. October 30–November 1, 1990, Memphis, TN. USDA For. Serv. Gen. Tech. Rep. SE-70. Brand, D.G. 1991. Biological and economic productivity of Canadian silvicultural regimes. pp. 324–332. In: Menzies, M.I., Parrott, G.E. and Whitehouse, L.J. (Eds.) Efficiency of Stand Establishment Operations. NZ Forest Research Institute. FRI Bulletin 156. Dorman, K.W. and Sims, I.H. 1949. Loblolly pine bibliography. USDA Forest Service Southeastern Forest Experiment Station. Station Paper No. 6. 65 pp. Grossnickle, S.C. and Reid, C.P.P. 1982. The use of ecotomycorrhizal conifer seedlings in the revegetation of a high-elevation mine site. Can. J. For. Res. 12: 354–361.

57 Hatcher, J.B. 1957. Prescription planting. Forest Farmer June: 4–6. Hitch, K.L., Shiver, B.D. and Borders, B.E. 1996. Mortality models for newly regenerated loblolly pine plantations in the Georgia Piedmont. Southern Journal of Applied Forestry 20: 197–202. Irwin, K.M., Duryea, M.L. and Stone, E.L. 1998. Fall-applied nitrogen improves performance of 1-0 slash pine nursery seedlings after outplanting. Southern Journal of Applied Forestry 22: 111–116. Johnson, P.S. 1984. Responses of planted northern red oak to three overstory treatments. Can. J. For. Res. 14: 536–542. Johansen, R.W. 1955. The effects of certain nursery treatments and site preparations on the vigor and survival of longleaf pine (Pinus palustris Mill.) in the deep sands of the Carolinas. M.S. thesis. North Caroina State College, School of Forestry. 46 pp. Kimmins, J.P. 1989. Ecological implications of successional manipulation. pp. 9–16. In: Schivener, B.A. and MacKinnon, J.A. (Eds.) Learning from the Past, Looking into the Future. B.C. Ministry of Forests. FRDA report 030. Land, S.B. 1983. Performance and G-E interactions of sycamore established from cuttings and seedlings. pp. 431–440. In: Jones, Earle P. (Ed.) Proceedings of the Second Biennial Southern Silvicultural Research Conference. November 4–5, 1982, Atlanta, GA. USDA For. Serv. Gen. Tech. Rep. SE-24. McDonald, P.M., Mori, S.R. and Fiddle, G.O. 1999. Effect of competition on genetically improved ponderosa pine seedlings. Can. J. For. Res. 29: 940–946. McKeand, S.E., Crook, R.P. and Allen, H.L. 1997. Genotypic stability effects on predicted family responses to silvicultural treatments in loblolly pine. Southern Journal of Applied Forestry 21: 84–89. Mitchell, R.J., Zutter, B.R. and South, D.B. 1988. Interaction between weed control and loblolly pine, Pinus taeda, seedling quality. Weed Technology 2: 191–195. Moulton, R.J. 1999. Tree planting in the United States – 1997. Tree Planters’ Notes 49: 1–15. Park, Y.S., Gerhold, H.D. and Palpant, E.H. 1973. Genotype × environment interactions of Douglas-fir provenances in Pennsylvania nurseries. School of Forest Resources, Pennsylvania State University, Research Briefs 7(2): 14–17. Payandeh, B. 1996. Growth and survival functions for three planted species in northern Ontario. North. J. Appl. For. 13: 19–23. Payandeh, B., Punch, M. and Basham, D. 1992. User’s manual for “Plant-PC”: a model for forest plantation establishment in Ontario. For. Can., Ont. Region, Sault Ste. Marie, Ont. COFRDA Rep. 3319. 46 p. Pitt, D.G., Krishka, C.S., Bell, F.W. and Lehela, A. 1999. Five-year performance of three conifer stock types on fine sandy loam soils treated with hexazinone. North. J. of Applied Forestry 16: 72–81. Powers, H.R. and Rowan, S.J. 1983. Influence of fertilization and ectomycorrhizae on loblolly pine growth and susceptibility to fusiform rust. South. J. Appl. For. 7: 101–103. Rose, R. and Ketchum, S. 1999. Vegetation Management Research Cooperative. 1998–1999 Annual Report. Forest Research Laboratory, Oregon State University, College of Forestry, Corvallis, Oregon 39 pp. Shipman, R.D. 1958. Planting pine in the Carolina Sandhills. USDA Forest Services Southeastern Forest Experiment Station. Station Paper No. 96. 43 pp. Snyder E.B. and Allen, R.M. 1963. Sampling, nursery, and year-replication effects in a longleaf pine progeny test. pp. 26–27. In: Proceedings of a Forest Genetics Workshop. Southern Forest Tree Improvement Committee. Publication No. 22. Macon, GA.

58 South, D.B. 1996. Top-pruning bareroot hardwoods: a review of the literature. Tree Planters’ Notes 47: 34–40. South, D.B. 1998. Needle-clipping longleaf pine and top-pruning loblolly pine in bareroot nurseries. South. J. Appl. For 22: 235–240. South, D.B. and Mitchell, R.J. 1999. Determining the “optimum” slash pine seedling size for use with four levels of vegetation management on a flatwoods site in Georgia, U.S.A. Can. J. For. Res. 29: 1039–1046. South, D.B., Mitchell, R.J., Zutter, B.R., Balneaves, J.M., Barber, B.L., Nelson, D.G. and Zwolinski, J.B. 1993. Integration of nursery practices and vegetation management: Economic and biological potential for improving regeneration. Can. J. For. Res. 23: 2083–2092. South, D.B., Zwolinski, J.B. and Allen, H.L. 1995. Economic returns from enhancing loblolly pine establishment on two upland sites: Effects of seedling grade, fertilization, hexazinone, and intensive soil cultivation. New For. 10: 239–256. Stein, W.I. 1984. The coastal reforestation systems study – five year results. USDA For. Serv., Pac. Northwest For. Range. Exp. Sta., Portland, OR. Res. Prog. Rep. 10 p. Sutherland, D.C. and Day, R.J. 1988. Container volume affects survival and growth of white spruce, black spruce, and jack pine seedlings: A literature review. N. J. Appl. For. 5: 185– 189. Todd, A.M.D. 1989. Stock types for planting – what’s available, what’s required. pp. 24–32. In: Schivener, B.A. and MacKinnon, J.A. (Eds.) Learning from the Past, Looking into the Future. B.C. Ministry of Forests. FRDA report 030. Toumey, J.W. 1916. Seeding and Planting. John Wiley and Sons, Inc. New York. 455 pp. Vardaman, J.M. 1998. The Vardaman 1998 seminar and the 1998 hypothesis. Vardaman’s Green Sheet October 15: 1–2. Wakeley, P.C. 1954. Planting the southern pines. USDA Agric. Monograph 18. 233 pp. Wakeley, P.C. 1963. Reducing the effects of nursery influences upon provenance tests. pp. 28– 32. In: Proceedings of a Forest Genetics Workshop. Southern Forest Tree Improvement Committee. Publication No. 22. Macon, GA. Ward, J.S. 1997. Influence of initial seedling size and browse protection on height growth: 5-year results. pp. 127–134. In: Landis, T.D., South, D.B. (Eds.) Tech. Coords. National Proceedings, Forest and Conservation Nursery Associations. Gen. Tech. Rep. PNW-GTR389. Wood, J.E. 1990. Black spruce and jack pine plantation performance in boreal Ontario: 10 year results. N. J. of Appl. For. 7: 175–179.