New Forests 13: 191–206, 1996. c 1996 Kluwer Academic Publishers. Printed in the Netherlands. Sowing methods and mulch affect 1+0 northern red oak see...
5 downloads
19 Views
653KB Size
New Forests 13: 191–206, 1996. c 1996 Kluwer Academic Publishers. Printed in the Netherlands.
Sowing methods and mulch affect 1+0 northern red oak seedling quality PATRICIA T. TOMLINSON, GARY L. BUCHSCHACHER and RONALD M. TECLAW US Department of Agriculture, Forest Service, North Central Forest Experiment Station, 5985 Highway K, Rhinelander, WI 54501 USA Received 1 January 1995; accepted in revised form 23 March 1996
Key words: corncob mulch, lateral roots, nursery practices, root collar diameter, sawdust mulch, seedling density, sowing depth Application. This study suggests that sowing depth, seedling density, and mulch can be used to manipulate morphological characteristics of 1+0 northern red oak seedlings. Root or shoot development can be favored depending upon the characteristics required for successful outplanting. Further, use of mulch can delay seedling emergence and increase survival. In our study, use of 2-year-old hardwood sawdust mulch decreased root collar diameter and number of first-order lateral roots. Consequently, we do not recommend this as a mulch material for production of northern red oak seedlings. Abstract. The effects of sowing depth, seedling density and mulches on northern red oak seedling survival and growth were evaluated in Wilson State Forest Nursery in southwest Wisconsin, USA. Sowing depths between 2.2 and 6.3 cm, combined with sowing densities of 75 and 150 acorns m 2 , made up five sowing method plots. Mulch treatments of ground corncobs aged 1 year, hardwood sawdust aged 2 years, and no mulch made up 3 subplot treatments. Treatments resulted in a range of densities from 18 to 148 seedlings m 2 . Mulch delayed emergence and increased seedling survival. Increasing sowing depth also delayed emergence. Corncob mulch increased root collar diameter; however, hardwood sawdust, aged for 2 years, decreased both root collar diameter and the number of permanent first-order lateral roots. Increasing sowing depth decreased root dry mass but increased shoot dry mass. Increasing density from 18 to 148 seedlings m 2 decreased root dry mass in this study. Abbreviations: DM, dry mass; FOLR, first-order lateral roots; RCD, root collar diameter; GenSP-S, general-seed planter – shallow sowing; Hand-S and Hand-D, hand-template sowing – shallow and deep sowing, respectively; LrgSP-S and LrgSP-D, large-seed planter – shallow and deep, respectively.
Introduction Oak (Quercus spp) regeneration is often difficult (Lorimer 1989). Mortality of planted oak seedlings is low but growth is slow (Lorimer 1989, Kaczmarek
The U.S. Government right to retain a non-exclusive, royalty-free license in and to any copyright is acknowledged.
VICTORY PIPS 111924 BIO1KAP nefo005.tex; 14/03/1997; 3:22; v.5; p.1
192 and Pope 1993) resulting in seedlings becoming suppressed by other species. Thus, a critical component in the successful artificial regeneration of oak is having seedlings capable of rapidly initiating new growth after planting. Many factors influence growth after outplanting. Some are associated with nursery practices while others are associated with regeneration practices. At least some of the variation in height growth is associated with root growth (Farmer 1975). It is clear that the characteristics used to define a quality seedling must be based on outplanting performance. Several potentially important factors in seedling quality are within the control of the nursery manager, including: sowing depth and density; cultural practices including mulching, fertilization and watering; seed genetics; lifting procedures; and seedling age. Decreased sowing densities in coniferous seedbeds resulted in increased shoot and root mass, increased root collar diameters, and occasionally increased height (Barnett 1991, Simpson 1991). Decreased sowing density increased the number of first-order lateral roots, height, and root collar diameter of 1+0 northern red oak (Quercus rubra L.) seedlings (Schultz and Thompson 1993). Seedling height and root collar diameter increased in 1+0 and 2+0 nuttall oak seedlings (Quercus nuttallii Palmer) without affecting shoot to root dry mass ratios when sowing densities decreased from 90 to 20 acorns m 2 . Kormanik and Sung (1993) recommended a seedbed density of 54 to 57 acorns m 2 for 15 species of oak, but did not supply any supporting data. Most of the literature on mulch has focused on research with outplanted seedlings. Mulching seedbeds, although common practice, has been studied primarily for effects on disease relationships (Barnard et al. 1993), or on soil characteristics (Van Nierop and White 1958). Increased plant growth and survival have been shown for many types of mulching materials on seedlings although there are some reports of increased plant mortality or lower growth rates due to mulching (Litzow and Pellett 1983). Current practices in Wilson State Forest Nursery are to sow acorns in the fall with a “general-seed planter” (which has no mechanism to control sowing density) at approximately 150 acorns m 2 in five furrows across a 1.2 m wide row (24 cm 3 cm spacing) and to not mulch after sowing. In our research plots, we typically sow at approximately half that density with seed uniformly spaced (12 cm square spacing) and with aged, ground corncob mulch being applied after sowing and then removed in early spring. To provide an operational method for decreased seedbed density, the Forestry Sciences Laboratories in Houghton, MI and Rhinelander, WI (USDA Forest Service, North Central Forest Experiment Station) in conjunction with Toumey Nursery in Watersmeet, MI (USDA Forest Service, Ottawa National Forest)
nefo005.tex; 14/03/1997; 3:22; v.5; p.2
193 developed a large-seed planter, which can mechanize controlled space sowing at densities of 50 to 150 acorns m 2 (Dinkel and Kangas 1991). Current corn harvesting systems leave the corncob in the field making corncobs less available as a mulch material; therefore, interest in other mulching materials, such as aged hardwood sawdust mulch, has arisen. The objectives for the current study were: (1) to examine the effects of sowing depth, density, spacing, and mulch on emergence and growth of 1+0 northern red oak seedlings, and (2) to compare operational practices used in Wilson State Nursery to grow oak seedlings with several experimental practices. Because both root and shoot characteristics may influence outplanting performance, various measures of both were recorded for this study.
Materials and methods Seeds Northern red oak acorns from a small, high-quality stand in northern Wisconsin were collected in September 1991. Acorns were floated to remove damaged acorns, and the sinkers were hydrated overnight before sowing (Teclaw and Isebrands 1986). Experimental design The study was conducted at the Wilson State Forest Nursery, Boscobel WI, USA (43N lat., 90W long.). Seeds were sown in October 1991, seedlings grew during 1992 and were lifted in April 1993. The experiment was installed in a randomized block design with split plots; sowing methods comprised the main plots with mulch treatments being the subplots. Five blocks were set up along a 150-m 1.2-m row. Sowing methods were randomly assigned to 2.7-m plots in the blocks. Space was allowed between sowing method treatment plots to facilitate uniform and representative operation of the machine planters within the sowing method plots. Mulch treatments were randomly assigned to 0.9-m subplots within each sowing method plot. Seed was sown by three types of planters and resulted in five treatments representing four sowing depths, two sowing densities, and two spacing patterns (Table 1). The general-seed planter did not operate correctly in wet, freshly-tilled soil. Rain before and during sowing precluded its use in this study. Rather, acorns were hand-seeded to provide a simulation of the depth and density routinely obtained with this planter. After seeding was complete, the entire bed was drum-rolled and sowing depths measured before mulch
nefo005.tex; 14/03/1997; 3:22; v.5; p.3
194 Table 1. Acorn depth, sowing density, and sowing pattern components of the sowing methods evaluated at Wilson State Forest Nursery for 1+0 northern red oak seedling emergence and growth responses. Sowing Method
Planter Type
Targeted Depth1
Density (acorns m
GenSP-S LrgSP-S LrgSP-D Hand-S Hand-D
General-seed planter Large-seed planter
S S D S D
00
Hand-template planter
00
Spacing Pattern3
Depth (cm)4
150 75
RECT.
75
SQR.
3.4 0.2 b5 5.0 0.2 c 5.8 0.3 d 2.2 0.1 a 6.3 0.5 d
00 00
2 2
)
00 00 00
1
S, Shallow; D, Deep. Based on machine settings and preliminary tests. 3 RECT., rectangular spacing (24 cm 3 to 6 cm dependent on density); SQR, square uniform spacing (12 cm 12 cm). 4 Measured from the soil surface to the top of the acorn; mean SEM of 3 replicates of 5 measurements each. 5 Depths followed by different letters are significantly different by ANOVA plus Tukey’s honestly significant difference analyses (p < 0.05). 2
treatments were applied. Depth was measured for 5 acorns in plots of 3 randomly selected blocks for each sowing method; soil covering the seed was carefully removed and the depth was measured from the top of the acorn to the soil surface. Mulch treatments were hardwood sawdust (aged 2 yrs), ground corncobs (aged 1 yr), or no mulch application. Mulches were applied by hand approximately 2.5 cm deep and held in place with a lightweight mesh designed for frost protection. The mesh was removed in April 1992, but the mulch was left in place. Seedling assessment Emergence measurement areas, 0.3 m long by 1.2 m wide, were marked in the center of each mulch subplot during early spring. Starting May 7, 1992, the number of seedlings in each measurement area was recorded weekly for 5 weeks and at 8 weeks. Weeks to 50% emergence was calculated by linear interpolation between these weekly counts after conversion to a percent of the maximum number of seedlings in the subplot. Survival was based on counts recorded after 8 weeks. Seedlings were lifted by undercutting at 20 cm depth in April 1993. Ten seedlings from each subplot were randomly sampled. Seedling height, root collar diameter (RCD), number of flushes of growth, and number of permanent first-order lateral roots (FOLR) were recorded. Permanent FOLR have been defined as those greater than 1 mm in diameter at the tap root (Kormanik and
nefo005.tex; 14/03/1997; 3:22; v.5; p.4
195 Muse 1986). The seedlings were separated into shoot and root; after drying, the masses were recorded and used to calculate the shoot to root ratio and total seedling mass. Statistical analyses Density data were analyzed using a generalized linear model (ANOVA) in the MGLH module of SYSTAT for Windows software version 5.0 (Wilkinson 1992). Residuals plots indicated that data transformations were not required to meet the assumptions of ANOVA. Mean separations were performed using Fisher’s least significant difference. Where interactions between sowing methods and mulch were significant, differences due to mulch within a sowing method and differences among sowing methods within a mulch were tested separately. Because of the large number of zeros in the initial emergence data, this data set was analyzed with small-sample categorical data methods using StatXact Turbo software version 2.11 (Mehta and Patel 1992). Because sowing methods involved several different variables, it was of interest to determine the influence of each of the known variables on the measured growth responses. Multiple linear regression was used to assess the relative importance of seedling density, sowing depth, spacing pattern, mulch, mulch type, and the interactions of these variables on seedling growth response. A linear model was chosen because the variables displayed little to no continuous representation on which to base any other model description. Models were sequentially simplified to eliminate variables with p-values greater than 0.15 starting with those of the highest-order interaction.
Results Seedling emergence Emergent seedlings on May 7 (Figure 1A) were present in nonmulched subplots and in mulched subplots of the shallow sowing method plots (Hand-S and GenSP-S). Analysis of these data showed a significant interaction between sowing method and mulch treatments (p < 0.01); therefore, simple effects were examined. Nonmulched subplots showed greater emergence on May 7 than mulched subplots for each sowing method. Emergence in nonmulched subplots of Hand-S and GenSP-S sowing method treatments were similar (p = 0.39) and greater than in nonmulched subplots of LrgSP-S, LrgSP-D, or Hand-D plots (p < 0.01). The nonmulched sub-plots of these latter three sowing methods were similar to each other (p = 0.24). For mulched
nefo005.tex; 14/03/1997; 3:22; v.5; p.5
196
Figure 1. Effects of sowing methods and mulches on emergence of northern red oak seedlings. The percent of maximum emergence observed on May 7 (A), and the time after May 7 required for half maximum emergence to be achieved (B) are shown as a function of sowing method. Sowing methods are listed from most shallow sowing depth on the left to deepest on the right; see Table 1 for definitions and characteristics of sowing methods used. The number of weeks after May 7 required to achieve 50% maximum emergence was linearly interpolated from weekly counts. Data are the mean of 5 blocks; the error bars given represent twice the average standard errors of the means. Simple effects of mulch within each sowing method and sowing methods for each mulch treatment were separated using ANOVA plus Fisher’s least significant difference methods. Differences due to mulch within each sowing method are designated by different lowercase letters above each bar. Differences due to sowing method for each mulch treatment are designated by different uppercase letters above the bars.
subplots, Hand-S plots showed more emergent seedlings (p < 0.01) than any of the other sowing methods; the other sowing methods were similar to each other. Within mulched subplots, sawdust and corncob mulches were not distinguishable by emergence on May 7 for any sowing method. Acorns sown at more shallow depths emerged earlier as suggested by the number of weeks after May 7 required to achieve 50% of maximum emergence (Figure 1B). An interaction between sowing method and mulch
nefo005.tex; 14/03/1997; 3:22; v.5; p.6
197 Table 2. Summary statistics (p-values) from ANOVA of northern red oak seedling survival as a function of sowing method (main plots) and mulch treatment (subplots). Source
Seedling Density (m 2 )
Seedling Emergence (% of sown)
Sowing Method Mulch Block Block Sowing Method Sowing Method Mulch
<0.01 <0.01
<0.01
0.02 0.77 <0.01
0.01 0.57 0.03
0.13
was observed (p = 0.02); however, for mulched subplots, weeks from May 7 to 50% of maximum emergence increased from an average of 1.6 weeks in Hand-S plots (2.2 cm sowing depth) to an average of 2.7 weeks in Hand-D plots (6.3 cm sowing depth). In nonmulched subplots, the response was more complex; the longest time to achieve 50% emergence occurred in both the shallowest and deepest sowing method plots (2.1 weeks after May 7 for HandS and Hand-D plots) while plots with intermediate sowing depths (GenSP-S, LrgSP-S, and LrgSP-D plots) took an average of 1.6 weeks. Seedling density Seedling density (Figure 2A) was affected by sowing method, mulch, and mulch by sowing method interactions (Table 2). Despite a sowing density nearly double that of other sowing method treatments, nonmulched subplots of GenSP-S contained about the same seedling density as mulched subplots of the other sowing methods (Figure 2A). However, mulched subplots of GenSP-S contained more seedlings per unit area than any other subplot. When expressed as a percent of the sown seed (Figure 2B), sowing methods still altered the response to mulch treatment as indicated by the presence of a significant mulch by sowing method interaction (Table 2). In mulched subplots, percent emergence in plots with sowing depths 5 cm was generally lower than in more shallowly sown plots. In nonmulched subplots, emergence was generally greater in plots with sowing depths 3.4 cm. This suggested that sowing depth, one of the variables known to differ among sowing methods, altered the effect of mulch on emergence. No differences were found between sawdust and corncob mulches. The presence of mulch increased emergence, although the benefit of mulch decreased somewhat for sowing methods with sowing depths 5 cm (LrgSP-S, LrgSP-D, and Hand-D).
nefo005.tex; 14/03/1997; 3:22; v.5; p.7
198
Figure 2. Effects of sowing methods and mulches on northern red oak seedling density and emergence. (A) Seedling density was counted 8 weeks after the beginning of emergence. (B) Emergence was expressed as a percent of the seed sown. Sowing methods are listed from most shallow sowing depth on the left to deepest on the right; see Table 1 for definitions and characteristics of sowing methods used. Data are the mean of 5 blocks for each mulch and sowing method treatment; the error bars represents twice the average standard errors of the means. Simple effects of mulch within each sowing method and sowing methods for each mulch treatment were separated using ANOVA plus Fisher’s least significant difference methods. Differences due to mulch within each sowing method are designated by different lowercase letters above each bar. Differences due to sowing method for each mulch treatment were tested only for the emergence data and are designated by different uppercase letters above the bars.
Seedling characteristics Seedling height ranged from 28 to 36 cm; the tallest seedlings were found in the highest density treatment (mulched GenSP-S subplots) and the most deeply sown plots (Hand-D and LrgSP-D) (Figure 3A). However, neither density nor depth appeared in the final regression model (Table 3) which contained only a constant.
nefo005.tex; 14/03/1997; 3:22; v.5; p.8
199
Figure 3. Effects of sowing methods and mulches on 1+0 northern red oak seedling characteristics of (A) height, (B) root collar diameter (RCD), and (C) number of first-order lateral roots (# FOLR). Sowing methods are listed from most shallow sowing depth on the left to deepest on the right; see Table 1 for definitions and characteristics of sowing methods used. Data are the mean of 5 blocks (N = 5) each represented by 10 seedlings (n = 10). The standard error bars represent twice the average standard errors of the means.
RCD ranged from 7.3 to 9.1 mm; the larger diameter seedlings were found in the mulched Hand-S subplots, and in the LrgSP-D and Hand-D plots (Figure 3B). The smallest diameter seedlings were found in sawdust mulched
nefo005.tex; 14/03/1997; 3:22; v.5; p.9
200 subplots of the GenSP-S plots. Sawdust mulched subplots generally showed smaller diameters than corncob mulched subplots. The final linear regression model associated greater RCD with mulch presence, corncob mulch type, and rectangular spacing pattern when mulch was present (space mulch interaction) (Table 3). Further analysis of each mulch treatment separately resulted in models which each contained density as a variable; depth was a variable only for the nonmulched subplot model. Differences due to the type of mulch were also observed for the number of FOLR (Figure 3C); FOLR number was generally greater for corncob and nonmulched subplots than for sawdust mulched subplots. Greater sowing depth increased FOLR number in sawdust mulched subplots. Regression analysis associated decreased FOLR number with the use of sawdust mulch; although the response is offset somewhat by increased sowing depth as indicated by the mulch type depth interaction term (Table 3). Increasing density also tended to result in decreased FOLR number. Examination of data for each mulch treatment separately discerned no obvious general trend of FOLR number with either sowing depth or seedling density. The final regression model for nonmulched subplots contained only a constant. Sowing depth was the only variable remaining in the regression model for sawdust mulch (p = 0.01); increasing depth increased FOLR number. Seedling density was the only variable in the model for corncob mulched subplots and it was not significant (p = 0.06). Average shoot dry mass varied between 3.8 g and 5.8 g and fell into two groups: those in the three shallow (-S) sowing depth treatments averaged 4.4 g while those in the deep (-D) sowing depth treatments averaged 5.6 g (Figure 4A). For sowing depths 3.4 cm, sawdust mulch decreased shoot dry mass compared to corncob mulch. Sowing depth was the only variable in the linear regression model of shoot dry mass (Table 3) with increased sowing depth associated with increased shoot DM. Average root dry mass ranged from 6.7 g to 10.4 g. The greatest root dry mass was found in the Hand-S plots, the lowest was found in the highest density subplots (mulched GenSP-S) (Figure 4B). Linear regression associated both increased seedling density and increased sowing depth with decreased root dry mass (Table 3). Mulch type (sawdust) was also a variable in the final model. Increased total seedling dry mass (data not shown) was associated with decreased seedling density (Table 3). This response reflects the importance of seedling density in the root DM response as well as the cancellation of opposite responses of shoot and root dry mass to sowing depth. Shoot to root dry mass ratio (data not shown) was associated with changes in sowing depth, and mulch (Table 3). Several variables remained in the model
nefo005.tex; 14/03/1997; 3:22; v.5; p.10
Height RCD # FOLR4 Shoot DM Root DM Seedling DM Shoot/Root5 0.82(+)
0.02(+)
0.43(+)
0.01(+) <0.01(+) <0.01(+) <0.01(+) 0.24(+)
0.77(–) 0.12(+) 0.03(–)
<0.01(+) <0.01(+)
0.10(–) 0.07(–) 0.01(+)
0.01(+)
p-values for and signs (–/+) of regression coefficients2 constant depth space density mulch depth type
0.30(+)
0.09(–) 0.94(–)
<0.01(+) 0.02(–)
0.04(+)
mulch type 3 0.04(–)
space mulch
Mean square for error (MSE) and squared multiple regression coefficient (R2 ) for the final regression model calculated by SYSTAT. 2 P-values for and sign of final model variable coefficients. P-values indicate the importance of the variable to the measured growth response. The sign indicates whether increases in the treatment variable generally increased (+) or decreased (–) the growth variable response. Final model variables listed resulted from the sequential elimination of treatment variables with p-values > 0.15 starting with the highest-order interaction terms. 3 Mulch type codes sawdust mulch as distinct from corncob or no mulch; mulch codes sawdust mulch as similar to corncob and distinct from no mulch. 4 Block was also in the final model (p = 0.11[–]). 5 Depth space mulch type (p = 0.10[–]), space mulch type (p = 0.16[+], and depth space (p = 0.96[–]) were also in the final model.
1
11.71 0.553 3.23 0.66 1.17 2.73 0.006
Variable 0.32 0.38 0.46 0.42 0.51 0.35 0.71
Regression1 MSE R2
Table 3. Multiple linear regression results indicating the significance (p-values) and sign of coefficients for treatment variables on 1+0 northern red oak seedling growth response.
201
nefo005.tex; 14/03/1997; 3:22; v.5; p.11
202
Figure 4. Effects of sowing methods and mulches on dry mass of (A) Shoots, and (B) Roots of 1+0 northern red oak seedlings. Sowing methods are listed from most shallow sowing depth on the left to deepest on the right; see Table 1 for definitions and characteristics of sowing methods used. Data are the mean of 5 blocks (N = 5) each represented by 10 seedlings (n = 10). The standard error bars represent twice the average standard errors of the means.
because the p-value for depth space mulch type was less than 0.15; this interaction term suggests that the shoot to root ratio responds to depth under particular circumstances of spacing pattern and mulch type. The number of flushes of growth (data not shown) was not significantly different among treatments and averaged 2.8.
nefo005.tex; 14/03/1997; 3:22; v.5; p.12
203 Discussion Emergence and density Mulch strongly influenced seedling emergence as well as seedling density. Mulch delayed emergence but improved survival. Frosts during early spring are common in this area, and seedling damage typical of that induced by frosts were observed during this study. Because the likelihood of spring frosts decreases as the season progresses, mulch would decrease the risk of seedling damage due to these frosts by delaying seedling emergence. The apparent interaction between sowing depth and mulch could be explained, at least in part, because sowing depth was measured prior to mulch application. Thus, the total depth on mulched subplots was approximately 2.5 cm deeper than for nonmulched subplots of the same sowing depth. However, nonmulched subplots when compared to mulched subplots of similar total depth (sowing depth plus 2.5 cm mulch depth) showed lower frequency of survival. Therefore, other attributes of mulch, such as effects on soil moisture retention and on soil temperature, could be involved. Survival rates of blue (Q. douglasii) and valley (Q. lobata) oaks sown at depths of 1.3 and 5.2 cm were similar (Tietje et al. 1991) while those of four southern oak species (Q. alba, Q. falcata, Q. nigra, and Q. prinus) sown 0.6 cm deep was lower than when sown 1.2 and 2.5 cm deep (Shipman 1962). Depth of sowing and mulch, by affecting earliness of emergence, might be expected to be of more importance in the current study which was conducted at a more northern latitude than either of these referenced studies. Seedling characteristics Increased height appeared to be associated with increased seedling density. Seedling density was not well represented, particularly at higher values. Only the two mulched GenSP-S treatments provided densities greater than 75 seedlings m 2 (see Figure 1A). In addition, only low densities were represented for the square spacing pattern making distinction between spacingand density- related responses difficult. Decreased (Kennedy 1988, Schultz and Thompson 1993), increased and no height responses (Shipman 1962) have been reported for oak seedlings in response to increasing bed densities. Increased height with increased density has been observed with other large-seeded deciduous species such as walnut (Juglans spp) (personal communication, T. H. Hill II, Nursery Manager, Wilson State Forest Nursery, 1994). In conifer seedlings, increased height is occasionally observed as a result of increased density (Simpson 1991).
nefo005.tex; 14/03/1997; 3:22; v.5; p.13
204 RCD response to sowing methods and mulch was complex. In general, use of the hardwood sawdust mulch decreased RCD. The use of corncob mulch, particularly for the square spacing pattern, increased RCD. Increased density was associated with decreased RCD in other studies on northern red oak (Schultz and Thompson 1993), as well as on 1+0 and 2+0 nuttall oak (Kennedy 1988), longleaf pine (Pinus palustris Mill.) (Barnett 1991) and interior spruce (Picea glauca and P. englemannii) (Simpson 1991) seedlings. Additional data is needed to determine the important cultural factors affecting RCD of northern red oak. The response of FOLR number to sowing methods and mulch was even more complex than was response of RCD. Sawdust mulch generally decreased FOLR number, providing further evidence against the use of this mulch material. Increasing density has been reported to decrease FOLR number (Schultz and Thompson 1993). In our study, density could not be associated with major changes in FOLR number. Further study, perhaps over a wider range of densities, is needed to identify the importance of seedling density as well as depth and mulch to FOLR number response. The increase in shoot DM with increasing sowing depth appears to offset the decrease in root DM with increasing sowing depth such that the regression model for seedling DM did not contain depth, only density. Shoot to root ratio was increased by increased sowing depth, consistent with the relative influences of depth on root and shoot DM. The influence of mulch on shoot DM was also apparent in shoot to root ratios.
Conclusions Mulch was important in seedling emergence and survival. It delayed onset of emergence and increased seedling survival. Increased sowing depth could offset some but not all of these mulch effects. No differential influence due to the type of mulch (sawdust vs. corncob) was observed on emergence and survival variables. Increasing seedling density appeared to increase seedling height, decrease RCD, and decrease root and seedling DM. Increasing sowing depth generally increased height and shoot DM while it decreased root DM. Shallow sowing depth (2.2 cm) increased RCD and root DM although the response to sowing depth appeared nonlinear. Interactions with other variables of the sowing methods and mulch treatments used could be involved in the non-linearity. Mulch influenced RCD, FOLR number, shoot DM, and shoot to root ratio. For FOLR number and RCD, this effect appeared to be related to a negative influence of sawdust mulch.
nefo005.tex; 14/03/1997; 3:22; v.5; p.14
205 There is general agreement in the literature that a large seedling with a large root system is needed for successful regeneration (Farmer 1975, Larson 1975, Thompson and Schultz 1993) although regeneration practices, such as procedures used to plant stock, are also important (Schultz and Thompson 1993). The need for a large root system with many FOLR has been highlighted by Kormanik and Muse (1986). If a large root system is used as the basis for defining a high-quality seedling, sowing 5 cm or less deep, controlling seedling density, and mulching were important conditions for the production of northern red oak seedlings in this study. Hardwood sawdust that has been aged only 2 years is not recommended as a mulch material because of its negative influence on root development. Additional study over a broader range of seedling densities, depths, and spacings is needed to more clearly define the influences of these variables on northern red oak seedling growth and morphology.
Acknowledgments The large-seed planter used in this study was designed and engineered by the Engineering Technology for Managing Northern Forest Stands project of the North Central Forest Experiment Station, USDA Forest Service. The authors wish to acknowledge the use of this prototype planter and the efforts of Mr. Roy Kangas (Engineering unit) and Mr. Gary Dinkel (Toumey Nursery, Ottawa National Forest) in the use of this equipment. We wish to thank Mr. David Rugg for his assistance with the experimental design and statistical analyses shown in this paper. Partial support for this work was provided by the Wisconsin Department of Natural Resources, Bureau of Forestry, which provided both financial support and in-kind support through the use of the facilities at the Wilson State Forest Nursery in Boscobel, Wisconsin.
References Barnard, E. L., Fraedrich, S. W. and Gilly, S. P. 1994. Interactions between seedbed mulches and seedling disease development, pp 97–104. In: Proceedings of the Northeastern and Intermountain Forest and Conservation Nursery Association Meeting, St. Louis, Missouri, August 2–5, 1993. USDA-Forest Service Gen. Tech. Report RM-243. Barnett, J. P. 1991. Seedbed densities and sowing and lifting dates affect nursery development and field survival of longleaf pine seedlings. Tree Planters’ Notes 42(3): 28–31. Dinkel, G. and Kangas, R. 1991. New Style Acorn Seeder. Tree Planters’ Notes 42(3): 16–17. Farmer, R. E. 1975. Dormancy and root regeneration of northern red oak. Can. J. Forest Res. 5: 176–185. Kaczmarek, D. J. and Pope, P. E. 1993. Seedling morphology related to growth and survival of northern red oak, pp. 11. In: Thompson, J. R., Schultz, R. C. and Van Sambeek, J. W. (Eds)
nefo005.tex; 14/03/1997; 3:22; v.5; p.15
206 Fifth Workshop on Seedling Physiology and Growth Problems in Oak Plantings, March 4–5, 1992, Ames, IA (abstracts). USDA Forest Service Gen. Tech. Rep. NC-158. Kennedy, H. E., Jr. 1988. Effects of seedbed density and row spacing on growth and nutrient concentrations of nuttall oak and green ash seedlings. USDA Forest Service Res. Note SO-349. 5 pp. Kormanik P. P. and Muse, H. D. 1986. Lateral roots a potential indicator of nursery seedling quality, pp. 187–190. In: TAPPI Proceedings 1986 Research and Development Conference, Raleigh, NC. Kormanik, P. P. and S. S. Sung 1993. Toward a single nursery protocol for oak seedlings, pp. 89–98. In: Proceedings of the 22nd Southern Forest Tree Improvement Conference, June 14–17, 1993, Atlanta, GA. Larson, M. M. 1975. Pruning northern red oak nursery seedlings: effects on root regeneration and early growth. Can. J. For. Res. 5: 381–386. Litzow, M. and Pellett, H. 1983. Influence of mulch materials on growth of green ash. J. Arboric. 9: 7–11. Lorimer, C. G. 1989. The Oak Regeneration Problem: New Evidence on Causes and Possible Solutions. Forest Resource Analyses No. 8, Agricultural Bulletin R3484. Department of Forestry, School of Natural Resources, College of Agricultural and Life Sciences, University of Wisconsin-Madison, USA, 34 pp. Mehta, C. and Patel, N. 1992. StatXact-Turbo: Statistical Software for Exact Nonparametric Inference. User manual. CYTEL Software Corp. Cambridge, MA, 376 pp. Schultz, R. C. and Thompson, J. R. 1993. Northern red oak seedlings produced at three nursery bed densities with and without undercutting: 4-year field results, pp. 15. In: Thompson, J. R., Schultz, R. C. and Van Sambeek, J. W. (Eds) Fifth Workshop on Seedling Physiology and Growth Problems in Oak Plantings, March 4–5, 1992, Ames, IA (abstracts). USDA Forest Service Gen. Tech. Rep. NC-158. Shipman, R. D. 1962. Nursery-seeded hardwoods – influenced by depth and density of sowing. Tree Planters’ Notes 54: 27–32. Simpson, D. G. 1991. Growing density and container volume affect nursery and field growth of interior spruce seedlings. North. J. Appl. For. 8: 160–165. Teclaw, R. M. and Isebrands, J. G. 1986. Collection procedures affect germination of northern red oak (Quercus rubra L.) acorns. Tree Planters’ Notes 37(3): 8–12. Thompson, J. R. and Schultz, R. C. 1993. Three-year survival and growth of northern red oak seedlings with respect to root grade, pp. 9. In: Thompson, J. R., Schultz, R. C. and Van Sambeek, J. W. (Eds) Fifth Workshop on Seedling Physiology and Growth Problems in Oak Plantings, March 4–5, 1992, Ames, IA (abstracts). USDA Forest Service Gen. Tech. Rep. NC-158. Tietje, W. D., Nives, S. L., Honig, J. A. and Weitkamp, W. H. 1991. Effect of acorn planting depth on depredation, emergence, and survival of valley and blue oak, pp 14–20. In: USDA Forest Service Gen. Tech. Rep. PSW-126. Van Nierop, E. T. and White D. P. 1958. Evaluation of several organic mulching materials on a sandy loam forest nursery soil. J. For. 56: 23–27. Wilkinson, L. 1992. SYSTAT for Windows: Statistics, Version 5 Edition. Evanston, IL, SYSTAT, Inc., 750 p.
nefo005.tex; 14/03/1997; 3:22; v.5; p.16