Summary Fertilizer was applied annually for eight years to individual ramets in a loblolly pine (Pinus taeda L.) seed orchard at rates ranging from 0 to 448 kg nitrogen (N) ha−1 year−1. Clonal effects accounted for a major source of variation in both flowering and foliar nutrient concentrations. Foliar N concentrations were generally correlated with the intensity of fertilizer application, but were only weakly correlated with flowering. There was a long-term trend for increasing concen-trations of foliar manganese (Mn) and boron (B), and decreas-ing concentrations of magnesium (Mg) and zinc (Zn) with increasing fertilizer rates, although only the differences in Mn concentration were statistically significant. Fertilizer had little effect on the concentrations of other foliar macro- or micronu-trients during the study. The optimum fertilizer rate for flow-ering was 224 kg N ha−1 year−1.
Keywords: boron, clonal effects, flowering, magnesium, man-ganese, nitrogen, Pinus taeda, zinc.
Introduction
Southern pine seed orchards are routinely fertilized with nitro-gen (N) to enhance and maintain cone and seed production (Jett 1986), because N is considered the critical nutrient in flower stimulation (Schmidtling 1974, Sprague et al. 1978). However, current fertilization rate recommendations are based on a limited data set.
Fertilization rates and formulations are sometimes based on soil analysis (Sprague et al. 1978), but the procedure is not applicable for N because there is no reliable test for available N, and it is only marginally satisfactory for other nutrients.
Foliar analysis seems to be a better method for diagnosing the mineral requirements of pine seed orchards than soil analy-sis (Jett 1987). Minimum standards of foliar nutrients for some minerals have been established for forest stands of some south-ern pines (Leaf 1968, Pritchett 1968, Wells 1968), but little information has been published relating to seed production.
Based on a fertilizer rate experiment in slash pine (Pinus elliottii Engelm. var. elliottii) and longleaf pine (P. palustris Mill.) stands, Shoulders (1968) observed the best flowering and cone production response at the highest rate tested, i.e., 168 kg N ha−1 year−1. This rate is lower than the routinely used rates of 202 to 224 kg N ha−1 year−1, which were originally
recommended for loblolly pine seed orchards by Webster (1974).
Thus, although fertilizers have been applied to many or-chards at high rates for a number of years, optimum fertilizer rates have rarely been determined, and the long-term effects of high fertilizer rates on cone and seed production as well as foliar micronutrient concentrations are not known. In addition, clonal effects on flowering and foliar nutrient concentrations are also important, but have rarely been considered in orchard fertilization studies (Schmidtling 1988, 1991). The objective of this experiment was to examine environmental and genetic variation in flowering and foliar concentrations of macro- and micronutrients of orchard ramets fertilized at various rates.
Materials and methods
The study was established at the US Forest Service Erambert Seed Orchard in south Mississippi. In most production seed orchards, age of ramets and clonal composition vary consider-ably. Because the broad-sense heritability of flowering traits is high, it is important to balance the clonal composition care-fully (Schmidtling 1974). Experiments in which whole or-chard blocks are treated and compared are often difficult to interpret because of high error values, therefore, in this experi-ment, individual ramets were treated.
The experiment was installed in the Alabama seed source orchard because it had not been fertilized for the previous three years. Ten clones were chosen from the 50 available on the basis of having a minimum of 10 ramets, each 15 years of age, when the experiment was initiated. The ramets had been grafted on woods-run nursery stock and averaged 25 cm in diameter at the start of the study in 1982. Spacing in the orchard was 4.6 × 9.1 m. Soils were moderately well-drained sandy loams.
Beginning in summer 1982, five fertilization treatments were applied yearly in a factorial design to two ramets each of the 10 clones. The treatments consisted of a control (no fertil-izer application), and 112, 224, 336 and 448 kg N ha−1 year−1. A total of 100 ramets were included in the experiment. Be-cause severe phosphorus (P) or potassium (K) deficiency may inhibit flowering, a 50/50 fertilizer mix of ammonium nitrate and 13/13/13 N,P,K was applied. Thus, the 112 kg N ha−1 year−1 treatment comprised 112 kg N ha−1, 32 kg P2O5ha−1 and
Genetic and environmental variation of foliar nutrient concentrations
and strobilus initiation in fertilized loblolly pine seed orchard ramets
R. C. SCHMIDTLING
USDA Forest Service, Southern Forest Experiment Station, Gulfport, MS 39503, USA
Received March 11, 1994
32 kg K2O ha−1. This mixture approximates current orchard fertilization practice.
Fertilizer was broadcast within the drip line (beneath the crown) of the individual ramets each July from 1982 through 1989. Applying fertilizers in July increases both male and female flowering in loblolly pine (Schmidtling 1983b).
From 1983 through 1986, total male and female strobili were estimated by counting reproductive structures in the southeast quadrant of the crowns and multiplying by four. This was done in the spring, just after the terminal buds elongated, but before needle elongation had occurred. Foliar samples were collected in the fall of 1982, 1983, 1984, 1985 and 1989, and oven dried to constant weight. Foliar samples consisted of the most recently formed needles from at least three branch tips from the upper crown. The samples were collected in late September, just after formation of female strobili initials (Schmidtling 1975). Foliar nutrients were analyzed by a com-mercial agricultural laboratory. Nitrogen was determined by traditional Kjeldahl methods. For all other elements, the sam-ples were prepared by acid digestion with nitric and perchloric acid. Analysis for K, calcium (Ca), magnesium (Mg), iron (Fe), boron (B), manganese (Mn), aluminum (Al), zinc (Zn) and copper (Cu) was by atomic absorption spectrometry. Phos-phate was analyzed by the ammonium molybdate method. Sulfur (S) was analyzed by adding barium chloride and meas-uring turbidity spectrophotometrically. Boron was determined with azomethine-H.
The SAS General Linear Model (GLM) procedure (SAS Institute, Cary, NC) was used to test differences among treat-ment means in the analysis of variance for the completely random, single-tree plot design. The sources of variation (and degrees of freedom) of the initial analyses were clone (9),
treatment (4), year (3), clone × treatment (36), clone × year (27), treatment × year (12) and residual (error) (308). The overall analyses showed that there was a significant year × clone or year × treatment interaction for nearly every trait analyzed, so analyses were done separately for each year.
Simple and multiple regression was used to test relation-ships between flowering (dependent variables) and foliar nu-trients (independent variables). Probabilities of less than 0.05 for no difference were considered significant.
Results and discussion
Flowering
Clonal effects on both male and female flowering were usually large and significant (Table 1). The four-year average for number of female strobili was 321 and ranged from 101 for Clone 22 to 787 for Clone 8, nearly an 8-fold difference (Table 2). There was a 5-fold difference in average numbers of male strobili clusters, ranging from 400 for Clone 15 to 2031 for Clone 2. There was only a weak relationship between male and female flowering on a clonal basis (r = 0.438), which agrees with previous observations (cf. Schmidtling 1983a).
Fertilizer effects on both male and female flowering were significant each year except 1985 for female flowering and for the first two years for male flowering (Figure 1, Table 1). Clone
× treatment interactions were not significant. In 1983 and 1984, the 224 kg ha−1 year−1 rate was optimum for both male and female flowering (Figure 1). In 1983, controls averaged around 160 strobili per ramet. Fertilizing with 112 kg ha−1 year−1 increased flowering to over 180 strobili; fertilizing with 224 kg ha−1 year−1 increased flowering to over 220 strobili per ramet. Higher fertilizer rates did not significantly increase
Table 1. Probabilities from the analysis of variance of foliar nutrient concentration and flowering. The probability of no effect is less than the values shown. In 1982 through 1985, a 100-tree sample (10 clones × 5 treatments × 2 ramets) was used in the foliar analyses. In 1989, samples from only 47 trees were available (6 clones × 5 treatments with one or two ramets per clone per treatment combination). Foliar samples were taken in mid-to late September. The clone × treatment interaction is not shown because it was never statistically significant at the 0.05 level.
Variable 1982 1983 1984 1985 19891
Clone Treatment Clone Treatment Clone Treatment Clone Treatment Clone Treatment
Female2 0.000*3 0.004* 0.000* 0.018* 0.000* 0.150 0.000* 0.006* -- --Male2 0.000* 0.049* 0.000* 0.001* 0.000* 0.146 0.000* 0.316 -- --N 0.002* 0.018* 0.000* 0.000* 0.001* 0.000* 0.045* 0.392 0.000* 0.009*
P 0.000* 0.163 0.256 0.431 0.369 0.004* 0.173 0.146 0.174 0.681
K 0.001* 0.517 0.004* 0.845 0.001* 0.339 0.107 0.605 0.002* 0.961
S 0.068 0.456 -- -- 0.010* 0.476 0.397 0.991 0.218 0.863
Mg 0.000* 0.277 0.005* 0.148 0.000* 0.348 0.000* 0.944 0.016* 0.176 Ca 0.070 0.739 0.001* 0.162 0.014* 0.389 0.001* 0.514 0.221 0.225
Fe 0.002* 0.332 0.608 0.837 0.005* 0.261 0.576 0.067 0.056 0.072
Al 0.027* 0.820 0.000* 0.422 0.012* 0.541 0.133 0.101 0.138 0.023* Mn 0.002* 0.596 0.048* 0.892 0.002* 0.257 0.043* 0.218 0.256 0.004*
B 0.043* 0.501 0.028* 0.359 0.000* 0.941 0.113 0.075 0.125 0.192
Cu 0.399 0.609 0.017* 0.667 0.030* 0.089 0.080 0.119 0.069 0.263
Zn 0.309 0.717 0.000* 0.602 0.001* 0.101 0.001* 0.325 0.037* 0.086
1 Based on type III sums of squares from SAS GLM analysis.
2 Flower counts for the spring following foliar sampling.
flowering over the 224 kg ha−1 year−1 rate, and the highest fertilizer rate was not as effective as the 224 kg ha−1 year−1 rate. The response in number of male strobili per ramet followed a
similar pattern to that of female strobili.
The flowering response pattern changed somewhat in 1985 and 1986, when it appeared that fertilizer rates higher than 224
Table 2. Clonal means for flowering and foliar nutrients. Four-year averages over all treatments (1989 data not included).
Variable Clone Mean
2 4 8 11 15 18 20 22 24 32
Female strobili 385 349 787 218 173 387 251 101 333 224 321
Male strobili 2031 1073 1544 543 400 1425 1598 1345 1471 1040 1247
Nitrogen (%) 1.64 1.62 1.51 1.57 1.66 1.64 1.64 1.65 1.53 1.66 1.61
Phosphorus (ppm) 1800 1845 1662 1765 1785 1853 1823 1890 1610 1805 1784
Potassium (ppm) 9440 8375 8613 8655 8730 7963 8150 9160 9013 8038 8614
Sulfur (ppm) 853 670 727 940 910 937 877 893 853 977 864
Calcium (ppm) 1968 1968 2060 1748 1818 2140 1788 2215 1820 1833 1936
Magnesium (ppm) 1423 1520 1645 1548 1290 1848 1683 1510 1405 1528 1540
Iron (ppm) 59.8 59.5 52.4 51.7 53.9 61.6 57.1 60.5 57.7 58.6 57.3
Aluminum (ppm) 495 462 439 386 509 489 483 554 437 539 479
Manganese (ppm) 323 330 392 330 356 476 325 308 299 341 348
Boron (ppm) 13.9 18.9 15.5 16.7 18.8 17.6 16.3 16.1 14.5 15.4 16.4
Copper (ppm) 5.32 4.55 5.38 5.20 5.00 4.38 4.40 4.73 4.55 4.23 4.77
Zinc (ppm) 28.7 25.1 29.7 30.4 28.0 24.3 28.0 32.1 24.9 27.8 27.9
kg ha−1 year−1, especially the 448 kg ha−1 year−1 rate, en-hanced flowering. It is hypothesized that rather than indicating a fundamental change in the flowering response, this change probably resulted from root growth into fertilized areas by non-study trees. Because most feeder roots occur within the drip line of the crown, intensive root growth probably occurred in response to nutrient gradients between fertilized and unfer-tilized trees. Nitrogen fertilization increases root growth in slash pine (Schultz 1969), so it is likely that roots from adja-cent, unfertilized trees extended and proliferated into the area beneath the fertilized trees, and absorbed nutrients, thus giving the appearance of a reduction in the effectiveness of the fertil-izer treatments for the study trees. One of the striking effects of N fertilization in agronomic crops is increased growth of roots, which brings the plant into contact with a greater quan-tity of nutrients (Grunes 1959).
Subsoiling between the ramets to minimize this effect was
not carried out because of the risk of increased root diseases (Webb and Alexander 1982, 1983). However, such risks appear to have been exaggerated (Schmidtling 1986), and subsoiling is now common practice (Jett 1987).
Foliar nutrients
Overall concentrations of foliar macro- and micronutrients varied considerably from year to year (Figures 2 and 3) as a result of differences in the growth phase when samples were taken (cf. Wells 1968, Smith et al. 1970). Although the timing of sampling was consistent from year to year according to the calendar, growth activity in September can range from com-pletely dormant to actively growing in southern pines, depend-ing on climatic conditions (Allen 1964, Griffdepend-ing and Elam 1971). More consistent results would have been obtained if sampling had been done during the dormant season, but such
samples would not necessarily reflect conditions during stro-bilus initiation.
Analysis of foliar nutrients showed that clonal effects were significant for N, K, Mg, Ca, Al, Mn, B and Zn for at least three of the first four years sampled, and for two out of four years for Fe and Cu (Table 1). The 1989 foliar analyses were based on only six clones and 47 ramets because of losses during 1986--1989. Despite the limited sample, clonal effects were signifi-cant for N, K, Mg and Zn in 1989 (Table 1). Clonal effects for P and S were significant in only one year in four, perhaps because of differences in mycorrhizal associations, or because the analyses were not sensitive enough.
Clonal effects on foliar nutrients were smaller than those on flowering (Table 2). For instance, N concentrations only varied from 1.51% for Clone 8 to 1.66% for Clone 15. There was no obvious relationship between flowering and any of the foliar nutrients on a clonal basis. For instance, Clone 15 had the
highest N concentration but ranked only ninth in flowering, whereas Clone 8 had the lowest N concentration but ranked first in female flowering.
In previous studies involving the use of different types of rootstocks for grafting southern pines, foliar nutrient concen-trations were affected by rootstock type, but the effect of the scion clone was much greater than that of the rootstock (Schmidtling 1988, 1991). Clonal effects on foliar nutrient concentrations were smaller than those on flowering as a result of genetic differences in the aboveground portions, because these orchard ramets were grafted on unselected, woods-run seedlings. It appears that a major factor in determining nutrient uptake in pines is the nutrient gradients produced by incorpo-ration of the nutrients into foliar tissue.
The only consistent effect of the fertilizer treatments on foliar nutrients was for N (Table 1, Figures 2 and 3). Although only 47 of the original 100 ramets survived until the 1989
samplings (fertilizer treatments were continued through 1989, but no measurements were taken), and clone × treatment com-binations were poorly balanced, treatment effects were statis-tically significant, and the highest N concentrations were in foliage from trees from the highest fertilizer rate and the lowest from control trees that were not fertilized (Figure 2A).
In the earlier samplings, there was a decrease in foliar P concentrations at the higher fertilizer rates, even though P was included in the fertilizer mix. A similar response was observed by Webster (1974). The difference was statistically significant in 1984, but not in other years (Figure 2B). This difference was not evident in the final sampling in 1989. The only other macronutrient for which there may be a fertilizer effect is Mg (Figure 2E). Foliar concentrations of Mg were inversely re-lated to fertilizer rate in the 1989 sampling, but the differences were not statistically significant.
There were no significant fertilizer treatment effects on the concentrations of micronutrients for the first four years (Fig-ures 3A to 3F, Table 1), although there were significant treat-ment effects on Al and Mn concentrations (Figures 3B and 3C) in 1989. The effects of fertilizer on the concentration of Al are difficult to interpret, because foliage from the control and the highest fertilizer rate had the highest concentrations of Al and did not differ from each other (Figure 3B). Foliage with the highest foliar concentration of Al received the lowest fertilizer rate, 112 kg N ha−1 year−1.
In 1989, foliar Mn concentrations increased significantly with increasing fertilizer rates (Figure 3C). The same trend was apparent in the concentration of B (Figure 3D), but the differ-ences were not statistically significant. The only micronutrient showing a possible decrease in foliar concentration in response to fertilization was Zn (Figure 3F). In 1989, the concentration of foliar Zn was inversely related to rate of fertilization, but the differences were not statistically significant. The results of the 1989 samplings were not conclusive because of the loss of ramets. A general decrease in foliar micronutrients due to heavy N fertilization, which was observed in Pinus radiata (D. Don) growing on deep sand (Woods 1983), did not occur in this experiment, with the possible exception of Mg and Zn.
Flowering was not significantly related to concentrations of macro- or micronutrients in simple or multiple regressions on an individual-tree basis. Wilcox et al. (1991) reported that ratios of nutrients, especially Ca/P and Ca/K, were related to flowering, but no significant relationship was found in this experiment for these ratios, or for the nutrient ratios N/P, Mg/P and Mg/N. There was no relationship between concentration of foliar N and flowering at the individual-tree level.
Conclusions
Considering the cost of fertilizers and possible environmental effects, such as runoff and groundwater contamination, the optimum fertilization rate for flowering and seed production in loblolly pine seed orchards is about 224 kg N ha−1 year−1, which is identical to the optimum rate determined for vegeta-tive growth of loblolly pine in plantations (Ballard 1981).
The usefulness of foliar analysis as a basis for making
fertilizer recommendations for seed orchards depends on the timing of sample collection and the accuracy of the analysis used. Collecting samples during the dormant season would probably result in better control of year by year variation. Although high rates of fertilization had only minor effects on the concentration of foliar micronutrients in the soils of the Erambert Seed Orchard, trends over the eight years of the study indicate possible problems in older orchards.
Acknowledgments
The author is indebted to Jim McConnell, Jerry Windham and other members of the Tree Improvement Program of the Southern Region, USDA Forest Service, for help in carrying out this study. Helpful reviews were provided by Drs. T. Blush, M. Rutter, J.B. Jett and an anonymous referee.
References
Allen, R.M. 1964. Contributions of roots, stems, and leaves to height growth of longleaf pine. For. Sci. 10:14--16.
Ballard, R. 1981. Optimum nitrogen rates for fertilization of loblolly pine plantations. South. J. Appl. For. 5:212--216.
Griffing, C.G. and W.W. Elam. 1971. Height growth patterns of lob-lolly pine saplings. For. Sci. 17:52--54.
Grunes, D.L. 1959. Effect of nitrogen on the availability of soil and fertilizer phosphorous to plants. Adv. Agron. 11:369--396. Jett, J.B. 1986. Reaching full production: a review of seed orchard
management in the southeastern United States. In Proc. IUFRO Conf. on Breeding Theory, Progeny Testing and Seed Orchards. Williamsburg, VA, pp 34--58.
Jett, J.B. 1987. Seed orchard management: something old and some-thing new. In Proc. 19th South. For. Tree Improv. Conf. College Station, TX, pp 160--171.
Leaf, A.L. 1968. K, Mg, and S deficiencies in forest trees. In Forest Fertilization: Theory and Practice. Tennessee Valley Authority, Muscle Shoals, AL, pp 88--122.
Pritchett, W.L. 1968. Progress in development of techniques and standards for soil and foliar diagnosis of phosphorus deficiency in slash pine. In Forest Fertilization: Theory and Practice. Tennessee Valley Authority, Muscle Shoals, AL, pp 81--87.
Schmidtling, R.C. 1974. Fruitfulness in conifers: nitrogen, carbohy-drate, and genetic control. In Proc. 3rd North Am. For. Biol. Work-shop. Ft. Collins, CO, pp 148--164.
Schmidtling, R.C. 1975. Fertilizer timing and formulation effect flow-ering in a loblolly pine seed orchard. In Proc. 13th South. For. Tree Improv. Conf. Raleigh, NC, pp 153--160.
Schmidtling, R.C. 1983a. Genetic variation in fruitfulness in a loblolly pine (Pinus taeda L.) seed orchard. Silvae Genet. 32:76--80. Schmidtling, R.C. 1983b. Timing of fertilizer application important
for management of southern pine seed orchards. South. J. Appl. For. 7:76--81.
Schmidtling, R.C. 1986. Long-term effects of subsoiling and fertiliza-tion on growth and flowering in a Virginia pine seed orchard. In
Proc. 9th North Am. For. Biol. Workshop. Stillwater, OK, pp 267--273.
Schmidtling, R.C. 1988. Influence of rootstock on flowering, growth, and foliar nutrients of slash pine grafts. In Proc. 10th North Am. For. Biol. Workshop. Vancouver, BC, Canada, pp 120--127.
Schultz, R.P. 1969. Effect of seed source and fertilization on slash pine seedling growth and development. USDA For. Serv. Res. Paper SE-49. Southeastern For. Exp. Stn., Asheville, NC, 8 p.
Shoulders, E. 1968. Fertilization increases longleaf and slash pine flower and seed crops in Louisiana. J. For. 66:193--197.
Smith, W.H., G.L. Switzer and L.E. Nelson. 1970. Development of the shoot system of young loblolly pine. 1. Apical growth and nitrogen concentration. For. Sci. 16:483--490.
Sprague, J., J.B. Jett and B. Zobel. 1978. The management of southern pine seed orchards to increase seed production. In Proc. Flowering and Seed Development in Trees. Starkville, MS, pp 145--162. Webb, R.S. and S.A. Alexander. 1982. Subsoiling and reduced radial
growth in seed orchard loblolly pine established on sandy soils. South. J. Appl. For. 6:163--167.
Webb, R.S. and S.A. Alexander. 1983. Incidence of resin-soaked roots in subsoiled loblolly pine seed orchards in sandy soils. South. J. Appl. For. 7:104--107.
Webster, S.R. 1974. Nutrition of seed orchard pine in Virginia. Ph.D. Dissertation, North Carolina State Univ., 185 p.
Wells, C.G. 1968. Techniques and standards for foliar analysis of N deficiency in loblolly pine. In Forest Fertilization: Theory and Practice. Tennessee Valley Authority, Muscle Shoals, AL, pp 72--85.
Wilcox, P.L., R.L. Allen and J.B. Jett. 1991. Foliar nutrient variation in loblolly pine seed orchards. In Proc. 21st South. For. Tree Im-prov. Conf. Knoxville, TN, pp 120--123.
