Nitrogen Turnover In Forest Floors Of Coastal Douglas-Fir At Sites Differing In Soil Nitrogen Capital
C. E. Prescott
C. E. PRESCOTT [1,4]
H. N. CHAPPELL [2,5]
L. VESTERDAL [3]
Abstract. Nitrogen cycling is generally considered to be more rapid on sites with high availability of N; this is usually associated with differences in tree species composition. We tested whether N cycling in stands of a single tree species increased with increasing mineral soil nitrogen capital. Rates of N cycling in nine stands of Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) were estimated by measuring annual N input in litter, N content of the forest floor, and net N mineralization rate in the forest floor. Rates of C and N turnover were estimated from the litterfall:forest floor ratio. The amount of N returned in litter increased as soil N capital increased. The increase in litter N content resulted from both increased litter mass (C content) and increased N concentration in litter. Despite the greater litter input, forest floor mass was smaller at N-rich sites, indicating more rapid turnover of the forest floor on N-rich sites. There was a positive relationship between fractional annual loss of N from the forest floor and soil N capital. Therefore, even without changes in tree species composition, sites with greater soil N capital returned more N in annual litterfall and had faster turnover of N in the forest floor. Fractional annual loss of C also increased with increasing soil N capital, indicating faster decomposition on N-rich sites. The rate of net N mineralization during laboratory incubations of the forest floor was not correlated with soil N capital or N concentrations of litter, but was related to the C:N ratio of the forest floor. Net N mineralization was appreciable only at two sites where forest floor C:N ratios were [less than]35. The rate of net N mineralization or C:N ratio of the forest floor were not good indicators of N availability at these sites. The results of this study are consistent with the hypothesis that rates of N cycling are faster on N-rich sites, and that this effect can occur in the absence of changes in tree species composition.
Key words: carbon turnover; decomposition; Douglas-fir; forest floor; litterfall; net nitrogen mineralization; nitrogen availability; nitrogen turnover; soil nitrogen capital.
INTRODUCTION
It is generally considered that nitrogen cycling is more rapid on sites with high availability of N. As Gosz (1981) outlined, vegetation on N-rich sites produces litter with high N concentrations and low amounts of phenolics, leading to rapid decomposition and mineralization of N. On N-poor sites, the litter has low N concentrations and higher amounts of phenolic compounds, decomposes slowly and releases the N slowly. These differences in litter chemistry among sites thus create a feedback which increases N availability on rich sites but decreases it on poor sites. Support for this hypothesis has come from studies of N cycling along N availability gradients in Wisconsin. Pastor et al. (1984) reported higher N return in litter at eight sites along a gradient in soil N mineralization. Tree species composition changed along the gradient, with species that produce higher quality litter replacing others as N availability increased. Decomposition rates of the native dominant foliar litter were positively correlate d with soil N mineralization rates (McClaugherty et al. 1985). Reich et al. (1997) found positive correlations between net N mineralization, aboveground net primary production and N return in litter at 50 sites across a range of soils and forest types in Wisconsin and Minnesota.
In these studies the gradients in N availability were associated with changes in tree species composition; it is less clear if differences in N cycling occur if the tree species remains the same. In loblolly pine (Pin us taeda L.) forests, Birk and Vitousek (1986) compared N content of litterfall with N availability (based, on net N mineralization of forest floor and soil during lab incubations) and found that increased soil N availability was associated with increased mass and concentrations of N in litterfall. However, this relationship was largely driven by data from sites that had been heavily fertilized with sewage sludge; there was no apparent relationship along the natural N availability gradient (i.e., among unfertilized sites). Lamb (1975) found lower litter N concentrations, higher forest floor accumulation and slower N mineralization in radiata pine (Pin us radiata D. Don) plantations on poor sandy soils compared with richer sites. He attributed these effects to lower nutrient concentrations and h igher concentrations of polyphenols in the litter on poor sites, rather than to site N availability. In contrast, studies of the effects of different plant species have consistently shown differences in rates of N mineralization (Wedin and Tilman 1990, Gower and Son 1992, Binkley and Giardina 1998). It is therefore possible that differences in N cycling between N-rich and N-poor sites arise entirely from differences in tree species composition, and that differences in N cycling do not occur at sites with the same tree species.
In this study we examined the influence of mineral soil N capital on rates of N cycling in forests of a single tree species (Douglas-fir [Pseudotsuga menziesii (Mirb.) Francol) at nine sites in coastal Washington and Oregon naturally differing in soil N capital. Rates of N cycling were estimated by measuring annual N input in litter, amount of N in the forest floor and rates of net N mineralization in the forest floor. Rates of N turnover in the forest floor were estimated from the litterfall:forest floor ratios. If N turnover is influenced by site N availability in the absence of changes in tree species composition, then rates of N input in litter, N turnover in the forest floor and net N mineralization would increase as soil N capital increased. The underlying causes of the differences in N capital among the sites were explored by comparing soil N capital with other site properties.
MATERIALS AND METHODS
Study sites
Nine coastal Douglas-fir stands with site indices ranging from 26 to 42 m at 50 yr were selected to provide a range in N availability, based on the established relationship between N availability and productivity of Douglas-fir stands in coastal Washington and Oregon (Chappell et al. 1991). The nine fully-stocked second growth stands were control plots in a fertilization trial established in 1969-1970. The stands were between 42 and 70 years of age, and standing volume ranged from 613 to 1018 [m.sup.3]/ha (Table 1). Soil properties for the surface 15-cm layer are given in Table 2 together with estimated total contents of C and N in the soil profile. Total C and N contents in the soils were estimated from profiles ranging from 63 to 152 cm in depth. The differences in profile depth were not related to soil C and N capitals and were considered to be of little importance due to very low concentrations in the deep soil horizons. Total C contents of mineral soils ranged from 83 to 220 Mg/ha, and total N contents ranged from 2.5 to 12.1 Mg/ha.
Sampling and analyses
The rate of N turnover in the forest floor in each stand was estimated by measuring: (1) N concentration in needle litter and annual litterfall mass, (2) N concentration and mass of the forest floor, and (3) net N mineralization in the forest floor. All samples were taken from a single 20 X 20-m plot in each stand. Carbon concentrations in litterfall and forest floor were measured to allow us to distinguish the effects of mass and N concentration on N content.
To estimate the mass of annual litterfall, ten 0.135[m.sup.2] plastic trays with fiberglass screens in the bottom and holes for drainage were randomly placed in each measurement plot in April 1993. Fallen litter was collected from each tray at two-month intervals for one year; dried at 70[degrees]C; sorted into brown needles, green needles, and other material; and weighed. Concentrations of C and N were measured in litter collected in October following the annual peak in litterfall, using a CHN analyzer (Perkin Elmer Series II CHNS/O Analyzer 2400, Perkin Elmer, Norwalk, Connecticut, USA). The total C and N contents of annual litterfall were estimated by multiplying the mass of litter in each tray by the concentrations of C and N, and were averaged for each site.
The mass of the forest floor at each site was estimated from five 0.093-[m.sup.2] samples of the pooled Oi, Oe, and Oa layers collected from each plot in October 1993. Samples were dried at 70[degrees]C and woody debris larger than 1 cm in diameter was removed. The remaining material was weighed, and total concentrations of C and N were measured with the CHN analyzer. The total amounts of C and N in the forest floor were estimated by multiplying the mass of the forest floor by the concentrations of C and N, and were averaged for each site.
The fractional annual losses of C and N from the forest floor were calculated by dividing the annual inputs of C and N in litter by the total amounts of C and N in the forest floor (Gosz et al. 1976). Turnover estimates were based on brown-needle litter and all forest floor material except woody debris larger than 1 cm.
Potential rates of net N mineralization in the forest floors were estimated from the amounts of [NH.sub.4]-N and [NO.sub.3]-N produced during a 24-d aerobic incubation in the laboratory (Vitousek et al. 1982). The laboratory incubation was preferable to in situ incubations for assessing the effect of differences in the chemical nature of the forest floors, without the confounding influence of climatic differences among the sites. Ten samples of the Oe and On layers in each plot were collected in October 1993 separately from the forest floor mass samples described above. A 5 g subsample (dry weight equivalent) was extracted with 2 mol/L KCl and concentrations of [NH.sub.4]-N and [NO.sub.3]-N were measured on an Alpkem RFA 300 AutoAnalyzer (Alpkem Corporation, Wilsonville, Oregon, USA; Page et al. 1982). A second 10 g subsample (dry weight equivalent) was placed in a 0.6 [dm.sup.3] glass jar. Distilled water was sprayed into each subsample to bring the moisture contents to 75% (wet weight basis), and the jars were incubated in the dark at [sim]20[degrees]C. Each week, the jars were opened to outside air for 15 mm. After 24 d, each sample was extracted with 2 mol/L KCI. Differences between the amounts of extractable N before and after incubation were used to estimate the net N mineralization rate in each forest floor sample.
Data analysis
Relationships between soil N capital and indices of N cycling were explored by simple linear regression (n = 9). Residual plots were examined for fit and variance homogeneity, and transformations were not necessary. The curvilinear relationship between forest floor net N mineralization and C:N ratio was fitted by nonlinear regression analysis. Intercorrelations between soil N capital and other site characteristics (n = 9) were tested by correlation analysis and calculation of Spearman’s nonparametric rank-order correlation coefficients. Multiple regression models were used to explore relationships between fractional loss of C and N and soil N capital and soil texture variables (percentage clay or sand). SAS (SAS Institute 1993) was used for all analyses. The accepted level of significance was P [less than] 0.05, but given the small number of sites (nine), some weaker relationships (P [less than] 0.10) are also reported.
RESULTS
Litterfall and forest floor C and N contents for the nine stands are shown in Table 3 together with net N mineralization rates and the annual fractional losses of C and N. The total amount of N returned in annual litterfall was positively correlated with soil N capital (Fig. la). This was the result of increased litterfall mass (C content; Fig. 1b) and N concentration (Fig. 1c), which were also positively correlated with soil N capital.
Despite the greater litter input, forest floor mass was smaller at N-rich sites. Forest floor N content was negatively correlated with soil N capital (Fig. 1d), and forest floor C content and N concentration were weakly (0.05 [less than] P [less than] 0.10) negatively correlated with soil N capital (Figs. le and f). Fractional annual losses of C and N were both positively correlated with soil N capital (Fig. 2a), indicating more rapid turnover of the forest floor on N-rich sites. Fractional annual losses of C and N from the forest floor were negatively correlated with mineral soil C:N (Fig. 2b), suggesting that soil N concentrations influence litter turnover rates.
Rates of net N mineralization in forest floors were highest at the two sites with the lowest forest floor C: N ratios, and mineralization was generally low above a C:N ratio of 35 (Table 3). This resulted in a curvilinear relationship between net N mineralization and forest floor C:N ratio (Fig. 3). However, net N mineralization was not correlated with either soil N capital, litter N content, or litter C:N ratio. Thus, compared with rates of C and N turnover, net N mineralization was most affected by the N concentration in the forest floor and less tightly linked to soil N capital and associated site factors.
The relationships between soil N capital and other site properties listed in Tables 1 and 2 were explored in order to identify site factors associated with the gradient in soil N. Soil N capital was positively correlated only with soil C capital (P [less than] 0.05, [r.sup.2] = 0.61). Soil C capital was also positively correlated with both litter N concentration (P [less than] 0.05, [r.sup.2] = 0.52) and litter N content (P [less than] 0.05, [r.sup.2] = 0.61). Of the other soil properties listed in Table 2, only percentage clay and percentage sand were significantly correlated with soil N and C capitals (Table 4). This suggests that soil N capital. is associated with differences in soil texture among the nine sites. However, multiple regression models including percentage sand or clay did not explain the variation in fractional loss of N and C better than the simple regression models with soil N capital. This may be related to the small overall sample size (nine). Site index was positively correlated with soil N and C capitals (Table 4).
DISCUSSION
The greater amount of litter returned with increasing site N capital probably reflects higher productivity at N-rich sites; this is consistent with the positive correlation between soil N capital and site index at these sites. Leaf area is generally positively correlated with aboveground net primary production (Fassnacht and Gower 1997), and leaf area of Douglas-fir stands increases in response to N fertilization (Brix 1983, Binkley and Reid 1984). Therefore, leaf area probably increased with increasing soil N capital in this study, contributing to greater litterfall. Reich et al. (1997) found a positive correlation between aboveground net primary production and litterfall N content. The higher N concentration in the litter at N-rich sites in addition to the greater mass of litter suggests that N was less limiting to tree growth at N-rich sites. The positive correlations between soil N capital, site index, and litter N content support the suggestion of Vitousek (1982) that the amount of N circulating in annu al litterfall provides a good indication of N availability.
The positive correlation between fractional annual losses of C and N in the forest floor and soil N capital indicate that decomposition is faster on N-rich sites. This could result from either faster decomposition of Douglas-fir litter or differences in the ground vegetation at the sites. It is not clear if there was sufficient difference in the quality of the Douglas-fir litter among the sites to influence its decomposition rate. Although decomposition rates are generally correlated with litter N concentration; a consistent relationship between litter N concentration within a species and decomposition rate has not been demonstrated (Fog 1988, Prescott 1995). A litterbag experiment with needles from the nine sites is necessary to determine if the N concentration in the Douglas-fir needles influenced their decomposition rates.
Ground vegetation was dominated by salal (Gaultheria shallon Pursh) and Oregon grape (Mahonia nervosa (Pursh) Nutt.) at the N-poor sites and by sword fern (Polystichum munitum (Kaulf.) Presl.) at the N-rich sites. Differences in decomposability of the ground vegetation at the nine sites may have contributed to the differences in turnover rates of forest floors. Biomass and litter input from understory vegetation were not quantified in this study, so we cannot determine the extent to which the understory contributed to differences in N cycling among the sites. However, the strong correlations between soil N capital, and C and N contents of overstory foliar litter suggest that the influence of soil N capital was expressed largely though its influence on overstory litter input.
We estimated turnover rates by the litterfall-forest floor method which assumes that the forest floors are in steady state, i.e., that annual decomposition in the forest floor equals annual litter input. The stands were more than 40 yr old, and closed canopies many years earlier, so rates of litter production should have stabilized. Forest floor masses would probably still be increasing at these sites, but it is unlikely that the trend for smaller forest floor masses at N-rich sites would be reversed. It is more likely that forest floor mass at the N-poor sites will increase relative to the forest floor mass at the richer sites. Forest floors at N-rich sites with fast decomposition will likely reach steady state sooner than forest floors at N-poor sites with slow decomposition. This would lead to greater differences in turnover rates between N-rich and N-poor sites over time. In any event, our finding of a decline in forest floor mass despite increasing litter input as soil N capital increased is strong evid ence of faster decomposition at sites with larger soil N capital.
The nine stands in this study varied in several properties in addition to soil N capital, which could influence the rate of N cycling. The stands ranged in age from 42 to 70 years, which might influence the amount of litter returned annually and the amount of organic matter accumulated in the forest floor. However, there were no significant correlations between age and C or N contents of the forest floor or litterfall, nor did age improve the strength of the significant relationships with soil N capital. Litterfall rates might have been influenced by different climatic conditions arising from differences in elevation (162-945 m) or precipitation (152-292 cm/yr) among the nine sites. However, there were no significant relationships between litterfall C or N content and either elevation or precipitation while relationships with soil N capital were consistently significant. Therefore, although these other factors probably influenced rates of N cycling, their influence was much less apparent and so was either sm aller or less direct than the influence of soil N capital.
Litter input can vary according to climatic conditions during the year the needles were formed or the year they fell, so measurement of annual litterfall for more than one year would have been preferable. Given the proximity of the sites, year to year variations in climate would be similar, so litter input rates in any given year should be comparable among sites. Using only brown needles to estimate litter input and the entire forest floor to estimate litter accumulation raises the possibility that the trend of slower litter turnover at N-poor sites could have resulted from larger inputs of other materials at N-poor sites. The mass of other materials (largely branches and green needles) in litterfall showed no correlation with soil N capital (P [less than] 0.42, [r.sup.2] = 0.10). Therefore, including other types of aboveground litter in the turnover estimates resulted in the same trend of increased turnover estimates with increasing soil N capital. Branch litter is primarily associated with winter storms, and is highly variable among sites and among years, so litter estimates including this material are less preferable than brown needles for gauging site N status or estimating litter turnover rates.
Although rates of net N mineralization in the forest floor were positively correlated with the N content of litterfall across a range of forest types in the United States (Vitousek et al. 1982), we did not find a significant relationship at our sites. The amount of N in annual litterfall at our sites (5-16 kg N/ha) was at the very low end of the range of values in their study, at which they found negligible rates of net N mineralization. Net N mineralization was also not significantly correlated with the N concentration or C:N of foliar litter, unlike the relationship that Scott and Binkley (1997) found between annual net N mineralization and the lignin:N of litter from diverse forest types. We did, however, find a curvilinear relationship between net N mineralization and the C:N of the forest floor. There was appreciable net N mineralization only in the two forest floors with C:N [less than]35, corresponding to 1.35% N. The critical C:N ratio below which net N mineralization occurs is commonly quoted as being in the range of 25-30, or 1.7-2.5% N (Haynes 1986). A relationship between net nitrification and the C:N ratio of the forest floor has been reported in other studies, with critical ratios of 25-27 (Gundersen and Rasmussen 1990) and 31(1.4% N; McNulty et al. 1991). Net nitrification was negligible in the nine forest floors in our study (data not shown).
Rates of net N mineralization in the forest floors during the laboratory incubation were not related to the soil N capital whereas turnover rates of C were, indicating a weak relationship between C and net N mineralization rates. Poor relationships between rates of C mineralization and net N mineralization are commonly observed during laboratory incubations (Harris and Riha 1991), and is explained as the effect of N immobilization into microbial biomass. Rates of gross N mineralization are correlated with rates of C mineralization, but a very substantial proportion of the mineralized N is re-immobilized into microbial biomass. This results in low rates of net N mineralization that are not closely related to gross N mineralization or C mineralization (Davidson et al. 1992, Stottlemeyer and Toczydlowski 1999). The relationship between net N mineralization and forest floor C:N in our study suggests that the amount of N re-immobilized was most closely related to the N concentration of the forest floors. The poor relationships between net N mineralization and the soil N capital and rates of N turnover suggest that laboratory measures of net N mineralization in the forest floor may not provide a good indication of site N availability.
Forest floor N concentrations were weakly negatively correlated with site N capital, which was positively correlated with the other indices of N cycling rates (litter N content and N turnover of the forest floor). This unexpected result suggests that on N-poor sites, not only is decomposition slower, but a smaller proportion of the N in litter is mineralized and then removed from the forest floor (taken up by plants or otherwise lost). This may be related to greater production of polyphenols on N-poor sites, which would cause more N to remain in the forest floor, bound in polyphenol-protein complexes (Gosz 1981). Therefore, although forest floor N concentration or C:N may be related to rates of N mineralization rates in the forest floor, they may not be reliable indicators of site N availability.
These analyses indicate that soil N capital exerted the strongest control on N turnover. The remaining question is: What factors are responsible for the differences in soil N capital among these sites, and therefore exert the ultimate control on N turnover in these forests? Our exploration of site factors associated with soil N capital did not indicate that climatic factors such as precipitation or elevation (a surrogate for temperature) were related to soil N capital or N turnover. The only soil factors that were significantly correlated with soil N capital were percentage clay and percentage sand. These soil factors were not significantly correlated with N turnover, suggesting that their influence is either weaker or less direct. However, over time, the differences in soil texture among the sites may have contributed to the differences in soil N capital, according to the following scenario. At sites with finer textured soils, the higher moisture holding capacity and higher content of base cations promote p lant growth and organic matter production. Over time these sites accumulate more organic matter, C, and N in the soil. The feedback through increased litter N contents identified in this study further increases turnover and N availability at these sites. On coarser textured soils, accumulation of organic matter, C, and N is slower, and is further constrained by slower N cycling in litter. Thus, although soil texture does not appear to be tightly coupled with rates of N turnover in these stands, it may ultimately control N turnover by controlling the buildup of soil N. A similar scenario was proposed by Reich et al. (1997) to explain correlations between geologic substrate, productivity, and N cycling on their sites.
Other studies have indicated an indirect role of soil texture in controlling N cycling. In the study of radiata pine stands by Florence and Lamb (1974), litter accumulation was greater on sandy podzols than on finer textured soils, despite similar rates of litter input. This was attributed to slower decomposition as a consequence of lower moisture and poorer quality litter on the coarse-textured soils. Net N mineralization was also slower in forest floors from the sand podzol sites (Lamb 1975). Reich et al. (1997) found significant correlations between net N mineralization, net primary productivity, and litterfall N content in 50 hardwood and conifer stands on diverse soils. Rates were generally higher on finer textured Alfisols than on coarser textured Entisols, suggesting an influence of soil texture. Soil texture (percentage silt + clay) along with mean annual temperature and litterfall N explained 81% of the variance in annual soil N mineralization in the 31 natural stands. They concluded that N status w as an important proximate control of productivity, but that the geological substrate may also exert strong influence on N cycling. Our results support their conclusion and further indicate that even within stands of the same species, differences in soil texture may lead to measurable differences in N cycling by influencing the buildup of soil N.
Other factors such as disturbance history also influence soil N capital, but could not be addressed in this study because there is little known about the history of the nine stands. Forests of the Pacific Northwest are characterized by infrequent catastrophic disturbances in the forms of wildfire and wind (Franklin 1988). Wildfires would volatilize a portion of the surface organic matter and N, thereby slowing their buildup. Variation in the frequency and intensity of fires among the nine sites may also have contributed to differences in soil N capital.
In conclusion, our results were for the most part consistent with the hypothesis that rates of N cycling in forests of a single tree species would increase with increasing soil N capital. The amounts of N returned in litter and the rate of N turnover both increased with increasing soil N capital. Turnover rate of C also increased with increasing soil N capital. Forest floor net N mineralization was poorly correlated with soil N capital and N turnover rates and was appreciable only at two sites where forest floor C:N ratios were [less than]35. The study provided further evidence of a positive feedback increasing N availability on N-rich sites and decreasing N availability on N-poor sites. Differences in soil texture among the sites may be partly responsible for the development of differences in soil N capital and the resulting patterns of N cycling.
ACKNOWLEDGMENTS
The support of Stand Management Cooperative member organizations and the Natural Sciences and Engineering Research Council of Canada is gratefully acknowledged. We thank D. Bradley, R. Gonyea, B. Hasselberg, M. Holmes, C. Staley, and K. Thomas for field and laboratory assistance.
(1.) Department of Forest Sciences, University of British Columbia, 3041-2424 Main Mall, Vancouver, British Columbia, Canada V6T 1Z4
(2.) College of Forest Resources, University of Washington, Seattle, Washington 98195 USA
(3.) Department of Forest Ecology, Danish Forest and Landscape Research Institute, H[oslash]rsholm Kongevej 11, DK-2970 H[oslash]rsholm, Denmark
(4.) E-mail: cpres@unixg.ubc.ca
(5.) Present address: Potlatch Corporation, P.O. Box 390, Warren, Arkansas 71671 USA.
LITERATURE CITED
Binkley, D., and C. Giardina. 1998. Why do species affect soils? The warp and woof of tree–soil interactions. Biogeochemistry 42:89-106.
Binkley, D., and P. Reid. 1984. Long-term responses of stem growth and leaf area to thinning and fertilization in a Douglas-fir plantation. Canadian Journal of Forest Research 14:656-660.
Birk, E. M., and P. M. Vitousek. 1986. Nitrogen availability and nitrogen use efficiency in loblolly pine stands. Ecology 67:69-79.
Brix, H. 1983. Effects of thinning and nitrogen fertilization on growth of Douglas-fir: relative contribution of foliage quantity and efficiency. Canadian Journal of Forest Research 13:167-175.
Chappell, H. N., D. W. Cole, S. P. Gessel, and R. B. Walker. 1991. Forest fertilization research and practice in the Pacific Northwest. Fertilizer Research 27:129-140.
Davidson, E. A., S. C. Hart, and M. K. Firestone. 1992. Internal cycling of nitrate in soils of a mature coniferous forest. Ecology 73:1148-1156.
Fassnacht, K. S., and S. T. Gower. 1997. Interrelationships among the edaphic and stand characteristics, leaf area index, and aboveground net primary production of upland forest ecosystems in north central Wisconsin. Canadian Journal of Forest Research 27:1058-1067.
Florence, R. G., and D. Lamb. 1974. Influence of stand and site on radiata pine litter in South Australia. New Zealand Journal of Forestry Science 4:502-510.
Fog, K. 1988. The effect of added nitrogen on the rate of decomposition of organic matter. Biological Reviews 63:433-462.
Franklin, J. F. 1988. Pacific Northwest Forests. Pages 103-130 in M. G. Barbour and W. D. Billings, editors. North American terrestrial vegetation. Cambridge University Press, Cambridge, UK.
Gosz, J. R. 1981. Nitrogen cycling in coniferous ecosystems. In F. E. Clark and T. Roswall, editors. Terrestrial nitrogen cycles. Ecological Bulletins-NFR 33:405-426.
Gosz, J. R., G. E. Likens, and F. H. Bormann. 1976. Organic matter and nutrient dynamics of the forest and forest floor in the Hubbard Brook Forest. Oecologia (Berlin) 22:305-320.
Gower, S. T., and Y. Son. 1992. Differences in soil and leaf litterfall nitrogen dynamics for five forest plantations. Soil Science Society of America Journal 56:1959-1966.
Gundersen, P., and L. Rasmussen. 1990. Nitrification in forest soils: effects of nitrogen deposition on soil acidification and aluminium release. Reviews of Environmental Contamination and Toxicology 113:1-45.
Harris, M. M., and S. J. Riha. 1991. Carbon and nitrogen dynamics in forest floor during short-term laboratory incubations. Soil Biology and Biochemistry 23:1035-1041.
Haynes, R. J. 1986. Mineral nitrogen in the plant-soil system. Academic, Orlando, Florida, USA.
Lamb, D. 1975. Patterns of nitrogen mineralization in the forest floor of stands of Pinus radiata on different soils. Journal of Ecology 63:615-625.
McClaugherty, C. A., J. Pastor, J. D. Aber, and I. M. Melillo. 1985. Forest litter decomposition in relation to soil nitrogen dynamics and litter quality. Ecology 66:266-275.
McNulty, S. G., J. D. Aber, and R. D. Boone. 1991. Spatial changes in forest floor and foliar chemistry of spruce–fir forests across New England. Biogeochemistry 14:13-29.
Page, A. L., R. H. Miller, and D. R. Keeney. 1982. Methods of soil analysis. Part 2: chemical and microbiological properties. Soil Science Society of America, Madison, Wisconsin, USA.
Pastor, J., J. D. Aber, C. A. McClaugherty, and J. M. Melillo. 1984. Aboveground production and N and P cycling along a nitrogen mineralization gradient on Blackhawk Island, Wisconsin. Ecology 65:256-268.
Prescott, C. E. 1995. Does nitrogen availability control rates of litter decomposition in forests? Plant and Soil 168-169: 83-88.
Reich, P. B., D. F. Grigal, J. D. Aber, and S. T. Gower. 1997. Nitrogen mineralization and productivity in 50 hardwood and conifer stands on diverse soils. Ecology 78:335-347.
SAS Institute. 1993. SAS/STAT user’s guide. Release 6.07. SAS Institute, Cary, North Carolina, USA.
Scott, N. A., and D. Binkley. 1997. Foliage litter quality and annual net N mineralization: comparison across North American forest sites. Oecologia 111:151-159.
Stottlemeyer, R., and D. Toczydlowski. 1999. Nitrogen mineralization in a mature boreal forest, Isle Royale, Michigan. Journal of Environmental Quality 28:709-720.
Vitousek, P. M. 1982. Nutrient cycling and nutrient use efficiency. American Naturalist 119:553-572.
Vitousek, P. M., J. R. Gosz, C. C. Grier, J. M. Melillo, and W. A. Reiners. 1982. A comparative analysis of potential nitrification and nitrate mobility in forest ecosystems. Ecological Monographs 52:155-177.
Wedin, D. A., and D. Tilman. 1990. Species effects on nitrogen cycling: a test with perennial grasses. Oecologia (Berlin) 84:433-441.
COPYRIGHT 2000 Ecological Society of America
COPYRIGHT 2000 Gale Group
