| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Virginia Polytechnic Institute and State Univ., Dep. of Forestry, 228 Cheatham Hall, Blacksburg, VA 24061
* Corresponding author (trfox{at}vt.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: N0, nitrogen mineralization potential determined from 15-wk aerobic incubation • TKN, total Kjeldahl nitrogen
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The longevity of the growth response following N fertilization in most Douglas-fir stands is short lived, usually lasting <10 yr (Peterson and Gessel, 1983; Miller and Tarrant, 1983). Consequently, repeated applications of N are usually required to maintain rapid growth rates throughout the rotation in forest ecosystems. Current operational fertilizer practices in Douglas-fir stands often apply urea two or three times on 5- to 10-yr intervals during the mid- to late-rotation. The growth response of Douglas-fir following N fertilization on poor quality sites may be much longer. Binkley and Reid (1985) observed an extended growth response lasting longer than 15 yr in a Douglas-fir stand on a poor quality site in Washington. Long-term growth response following N fertilization in forest ecosystems is more likely to occur on sites where the amount of N applied in the fertilizer is large relative to the mineralizable N reserves (Miller, 1981; Proe and Williams, 1989). Basal area growth of Corsican pine [Pinus nigra subsp. laricio (poiret) Maire] growing on a relatively poor, sandy soil receiving 1512 kg N ha–1 was significantly greater than unfertilized plots for almost 20 yr (Proe and Williams, 1989). Stands on low quality sites fertilized with N tend to produce litter with lower C/N ratios, which should decompose and mineralize more rapidly and thus maintain higher levels of N availability (Vitousek and Matson, 1984; Binkley, 1986). This seems to have occurred at the site studied by Binkley and Reid (1984) where long-term responses in volume growth were observed. At this site, soil N availability in the fertilized plots was twice that in the unfertilized plots 15 yr after fertilization (Binkley and Reid, 1985). Mineralization rates in these fertilized stands were still higher than in non-fertilized stands 22 yr after treatment (Strader and Binkley, 1989).
The growth response of trees to added N may be greater when several small applications of N are made rather than a single large dose. For example, in a study with American sycamore (Platanus occidentalis L.), growth response over 3 yr was greater following annual applications of 150 kg N ha–1 than that following a single applications of 450 kg N ha–1 (Van Miegroet et al., 1994). It has also been hypothesized that long-term increases in N mineralization may follow repeated applications of small amounts of N fertilizer (Adams and Attiwill, 1984; Snowdon and Khanna, 1989; Tschaplinski et al., 1991). Repeated fertilization with urea over a 16-yr period increased N mineralization in the forest floor more than a single application in a mixed Scots pine (Pinus sylvestris L.) and Norway spruce (Picea abies L.) forest in Finland studied by Aarnio and Martikainen (1992). They attributed this to changes in soil microbial populations, including autotrophic ammonium and nitrate oxidizers, caused by the initial fertilization. Likewise, Prescott et al. (1995) found higher levels of mineralizable N in the forest floor of jack pine (Pinus banksiana Lamb.) stands that were fertilized six times receiving a total of 672 kg N ha–1. In contrast, Chappell et al. (1999) found that N fertilization did not increase N mineralization in the forest floor of nine Douglas-fir stands in coastal Washington and Oregon that received between 898 and 1120 kg N ha–1 from 1969 to 1985. These results were similar to those reported by Prescott et al. (1993) who also found that repeated N fertilization did not increase N availability in the forest floor in coastal stands of Douglas-fir.
Maintaining soil productivity is a central tenant of sustainable forest management (Nambiar, 1996), serving as one of the criteria of sustainable forest management outlined in the Santiago Declaration of the Montreal Process countries (Anonymous, 1995). Unfortunately, it is difficult to identify specific indicators of soil quality that can be used to judge sustainability (Burger and Kelting, 1998; Powers, 1999). Because most forest ecosystems are limited by N, indices of soil N availability may be useful indicators of soil quality and sustainability (Gregorich and Carter, 1997). However, extractable mineral soil N may not accurately reflect long-term trends in soil quality. It is possible that short-term increases in extractable soil N may mask potential decreases in soil quality associated with decreased organic matter and long-term N availability (Burger and Kelting, 1998). Likewise, total soil N may not be a good indicator of soil quality because much of the N in soil is tied up with soil organic matter that is relatively stable. Total soil N is therefore not a very sensitive indicator of management-induced changes in soil quality. Mineralizable N has been proposed as a more useful indicator of soil quality (Powers et al., 1998; Knoepp et al., 2000). Mineralizable N is the fraction of soil N that actively cycles through the ecosystem and thus is an index of available N. Stable or increasing N mineralization potential in forest soils would suggest that the forest management practices sustain soil quality. Decreases in N mineralization potential would suggest that management practices are decreasing soil quality and thus are not sustainable. Because forest fertilization is a widespread forest management practice, it is important to understand the effects of N fertilization on N mineralization, soil quality, and forest sustainability.
In this study, soil mineralizable N was evaluated in two Douglas-fir stands located in the Cascades Mountains of western Washington that were fertilized with N using various doses and frequencies of urea fertilizer over a 6-yr period. The purpose of the experiment was to examine the effects of repeated applications of N fertilizers on N mineralization and the potential impacts this had on soil quality. The specific objectives of the study were to determine if: (i) N fertilization increased N mineralization potential in the soil; (ii) N mineralization potential was higher following annual applications of N compared with applications at 5-yr intervals; (iii) N mineralization potential was related to the total amount of N fertilizer added; and (iv) N fertilization affected other soil chemical properties related to soil quality, such as pH and extractable cations.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The first site, South Toutle, was located at 46°13'48''N, 122°30'36''W at an elevation of 472 m. The site had a northerly aspect with an average slope of 8%. Site index base age 50 for Douglas-fir ranged from 34 to 40 m. Soils at the South Toutle site were mapped by Weyerhaeuser Company as Dunnigan and Beigle series. The second site, Coweeman, was located at 46°10'12''N, 122°28'48''W at an elevation of 732 m. The site had an easterly aspect with an average slope of 15%. Site index base age 50 for Douglas-fir ranged from 30 to 34 m. Soils at the Coweeman site were mapped by Weyerhaeuser as Beigle and Wolpoint series. The soils at both locations were well drained and were derived from andesite, tuff, and volcanic ash parent material. The Dunnigan series is an Ultisol with a loamy A horizon and a clay loam argillic horizon at 35 to 45 cm. Total soil depth is >150 cm with very few rocks throughout the profile. The Beigle and Wolpoint series are Inceptisols that are loamy throughout the profile. In the Beigle series, a Bw horizon occurs at 40 to 60 cm and soft, fractured andesite occurs at 100 to 150 cm. The Wolpoint series has a Bw horizon at 36 to 65 cm and total soil depth is approximately 150 cm. Rocks are present throughout the profile in both soils with rock content increasing with depth. Rock content ranges from 10 to 40% on a volume basis in the Beigle series and from 20 to 90% in the Wolpoint series.
Dose x Frequency Fertilizer Study Treatments
The study was initially designed as a response surface to evaluate N rates and application frequencies using a central-composite rotatable design with incomplete blocks (Myers, 1976). The original objective of the study was to determine if the growth response following small, frequent applications of N would be superior to that following larger more infrequent applications. Individual N application rates ranged from 0 to 600 kg N ha–1 and fertilizer was applied at annual, 3-yr, or 5-yr intervals to 0.16-ha plots. Fertilizer was applied by hand starting in January of 1980 using urea as the N source. Subsequent applications were also made in January of the appropriate year depending upon the application interval. An initial dose of fertilizer supplying 225 kg N ha–1 was made in a portion of the treatments in Year 1 and the application rates and intervals started in Year 2. This was done because it was hypothesized that small, frequent applications would maintain rapid growth rates, but that a larger rate was needed initially to obtain a significant response. In the rest of the treatments, the fertilizer treatments started in Year 2 without the initial plateau dose of 225 kg N ha–1. A fertilizer treatment supplying P, K, and S in addition to N was also used in one treatment. Fertilizer rates in this treatment were 70 kg P ha–1 as triple superphosphate, 65 kg K ha–1 as potassium sulfate, and 65 kg S ha–1 as potassium sulfate and magnesium sulfate. Nitrogen application rates and intervals were selected so that it would be possible to compare the growth response of stands that received the same total amount of N applied at rates and intervals over the study period. Each combination of application rate and application interval was replicated four times at each site.
In the summer of 1985, 6 yr following the start of the study, a subset of the treatments were selected to study the effect of N application rates and frequencies on N mineralization potential in the soil. Eight treatments were selected from the annual and the 5-yr application interval that supplied from 0 to 825 kg N ha–1 over the 6-yr period (Table 1). The treatments were selected to compare the same total amount of N applied in small doses at annual intervals or large doses at 5-yr intervals. Four replications of each of the eight selected treatments were sampled at each of the two locations.
|
Nitrogen mineralization potential (N0) was determined using an aerobic incubation procedure (Stanford and Smith, 1972; Keeney, 1982). Forty grams of soil were thoroughly mixed with 40 g of acid washed quartz sand. The 1:1 mixture was placed in a 3.8-cm diam. polyvinyl chloride (PVC) tube 20 cm long and closed at one end with a one-hole rubber stopper, glass wool, and a glass filter. A plug of glass wool was placed on top of the soil column and it was then closed with a one-hole rubber stopper. The completed columns were incubated at 25°C for 15 wk. The columns were leached of accumulated mineral N at 0, 1, 3, 5, 8, 11, and 15 wk with 250 mL of 0.01 M CaCl2 under a vacuum pressure of 0.033 MPa to approximate field capacity. The leachate collected was weighed and a 100-mL aliquot was analyzed for inorganic ammonium and nitrate using rapid flow analysis (USEPA, 1979) on a Technicon Autoanalyzer (Technicon Instrument Corp., Tarrytown, NY). After each CaCl2 leaching, 100 mL of minus-N nutrient solution [0.002 M CaSO4; 0.002 M MgSO4; 0.005 M Ca(H2PO4); 0.0025 M K2SO4] were passed though the column under vacuum pressure of 0.033 MPa to approximate field capacity and thus ensure that aerobic conditions were established in the soil. Nitrogen mineralization potential (N0) and the mineralization rate constant (k) were determined based on the cumulative amount of N mineralized during the 15-wk incubation using the following nonlinear equation (Smith et al., 1980; Deans et al., 1986):
 |
Anaerobic N mineralization (anaerobic N) was also estimated using a short-term anaerobic procedure (Waring and Bremner, 1964; Keeney, 1982). The analyses were done in triplicate. Five grams of soil from each plot were placed into 100-mL plastic centrifuge tubes and 15 mL of distilled water was added to each tube. The tubes were tightly capped with rubber stoppers and incubated at 40°C for 7 d. At the end of the incubation period, the samples were shaken with 25 mL of 1 M KCl for 1 h. The KCl extracts were filtered through Whatman #1 filter paper. Ammonium in the samples was analyzed using rapid flow analysis (USEPA, 1979) on a Technicon Autoanalyzer (Technicon Instrument Corp., Tarrytown, NY).
Extractable soil nutrients were determined in duplicate using the modified Bray extraction procedure (Olsen and Sommers, 1982). Two grams of soil were mixed with 20 mL of 0.03 M NH4F + 0.025 M HCl in a 100-mL plastic centrifuge tube, shaken on a reciprocating shaker for 1 min, and then filtered through Whatman # 42 filter paper. Extracts were analyzed for P, K, Ca, Mg, Fe, Al, and Mn using inductively coupled plasma emission spectrophotometry (Soltanpour et al., 1996) on a ThermoElectron ICP-OES spectrometer (Thermo Electron Corp., Woburn, MA). Soil organic C was determined using the Walkley–Black wet oxidation method (Nelson and Sommers, 1982). Total N in the soil was determined by the Kjehldahl digestion procedure (Bremner and Mulvaney, 1982). Soil pH was measured using a glass electrode in a 2:1 water/soil mixture.
Statistical Analysis
The N mineralization data was analyzed as a random complete block design with four replications because, although the original study was an incomplete block design, the subset of treatments selected for the N mineralization study were all equally represented. Analysis of variance using the GLM procedure in SAS (SAS Institute Inc., Cary, NC) was used to test the following hypotheses with respect to extractable soil nutrients, soil organic matter, total Kjeldahl N (TKN), anaerobic N, and N0 (Gomez and Gomez, 1984).
Correlation coefficients were calculated (Gomez and Gomez, 1984) among anaerobic N, N0, soil TKN, soil C, C/N ratio, and the total amount of N applied in fertilizer during the study using the CORR procedure in SAS (SAS Institute Inc., Cary, NC). Regression techniques were used to model the relationship between N0 and the cumulative amount of N added in fertilizer over the 6-yr treatment period. In addition to simple linear regressions, the data were fit using quadratic equations because the data appeared to be curvilinear (Montgomery and Peck, 1982). All equations were fit using the REG procedure in SAS (SAS Institute Inc., Cary, NC).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
Significant differences in extractable soil P and K were found among the fertilizer treatments (Table 2). Higher concentrations of P and K were found in the plots fertilized with NPKS compared with the plots receiving N only at the same dose and frequency (Table 3). Extractable P and K levels increased from 100.7 and 104.1 mg kg–1, respectively in the N only plots to 151.4 and 121.6 mg kg–1 in the NPKS fertilized plots. Extractable soil Ca and Mg were also affected by the N fertilization treatments (Table 2). Linear contrasts indicate that extractable soil Ca and Mg concentrations were lower in the treatments receiving annual N doses following the initial plateau dose (Ca = 493 mg kg–1, Mg = 85.8 mg kg–1) compared with the treatments fertilized with N at 5-yr intervals (Ca = 586.4 mg kg–1, Mg = 105.3 mg kg–1) (Table 3). The most notable effect were found in Treatment 6 where annual applications of 120 kg ha–1 of N were made following the initial plateau dose of 225 kg ha–1 in Year 1 (Table 2). Soil Mg concentrations in this treatment declined to 49 mg kg–1 while soil Ca levels declines to 338.9 mg kg–1. There were no differences in extractable Fe, Al, or Mn among the N fertilizer treatments.
Nitrogen mineralization potential was significantly correlated with the anaerobic N mineralized (Table 4). Both measures of N mineralization were positively correlated with soil TKN and soil C while N0 was negatively correlated with soil C/N ratio. The N fertilization treatments in this study added from 0 to 825 kg ha–1 over the 6-yr treatment period applied either in large infrequent doses or small annual doses (Table 1). Plotting the relationship between N0 and total N applied during the study period revealed a quadratic relationship (Fig. 1) . Nitrogen mineralization potential reached a maximum at N fertilization rates of 450 kg ha–1 and then declined. This may explain why neither measure of N mineralization was linearly correlated with the total amount of N applied during the 6-yr treatment period (Table 4). At equal N application rates, N0 values were consistently higher at South Toutle than at Coweeman (Fig. 1).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
However, several studies on the effects of N fertilization on N mineralization in Douglas-fir forest ecosystems have not found any effect (White et al., 1988; Prescott et al., 1993; Chappell et al., 1999). Some of the variation observed in N mineralization following N fertilization in various studies may to be related to the soil horizon sampled and the method used to evaluate N mineralization. In the present study, the A horizon was sampled and N0 increased following urea fertilization. White and coworkers (1988) concluded from their study of N mineralization in both the forest floor and mineral soil of a Douglas-fir stand that the surface soil was a more important source of mineral N than the forest floor. The impact of N fertilization on N mineralization appears to be smaller in the forest floor of Douglas-fir stands than the mineral soil. In a recent study of the long-term effects of N fertilization on N mineralization in Douglas-fir forests, Chappell et al. (1999) found that N mineralization rates in the forest floor of Douglas-fir forests in western Washington and Oregon was not affected by repeated N fertilization over a 16-yr period. In another study examining the long-term effects of inorganic fertilization on litter turnover in Douglas-fir, Prescott et al. (1993) also found no effect of N fertilization on N mineralization. Greater variability seems to occur in studies that have examined forest floor N dynamics compared with those that examine N mineralization in the mineral soil. Polglase et al. (1992b) examined the impact of annual N fertilization on release of N from forest floor in loblolly pine stands. They found that the Oi horizon immobilized N while net mineralization of N occurred in the Oe and Oa horizons.
The relationship between N0 and the total amount of N applied over the 6-yr period was best modeled as a quadratic function, increasing up to 450 kg ha–1 and then decreasing at higher total N application amounts. Several studies have demonstrated a similar relationship where N mineralization rates decreased following N fertilization with high rates of N (Baath et al., 1981; Soderstrom et al., 1983; Fog, 1988). McNulty and Aber (1993) found that net N mineralization increased on plots receiving annual applications of <25 kg ha–1 yr–1 but decreased on plots receiving higher annual applications. Biederbeck et al. (1996) examined N mineralization rates following 10 yr of annual application of urea at rates of 0 to 180 kg ha–1. Nitrogen mineralization rates followed a quadratic relationship with N applied, with the greatest rate in the 45 kg ha–1 yr–1 treatment, followed by a decrease at higher rates. Results from studies of N deposition in forest ecosystems in Europe have shown that N mineralization declines in N saturated systems. For example, Gunderson et al. (1998) studied N mineralization at a number of sites subjected to N deposition and found that at sites with high initial N availability, further additions of N decreased net N mineralization. In a study of the impacts of N deposition in forest ecosystems across Europe, Tietema (1998) also found that N mineralization peaked at intermediate N deposition rates.
Aber et al. (1998) have proposed two hypotheses to explain the decline in N mineralization rates following application of additional N to N saturated ecosystems. The first explanation is that the addition of N alters the chemical bond structure of soil organic matter, which reduces the efficiency of extracellular enzymes and thus decreases decomposition and mineralization. High rates of N fertilization can lead to the formation of heterocyclic ring structures between lignin and NH4 that are highly resistant to microbial decomposition (Berg and Staaf, 1981; Berg, 1986). This nonbiological fixation of NH4 can be an important process of N immobilization in forest ecosystems following urea fertilization (Johnson, 1992). The decline in N mineralization at high N fertilization rates may thus be the result of a lack of labile C that serves as an energy source for soil microbes. Microbial populations may be more limited by C than by N under these conditions (Flanagan and Van Cleve, 1983; McNulty and Aber, 1993). The data of Biederbeck et al. (1996) supports this hypothesis. They found that the decrease in N mineralization at the higher N application rates they tested was accompanied by a decrease in microbial biomass and actinomycete populations. The second potential explanation for decreased N mineralization following high N applications is that the production of extracellular enzymes by soil microbes, particularly fungi, is suppressed in the presence of elevated concentrations of mineral N, thus reducing N mineralization (Fog, 1988; Aber et al., 1998). It seems likely that the first explanation relating to changes in the quality of soil organic matter in plots receiving high N additions is a more plausible explanation for the decreased N mineralization observed in the controlled laboratory conditions in the present study. The initial leaching of accumulated mineral N during the aerobic incubation procedure would most likely have eliminated the inhibitory effect of mineral N on extracellular enzyme production by soil microbes. Furthermore, the microbial community present in the soil cores during the lab incubation procedure would likely have been similar. Any differences in the microbial community associated with site or treatments in the field would have probably have been lost during sample preparation and storage. Thus, it was more likely that changes in the quality of the soil organic matter due to N fertilization caused the quadratic response observed with N0 in this study. Additional detailed studies of the composition of soil organic matter at these sites would be required to confirm this conclusion.
The pH of the soil decreased following urea fertilization in this study. Numerous studies have documented decreases in forest soil pH following urea fertilization (Johnson, 1992). Hydrolysis of urea initially results in a temporary rise in soil pH. Urea also provides a C source for heterotrophic nitrifiers (Killham, 1990). Subsequent nitrification of the ammonium produced during urea hydrolysis produces H+ ions, which contributes to a decrease in soil pH (Killham, 1990). Johnson and Todd (1988) found that the production of NO3 and the decrease in soil pH is greater when frequent applications of urea are made, as in the present study. McNulty and Aber (1993) also found that repeated applications of small amounts of urea fertilizer over several years to a spruce-fir forest in Vermont significantly increased nitrate production and potential leaching. However, repeated applications of urea at low rates may not stimulate nitrate production, at least initially. Tschaplinski et al. (1991) found lower nitrate production and leaching during the first year following fertilization when three small applications of urea at 37.5 kg N ha–1 were applied over the growing season compared with a single, large dose delivering 450 kg N ha–1. However, following N fertilization for 3 yr at this same site, Van Miegroet et al. (1994) observed that nitrate leaching was greatest in plots fertilized annually with 150 kg N ha–1.
Excess N additions to forest soils either as fertilizer or through atmospheric deposition, leads to N saturation and increased production and leaching of nitrate in most forest ecosystems (Aber et al., 1998). This is accompanied by the loss of base cations such as Ca and Mg from the cation exchange sites in the soil (McColl and Cole, 1968; Johnson and Cole, 1980). Nitrate mediated leaching of cations has been observed in the Pacific Northwest in stands where red alder (Alnus rubra Bong.) has invaded following disturbances such as timber harvest (Van Miegroet and Cole, 1984). Acid deposition in the form of nitric acid has also led to N saturation in many forests in the northeast United States and in Europe (Aber et al., 1998). In these ecosystems, high level of nitrate leaching through the soil have depleted base cations, such as Ca and Mg which has lead to nutrient deficiencies and has been implicated as a cause of forest decline (Van Dijk and Roelofs, 1988; Zoettle and Huettle, 1986; Huttl, 1990). This phenomenon was evident in the present study, where extractable Ca, Mg and K all decreased in the plots receiving annual applications of urea at a rate of 120 kg ha–1. However, current operational N fertilization programs in Douglas-fir forests do not apply N annually. Typical operational N fertilization programs in the region apply urea two or three times at 5- to 10-yr intervals during the mid- to late-rotation. Therefore, it is not likely that current N fertilization applications result in significant cation loss induced by nitrate leaching in most managed Douglas-fir forest ecosystems. The results from this study suggest that applications of fertilizer containing base cations can be used to replace cations lost through nitrate mediated leaching. In the plots receiving NPKS fertilizer, K levels were significantly higher than in the other plots. Applications of lime have also been shown to improve base saturation of forest soils subject to acidic deposition and high rates of nitrate leaching induced cation loss (Yavitt and Newton, 1990; Derome, 1990; Misson et al., 2001).
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Received for publication July 29, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. D. Ritchey, D. G. Boyer, K. E. Turner, and J. D. Snuffer Goat Urine and Limestone Affect Nitrogen and Cation Distributions in an Acidic Grassland Soil Sci. Soc. Am. J., May 1, 2008; 72(3): 626 - 633. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
