Species Effects of Ceanothus Velutinus Versus Pseudotsuga Menziesii, Douglas-®R, on Soil Phosphorus and Nitrogen Properties in the Oregon Cascades Julie D.H

Forest Ecology and Management 149 32001) 205±216

Species effects of Ceanothus velutinus versus Pseudotsuga menziesii, Douglas-®r, on soil phosphorus and nitrogen properties in the Oregon cascades Julie D.H. Spearsa,*, Kate Lajthaa, Bruce A. Caldwellb, Shana B. Penningtona, Kristin Vanderbiltb aBotany and Plant Pathology Department, Oregon State University, Cordley Hall 2082, Corvallis, OR 97331-2902, USA bDepartment of Forest Science, Oregon State University, 321 Richardson Hall, Corvallis, OR 97331-5752, USA Received 14 February 2000; accepted 18 June 2000

Abstract

Many authors have hypothesized that nitrogen-®xing species, as a functional group, would express different controls on soil properties and ecosystem development than non-nitrogen-®xing species. Although nitrogen 3N) accretion under nitrogen- ®xing tree species has been well studied, the effect of nitrogen-®xing species on other soil nutrients, such as phosphorus 3P), has received less attention. We studied differences in soil phosphorus and nitrogen properties beneath Ceanothus velutinus 3Ceanothus), a nitrogen-®xing species, and Psuedotsuga menziesii 3Douglas-®r), a non-®xing species, in a high elevation successional watershed in the H.J. Andrews Experimental Forest in Oregon. Total P was 20% greater in Douglas-®r soils than Ceanothus soils in surface horizons, but there was no signi®cant difference in deeper soil horizons. Surface soils 35 and 15 cm) under Douglas-®r generally had higher concentrations of speci®c P fractions than surface soils under Ceanothus, but this difference either disappeared or was not as apparent at greater soil depths 330 and 60 cm). Total nitrogen, and extractable ammonium and nitrate were greater in surface soils under Ceanothus than under Douglas-®r. d15N values of leaves and litter differed between Ceanothus and Douglas-®r 3p-value 0:0001 and 0.03, respectively), but the d15N of bulk soil and KCl extracted nitrate and ammonium did not differ. Soil enzyme activities suggested greater mineralization of organic P 3phosphatase activity) under Ceanothus in summer, but not in fall, while no signi®cant differences in general decomposition 3b-glucosidase activity) were found in soils between the two species. # 2001 Elsevier Science B.V. All rights reserved.

Keywords: Phosphorus fractions; Nitrogen-®xation; Soil enzymes; d15N natural abundance; Ceanothus; Douglas-®r

1. Introduction 1994; Van Breemen and Finzi, 1994;Berendse, 1998). Trees may affect soils through soil solution uptake, Tree species can in¯uence the size and distribution root production and turnover, mycorrhizal activity, of nutrient pools across soil horizons, and soils often organic compound exudation, and the quantity and differ under different types of vegetation 3Binkley, quality of litter produced for decomposition 3Angers and Caron, 1998). Tree species might also differ in their in¯uence on soil biologic activity, such as the * Corresponding author. Tel.: 1-541-737-3451; fax: 1-541-737-3573. presence of mycorrhizae or the composition of micro- E-mail address: [email protected] 3J.D.H. Spears). bial communities, that in turn may affect soil chemistry

3Binkley, 1994; Saetre et al., 1999). Furthermore, 3Lajtha and Schlesinger, 1988; Frossard et al., 1995). some plants have the ability to modify the rhizosphere In surface soils, mycorrhizal symbionts and other to access previously unavailable nutrients, such as microorganisms closely couple decomposition and phosphorus 3P) 3Ae et al., 1990; Ae and Otani, uptake processes, thus contributing to surface soil P 1997; Jones, 1998). retention. Biological controls on P availability include Many authors have hypothesized that nitrogen-®x- root growth patterns, amounts and quality of detrital ing species, as a functional group, would express inputs, extracellular enzyme activity, production of different controls on soil properties and ecosystem organic chelates, and mycorrhizal activity 3Binkley, development than non-nitrogen-®xing species 3Van 1992; Zou et al., 1995). In lower soil horizons, Breemen and Finzi, 1998). Nitrogen 3N) accretion decreased root and microorganism density may lower from symbiotic nitrogen ®xation in early successional the potential for biological interception of soil solution ecosystems can range up to 160 kg haÀ1 per year P 3Wood et al., 1984). However, lower soil horizons, 3Boring et al., 1988). Indeed, many authors have found such as the B horizon, may have increased geochem- that nitrogen-®xing species can signi®cantly increase ical buffers which regulate soil solution P at low soil N levels; e.g. Alnus rubra Bong., red alder, can levels. Aluminum and iron hydroxides tightly control accrete up to 130 kg N haÀ1 per year 3Binkley, 1992), P retention by ®xation, and thus very little P is lost to while Ceanothus spp. can accrete up to 101 kg N haÀ1 deeper horizons by leaching. Additional geochemical per year 3Youngberg and Wollum, 1976; Binkley, controls on P availability include surface adsorption± 1982; Conard et al., 1985; McNabb and Cromack, desorption on organic matter, soil acidi®cation 3Bink- 1983). In contrast, Walker 31993) found no correlation ley and Sollins, 1990; Frossard et al., 1995), and between the presence of nitrogen-®xing species and dissolution and precipitation of minerals 3Gobran total N accumulation in surface soil in a review of et al., 1998). primary succession studies. Cromack et al. 31999) Although nitrogen-®xing species may in¯uence the found no correlation between red alder basal area biological and geochemical soil P controls differently and the amount of soil N in a coastal Oregon than non-®xing species, few studies have looked at Douglas-®r plantation. species effects on soil biological or geochemical P Besides N accretion and N-related properties, nitro- sinks 3Compton and Cole, 1998). Nitrogen-®xing gen-®xing plants may affect other soil properties species have been shown to increase soil acidity in including soil pH 3Binkley and Sollins, 1990), enzyme some studies, which may increase the lability of soil P. activity 3Giardina et al., 1995; Zou et al., 1995), Recent studies on red alder have shown that the carbon accumulation 3Cole et al., 1995), soil phos- process of nitrogen-®xation can acidify soil 3Binkley phorus fractions and P cycling 3Giardina et al., 1995; and Sollins, 1990), decrease soil P and base cation Compton and Cole, 1998; Binkley et al., 2000). Nitro- content, and increase soil organic matter 3Cole et al., gen-®xing species have high P requirements, which 1995). Decreasing soil pH converts primary phosphate are essential for the high levels required during N- minerals into secondary Fe and Al-phosphate minerals ®xation 3Grif®th, 1978; Spent, 1988; Gressel et al., 3Lajtha and Schlesinger, 1988), while maximum P 1996), and therefore lower available soil P concentra- availability occurs at intermediate pH values. Yet in tions may result under nitrogen-®xing species than Nevada, Johnson 31995) found that the nitrogen-®xing under non-®xing species. This high requirement for P, species Ceanothus did not acidify soil, decrease base- in effect, closely links the N and P cycles in soils 3Cole saturation or lower extractable P. Although Ceanothus and Heil, 1981). Due to this greater P demand, ter- velutinus is the primary nitrogen ®xer in the higher restrial nitrogen-®xing species are believed to cycle P elevation xeric sites in the PACIFIC NORTHWEST at faster rates than non-®xing species. However, a 3PNW), few studies have looked at the effect of this consistent pattern of nitrogen-®xing species increas- species on available soil P. ing or decreasing available soil P has not been estab- Nitrogen-®xing species may not only affect the total lished. amount and availability of N, but may also affect the Differences in soil P supply may result from bio- isotopic composition of soil N. The mean d15Nof logical or geochemical processes at different depths nitrogen-®xing species' foliage has been shown to be J.D.H. Spears et al. / Forest Ecology and Management 14942001) 205±216 207 signi®cantly lower than the mean of non-®xing spe- ence would be evident in surface soils where biolo- cies, who derive their N from the soil 3Kohl and gical mechanisms control P, but not at depth where Shearer, 1980; Vitousek et al., 1989). Isotopic signals geochemical mechanisms would dominate. from nitrogen-®xing plants re¯ect the atmospheric We further hypothesized that there should be an 15N abundance, while leaves from plants that rely enriched Ceanothus d15N isotopic signal that could be on soil N are more enriched in 15N 3Shearer et al., used to distinguish nitrogen ®xation-N from soil N, 1980; Lajtha and Marshall, 1994; Handley and Scrim- and that this difference would be evident in the surface geour, 1997; HoÈgberg, 1997). soil N pools. We hypothesized that the pH of the soils It has been hypothesized that the d15N signal from a would be lower under the Ceanothus due to nitrogen nitrogen-®xing species could be used to trace atmo- ®xation. We also hypothesized that phosphatase activ- spherically derived N through soil N pools beginning ity would be greater under Ceanothus, because Cea- with leaf litter. In an experiment to test this hypothesis, nothus may have an increased need for organic forms Binkley et al. 31985) concluded that 15N isotopes are of soil P. Enzymes catalyze the rate limiting steps of not a simple method of tracing alder ®xed-N in the nutrient cycling and often begin the transformation of soil. Yet his study was at the stand-level scale and the organic forms of nutrients into available inorganic d15N values varied inconsistently from stand to stand. forms; and therefore phosphatase activity can be used Lajtha and Schlesinger 31986) suggested that, as an indicator of P availability in soils. We also although the d15N value of the desert shrub Prosopis hypothesized that b-glucosidase enzyme activity is unique, the values could be lost upon entering the would also be greater under Ceanothus than Dou- general soil pool. However, as Lajtha and Schlesinger glas-®r, because decomposition of N-rich Ceanothus 31986) concluded, the distance from a nitrogen-®xing litter may be faster than N-poor Douglas-®r litter. tree or shrub may in¯uence the d15N of the soil. Most previous studies examining the effects of nitrogen-®xers on soil biogeochemistry have been at the 2. Methods and materials stand level, e.g. comparing the difference between a conifer stand, a pure nitrogen-®xer stand, or a mixed 2.1. Site description stand. However, such studies may be inherently proble- matic: there may be differences in stocking density, site The study site is located in Watersheds 6 and 7 at the characteristics, or soils among stands that make these H.J. Andrews Experimental Forest in the Cascade results dif®cult to interpret 3Compton and Cole, 1998). Mountains of western Oregon. Although delineated Indeed, Kohl et al. 31981) have shown that there is a as two watersheds, the site is one contiguous stand of large site-to-sitevariation in d15Nvalues, and to date, no Douglas-®r and Ceanothus. Soils are frigid Andic 15N studies have been conducted on nitrogen-®xing Dystrudepts, derived from volcanic ash and pumice, trees at the single tree scale to test the above hypothesis. deep, and well-drained with a 35% southern slope. The overall objective of this study was to determine The climate is quasi-Mediterranean with cool, wet if C. velutinus, as a nitrogen-®xing species, has dif- winters and warm, dry summers. Precipitation ferent effects on soil nutrient availability than a non- averages 2500 mm per year, and mean January and ®xing species, Douglas-®r. This study was conducted July temperatures are 2 and 118C, respectively. The within a single stand to reduce variation in abiotic stand was clear-cut and burned under low intensity in factors, and possible inter-stand variation. Speci®c 1974, and planted with Douglas-®r in 1975 and 1976. objectives were to characterize soil P, and to examine Ceanothus germinated from the soil seed bank after relationships between species for soil P, N, and soil ®re 3Binkley et al., 1982) and are now beginning to be phosphatase and b-glucosidase activities. We overtopped by the Douglas-®r canopy. hypothesized that the nitrogen-®xing species, Cea- nothus, would decrease soil P concentrations due to 2.2. Field sampling the increased demand for P by nitrogen-®xation and therefore soils beneath Ceanothus would have lower Field sampling consisted of randomly placed trans- labile P concentrations than Douglas-®r. This differ- ects from the road into the stand. Soil pits were dug 208 J.D.H. Spears et al. / Forest Ecology and Management 14942001) 205±216

15 À under 12 Ceanothus and 12 Douglas-®r trees on the extracted in 2 M KCl and analyzed for d NasNO3 - transects, which were selected based on the criteria N and NH4 -N 3Sorenson and Jensen, 1991; Stark and that the litter beneath the tree was speci®c to that Hart, 1996; Sigman et al., 1997). species and did not overlap with the opposite species. Soil pH was analyzed in 0.01 M CaCl2 3McLean, Soil was collected at four discrete depths 35, 15, 30 1982). Soil organic matter was estimated by loss on and 60 cm) for P fractionation in August 1996. Sam- ignition 3Lim and Jackson, 1982). ples for enzyme analysis and N analysis were collected from 0 to 5 cm beneath 10 Ceanothus and 10 2.2.2. Statistical analysis Douglas-®r in August 1998, and at two depths 35 and All statistical analyses of data were performed 15 cm) in November 1998. using SAS 3SAS Institute, Cary, NC, 1996). A Stu- dent's t-test was used to compare the differences in À 2.2.1. Chemical analysis soil pH, NO3 ,NH4 , phosphatase activity, b-gluco- Soil samples were analyzed for phosphorus using a sidase activity, and d15N isotope values between Cea- modi®ed Hedley fractionation, which separates P into nothus and Douglas-®r soils. Although correlation fractions based on their biological availability 3Hedley exists between depths, examining the data at each et al., 1982; Tiessen et al., 1983; Lajtha and Schle- depth individually still results in a valid test. There- singer, 1988). Labile phosphorus in soil solution was fore, P fractions were also analyzed by t-tests. All removed using an anion exchange resin. Adsorbed hypotheses were tested at a 0:05. labile P was extracted with a 0.5 M sodium bicarbonate solution. Amorphous, crystalline aluminum and iron phosphate P was extracted with NaOH. This 3. Results fraction can also contain organic phosphorus associated with humic and fulvic acids, and is therefore The sum of the Hedley P fractions and total P by the separated into organic and inorganic fractions. The fusion method were highly correlated 3R2 0:91)

NaOH Po fraction was calculated by subtracting suggesting that sample extraction ef®ciency and accu- NaOH Pi from total NaOH P after digestion. Insoluble racy remained relatively constant between the two calcium bound phosphorus was extracted with HCl. techniques. Total P was highest in the surface soils

Residual P was analyzed by H2SO4±H2O2 digestion and decreased with depth ranging from 518 to À1 3Lowther, 1980), or by LiBO4 fusion 3Thompson and 1961 mgPg in soils under Ceanothus, and from Walsh, 1983). Total soil P was measured on 16 622 to 3308 mgPgÀ1 in soils under Douglas-®r. Total replicate samples by fusion to compare with the soil P was 22% greater in Douglas-®r soils at 5 cm sum of the Hedley fractions. All samples were ana- 3p 0:05) and 26% greater at 15 cm 3p 0:04). lyzed on an Orion Scienti®c AC 100 Autoanalyzer Although there were no signi®cant differences in total using the ascorbic acid method for ortho-P 3Lennox, P between the two species deeper in the soil pro®le, 1979). the mean soil concentration of total P under Douglas- Soil enzyme assays were adopted from Caldwell ®r was higher than Ceanothus at 30 and 60 cm et al. 31999) using 5 g 3surface) soil slurried with 3Fig. 1a±h). 25 ml deionized water. Phosphatase assays were incu- In general, each phosphorus fraction was higher for bated for 1 h and b-glucosidase assays for 3 h. Results Douglas-®r soils than for Ceanothus soils, while the were calculated as umol p-nitrophenol released per mean concentration of the P fractions generally gram dry weight per hour. decreased with depth for both species. Resin P was Surface mineral soils collected in summer and fall signi®cantly lower in Ceanothus soils than Douglas-®r of 1998 were analyzed for total nitrogen after H2SO4± soils at 5 cm 3p 0:006), and marginally lower at À H2O2 digestion 3Lowther, 1980), and for NO3 -N and 15 cm 3p 0:07) and 60 cm 3p 0:07). There were NH4 -N after 1 M KCl extraction. no signi®cant differences between Ceanothus soils Soil, litter, and leaves from Ceanothus and Douglas- and Douglas-®r soils at 30 cm 3p 0:16). ®r were sent to the Boston University Stable Isotope Bicarbonate P was lower in Ceanothus soils than Lab for analysis of d15N. Additional soil samples were Douglas-®r soils at 5 cm 3p 0:07), but was more J.D.H. Spears et al. / Forest Ecology and Management 14942001) 205±216 209

Fig. 1. Mean soil phosphorus concentrations in ug P gÀ1 air-dry soil from four depths under Douglas-fir and Ceanothus 3n 12) for each Hedley fraction 3a±g) and total P 3h). 210 J.D.H. Spears et al. / Forest Ecology and Management 14942001) 205±216

Fig. 1. 3Continued ) clearly lower at 30 cm 3p 0:02). There were no although the trend of more P under Douglas-®r signi®cant species differences at 15 cm 3p 0:16) remained constant. and 60 cm 3p 0:22). Phosphatase activity was marginally greater in soils

NaOH Pi was signi®cantly greater in soil under under Ceanothus 3p 0:09) 3Table 1). There were no Douglas-®r than soil under Ceanothus at 5 cm differences in phosphatase activity between Cea- 3p 0:045) and marginally greater at 30 cm nothus and Douglas-®r soils in fall 3p 0:8862). 3p 0:08). There was no difference between species No difference was found in b-glucosidase activity at 60 cm 3p 0:71). NaOH Po was signi®cantly for either summer or fall 3p 0:16 and 0.44, respec- greater for Douglas-®r than for Ceanothus only at tively). There were no signi®cant pH differences 15 cm depth 3p 0:04). between soils for these species. Total NaOH P was signi®cantly higher under Dou- Total N was greater in soil under Ceanothus, glas-®r than Ceanothus at 5 cm 3p 0:04) and at 2.60 mg gÀ1, than in soil under Douglas-®r, 15 cm 3p 0:04), marginally higher at 30 cm 1.99 mg gÀ1 3p 0:0283). Ceanothus litter was 3p 0:09), and not signi®cantly different between higher in total N, 17.94 mg gÀ1, than Douglas-®r litter, À1 À the two species at 60 cm 3p 0:81). Residual and 5.34 mg g 3p 0:0001). Extractable NO3 -N was HCl P fractions were not signi®cantly different signi®cantly greater under Ceanothus soils in the between Ceanothus and Douglas-®r soils at any depth, summer 3p 0:0355) although extractable NH4 -N

Table 1 Mean b-glucosidase and phosphatase enzyme activity in mmol gÀ1 soil 3n 10)a

Ceanothus 3umol gÀ1 hÀ1) Douglas-fir 3umol gÀ1 hÀ1) p-value

Summer 1998 Phosphatase 14.47 32.60) 8.89 31.75) 0.09 b-Glucosidase 5.29 30.69) 3.93 30.64) 0.17 Organic matter 0.40 30.03) 0.42 30.05) 0.69 Fall 1998 Phosphatase 15.37 33.32) 15.90 31.61) 0.89 b-Glucosidase 4.03 30.45) 3.59 30.36) 0.44 a S.E. are given in parentheses. p-values are from a t-test for a difference between species. J.D.H. Spears et al. / Forest Ecology and Management 14942001) 205±216 211

Table 2 À a Mean NO3 -N and NH4 -N concentrations in KCl extracts for Ceanothus and Douglas-fir soils 3n 10) Depth 3cm) Ceanothus 3mg N gÀ1 soil) Douglas-fir 3mg N gÀ1 soil) p-value

Summer 1998 À NO3 -N 5 1.59 30.13) 1.22 30.10) 0.04 NH4 -N 5 22.70 33.44) 15.53 31.38) 0.08 Fall 1998 À NO3 -N 5 5.72 30.54) 5.24 30.57) 0.55 NH4 -N 5 8.04 31.13) 4.18 30.88) 0.01 À NO3 -N 15 5.60 30.39) 3.92 30.56) 0.02 NH4 -N 15 2.99 30.38) 1.74 30.80) 0.02 a S.E. are given in parentheses. p-values are from a t-test for a difference between species.

Table 3 Mean delta 15N values for Ceanothus and Douglas-fir leaves, litter and soil 3n 10)a

Ceanothus Douglas-fir p-value

Leaf 15N À0.327 30.05) À4.17 30.22) 0.0001 Litter 15N À0.86 30.17) À1.57 30.24) 0.03 Soil 15N 3.845 30.68) 3.915 30.36) 0.93 15 NH4 N in KCl extract 4.7 30.44) 6.31 30.56) 0.04 À 15 NO3 N in KCl extract À2.212 30.76) À0.168 30.40) 0.03 a S.E. are given in parentheses. p-values are from a t-test for a difference between species. was only marginally greater under Ceanothus 4. Discussion 3p 0:08) 3Table 2). NH4 -N was signi®cantly greater in soil under Ceanothus in the fall 4.1. Soil P 3p 0:015), although there was no species difference À for NO3 -N 3p 0:55). There was a signi®cant sea- Wood et al., 1984 suggested that biological activity À sonal difference for NO3 -N 3p 0:0001) and for controls the retention of P in surface soils, while NH4 -N 3p 0:0001) in soils. For both species' soils, geochemical activity controls the retention of P in À NO3 -N was lower in the summer than in the fall, lower soil horizons. Our results are consistent with while NH4 -N was lower in the fall than in the Wood's hypothesis. We found no signi®cant differ- summer. ences between species for residual or HCl P fractions Although the d15N of Ceanothus leaves 3À0.327) at any depths, suggesting that these fractions are under and litter 3À0.86) were signi®cantly higher than geochemical rather than biological control. We found Douglas-®r needles 3À4.17, p < 0:0001) and litter signi®cant differences between species for total P, 3À1.57, p 0:025), total soil d15N did not differ resin P, and bicarbonate P at shallow soil depths, between species 3p 0:92) 3Table 3). KCl-extracted but not in lower soil horizons, which suggests that 15 d NofNH4 -N values were signi®cantly lower for biological activity is more prevalent in the surface soil the Ceanothus soil than for the Douglas-®r soil horizons and that Ceanothus may affect soil differ- 15 À 3p 0:04). The d N of KCl-extracted NO3 -N was ently than Douglas-®r. signi®cantly lower in Ceanothus soils than in Douglas- The differences in soil P fraction concentrations ®r soils 3p 0:0325). No correlation was found between species may be in¯uenced by soil biological À between extractable levels of NH4 or NO3 and activity. Microbial activity may increase soil weath- d15N. ering and thus play a role in P availability. Dissolution 212 J.D.H. Spears et al. / Forest Ecology and Management 14942001) 205±216 of inorganic phosphates by organic acids or H may Compton and Cole 31998) hypothesized that lower stimulate weathering of adjacent minerals 3Ban®eld soil pH under nitrogen-®xing alder would increase the et al., 1999). Several aliphatic acids such as oxalate anion sorption capacity of the soil and therefore, more and citrate are very effective in the dissolution of P- Pi would be adsorbed to Fe and Al hydrous oxides. containing minerals, as well as in the mobilization of P Instead, they found that NaOH Po was the largest pool held in humic-metal complexes 3Frossard et al., 1995; under red alder and NaOH Pi was the largest pool Jones, 1998). Mycorrhizal fungi can secrete organic under Douglas-®r. Our study found that for both acids, which may dissolve insoluble P minerals or they species, NaOH Po increased from 5 to 15 cm but then may secrete phosphatases which hydrolyze Po 3Cro- declined with increasing depth, while NaOH Pi mack et al., 1979; Frossard et al., 1995). Oxalic acid decreased from depth 5 to 15 cm. The decrease in has been shown to react with Al and Fe in andesite, NaOH Pi is indicative of root uptake; the increase in increasing soil weathering under Douglas-®r 3Cro- NaOH Po is not as easily explained, but might be mack et al., 1979). These reactions may have caused indicative of root exudation or high microbial activity the distinct decrease in NaOH Pi in the rhizosphere and/or microbe and root turnover. under Douglas-®r 3Fig. 1c). However, desorption and Although both Douglas-®r and Ceanothus have release of P by organic acids is soil dependent, and mycorrhizal symbionts, one mycorrhizae may be more generally requires high concentrations of organic effective than another at ®nding and absorbing P in acids 3Jones, 1998). soil. Douglas-®r are known to have ectomycorrhizae, Because soil enzyme activities are sensitive to whereas Ceanothus have arbuscular mycorrhizae. changes in soil properties, assays of microbial Arbuscular mycorrhizae are not as effective as ecto- enzymes can be used as general measures of hetero- mycorrhizae in aggregate penetration 3Cromack, per- trophic microbial activity 3Sinsabaugh, 1994) and sonal communication). Although the Douglas-®r more speci®cally, as measures of nutrient cycling in symbiont may be more effective in P acquisition, the ecosystem 3Bergstrom et al., 1998). We found our results suggest that the Ceanothus and its symbiont marginally signi®cant differences between Ceanothus were more ef®cient in mining soil P because the soil P and Douglas-®r for phosphatase activity, but not for b- concentration beneath Ceanothus was generally lower glucosidase activity. Our results agree with Zou et al. than the Douglas-®r. However, the difference in the 31995), who did ®nd an increase in phosphatase total P between Ceanothus and Douglas-®r soil was activity in soils under both nitrogen-®xing alders 110 kg P haÀ1 for the top 15 cm of the soil and could in Oregon and Albizia falcataria in Hawaii, but not be accounted for in Ceanothus biomass. We concluded that the nitrogen-®xing species, A. falca- estimated the biomass of our stand by assuming that taria, had little effect on soil P transformation rates. our stand is similar to a 17-year-old stand at Giardina et al. 31995) found phosphatase activity to 42 680 kg haÀ1 measured by Cromack et al. 31979). be three times greater in an alder-Douglas-®r stand Assuming a mean P concentration in biomass of than a pure Douglas-®r stand, and they suggested that 0.08%, the amount of P in Ceanothus biomass could the alder-Douglas-®r stand's higher enzyme activity only be 28.33 kg haÀ1. may help to meet alder's high requirement for P. Other possible explanations for the large difference However, Giardina et al. 31995) also concluded that in P from soil under Ceanothus compared to Douglas- the correlation between resin P and phosphatase may ®r include leaching of organic and inorganic P from be due to soil differences between stand types. the soil, dispersion of P-rich litter from Ceanothus Indeed, McGill and Cole 31981) found an inverse through the area, non-random germination of Cea- relationship between phosphatase activity and labile nothus in the ®eld, or simply random ®eld sampling inorganic P. They concluded that the lack of consis- error. Leaching of both organic and inorganic P from tent relationships between available P and phospha- PNW soils are low 3Compton and Cole, 1998). tase activity is perhaps due to the complexity of Although it is possible that Ceanothus preferentially organic P polymers; additional enzymes may be germinated in low P microsites, this is highly unlikely needed to depolymerize the Po before it is available at the scale and density of Ceanothus in the ®eld. for phosphatase. Ceanothus appears to germinate and grow in dense J.D.H. Spears et al. / Forest Ecology and Management 14942001) 205±216 213 stands until outcompeted by planted Douglas-®r, and would tend to decompose faster than litter from a because Douglas-®r in these stands are planted, it is non-®xing species 3Melillo et al., 1982). unlikely that they were systematically planted in Although many authors have attempted to use high P microsites. Field sampling errors may occur natural abundance d15N isotopes as tracers of biolo- randomly. Therefore, we used retrospective power gically-®xed N in soils, in most cases it has been analysis to check the power of our statistics. Power largely unsuccessful 3Binkley et al., 1985; Lajtha and is the probability of rejecting the null hypothesis if Marshall, 1994). Indeed, we found that the d15Nin the null is false, while some alternative hypothesis is Ceanothus and Douglas-®r leaves were signi®cantly true. Aretrospective power test using observed var- different, with Ceanothus re¯ecting the atmospheric iances can evaluate whether the sample size and the signal while the Douglas-®r needles were depleted. a-level are suf®cient to detect a meaningful signi®- The differences in d15N ratios between species nar- cance given an observed level of variation 3Thomas, rowed in leaf litter. Ceanothus litter values became 1997). Our power to detect a meaningful difference more depleted while Douglas-®r values were slightly of 200 units 3in our case 200 ug P gÀ1) between more enriched. This may occur in Ceanothus, perhaps, species was 0.8, a value which indicates that our because the heavier isotope is discriminated against sample size was adequate to detect a statistical dif- during N resorption before leaf senescence and abscis- ference in total P between species amid a large amount sion. No study has determined if fractionation occurs of variation. during N resorption. Although understory plant biomass appears greater Soil d15N values were approximately equal for under Ceanothus as compared to Douglas-®r, it is Ceanothus and Douglas-®r 33.8 and 3.9%, respec- unlikely that understory species alone could account tively) and are in the range 35toÀ12%) commonly for such a large P discrepancy. Furthermore, litter found in mineral soils 3Nadelhoffer and Fry, 1994; dispersion could account for only 8.89 kg P haÀ1 over Stevenson, 1994). Mineralization and immobilization 9 years. This amount was calculated from discriminate against the heavier isotope, leaving it 157kgNhaÀ1 in litter 3Youngberg and Wollum, behind in residual organic matter 3Lajtha and Mar- 1976), and converted to P using a N:P value of shall, 1994), and thus the atmospherically-derived 15N 17.65 as found by in Valachovic 31998) for snowbrush. signal can be obscured in soil organic matter. The d15N Further studies in a broad range of sites will be needed values of NH4 became heavier still which is not as to test the robustness of our results and to elucidate the easily explained. These results do concur with HoÈg- mechanisms behind the large difference in phosphorus berg 31990) who found that large additions of fertilizer under Douglas-®r as compared to Ceanothus. N with a low d15N abundance increased the d15N values of the soil. 4.2. Soil nitrogen The d15N of available N became further convoluted À in NO3 . Nitri®cation may discriminate against 15 Previous work has shown that soil under nitrogen- NH4 and hence the soil nitrate became more ®xing alder has a greater organic matter content and depleted, but our values seemed extreme. Therefore, lower bulk density than under a non-nitrogen-®xing even at the single tree level, d15N isotopes were not species 3Tarrant and Miller, 1963; Binkley, 1994; Cole able to trace N ®xed by Ceanothus into the soil N pool et al., 1995). Soil C has been shown to parallel N as the signal was lost upon entry into the soil N pool. accumulation under alder 3Cole et al., 1991). In con- The progressive enrichment pattern found in this study trast, we found no difference in soil organic matter and by HoÈgberg et al. 31995) deserves more study. content in soils under Douglas-®r versus those under Ceanothus, but we did ®nd more available N under Ceanothus than Douglas-®r. This could be due to 5. Conclusion increased decomposition of Ceanothus litter, which is less recalcitrant than Douglas-®r needles, or it could Management concerns that N-®xing species decrease be due to lower amounts of litterfall under Douglas-®r. soil P availability and therefore, limit Douglas-®r In general, litter from the nitrogen-®xing species growth, remain unanswered 3Giardina et al., 1995). 214 J.D.H. Spears et al. / Forest Ecology and Management 14942001) 205±216

Although some studies have found that alder, in a Berendse, F., 1998. Effects of dominant plant species on soils mixed stand, can lower P nutrition of Douglas-®r, during succession in nutrient-poor ecosystems. Biogeochem- istry 42, 73±88. other studies have found nitrogen-®xing species to Bergstrom, D.W., Monreal, C.M., King, D.J., 1998. Sensitivity of have no affect on Douglas-®r nutrition 3Binkley and soil enzyme activities to conservation practices. Soil Sci. Soc. Husted, 1983; Binkley et al., 1991). Results of most Am. J. 62, 1286±1295. studies vary depending on the stand, soil, and species. Binkley, D., 1992. Mixtures of nitrogen-fixing and non-nitrogen- For example, a mixed stand of Sitka alder contained fixing tree species. In: Cannell, M.G.R., Malcolm, D.C., more P than a mixed stand of red alder suggesting that Robertson, P.A. 3Eds.), The Ecology of Mixed-Species Stands of Trees. The British Ecological Society, Oxford. P availability may be determined by site-speci®c Binkley, D., 1994. The Influence of tree species on forest soils: interactions 3Binkley, 1992). processes and patterns. In: Proceedings of the Trees and Soil The Hedley fractionation used in this study, is a Workshop, Lincoln University. Lincoln University Press, static measure of pool size, and therefore does not Cantesbury, pp. 1±33. Binkley, D., Husted, L., 1983. Nitrogen accretion, soil fertility and include a true assessment of availability 3¯ux). How- Dougas-fir nutrition in association with redstem ceanothus. ever, because plants may have different methods of Canadian J. For. Res. 13, 122±125. acquiring P, e.g. mycorrhizae and organic acids, ¯ux Binkley, D., 1982. Nitrogen fixation and net primary production in and bioavailability of P is not as easily measured as it a young Sitka alder ecosystem. Canadian J. Botany 60, 281± is for N. Our study has shown that surface soils under 286. Binkley, D., Sollins, P., 1990. Factors determining differences in individual Douglas-®r trees have more total P than soil pH in adjacent conifer and alder-conifer stands. Soil Soc. soils under individual Ceanothus trees. Further study Sci. Am. J. 54, 1427±1433. is necessary to clarify the spatial and temporal extent Binkley, D., Cromack Jr., K., Fredriksen, R.L., 1982. Nitrogen of a species effect in soil, i.e. how long does the accretion and availability in some snowbrush ecosystems. For. nitrogen-®xer effect concentrations of available soil Sci. 28, 720±724. P and how far 3in distance) from the tree does the Binkley, D., Sollins, P., McGill, W.B., 1985. Natural abundance of nitrogen-15 as a tool for tracing Alder-fixed nitrogen. Soil Sci. effect extend? Soc. Am. J. 49, 444±447. Binkley, D., Giardina, C., Bashkin, M.A., 2000. Soil phosphorus pools and supply under the influence of Eucalyptus saligna and Acknowledgements nitrogen-fixing Albizia facaltaria. For. Ecol. Manage. 128, 241±247. Boring, L.R., Swank, W.T., Waide, J.B., Hendersen, G.S., 1988. We would like to thank Sherman Bloomer for Sources, fates, and impacts of nitrogen inputs to terrestrial instruction on the P fusion, and Dan Binkley, Dave ecosystems: review and synthesis. Biogeochemistry 6, 119± Myrold, Mark Harmon, Kermit Cromack, and Pat 159. Muir for reviewing earlier drafts, and Scott Holub Caldwell, B.A., Griffiths, R.P., Sollins, P., 1999. Soil enzyme for support. We would also like to thank the two response to vegetation disturbance in two lowland Costa Rican soils. Soil Biol. Biochem. 31, 1603±1608. anonymous reviewers for their insights. Cole, D.W., Compton, J.E., Edmonds, R.L., Homann, P.S., Van Miegroet, H., 1995. Comparison of carbon accumulation in Douglas-fir and red alder forests. In: McFee, W.W., Kelly, J.M. References 3Eds.), Carbon Forms and Functions in Forest Soils. Soil Science Society of America, Madison, WI, pp. 527±546. Ae, N., Otani, T., 1997. The role of cell wall components from Cole, C.V., Heil, R.D., 1981. Phosphorous effects on terrestrial groundnut roots in solubilizing sparingly soluble phosphorus in nitrogen cycling. Terrestrial Nitrogen Cycles, in: Clark, F.E., low fertility soils. Plant Soil 196, 265±270. Rosswall, T. 3Eds.), Ecological Bull. Stockholm 33, 363±374. Ae, N., Arihara, J., Okada, K., Yoshihara, T., Johansen, C., 1990. Cole, D.W., Compton, J., Van Miegroet, H., Homann, P., 1991. Phosphorus uptake by pigeon pea and its role in cropping Changes in soil properties and site productivity caused by alder. systems of the Indian subcontinent. Science 248, 477±480. Water, Air, and Soil Pollution 54, 231±246. Angers, D.A., Caron, J., 1998. Plant-induced changes in soil Compton, J.E., Cole, D.W., 1998. Phosphorus cycling and soil P structure: processes and feedbacks. Biogeochemistry 42, 55±72. fractions in Douglas-fir and red alder stands. For. Ecol. Banfield, J.F., Barker, W.W., Welch, S.A., Taunton, A., 1999. Manage. 110, 101±112. Biological impact on mineral dissolution: application of the Conard, S.G., Jarmillo, A.E., Cromack Jr., K., Rose, S., 1985. The lichen model to understanding mineral weathering in the role of the genus Ceanothus in western forest ecosystems. US rhizosphere. Proc. Natl. Acad. Sci. 96, 3404±3411. Department of Agriculture, PNW-182. J.D.H. Spears et al. / Forest Ecology and Management 14942001) 205±216 215

Cromack Jr., K., Sollins, P., Graustein, W.C., Speidel, K., Todd, Lajtha, K., Schlesinger, W.H., 1986. Plant response to variation in A.W., Spycher, G., Li, C.Y., Todd, R.L., 1979. Calcium oxalate nitrogen availability in a desert shrubland community. Biogeo- accumulation and soil weathering in mats of the hypogenous chemistry 2, 29±37. fungus Hysterangium Crassum. Soil Biol. Biochem. 11, Lajtha, K., Schlesinger, W.H., 1988. The biogeochemistry of 463±468. phosphorus cycling and phosphorus availability along a desert Cromack Jr., K., Miller, R.E., Helgerson, O.T., Smith, R.B., soil chronosequence. Ecology 69, 24±39. Anderson, H.W., 1999. Soil carbon and nutrients in a coastal Lennox, L.J., 1979. An automated procedure for the determination Oregon Douglas-fir plantation with red alder. Soil Sci. Soc. of phosphorus. Water Res. 13, 1329±1333. Am. J. 63, 232±239. Lim, C.H., Jackson, M.L., 1982. Dissolution for total elemental Frossard, E., Brossard, M., Hedley, M.J., Metherell, A., 1995. analysis. In: Page, L., Miller, R.H., Keeney, D.R. 3Eds.), Reactions controlling the cycling of P in soils. In: Tiessen, H. Methods of Soil Analysis, Part 2. Soil Science Society of 3Ed.), Phosphorus in the global environment: transfers, cycles America, Madison, WI, pp. 1±11. and management. Chichester, NY. Lowther, J.R., 1980. Use of a single sulfuric acid±hydrogen Giardina, C.P., Huffman, S., Binkley, D., Caldwell, B.A., 1995. peroxide digest for the analysis of Pinus radiata needles. Alders increase soil phosphorus availability in a Douglas-fir Commun. Soil Sci. Plant Anal. 11, 175±188. plantation. Can. J. For. Res. 25, 1652±1657. McGill, W.B., Cole, C.V., 1981. Comparative aspects of cycling of Gobran, G.R., Clegg, S., Courchesne, F., 1998. Rhizospheric organic C, N, S, and P through soil organic matter. Geoderma processes influencing the biogeochemistry of forest ecosys- 26, 267±286. tems. Biogeochemistry 42, 107±120. McLean, E.O., 1982. Soil pH and lime requirement. In: Page, A.L., Gressel, N., McColl, J.G., Preston, C.M., Newman, R.H., Powers, Miller, R.H., Keeney, D.R. 3Eds.), Methods of Soil Analysis, R.F., 1996. Linkages between phosphorus transformation and Part 2. Madison, WI, pp. 199±223. carbon decomposition in a forest soil. Biogeochemistry 33, McNabb, D.H., Cromack Jr., K., 1983. Dinitrogen fixation by a 97±123. mature Ceanothus velutinus Dougl. stand in the western Oregon Griffith, W.K., 1978. Effects of phosphorus and potassium on cascades. Can. J. Microbiol. 29, 1014±1021. nitrogen fixation. In: Phosphorus for Agriculture Ð A Situation Melillo, J.M., Aber, J.D., Muratore, J.F., 1982. Nitrogen and lignin Analysis. Potash/Phosphate Institute, Atlanta, GA, pp. 80±94. control of hardwood leaf litter decomposition dynamics. Handley, L.L., Scrimgeour, C.M., 1997. Terrestrial plant ecology Ecology 63, 621±626. and 15N Natural abundance: the present limits to interpretation Nadelhoffer, K.J., Fry, B., 1994. Nitrogen isotope studies in forest of uncultivated systems with original data from a Scottish old ecosystems. In: Lajtha, K., Michener, R.H. 3Eds.), Stable field. Advances in Ecol. Res. 27, 133±212. Isotopes in Ecology and Environmental Science. Blackwell HoÈgberg, P., 1997. 15N natural abundance in soil-plant systems. Scientific Publications, London, pp. 22±44. The New Phytologist 137, 179±203. Saetre, P., Brandtberg, P.-O., Lundkvist, H., Bengtsson, J., 1999. Hedley, M.J., Stewart, J.W.B., Chauhan, B.S., 1982. Changes in Soil organisms and carbon, nitrogen, and phosphorus miner- inorganic and organic soil phosphorus fractions induced by alisation in Norway spruce and mixed Norway spruce-Birch cultivation practices and by laboratory incubation. Soil Sci. stands. Biol. Fertil. Soil 28, 382±388. Soc. Am. J. 46, 970±976. SAS Institute Inc., 1989±1996. Cary, NC, USA. HoÈgberg, P., 1990. Forests losing large quantities of nitrogen have Shearer, G., Kohl, D.H., Harper, J.E., 1980. Distribution of 15N elevated 15N:14N ratios. Oecology 84, 229±231. among plant parts of nodulating and nonnodulating isolines of HoÈgberg, P., Johannisson, C., HoÈgberg, M., HoÈgbom, L., NaÈsholm, soybeans. Plant Physiol. 66, 57±60. T., HaÈllgren, J.-E., 1995. Measurements of abundances of 15N Sigman, D.M., Altabet, M.A., Michener, R., McCorkle, D.C., Fry, and 13C as tools in retrospective studies on N balances and B., Holmes, R.M., 1997. Natural abundance-level measurement water stress in forests: Adiscussion of preliminary results. of the nitrogen isotopic composition of oceanic nitrate: an Plant and Soil 168, 125±133. adaptation of the ammonia diffusion method. Mar. Chem. 57, Johnson, D.W., 1995. Soil properties beneath Ceanothus and pine 227±242. stands in the Eastern Sierra Nevada. Soil Sci. Am. J. 59, 918±924. Sinsabaugh, R.L., 1994. Enzymatic analysis of microbial pattern Jones, D.L., 1998. Organic acids in the rhizosphere Ð a critical and process. Biol. Fertil. Soils 17, 69±74. review. Plant Soil 205, 25±44. Sorenson, P., Jensen, E.S., 1991. Sequential diffusion of ammo- Kohl, D.H., Shearer, G., 1980. Isotopic fractionation associated nium and nitrate from soil extracts to a polytetrafluoroethlyene À 15 with symbiotic N2 fixation and uptake of NO3 by plants. Plant trap for N determination. Anal. Chim. Acta 254, 201±203. Physiol. 66, 51±56. Spent, J., 1988. The Ecology of the Nitrogen Cycle. Cambridge Kohl, D.H., Bryan, B.A., Shearer, G., Virginia, R., 1981. University Press, Cambridge, UK. Concerning the heterogeneity of the natural abundance of 15N Stark, J.M., Hart, S.C., 1996. Diffusion technique for preparing in soil nitrogen. Soil Sci. Soc. Am. J. 45, 450±451. salt solutions, Kjeldahl digests, and persulfate digests for Lajtha, K., Marshall, J.D., 1994. Sources of variation in the stable nitrogen-15 analysis. Soil Sci. Soc. Am. J. 60, 1846±1855. isotopic composition of plants. In: Lajtha, K., Michener, R.H. Stevenson, F.J., 1994. Humus Chemistry: Genesis, Composition, 3Eds.), Stable Isotopes in Ecology Environmental Science. Reactions. Wiley, New York, NY, 496 pp. Blackwell Scientific Publications, London, pp. 1±21. Tarrant, R.F., Miller, R.E., 1963. Accumulation of organic matter 216 J.D.H. Spears et al. / Forest Ecology and Management 14942001) 205±216

and soil nitrogen beneath a plantation of red alder and Douglas- Vitousek, P.M., Shearer, G., Kohl, D.H., 1989. Foliar 15N natural fir. Soil Sci. Soc. Am. Proc. 27, 231±234. abundance in Hawaiian rainforest: patterns and possible Thomas, L., 1997. Retrospective power analysis. Conserv. Biol. 11, mechanisms. Oecologica. 78, 383±388. 276±280. Walker, L.R., 1993. Nitrogen fixers and species replacements in Thompson, M., Walsh, J.N., 1983. Handbook of Inductively primary succession. In: Miles, J., Walton, D.W.H. 3Eds.), Coupled Plasma Spectrometry. Blackie, Glasgow, Scotland, Primary Succession on Land, Special Publ. 12, British 273 pp. Ecological Society. Blackwell Scientific Publications, Tiessen, H., Stewart, J.W.B., Moir, J.O., 1983. Changes in organic London. and inorganic phosphorus composition of two grassland soils Wood, T., Bormann, F.H., Vogt, G.K., 1984. Phosphorus cycling in and their particle size fraction during 60±90 years of cultivation. a northern hardwood forests: biological and chemical control. J. Soil Sci., Blackwell Scientific Publications, 34, 815±823. Science 223, 391±393. Valachovic, Y.S., 1998. Leaf litter chemistry and decomposition in Youngberg, C.T., Wollum II, A.G., 1976. Nitrogen accretion in a Pacific Northwest coniferous forest ecosystem. Master's developing Ceanothus velutinus stands. Soil Sci. Soc. Am. J. Thesis, Oregon State University, Corvallis, OR, 74 pp. 40, 109±112. Van Breemen, N., Finzi, A.C., 1998. Plant±soil interactions: Zou, X., Binkley, D., Caldwell, B.A., 1995. Effects of dinitrogen- ecological aspects and evolutionary implications. Biogeochem- fixing trees on phosphorus biogeochemical cycling in contrast- istry 42, 1±19. ing forests. Soil Sci. Soc. Am. J. 59, 1452±1458.