Seasonal Patterns of Nutrient Deposition in a Quercus Douglasii Woodland in Central California

Plant and Soil 137: 209-222, 1991. Q 1991 Kluwer Academic Publishers. Printed in the Netherlands. PLSO 8842

Seasonal patterns of nutrient deposition in a Quercus douglasii woodland in central California

RAGAN M. CALLAWAY and NALINI M. NADKARNI' Department of Biological Sciences, University of California Santa Barbara, CA 93106, USA. 1 Present address: The Marie Selby Botanical Garden, 811 South Palm Avenue, Sarasota, FL 34236, USA

Received 14 August 1990. Revised March 1991

Key words: California, litterfall, mediterranean-climate, nutrient-cycling, oaks, Quercus douglasii, throughfall

Abstract

The monthly deposition of total nitrogen, phosphorus, potassium, calcium and magnesium via canopy throughfall, and various components of the litterfall was measured for 31 months under mature Quercus douglasii and in the bulk precipitation in the surrounding open grassland. Seasonal patterns of nutrient concentration in leaf litter, throughfall, and precipitation were also measured. Total annual subcanopy deposition exceeded open precipitation deposition by approximately 45-60X for nitrogen, 5-15X for phosphorus, 30-35x for potassium, 25-35~ for calcium, and 5-10~for magnesium. Total annual subcanopy deposition was low in comparison to other oak woodland sites reported in the literature. Throughfall and leaf litter were the primary sources of nutrients and thus determined the seasonal peaks of nutrient deposition. The first autumn rains and leaf fall were associated with one peak in nutrient deposition, and throughfall during early spring leaf emergence was associated with a second peak in potassium, magnesium and phosphorus. Non-leaf plant litter (excluding acorns) provided approximately 15-35% of most nutrients, with twigs and bark depositing over 12% of the annual calcium flux in 1987-1988, and flower litter depositing over 8% of the annual nitrogen flux in 1986-1987. Acorns had high concentrations of phosphorus and nitrogen and during the mast season of 1987-1988 they contained a large proportion of the total subcanopy annual flux of these elements. With acorns excluded, total annual nutrient deposition was similar between years, but timing of nutrient deposition differed. Late summer leaf fall associated with drought, variation in precipitation, and variation in deposition of non-leaf parts were associated with seasonal differences in nutrient deposition between years.

Introduction al., in press; Holland, 1973; Holland and Mor- ton, 1980; Kay 1987), and appears to be the Quercus douglasii H. and A. (blue oak) is the result of nutrient uptake by the deeper rooted dominant tree on an estimated 3 million ha in trees and the subsequent deposition of nutrients California woodlands and savannas, and has a at the soil surface via litterfall and throughfall. substantial effect on subcanopy plants, mi- High soil fertility associated with Q. douglasii crobiota, and soils. Total and available soil nu- has been reported to improve understory pro- trients are higher under Q. douglasii canopies ductivity under some trees (Callaway et al.. in than in surrounding open grassland (Callaway et press; Holland, 1973; 1980), and appeared to 210 Callaway and Nadkarni improve grassland productivity in some areas range of sizes of our sample trees was repre- when trees were removed (Jansen, 1987; Kay, sentative of trees at the study site. The site had 1987; Murphy and Crampton, 1956). not been grazed for over 60 years. Soils at the In addition to increasing the total annual nu- study site are coarse, loamy, mixed mesic Entic trient inputs, Q. douglasii canopies alter the Haploxerolls, with surface soil textures ranging seasonal patterns of nutrient deposition. Nu7 from sandy loam to sandy clay loam. Surface trient availability in mediterranean-climate soils under Q. douglasii canopies average 6.3 in ecosystems is largely dependent on favorable pH (unpub. data), 6.7 mg g-' total N, 0.3 mg g-' combinations of warm temperatures, high soil total P, 2.1 mg g-' total K, 4.5 mg g-' total Ca, moisture and available organic material, which and 2.0 mg g -' total Mg. (Callaway et al., in occur in the spring and fall. This is also the time press). In the open grassland, surface soils aver- of maximum nitrogen uptake (Jackson et age 6.0 in pH, (unpub. data), 2.9 mgg-' total a1.,1988) for the annual plants that dominate the N, 0.15 mgg-' total P, 1.9mgg-I total K, grassland. Oak canopy-mediated changes in nu- 3.0 mg g-' total Ca and 1.9 mgg-I total Mg trient inputs during these months may have a (Callaway et al., in press). Soils at the site are significant effect on subcanopy productivity and underlain by a Cretaceous quartz diorite parent species composition. material. Although the contribution of Q. douglasii to Q. douglasii, an oak in the Leucobalanus sub- nutrient cycles in California grasslands is signifi- genus, is a California endemic, but it is wide- cant, no quantitative analyses of nutrient fluxes spread within the state and is the dominant tree in litterfall or throughfall have been conducted. on much of the eastern side of the coastal ranges In this paper we describe the seasonal patterns of and the western foothills of the Sierra Nevada deposition of total nitrogen, phosphorus, potas- (Griffin and Critchfield, 1976). Its size is variable sium, calcium, and magnesium in various compo- and acorn-bearing individuals may range from nents of the litter fall, and in the throughfall 5 m to 20 m in height and 20 cm to over 100 cm beneath Q. douglasii trees and compare sub- in diameter at breast height. Although it occurs canopy nutrient deposition to that in the precipi- in a climate characterized by winter rain and tation in the open grassland. summer drought, Q. douglasii is winter-deciduous. Leaves usually emerge by mid-March and drop between October and January. Although Methods winter-deciduous oaks do exist in southern Europe, no other region with a mediterranean Study area and species climate has such large areas that are dominated by winter-deciduous perennials. We conducted this study at the University of California Hastings Natural History Reservation Data collection in the Santa Lucia Mountains of central Califor- nia. The climate is mediterranean, with 90% of Litterfall was collected monthly between June the average annual precipitation of 54 cm occur- 1986 and December 1988, which spanned three ring between November and April. Mean mini- autumn leaf-shedding events. Six trees, scattered mum temperatures range from 1.4"C in January throughout the woodland were chosen nonran- to 9.7"C in August, and mean maximum tem- domly to encompass the wide range of variation peratures range from 8.5"C in January to 20.1°C in size (Callaway, 1990). The litterfall from each in July. tree was collected using plastic buckets 25 cm in The study site was a Q. douglasii woodland diameter and 30cm deep attached to steel tri- with a density of approximately 690 stems ha-'. pods 1.5 m tall at five fixed stations that had Q. douglasii are difficult to age, primarily due to been randomly located under the canopy. Litter rotten centers in older trees. However, we esti- was dried at 60°C immediately after collection mated that our sample trees ranged from 80 to and divided into leaf, twig and bark, lichen, 250 years, based on trunk core samples. The acorn, flower, and miscellaneous categories. The Nutrient deposition in oak woodlands 211 miscellaneous category consisted mostly of un- analyses. Twenty-five mL of each sample was identifiable fragments, bud scales, and insect digested separately, and post-digest analyses fol- frass. Because virtually all acorns that fall to the lowed those for litter. Both litterfall and through- ground are removed by predators under natural fall were collected near the middle of each conditions (pers. obs., see Borchert et al., 1989) month, although the day of collection was not and are unlikely to contribute directly to soil always the same. Time between collections nutrient pools available to understory plants, ranged from 28 to 33 days. they were not included in the monthly totals of nutrient deposition. Nutrient concentrations were measured in all Results leaf litter bulked by month. In several winter months when leaf fall was very low, leaves could Over 80% of the annual canopy deposition (ex- not be collected from all of the sample trees, and cluding acorns) for each element was in the these concentrations are based on small amounts throughfall and leaf fall (Table 1). Thus, the of leaves. Initial analyses of lichen litter revealed seasonal variation in these sources generally de- that concentrations of nutrients were not differ- termined the seasonal trends in total deposition. ent between months or seasons, so total lichen Throughfall provided high percentages of total litter was combined by year and analyzed in phosphorus, potassium, and magnesium. Thus, eight subsamples. Each of the other litter frac- total deposition of these nutrients (Figs. 1, 2, tions were combined by the fall, winter, spring and 3) in winter and spring was much higher and summer seasons and analyzed for nutrient than for nitrogen and calcium (Figs. 4 and 5) concentration in eight subsamples for each cate- which were deposited primarily in leaf fall be- gory in each season. Total nitrogen, phosphorus. tween August and January. potassium, calcium, and magnesium concentra- Nutrient deposition via non-leaf plant parts tions were determined after samples were ion- contributed significantly to the total annual nu- ized in a Technicon BD/20/40 Block Digester. trient flux (Table I), and was the primary source Nitrogen was sampled using a modified in- of some nutrients in several months. Flowers dophenol method (Setaro and Jones, 1989). were the primary source of nitrogen and calcium phosphorus using a modified molybdenum blue in May 1987, a period when throughfall and leaf assay (Setaro and Jones, 1989), and total cations deposition were low. Twig and bark litter con- using a Perkin-Elmer atomic absorption spectro- tained over 12% of the total (excluding acorns) photometer. After concentrations were deter- calcium flux in 1987-1988, and was the primary mined, they were used with the litterfall weights non-leaf source of this element. The 1987-1988 to estimate monthly nutrient fluxes. season was a mast year for oaks in the region. Throughfall and incident precipitation were Thus, annual nutrient flux via acorn litter ranged collected monthly during the wet seasons of from 20% of total magnesium to 55% of total 1986-1987 and 1987-1988. Collectors consisted phosphorus. The total amounts of nutrients de- of 10-cm diameter funnels attached by Tygon posited by lichen litter did not exceed 5% annu- tubing to 1-L Nalgene bottles. Collectors were ally, and peaks of deposition were associated fastened to steel stakes 50 cm tall and the tygon with exceptionally large pieces falling into single was looped to slow evaporative loss. Litter was collectors and not with any seasonal or blocked from entering the funnels with phenological pattern. aluminum mesh and glass wool. Four collectors Excluding acorns, annual nutrient fluxes were were placed under each of the trees that were similar (Table 1). Monthly patterns of nutrient sampled for litter fall and four were placed in the deposition, however, varied widely between open. One mL of formaldehyde was added each years (Figs. 1-5). The autumn rains began in month to prevent bacterial and algal growth and October in 1987 but not until November in 1986, the bottles were scrubbed with a brush to collect and the former were more highly concentrated in any build-up of organic material. Collected water potassium and magnesium (Fig. 6). Early rains, was frozen for up to six months before nutrient with high nutrient concentrations, and more 212 Callaway and Nadkarni

Table 1. Mean annual biomass and nutrient deposition ( + one S.D., n = 6 trees) under Quercus douglusii via litterfall and throughfall and in open grassland via bulk precipitation. Data are total nutrients after Kjeldahl digests (kg ha-') Litter weight Subcanopy deposition June 1986 -May 1987 Leaves 2514 + 611 Twigs and bark 314 + 123 Lichen 166 + 168 Miscellaneous 316 r 214 Acorns 39 42 Flowers 146 2 139 Throughfall - Total - Open PPT deposition -

June 1987- May 1988 Leaves Twigs and bark Lichen Miscellaneous Acorns Flowers Throughfall Total Open PPT deposition

June 1988- December 1988" Leaves 1756 2 622 21.6 t 8.0 0.5 2 0.2 4.7 t 1.5 14.2 1'- 5.1 2.1 20.7 "Annual leaf fall was complete by this date, other litter components were not annual totals.

abundant rain in the autumn and winter resulted was probably due to the cumulative effects of the in much higher fall nutrient deposition via drought. throughfall during these months in 1987-1988 Seasonal nutrient concentrations in leaf litter than in 1986-1987. Conversely, spring rainfall varied differently for each nutrient (Fig. 7). Ni- was higher in 1986 than in 1987, and spring trogen concentrations dropped by 50% between throughfall nutrient deposition was also higher August and September of 1986, but in the fol- (Figs. 1-5). In general, high throughfall concen- lowing two drought years concentrations began trations were associated with the presence of decreasing earlier in the summer and decreased leaves and low monthly rainfall totals. at a lower rate. Calcium, which is not retranslo- Variation in the seasonal patterns of nutrient cated from leaf tissues, was at higher concen- depqsition between years was also associated trations in the winter than in the other seasons. with early leaf abscission in the summers and Phosphorus concentrations in the leaf litter de- autumns of 1987 and 1988 (Figs. 1-5), which creased slowly during the fall and winter and were the second and third years of a drought in increased sharply with leaf emergence. Potas- central California. The combination of earlier sium and magnesium concentrations began to leaf fall and higher nutrient concentrations of decrease later in the fall than nitrogen concen- leaf litter in the summer and early autumn than trations, which may explain the exceptionally in later months (Fig. 7) resulted in higher nu- high concentrations of these elements in the trient fluxes via leaves in the autumns of 1987 October 1987 throughfall (Fig. 6) in comparison and 1988 than in 1986. Total leaf fall decreased to the November throughfall collected during the in each of the consecutive years of the study, and first rains of 1986. Nutrient deposition in oak ,woodlands 213

- NITROGEN * TWIGS AND BARK - LEAVES ----. ACOFIJS

I I

- DEWSITION BY PRECIPITATION TOTAL SUBCANOPY DEPOSITION' "10 E - A

1986 1987 1988* Fig. 1. Total nitrogen deposition under Quercus douglasii canopies and in the open grassland. 1 =not including acorns, see Methods for explanation. * = throughfall and precipitation were not measured in 1988. Error bars show two standard errors on each side of means with the largest standard error for each litter fraction, n = 6. 214 Callaway and Nadkarni

TWIGS AND BARK I I PHOSPHORUS - LEAVES THROUGHFAU - - - -0- -- ACOFINS

J JASONDJ FMAMJ JASONDJ FMAMJ JASOND

0.06 ...... + 0.05 LICHEN IMISCELLANEOUS 0.04 . *- FLowE!4s 0.03 0.02 0.01 0.00 JJASONDJFMAMJJASONDJFMAMJJASOND

DEPOSITION BY PRECIPITATION 0.4 I-

Fig. 2. Total phosphorus deposition under Quercus douglasii canopies and in the open grassland. 1 =not including acorns. see Methods for explanation. * = throughfall and precipitation were not measured in 1988. Error bars show two standard errors on each side of means with the largest standard error for each litter fraction, n = 6. Nutrient deposition in oak woodlands 215

POTASSIUM 6 *TWIGS LEAMS

-l --f THROUGHFALL ---*--. laRNs 4

0 JJASONDJFMAMJJASONDJFMAMJJASOND

JJASONDJFMAMJJASONDJFMAMJJASOND

- DEPOSITION BY PRECIPITATION - TOTAL SUBCANOPY DEPOSITION'

JJASONDJFMAMJJASONDJFMAMJJASOND 1986 1987 1988. Fig. 3. Total potassium deposition under Quercus douglasii canopies and in the open grassland. 1 =not including acorns. see Methods for explanation. * = throughfall and precipitation were not measured in 1988. Error bars show two standard errors on each side of means with the largest standard error for each litter fraction. n = 6. 216 Callaway and Nadkarni

CALCIUM --fr--. TWIGS AND BARK - LEAVES

- DEP0SITK)N BY PRECIPITATION

SONDJ FMAMJ JASONDJ FM 1986 1987 1988* Fig. 4. Total calcium deposition under Quercus douglasii canopies and the open grassland. 1 = not including acorns. see Methods for explanation. * = throughfall and precipitation were not measured in 1988. Error bars show two standard errors on each side of means with the largest standard error for each litter fraction. n = 6. Nutrient deposition in oak woodlands 217

MAGNESIUM

* TWIGS AND BARK LEAVES ---c-. THROUGHFALL

JJASONDJFMAMJJASONDJFMAMJJASOND

0.3

MISCULANEOUSS =0 0.2 -

\ 2 0.1

0.0 JJASONDJFMAMJJASONDJFMAMJJASOND

- DEPOSITION BY PRECIPITATION - TOTAL SUBCANOPY DEPOSITION'

TTTTTTIIIIII ITVY~IIITIIITTTYITT. . ... JJASONDJFMAMJJASONDJFMAMJJASOND 1986 1987 1988* Fig. 5. Total magnesium deposition under Quercus douglasii canopies and in the open grassland. 1 = not including acorns. see Methods for explanation. * = throughfall and precipitation were not measured in 1988. Error bars show two standard errors on each side of means with the largest standard error for each litter fraction. n = 6. 218 Callaway and Nadkarni

80 -m 60 PRECIPITATION CY E --+- THROUGHFALL \ 40

0 JJASONDJFMAMJJASONDJFM

: PRECIPITATION 4 - NITROGEN 3 ; THROUGHFALL 2 - . 1 - 0 dbhI 1 I/ JJASONDJFMAMJJASONDJFMAMJJASOND

0.5 - - PHOSPHORUS 0.4 - 0.3 1 0.2 - 0.1 - I' {It 0.0 rl J JASONDJFMAMJ JASONDJFMAMJ JASOND

10 7 POTASSIUM 8 -

-'\ 61

cn 4: E 2; n" J JASONDJ FMAMJ JASONDJFMAMJ JASOND

4 - - CALCIUM 3 - 2 -

1 f " t I JJASONDJFMAMJJASONDJFMAMJJASOND

3 - . MAGNESIUM 2 -

1 -

n" JJASONDJFMAMJJASONDJFMAMJJASOND 1986 1987 1988 Fig. 6. Volume of throughfall and precipitation, and concentration of nutrients in throughfall and precipitation under Quercus douglasii canopies and in the open grassland. Error bars show two standard errors, n = 6. Nutrient deposition in oak woodlandr 219

"."".*"".. NmXXjOU

--+- POTASSIUM 30 I- CALCIUM "-7 it /-'".a*q\ ., f b...a.... ktxr~ 'i i

*\b+4 I I *+ -4k '*\ 0.--w*4-e 4. bud' I I +*\*\ +'cYC..w....

I . . 1- 1- 1111111111111111111IIIII~~III JJASONDJFMAMJJASO'NDJFMAMJJASOND

- Pm3SPm)RUS I - - +- MAGNESIUM

o~lllllllllllllllllllllllllllllll JJASONDJFMAMJJASONDJFMAMJJASOND 1986 1987 1988 Fig. 7. Concentrations of nutrients in leaf litterfall under Quercus douglasii canopies.

Discussion primary reason for the lower total deposition. Although throughfall nutrient concentrations Total nutrient deposition was lower than has were relatively high in this Q. douglasii wood- been reported for woodlands dominated by land, other winter-deciduous oak woodlands winter-deciduous oaks in northern England (Car- generally receive more total rainfall. Additional- lisle et al., 1966a; 1966b), several sites in Bel- ly, precipitation during our collection period was gium (Duvigneaud and Denaeyer-De Smet, below average, thus in other years total deposi- 1971), the eastern Soviet Union (Duvigneaud tion should be higher. In temperate climates, and Denaeyer-De Smet, 1971), and the eastern rainfall is common during the summer months (Monk and Day, 1985), central (Johnson and when leaves are present, and summer peaks of Risser, 1974; Rochow, 1974) and northern Un- throughfall nutrient deposition have been re- ited States (Killingbeck and Wali, 1978; Reiners, ported from other winter-deciduous oak wood- 1972). Total nutrient deposition was also lower lands (Reiners, 1972). In California oak wood- than in evergreen oak woodlands in the lands, a large percentage of total annual precipi- Himalaya (Mehra et al., 1985), and in southern tation falls when the trees are leafless. Thus, France (Rapp, 1969). In general, low throughfall leaves are rarely leached during summer. deposition in the woodland we studied was the The relative quantities of nutrients in litter in 220 Callaway and Nadkarni most winter-deciduous oak woodlands is Ca> of nutrients during mid-summer and mid-winter N > K > Mg > P (Duvigneaud and Denaeyer-De when leaf and throughfall deposition was low. Smet, 1971; Johnson and Risser, 1974; Killin- The importance of non-leaf parts to annual and gbeck and Wali, 19789; Monk and Day, 1985; seasonal fluxes of nutrients was also reported by Ostman and Weaver, 1982; Reiners, 1974; Roch- Carlisle et al. (1966a) who found that large ow, 1974). The relative quantities of nutrients in quantities of nitrogen, phosphorus and potas- the oak woodland studied here, however, were sium fell in flowers, twigs, bark, bud scales and N > Ca > K > Mg > P. The high quantities of ni- insect frass several months prior to leaf fall in a trogen in the litter and throughfall relative to winter-deciduous Q. petraea woodland in calcium (which is not retranslocated) suggests England. that nitrogen use in Q. douglasii woodlands may Nutrient deposition via Q. douglasii canopies be inefficient in comparison to deciduous oak has significant effects on understory soil nutrients woodlands in temperate climates. Leaf abscission and herbaceous productivity and appears to af- in Q. douglasii appears to occur over longer fect species composition. Callaway et al. (in periods of time than for oaks in habitats with press) examined light, temperature, soil mois- severe winters, which may affect the efficiency of ture, soil nutrients, and fine tree root distribu- nitrogen retranslocation. Although the relative tions under Q. douglasii canopies in order to amounts of nutrients may vary for many reasons, determine how variations in these factors were the relatively low amount of calcium in the litter associated with widely varying differences in sub- was probably not due to low availability because canopy herbaceous production. They found that calcium was abundant in the soil at our site soil under all trees sampled was higher in nu- under trees and in the open grassland (see trients than under treeless grassland, and had the Methods). potential to facilitate the growth of herbaceous Throughfall contributed only 3.7% and 1.7% understory plants. Facilitative effects in the field, of the canopy nitrogen deposition in 1986-87 however, were determined by tree root morphol- and 1987-1988, respectively. Low throughfall ni- ogy, with high understory productivity being as- trogen fluxes are common in winter-deciduous sociated with low biomass of fine oak roots, and oak woodlands (Killingbeck and Wali, 1978; low understory productivity being associated Monk and Day, 1985; Reiners, 1972) and net with high biomass of fine oak root. Under oaks canopy uptake of nitrogen was reported by Car- with low understory productivity, root exclosures lisle et al. (1966b) and Ostman and Weaver significantly improved herbaceous growth. (1982). Carlisle et al. (1966b) suggested that leaf Variation in the temporal pattern of nutrient and bark adsorption, microflora assimilation, availability also may affect the productivity and leaf absorption, and uptake by epiphytic mosses species composition of Q. douglasii understories. and lichens may have contributed to the net loss. High deposition in fall and spring corresponds The latter possibility is supported,by the findings with periods of maximal nitrogen uptake in Cali- of Lang et al. (1976) who fodd that several fornia grasslands (Jackson et al., 1988) and may species of lichens removed ammonium and ni- have substantial positive effects on overall pro- trate from experimental solutions, but leaked ductivity. Because most nitrogen deposition potassium, magnesium and calcium into the solu- under Q. douglasii canopies was in leaf litter, tions. Epiphytic lichens were common on all of nitrogen availability is dependent on mineraliza- our sample trees. tion rates, which may not co-occur with max- Non-leaf parts provided approximately 15- imum deposition rates. However, Callaway et al. 35% of the total nutrient budgets, which is com- (in press) found that in autumn, available am- parable to reports from other oak woodlands monium and nitrate were significantly higher (Carlisle et al., 1966a; Gosz et al., 1972; Killin- under Q. douglasii trees (NH, = 12-17 pg g - ', gbeck and Wali, 1978; Monk and Day, 1985). NO, = 21-23 pgg-') than in the open grassland The summer deposition of flowers and the non- (NH, = 7 pg g-L, NO, = 8 pg g-l). Growth seasonal deposition of twig and bark, lichen and phenologies of many California annual species miscellaneous litter provided important sources do not overlap completely (Chiarellio, 1989). Nutrient deposition in oak woodlands 221

Thus, early annual species may have an advan- idana in a sessile oak (Quercuspetraea) woodland. J. Ecol. tage in seasons with high deposition in the fall, 54, 65-85. whereas late annuals may benefit from high Carlisle A, Brown A H F and White E J 1966b The organic matter and nutrient elements in the precipitation beneath a spring deposition. Canopy-mediated soil fertility sessile oak (Quercus petraea) canopy. J. Ecol. 54, 87-98. may also contribute to the displacement of Carlisle A, Brown A H F and White E J 1967 The nutrient Avena fatua L. and other open grassland domin- content of tree stem flow and ground flora litter leachates ants by Bromus diandrus Roth. and other under- in a sessile oak (Quercus petraea) woodland. J. Ecol. 55, story dominants as has been suggested for Q. 615-627. Callaway R M 1990 Effects of Quercus douglasii on Grass- agrifolia by Parker and Muller (1982). land Productivity and Nutrient Cycling in Central Califor- We found that nutrient deposition in Q. doug- nia. Ph.D. Dissertation, University of California, Santa lasii woodlands was much higher under tree Barbara, CA. 1898 p. canopies than in the open grassland, However, Callaway R M, Nadkarni N and Mahall B E 1991 Facilitation most of the nutrient deposition was in organic and interference of Quercus douglasii on understory productivity in central California. Ecology. (In press.) form and not immediately available for plant Chiariello N R 1989 Phenology of California grasslands. In uptake. Additionally, only deposition via pre- Grassland Structure and Function: California Grassland. cipitation was measured for deposition in the Eds. L F Huenneke and H Mooney. pp 47-58. Kluwer open grassland, which underestimates total de- Academic publishers. Dordrecht, The Netherlands. position, and the contribution of the herbaceous Duvigneaud P and Denaeyer-De Smet S 1971 Cycle de elements biogenes dans les ecosystems forestiers d'Europe plants to the overall nutrient cycle (e.g. Carlisle (principalement forets caducifolees). In Productivity of et al., 1967) was not measured. Analysis of Forest Ecosystems. Ed. P. Duvigneaud. pp 527-542. UN- decomposition rates and non-plant sinks and los- ESCO, Paris. ses (e.g. Jackson et al., 1988) is required to fully Gosz J R, Likens G E and F H Borman 1972 Nutrient understand the effects of the fluxes described content of litterfall on the Hubbard Brook experimental forest, New Hampshire. Ecology 53, 769-784. here. Griffin J R and Critchfield W B 1976 The Distribution of Forest Trees in California. USDA Forest Research Paper, PSW-8211972, Berkeley, CA. Holland V L 1973 A Study of Vegetation and Soils under Acknowledgements Blue Oak Compared to Adjacent Open Grassland. Ph.D. Dissertation, Berkeley, CA. 390 p. We would like to thank Dr J R Griffin for advice Holland V L 1980 Effect of blue oak on rangeland forage and for support while at the Hastings Natural production in central California. In Ecology, Management and Utilization of California Oaks. Ed. T R Plumb. pp History Reservation. We also thank Dr Julia 314-318. USDA General Technical Report PSW-44, Ber- Jones and Frank Setaro for help with the nu- keley, CA. trient analyses, and Steve Gowler and Cathy Holland V L and Morton J 1980 Effect of blue oak on Callaway for field assistance. We thank Lisa nutritional quality of rangeland forage in central Califor- Thwing for computer assistance. Facilities for nia. In Ecology, Management and Utilization of California Oaks. Ed. T R Plumb. pp 319-322. USDA General Tech- nutrient analyses were supported financially by nical Report PSW-44, Berkeley, CA. the University of California Santa Barbara Aca- Jackson L E, Strauss R B, Firestone M K and Bartolome J W demic Senate and research grants from the Na- 1988 Plant and soil dynamics in California annual grass- tional Sciences Foundation (BSR-86-14935) and land. Plant and Soil 110, 9-17. the Whitehall Foundation. Jansen H C 1987 The effect of blue oak removal on herbaceous production on a foothill site in the northern Sierra Nevada. In Multiple Use Management of California's Oak Hardwood Resources. Eds. T R Plumb and H R Pillsbury. References pp 343-350. USDA General Technical Report PSW-100, Berkeley, CA. Johnson F L and Risser P G 1974 Biomass, net primary Borchert M, Davis F W, Michaelson J and Oyler L D 1989 production, and dynamics of six mineral elements in a post Interaction of factors affecting seedling recruitment of blue oak-blackjack oak forest. Ecology 55, 1246-1258. oak (Quercus douglasii) in California. Ecology 70, 389- Kay B L 1987 Long-term effect of blue oak removal on 404. forage production, forage quality, soil. and oak regenera- Carlisle A, Brown A H F and White E J 1966a Litterfall, leaf tion. In Multiple Use Management of California's Hard- production, and the effects of defoliation by Tortrix vir- wood Resources. Eds. T R Plumb and N H Pillsbury. pp 222 Nutrient deposition in oak woodlands

351-357. USDA General Technical Report PSW-100, Ber- transfers by retranslocation, leaching and litterfall in a keley, CA. chestnut oak forest in southern Illinois. Can. J. For. Res. Killingbeck K T and Wali M K 1978 Analysis of a North 12, 40-51. Dakota gallery forest: Nutrient, trace element and produc- Parker V T and Muller C H 1982 Vegetational and en- tivity relations. Oikos 30, 29-60. vironmental changes beneath isolated live oak trees (Quer- Lang G E, Reiners W A and Heler R K 1976 Potential cus agrifolia) in a California annual grassland. Am. Mid. alteration of precipitation chemistry by epiphytic lichens. Nat. 107, 69-81. Oecologia 25, 229-241. Rapp M 1969 Production de litiere et apport au sol d'ele- Mehra M S, Pathak P C and Singh J S 1985 Nutrient ments mineraux dans deux ecosystemes Mediterraneens: movement in litterfall and precipitation components for La foret de Quercus ilex L. et la garrigue de Quercus central Himalayan forests. Ann. Bot. 55, 153-170. coccifera L. Oecologia Plant. 4, 377-410. Monk C D and Day F P 1985 Vegetation analysis, primary Reiners W A 1972 Nutrient content of canopy throughfall in production and selected nutrient budgets for a southern three Minnesota forests. Oikos 23, 14-22. Appalachian oak forest: A synthesis of IBP studies at Rochow J 1975 Mineral nutrient pools and cycling in a Coweeta. For. Ecol. Manag. 10, 87-113. Missouri forest. J. Ecol. 63, 985-994. Murphy A H and Crampton B 1964 Quality and yield of Setaro F and Jones J A 1989 Methods of Nutrient Analysis forage as affected by chemical removal of blue oak (Quer- for Ecological Studies. University of California, Santa cus douglasii). J. Range Manag. 17, 142-144. Barbara, CA. 78 p. Ostman N L and Weaver G T 1982 Autumnal nutrient