| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Soil and Water Sciences Program, Dep. of Environmental Sciences, Univ. of California, Riverside, CA 92521-0424
b Dep. of Biological Sciences, Humboldt State Univ., Arcata, CA, 95521
* Corresponding author (robert.graham{at}ucr.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: AEC, anion exchange capacity • CEC, cation exchange capacity
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Most information regarding the function of arboreal soils in a canopy ecosystem is from studies done in Central and South American rain forests. These soils are important in the tropical canopy environment for several reasons. Primarily, arboreal soils can hold large amounts of water (Veneklaas et al., 1990; Bohlman et al., 1995) and can provide a more continuous moisture supply for epiphytes than the atmosphere or tree surfaces lacking soil accumulations (Benzing, 1990). Second, arboreal soils have cation exchange capacities (CECs) similar to those of the organic soil horizons on the terrestrial forest floor, which allows for a high potential for nutrient retention within the canopy (Nadkarni et al., 2002). Thus, these soils play an important role in plant and microbial nutrition. Furthermore, arboreal soils contain a diverse community of invertebrates (Nadkarni and Longino, 1990) and a microbial biomass that is comparable to that found within the organic horizons of the forest floor (Vance and Nadkarni, 1990).
Arboreal soils have formed within the crowns of large redwoods in the temperate rainforests of northern California. These soils are similar to the tropical arboreal soils in that they create forest floor conditions within the tree crown. Massive limbs (up to 2.8 m in diameter) and tree crotches provide platforms for soil accumulation (Sillett, 1999; Sillett and Bailey, 2003). Shrubs, ferns, and other vascular epiphytes grow in these soils (Sillett, 1999). An evergreen fern, Polypodium scouleri, is the most abundant vascular epiphyte in old-growth redwood forest canopies (Sillett, 1999). The dry mass of P. scouleri mats, including fern biomass and soil materials, can weigh up to 742 kg on individual trees (Sillett and Bailey, 2003). The largest and most complex redwood forest canopies contain enormous quantities of arboreal soil (up to 79 m3 or 2400 kg dry mass ha–1) and epiphytic vascular plants (up to 500 kg dry mass ha–1) (Sillett, unpublished data, 2005). Epiphytic mats and their associated soils provide habitat to a plethora of normally terrestrial animals, including arthropods, salamanders, and mollusks (Sillett, 1999; Sillett and Bailey, 2003). These animals, along with other resident epiphytic plants, may depend on water stored in the arboreal soils. The properties of the arboreal soils, how they form and how they influence the canopy environment, remain largely unexplored. In this study, we examined two arboreal soils, using pedological approaches, to assess their soil-like properties and to determine how they function as soils and how they formed. More specifically, the objectives of this research were (i) to gain insight into the water retention and nutrient content characteristics of arboreal soils in two large redwoods, and (ii) to interpret their genesis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Redwood trees dominate the forest at the study sites. Other tree species include big leaf maple (Acer macrophyllum), Douglas fir (Pseudotsuga menziesii), western hemlock (Tsuga heterophylla), and California bay (Umbellularia californica). Understory vegetation is primarily sword fern (Polystichum munitum), evergreen huckleberry (Vaccinium ovatum), and redwood sorrel (Oxalis oregana) (Sillett and Bailey, 2003). Epiphytes are abundant in the forest canopy of both sites. The leather-leaf fern, P. scouleri, is the most abundant species, but shrubs and a few other vascular plants are also common as epiphytes (Sillett, 1999). At the Jedediah Smith RSP site, the entire arboreal soil surface is covered by P. scouleri. An elderberry (Sambucus callicarpa) is also growing in the soil on this limb. The vegetation cover at the Prairie Creek RSP site is 65%, with P. scouleri covering 20% and V. ovatum (one individual) covering 45% of the arboreal soil surface. A white-flowered currant (Ribes laxiflorum) is growing near the base of the tree crotch, close to where the soil was sampled.
Field Methods
Samples from the arboreal soils were collected in September 2001, before the start of the rainy season. Both sites received no significant rain during the preceding 12 wk. Tree crowns were accessed by using standard rope techniques. In short, mechanical ascenders were used to climb a rope that was secured near the top of the tree. A handsaw was used to cut through the organic soil material, exposing a fresh cross-section along the edge of the pedon. Loose surface litter was collected before removing an intact soil column from the tree. Live fern fronds were also collected before the soil was sampled at the Jedediah Smith RSP site. An intact soil column (horizontal dimensions of approximately 45 by 30 cm) was cut out of the soil body and brought to the forest floor. A morphological description was then made of the profile, including the surface litter (Soil Survey Staff, 1999). Two to four intact core samples (54 mm in diameter by 60 mm long) were taken for each intact horizon described in the field. All of the core samples were vertically oriented within the soil profile. The remaining soil material was separated by morphologic horizon and stored at field moisture and –19°C. For the soil from the Jedediah Smith RSP site, two of the thicker horizons (Oe2 and Oa1, Table 1) were further subdivided for chemical analyses. The pedon sampled from Prairie Creek RSP is referred to as Pedon 1 and the one from Jedediah Smith RSP is Pedon 2.
|
Each morphological horizon was analyzed for fiber content (unrubbed and rubbed) and pyrophosphate color (Lynn et al., 1974; Soil Survey Laboratory Staff, 1996). An initial volume of 2.5 cm3 of moist soil material was prepared by packing it into a longitudinally halved syringe. This material was then placed in a saturated solution of sodium pyrophosphate for 24 h. A piece of chromatographic paper was placed in this solution for 3 to 5 min; the paper was compared with a Munsell chart to determine the sodium pyrophosphate extract color (SPEC). For the fiber content analyses, 2.5 cm3 of moist soil material was gently washed on a 100-mesh sieve to remove sapric materials. The sample was then repacked into the half-syringe to record its unrubbed fiber volume. The sample was returned to the sieve, rubbed between fingers under a stream of water, and returned to the syringe to determine rubbed fiber volume. Rubbed and unrubbed fiber content data were represented as a percentage of the initial 2.5 cm3 starting volume.
Fiber content and SPEC are used to distinguish fibric, hemic, and sapric materials as well as subordinate horizon labels (Soil Survey Staff, 1999). Fibric materials have 40% (v/v) or more fiber after rubbing or >75% (v/v) unrubbed fiber and a SPEC (value/chroma) of 7/1, 7/2, 8/1, 8/2, or 8/3. Sapric materials have <16% (v/v) rubbed fiber. For the samples reported in this study, the SPEC of sapric materials is 7/4 or 7/6. Hemic materials are determined as having an intermediate fiber content between fibric and sapric materials.
Field-moist soil samples collected from each horizon were physically fractionated through 6.3- and 2-mm sieves. Roots, alive and dead, and live P. scouleri rhizomes were removed from the >6.3-mm soil fractions. The <2- and 2- to 6.3-mm soil fractions contain a small amount of very fine root fragments. The >6.3-mm soil fraction was further separated into recognizable litter components (i.e., redwood leaves and bark). Each fraction as well as roots and live fern rhizomes were weighed to obtain a field-moist mass and a subsample was oven dried at 60°C to obtain moisture content. Fractions were then expressed as a percentage of the entire horizon dry mass with and without the presence of roots and live rhizomes. Loss on ignition (LOI) was calculated for all fractions by mass difference after placing the dried samples in a muffle furnace at 550°C for 20 h. The samples, after heating at 60 and 550°C, were examined for the presence of mineral materials under a dissecting microscope. Soil color was obtained on air-dry and moist samples of the <2-mm soil fraction using a Minolta CR-200 Chroma Meter (Minolta Corp., Ramsey, NJ).
The <2-mm fractions stored at field moisture were used to determine soil pH, CECs, and anion exchange capacities (AECs), and soil saturation paste chemistry. Soil pH was measured in water and 0.01 M CaCl2 by adding 20 mL of water or CaCl2 solution to 2.5 cm3 of moist soil material (Soil Survey Laboratory Staff, 1996). Cation exchange capacity and AEC were determined at pH 4 and pH 7 by saturating the soil with LiCl, rinsing the excess solution with deionized water, and then replacing the adsorbed Li+ and Cl– with K+ and NO3–. Exchangeable cations (Ca, Mg, K, Na, and NH4) were collected during the first extraction with LiCl. Cation exchange capacity was calculated from the displaced Li+ and AEC from the displaced Cl–. Cations were measured using inductively coupled plasma atomic emission spectroscopy (ICP–AES). Chloride and exchangeable ammonium were analyzed colorimetrically using a continuous flow analyzer. Exchangeable acidity (exchangeable Al3+ + H+) was determined by the KCl method (Thomas, 1982). Effective CEC was calculated as the sum of the exchangeable cations and exchangeable acidity (Bohn et al., 1985). Bulk density and the mass percentage of the <2-mm soil fraction were used to convert CEC measurement on a mass basis (cmolc kg–1) to a volume basis (cmolc L–1).
Aqueous soil extracts were obtained by using the saturation paste extract method (Rhoades, 1982) and were measured for electrical conductivity (EC), pH, and major solutes. Solutions were analyzed for metal cations using ICP–AES and for anions (Cl, NO3, SO4, PO4) using a Dionex 500 with an ASII HC column. Soluble ammonium was determined colorimetrically using a continuous flow analyzer. Dissolved organic C was measured on a Shimadzu TOC-V. Saturation extracts were not obtained for the surface horizons of either profile because a sufficient mass of soil material was not available.
The elemental composition of the <2-mm soil fraction from each horizon was determined, as was the composition of the redwood plant parts that serve as parent material. Redwood leaves and bark were chosen from the >6.3-mm soil fraction from the surface Oi horizons to represent the most recently deposited litter. Subsamples of redwood plant materials and <2-mm soil fractions were dried at 60°C and then finely ground. Total C and N on these samples were measured using a Europa Scientific ANCA G/S/L Preparation Module coupled to a Europa Scientific Tracer/20 Mass Spectrometer (PDZ Europa, Northwich, England). Concentrated nitric and hydrochloric acids were added to the soil samples and the redwood leaf and bark samples. These samples were then digested using a microwave-accelerated reaction system (MARS 5, CEM) (Milward and Kluckner, 1989). Plant and soil digests were analyzed for Ca, K, Mg, and Na by an ICP–AES.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Roots in both pedons include a mix of live and dead components. Live roots extend throughout Pedon 1, but only in the top 45 cm of Pedon 2. A cohesive network of P. scouleri roots and rhizomes dominate the soil structure of Pedon 2 (Table 1). Fern rhizomes (5–20 mm in diam.) occur from 5 to 36 cm. Pedon 1 has granular or subangular blocky structure from 25 to 70 cm. Roots did not form a cohesive network, but are most abundant in the deepest horizon. Some of these roots may be from R. laxiflorum, which is growing out of the side of the pedon near the base of the tree crotch. Over the entire soil depth, Pedon 2 contains a larger dry mass of roots per dry mass of soil than Pedon 1 (Table 2).
|
The soil size fractions of Pedons 1 and 2 have similar trends with depth (Fig. 2a and 2b). The <2-mm fraction increases with increasing depth, while the two coarse fractions generally decrease. The input of materials from the trunk interface may be significant in Pedon 1 since this soil profile has an increase in bark mass, as represented by the 2- to 6.3-mm fraction, within the deepest horizon. In general, changes in the proportions of the three size fractions with depth are related to soil morphology: Oi horizons have more >2-mm material whereas the Oe and Oa horizons are mainly composed of <2-mm soil material.
|
|
|
|
|
0.05 mmolc L–1 (data not shown). The sum of cations is higher than the sum of anions for the saturation paste extracts. The remaining negative charges are from organic acids. Dissolved organics, as measured by the organic C content of the saturation extracts, have higher concentrations in the near-surface horizons of both soils in comparison with the deeper horizons (Table 6). Anion exchange capacity and CEC were measured at pH 7 and at pH 4 (Table 7). The AEC values at pH 4 are around 1.2 to 2 times larger than those at pH 7 for both soils. Pedon 2 has higher AEC values than Pedon 1, a difference that is most prominent near the soil surface. The CEC values at pH 7 are similar to or larger than those at pH 4. Overall, CEC values increase with depth and, conversely, AEC values decrease with increasing depth for both profiles at the two pH values. In particular, at pH 4 the CEC values of Pedon 1 increase and AEC values decrease with increasing depth until near the bottom of the profile, where the trend is reversed. The CEC values at pH 4 are also represented on a volume basis, since roots inhabit a volume of soil. Reported CEC values (by mass and by volume) are based on the assumption that most of the charge comes from the <2-mm soil fraction. These values may be somewhat underestimated since the 2- to 6.3-mm fraction may have some exchange capacity.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Decomposition state influences soil bulk density (Boelter, 1969; Nichols and Boelter, 1984), water retention (Boelter, 1964, 1969), and CEC (Lévesque et al., 1980). As plant materials decompose in Histosols, the organic fibers decrease in size and in amount, yielding smaller pores and higher bulk densities (Boelter, 1969; Nichols and Boelter, 1984). Bulk density values in both profiles increase with increases in decomposition state (Oi and Oe horizons vs. Oa horizons, Table 3). The arboreal soils have bulk densities that are at the middle to lower end of the range typically found for Histosols (0.02– 0.30 g cm–3) (Rabenhorst and Swanson, 2000). More decomposed terrestrial Histosols can have a substantial mineral content, which may account for their higher range of bulk densities compared with the soils in the redwood canopy. Bulk densities and inorganic ash contents of the arboreal soils are similar to those found in a study of 47 young-growth redwood forest floor samples. Forest floor bulk density values ranged from 0.04 to 0.14 g cm–3, with mean bulk densities of 0.06 g cm–3 at the surface increasing to 0.10 g cm–3 at the 10-cm depth (Finney and Martin, 1993). In addition, the extensive framework of roots in Pedon 2 may influence soil bulk density more than the overall soil decomposition state. In particular, this may explain the lower bulk density value in the sapric horizons of Pedon 2 compared with Pedon 1.
The soil materials in the sapric horizons contain more water (by volume) at a given pressure than fibric or hemic materials (Table 3). More decomposed organic materials have a larger proportion of smaller pores and, therefore, hold more water at a given matric pressure than less decomposed materials (Boelter, 1969; Sharratt, 1997).
For both profiles, fibric surface horizons have the lowest CEC values, while hemic and sapric subsurface horizons have the highest values (Table 7). Cation exchange capacity of organic materials has been observed to increase with increasing degree of decomposition in organic soils (Lévesque et al., 1980) and in the forest floor (Wells and Davey, 1966; Van Cleve and Noonan, 1971). As the soils within the redwood canopy become more decomposed, it appears that their ability to retain plant available nutrients, as well as water, increases.
Water Relations
Water scarcity is a major abiotic limitation in the epiphytic habitat (Zotz and Hietz, 2001). Epiphytes depend on atmospheric sources for water and can become stressed when inputs are low. Hence, storage in soils must supply the epiphytes in the redwood tree crown with water between rain events. Soil moisture release curves indicate that the soil, regardless of decomposition state, contains large pores that drain freely at high potentials, from 0 to –5 kPa (Table 3). The soils also retain water at more negative potentials, and during dry periods, epiphytes must rely on this more strongly bound water held in the smaller pores. While –1500 kPa is commonly used as the "permanent wilting point," it is not appropriate for epiphytes that have evolved in a drought prevalent environment. The tropical fern Polypodium crassifolium can have leaf water potentials as low as –2500 kPa (G. Zotz, personal communication, 2003). Although no data are available, P. scouleri may reach similarly low potentials under drought conditions and thereby access more stored water.
The soils from both sites receive most of their yearly precipitation from October through April. This high amount of water input, coupled with shorter days and cooler air, implies that the epiphytes are not water stressed during this time period. In contrast, the summer season for the northern redwood region is characterized by low precipitation. Fog does not contribute appreciable water to the canopy at either study site (Sillett, unpublished data, 1996–2005). During the summer season, the main reserve of water for epiphytes is stored in the soils.
The field moisture measurements in Table 4 were taken in September when the soils are near their driest. Pedon 1 contains plant available water at the end of the dry season, with soil materials having 31 to 35% volumetric water at potentials ranging from –17 to –14 kPa. Since live fern roots are located throughout Pedon 1, the entire profile may have provided water to P. scouleri. The upper part (5–54 cm) of Pedon 2 was dry (–6732 to –1037 kPa) and offered very little plant-available water to P. scouleri, whereas the deeper Oa1 and Oa2 horizons had higher volumetric water contents at higher matric potentials. However, P. scouleri most likely did not directly access this deeper water, since live roots were observed only in the top 45 cm of Pedon 2. At the time of sampling, the P. scouleri growing in Pedon 2 were not wilting and had water storing rhizomes that contained 77 to 83% water by field-moist weight (data not shown). Rhizomes, as well as thick leaf cuticles that minimize water loss, contribute to the desiccation resistance of epiphytic ferns (Hietz and Briones, 1998). P. scouleri may depend on water stored in the soil earlier in the dry season and water stored in the rhizomes as the dry season progresses.
Water stored in arboreal soils is used by other canopy dwelling organisms, such as mollusks, invertebrates, and salamanders (Sillett 1999; Sillett and Bailey, 2003). Soil moisture contents of 20 to 30% during the dry season may produce a sufficiently high humidity to maintain a moist habitat for desiccation-sensitive animals.
The two soils had different field moisture potentials at the end of the dry season as a result of differences in their site topography and microclimate. Pedon 1 is in a bowl-shaped depression formed by the tree crotch, a situation that promotes the retention of water. Pedon 2 is on an open limb from which water can drain freely. Furthermore, the surface of Pedon 2 receives more photosynthetically active radiation compared with the surface of Pedon 1 (S. Sillett, unpublished data, 2003), which is in a shaded location within the interior of the tree, surrounded by Atlas' huge crown (Sillett, 1999). A potential result of this difference in microclimate between the two pedons was a lower plant cover in Pedon 1 and less water loss through evapotranspiration.
Dry soil conditions in the summer may not only limit plant-available water in the redwood canopy but also may impede microbial activity and slow organic matter (OM) decomposition (Wagner and Wolf, 1999). Since the Oi and Oe horizons of Pedon 2 were dry (< –1037 kPa) at the end of the dry season, OM decay may be more limited in these horizons compared with the rest of the profile. Bohlman et al. (1995) recorded periods of low soil moisture in arboreal soils in Costa Rica during the dry season over a 4-yr period. They suggested that this moisture pattern might affect the density and composition of the decomposer community. Forest floor litter decomposition studies in old-growth redwood stands in northern California have found that decomposition rates increased with increasing moisture (Pillers and Stuart, 1993).
Nutrient Availability
Cation exchange capacity is an important soil property since the exchange sites retain plant nutrient cations from sources such as OM decomposition and atmospheric deposition. The arboreal soils in this study have CEC values (Table 7) similar to those of acid forest humus layers in North America (Kalisz and Stone, 1980) and to arboreal soils in Costa Rica (Nadkarni et al., 2002). In general, the CEC values of organic soils on a mass basis are larger than those of mineral soils. However, roots experience a volume of soil not a mass, therefore, CEC on a volume basis is more appropriate for evaluating the nutrient holding-capacity of an organic soil (Rabenhorst and Swanson, 2000). Due to the low bulk density of arboreal soils, their CECs on a volume basis (Table 7) are similar to or lower than those typically found for mineral soils (4.2– 24 cmolc L–1, assuming a soil bulk density of 1.2–1.3 kg L–1 to convert a large mass-based soil CEC database by Holmgren et al., 1993).
The arboreal soils in this study are extremely acidic (pHCaCl2 < 4.2), however the exchangeable soil acidity (a measure of exchangeable H+ and Al3+) is lower than the sum of exchangeable base cations (Table 7). The base saturation of these arboreal soils is high, ranging from 55 to 76%. High base saturation percentages (>50%) have been found for acidic forest floors (pH 4–6) dominated by deciduous (Van Cleve and Noonan, 1971) and coniferous (Youngberg, 1966) trees.
The high base saturation of the two redwood arboreal soils indicates that they are able to buffer losses in base cations, especially Ca and Mg, from the soil solution. The relative concentrations of cations in the soil solution (e.g., Ca > Mg > K) mirror their relative concentrations on the exchange (Table 6, 7). The low concentrations of K compared with Ca and Mg may be due to losses from leaching, especially since this monovalent cation does not compete as well as divalent Ca and Mg for sites on the soil exchange. These organic soils also have an AEC, which may originate from the protonation of amine groups at the low soil pH. This soil characteristic may be important in retaining nitrate and phosphate anions within the canopy environment.
Sources of Nutrients
Nutrients may reach the canopy ecosystem from autochthonous sources, such as the decomposition of epiphytic biomass and host tree litter and from the leaching of live foliage by rain, or from allochthonous sources, primarily via atmospheric dry and wet deposition and N fixation (Nadkarni and Matelson, 1991). While we lack appropriate data to quantitatively distinguish between the different nutrient sources, interpretations can be made in light of relevant literature.
The C/N ratio of forest floor materials is a common parameter used for predicting N mineralization. Laboratory incubations of conifer litter from coastal British Columbia, Canada, have shown that significant N mineralization occurs below a C/N ratio of 35 (Prescott et al., 2000a). Using this C/N ratio as a guideline, immobilization of N may dominate the fibric surface horizons in Pedon 1 and the Oi1 horizon in Pedon 2, since both the <2-mm soil fraction and the >2-mm soil fractions (composed primarily of redwood litter) have a C/N above 35 (Table 5). Net N mineralization may occur within several of the subsurface horizons (i.e., Oa1 and Oa2 of Pedon 1 and Oe2 of Pedon 2) where the <2-mm soil materials, which compose 71 to 93% of the horizon dry mass, have a C/N ratio below 35. Since several of the subsurface soil horizons containing live roots have C/N ratios that promote N mineralization, litter decomposition may be a source of N to epiphytes growing in the redwood arboreal soils. Mineralization, however, does not always occur once the soil material reaches the critical C/N ratio because other factors, such as low soil moisture, may limit decomposition.
Precipitation and throughfall constitute an important source of plant-available nutrients into forest ecosystems. Biogeochemical cycling of Ca, Mg, K, and inorganic N has been studied in a Sitka spruce and western hemlock-Douglas fir ecosystem on the coast of southern Oregon (Bockheim and Langley-Turnbaugh, 1997), 100 to 170 km north of our study area. That study estimated that atmospheric deposition could supply up to 28% of the N and up to 100% of the K, Ca, and Mg required for the conifers' net primary production (Bockheim and Langley-Turnbaugh, 1997). In particular, an important source of nutrients to the forest floor is from throughfall. As precipitation moves through the forest canopy as throughfall, it leaches nutrients from leaf surfaces as well as washes off particles that have collected on the leaf surface (Edmonds et al., 1995). Studies of precipitation and throughfall chemistry in coastal forest ecosystems in Oregon and Washington found that throughfall was enriched in cations compared with precipitation (Edmonds et al., 1995; Bockheim and Langley-Turnbaugh, 1997). This nutrient-rich water from throughfall may partially account for the high base saturation of the arboreal soils.
Are Arboreal Soils Truly Soils?
From a pedological perspective, soil can be defined as a natural body composed of solids, liquids, and gases that occurs on the land surface, occupies space, and is characterized by horizons that are distinguished from the parent material by additions, losses, transfers, and transformations (Soil Survey Staff, 1999). Definitions of soil often refer to the development of pedogenic properties as a result of the integrated effects of environmental factors over time. Soils also serve as a medium for plant growth (Soil Survey Staff, 1999; Soil Science Society of America, 2004). The arboreal soils in the redwood canopy meet all of these criteria except they do not occur on the land surface in a strict sense. In a broader sense, they are a surficial feature, just as the vegetation itself is an earth surface feature. Soils have been described on isolated mesas, ancient buildings (Jenny, 1941), and the tops of giant boulders (Allen, 2005). In all of these cases, pedogenic processes have operated to produce soil features in parent materials that are elevated above the general landscape. We contend that those examples and the arboreal soils described here should be considered soils, in as much as they have the morphology of soils and they behave as soils within their ecosystems.
Genesis
Arboreal soils within old-growth redwood rainforests are entirely organic because the parent materials originate from redwood and fern biomass, with no measurable input from mineral dust. Soil formation was initiated with the accumulation of redwood litter, which then created a substrate for vascular epiphytes to colonize (Sillett, 1999). Inputs of organic parent materials continue to occur throughout the profile: at the surface from epiphyte biomass and intercepted redwood litter, at all depths from root turnover, and at the soil-limb interface from redwood bark (Table 1, Fig. 2). Arboreal soils of a tropical montane rainforest in Costa Rica are also highly organic (Nadkarni et al., 2002) and are derived primarily from epiphytic bryophytes (Clark et al., 1998). The soils within the crowns of large redwoods are heavily influenced by additions from epiphytic biomass, especially from fern roots. However, redwood materials are also significant as indicated by the thick accumulations of redwood litter at the soil surface and the presence of redwood bark throughout the profile.
Site location within the tree crown influences the parent material composition of the two pedons. The most recently deposited litter (Oi horizons) of Pedon 1 contain more redwood bark than redwood leaves and the opposite is observed for Pedon 2. Pedon 1 is located in a tree crotch; hence the soil material accumulates in a bowl formed by the surrounding trunks. In contrast, Pedon 2 is on a massive limb, which intercepts more leaves than bark. Furthermore, Pedon 2 has a higher plant cover and epiphyte root mass per mass of soil than Pedon 1 (Table 2).
Tree and limb ages for Atlas and Fangorn have not been measured, therefore the ages of the soils are also unknown. Atlas and Fangorn are large redwoods (652 and 580 m3 main trunk volume; Sillett and Bailey, 2003) and could be 1000- to over 2000-yr old (Sawyer et al., 1999). In order for stable platforms (limbs, crotches) to be large enough for litter accumulation, a redwood must already be large, so the age of the arboreal soils is considerably less than the age of the host tree. Nevertheless, the arboreal soils in this study are likely several centuries old.
Three major factors affecting decomposition and accumulation of soil organic layers are organisms involved in decomposition processes, parent material characteristics, and climate (Prescott et al., 2000b). Little research exists on litter decomposition within old-growth redwood forest floors and the organisms involved in this process. Soil invertebrates are present in old-growth redwood forest floors (Lattin, 1993; Hoekstra et al., 1995) and the forest canopy (Sillett and Bailey, 2003). In this study, two types of macrofauna, millipedes and centipedes, were observed within the soil samples. Studies of litter decomposition in coastal forests of British Columbia indicate that soil invertebrates, particularly millipedes, increase the rate of litter breakdown (Cárcamo et al., 2000, 2001). Millipedes fragment leaf litter, increasing the surface area available for microbial colonization (Anderson and Bignell, 1980). Invertebrates, such as millipedes, may initiate litter decomposition processes within the arboreal soils.
Since the arboreal soils contain entirely organic parent materials, the soils are extremely acidic. Fungi are the main decomposers of acidic conifer litter under oxic conditions (Millar, 1974; Donnelly et al., 1990) since they are tolerant of low pH (Tate, 1991; Brady and Weil, 1999). Although the composition of bacteria and fungi communities has not been studied in the arboreal soils, soil OM accumulation may be promoted if microbial diversity is insufficient for complete decomposition.
Climate affects soil formation by influencing OM translocation and decomposition processes. The climate of the redwood region promotes the loss of OM by leaching in the winter. Low soil moisture contents in the surface horizons of Pedon 2 may also limit decomposition processes in the summer. Jedediah Smith RSP receives approximately 30% more annual precipitation than Prairie Creek RSP, but differences in site topography (crotch verses limb) and location within the tree crown appear to affect soil moisture more than variations in annual precipitation.
The arboreal soils in this study are deeper than other soils elsewhere within the redwood canopy. One hundred and eighty-nine fern mats in 13 large redwood trees within Jedediah Smith and Prairie Creek Redwood State Parks have depths ranging from 1.0 to 90.0 cm (mean = 18.5 cm), with most >13 cm (Sillett, unpublished data, 2003). The variation in soil depth within the tree crown is largely due to the effect of site topography and tree limb diameter on soil accumulation. In a survey of 765 fern mats on 32 redwood and Sitka spruce trees, the largest mats are on thick limbs and in tree crotches (Sillett and Bailey, 2003). In general, tree crotches are better sites for soil accumulation than limbs, since limbs tend to shed materials off their sides. Thick limbs have broader platforms for soil to accumulate than thinner limbs. The topography of both pedons in this study favors thick accumulations of OM.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Received for publication July 6, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. L. Vaughan and P. A. McDaniel Organic Soils on Basaltic Lava Flows in a Cool, Arid Environment Soil Sci. Soc. Am. J., July 14, 2009; 73(5): 1510 - 1518. [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 | |||
