See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/228409652 Soil Chronosequences, Soil Development, and Soil Evolution: A Critical Review Article in Catena · June 1998 DOI: 10.1016/S0341-8162(98)00053-8 CITATIONS READS 222 1,367 1 author: Richard Huggett The University of Manchester 59 PUBLICATIONS 859 CITATIONS SEE PROFILE All content following this page was uploaded by Richard Huggett on 06 December 2016. The user has requested enhancement of the downloaded file.  Catena 32 Ž1998. 155–172 Soil chronosequences, soil development, and soil evolution: a critical review ) R.J. Huggett School of Geography, UniÕersity of Manchester, Oxford Road, Manchester M13 9PL, UK Received 26 May 1997; revised 16 April 1998; accepted 16 April 1998 Abstract Soils chronosequences are valuable tools for investigating rates and directions of soil and landscape evolution. Post-incisive chronosequences are the most common type of chronosequence. They are found in many landscapes, including sand dunes, glacial moraines, landslide scars, old pasture, burnt landscape patches, old mining areas, lava flows, alluvial fans, floodplains, river terraces, and marine terraces. They register pedogenic change over time-scales ranging from years to millions of years. Soil chronosequences help in testing rival theories of pedogenesis. Traditional soil formation theory sees a soil developing progressively under the influence of the environmen- tal state factors until it is in equilibrium with prevailing environmental conditions. This develop- mental view of pedogenesis is supported by the classic soil chronosequence studies. A new evolutionary view of pedogenesis, which was prompted by the omnipresent inconstancy of environmental conditions and the notions of multidirectional changes and multiple steady states Žas predicted by non-linear dynamics., proposes that environmental inconstancy and non-linear behaviour in soil-landscapes lead to soil evolution, rather than to soil development. Soils ‘evolve’ through continual creation and destruction at all scales, and may progress, stay the same, or retrogress, depending on the environmental circumstances. Some recent soil and vegetation chronosequence investigations support an evolutionary view of pedogenesis. It is concluded that soil chronosequences are still potent instruments for pedological investigations and that they have a starring role to play in the testing of pedological theories. q 1998 Elsevier Science B.V. All rights reserved. Keywords: Chronosequence; Pedogenesis; Soil chronosequence; Soil development; Soil evolution ) Corresponding author. Tel.: q44 161 275 3620; fax: q44 161 275 7878; e-mail: r.huggett@man.ac.uk 0341-8162r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 0 3 4 1 - 8 1 6 2 Ž 9 8 . 0 0 0 5 3 - 8 156 R.J. Huggettr Catena 32 (1998) 155–172 1. Introduction Soil chronosequences are genetically related suites of soils evolved under similar conditions of vegetation, topography, and climate ŽHarden, 1982.. They translate spatial differences between soils into temporal differences. The translation hints at the way that soils and soil properties in given regions change through time Žcf. Vincent et al., 1994.. Soil chronosequences were once relatively rare, but the number of well-documented and reliably dated examples is mounting very fast. Most of them show an exponential Žor sometimes linear. increase or decrease in soil properties with time, as would be expected from the laws of chemical kinetics. Soil chronosequences are excellent indicators of the rate and direction of pedogenic change, and they provide invaluable information for testing theories of pedogenesis. Also, they are central to much soil–geomorphic research ŽBirkeland, 1990, 1992., and they may be used Žwith caution. to date a range of Quaternary landforms, including stream terraces. Several excellent reviews of soil chronosequences are available ŽJenny, 1941, 1980; Burges, 1960; Rode, 1961; Walker, 1966; Stevens and Walker, 1970; Vreeken, 1975; Yaalon, 1975; Bockheim, 1980; Birkeland, 1990.. The present review focuses on three things. First, it briefly considers the construction and classification of soil chronose- quences. Second, it describes studies of historical, Holocene, and Quaternary soil chronosequences, mainly to illustrate the wide range of settings in which chronose- quences are now investigated. Third, it considers two rival theories of pedogenesis—the developmental and the evolutionary—and briefly explores the evidence for them in chronosequence investigations. 2. Constructing and classifying soil chronosequences One of the most significant discoveries made in pedology was that soil profiles record their own history ŽYaalon, 1983, p. 233.. However, the pedological ‘historical record’ is normally incomplete and ambiguous, and changes in soils are therefore difficult to reconstruct. Soil chronosequences greatly aid the reconstruction process. Most soil chronosequences are established using the comparative geographical technique ŽRode, 1961.. This involves arranging soils of different age, and so from different locations, into a time sequence. As a technique, it rests upon assumptions that are normally unstated Žcf. Vreeken, 1975.. A central assumption is that the soils in the identified sequence represent successive stages of one or several pedogenic processes. Further- more, it is assumed that the soils pass through stages characterized by some preceding member of the successional sequence Žcf. Rode, 1961, p. 28.. These assumptions underpin two different methods of constructing soil chronosequences. One method draws upon the traditional technique for identifying successional sequences in vegeta- tion. It involves looking for the various stages of a developmental soil sequence in modern soil–landscapes. By identifying the different stages and placing them in a supposed chronological order, a chronosequence is inferred. A well-known sequence of this kind is the development of a lithosol formed in marly limestone, through a rendzina, brunified rendzina, calcareous brown soil, and calcic brown soil, to a lessived brown  R.J. Huggettr Catena 32 (1998) 155–172 157 soil. This method rests on very shaky foundations and is no longer widely used. The chief weakness is that the age of different soils in the assumed sequence is at best educated guesswork. In addition, it supposes that, for example, rendzinas will evolve into brown soils under all circumstances. This supposition is suspect because rendzinas on steep slopes may be roughly in balance with local soil-forming factors and ‘progress’ no further. For these reasons, and others that will be elaborated later, strict adherence to a monosequential notion of soil chronosequences is unrealistic Žcf. Vreeken, 1975.. As Rode sagely said, ‘‘the bulk of pronouncements on soil evolution w qua developmentx advanced by the geographical method alone have to be considered as rather more or less probable hypotheses which require the application of other methods for rigid proof in their support’’ ŽRode, 1961, p. 30.. This admonitory statement is as relevant today as it was nearly 40 years ago. A far securer way of constructing geographical soil chronosequences is to use soils that have developed on surfaces of known age, and that either have persisted through to the present day or else have been buried beneath a sedimentary cover. Sequences of alluvial fans, alluvial terraces, coastal sand dunes, lava flows, and moraines have proved particularly fruitful sites for this purpose; abandoned pasture, earthflows, mudflows, fire-cleared areas, landslide scars, historically created polders, loess deposits, strand- lines, and other datable landscape features are also usable. Geographical soil chronosequences dominate the literature on the subject, but other time sequences of soils are possible. Vertical chronosequences ŽStevens and Walker, 1970., for instance, consist of buried soils sandwiched between successive sedimentary formations. Vreeken Ž1975. did pedology an enormous service by identifying four principal types of soil chronosequence. He argued that, if different soils should have evolved for different lengths of time, then they must have started evolving, or stopped evolving, or both, at different times. Soils that started evolving at the same time possess isochronous incipience; soils that started evolving at different times have time-transgres- sive incipience. Similarly, the times when soil evolution stopped may be isochronous or time-transgressive. Additionally, pedogenic start and stop times may, or else may not, overlap historically in different soils. These various possibilities produce the four principal kinds of soil chronosequence. First, post-incisive sequences occur where the start times are time-transgressive but the stop times are isochronous Žand are often the present day.. The term ‘incisive’ is used here to signify the geomorphic or human ‘incisions’ in the landscape that bury older surfaces, or create of new surfaces, or both. Post-incisive sequences contain soils that have evolved in sequence at successively later times. They are by far the most commonly studied kind of chronosequence, and result from post-incisive events producing a sequence of progressively younger surfaces in which soil evolution may begin. An example is soils formed in alluvial terraces. Commonly, alluvial terraces from a staircase within a river valley. In stepping down the staircase, soils become progressively younger. The soils forming in each tread form a chronosequence; at least, they do if it is assumed that the soils on the younger terraces will evolve into the soils on the older terraces. Second, pre-incisive sequences occur where the pedogenic start times are isochronous but the stop times time-transgressive. They contain soils that started to evolve at the same time, but which have been selectively and successively buried by pre-incisive events. An example would be soils 158 R.J. Huggettr Catena 32 (1998) 155–172 evolved in a newly exposed glacial till that has been gradually covered by another deposit, so stopping pedogenesis and burying soils at different stages of evolution. Third, fully time-transgressive sequences with partial historical overlap occur where the pedogenic start and stop times are both time-transgressive but coeval in part. They consist of a mixture of buried and relict soils produced by erosion and deposition. An example is a complex of buried and surface soils in lateral moraines, Bugaboo Glacier area, British Columbia, Canada ŽKarlstrom and Osborn, 1992.. And four, fully time- transgressive sequences with no historical overlap occur where the soils in question were never coeval. They are found in vertical sequences of soil–landscape units, such as those found between successive sedimentary units. In addition, the ‘geological palaeosols’ dating back to Precambrian times form a very disjointed chronosequence of sorts Že.g., Retallack, 1990, 1992.. Chronosequences may also be classed according to the time resolution on which they are based. At one extreme, the land surfaces Žpresent ground and buried. on which the soils are evolved are not well known and are given relative ‘ages’ by applying the principle of superimposition. These may be thought of as ‘ordinal’ chronosequences, by analogy with ordinal data—the sequence of events is known Žor at least guessed at., but the timing is not. An example is work on Holocene fan and floodplain deposits in Gandak megafan and adjoining Middle Gangetic Plains, India ŽMohindra and Parkash, 1990.. Clay minerals in these deposits reflect the duration of pedogenesis, but time is resolved on an ‘ordinal’ scale—young soils Žactive floodplain and young plain of the Gandak., older soils Žolder Gandak Plain., and still older soils Žoldest Gandak Plain.. At the other extreme lie ‘ratio’ soil chronosequences where the age of the landforms in which the soils have evolved is established by geochronometric techniques. Between these two extremes lie chronosequences calibrated using more sophisticated relative dating techniques. These methods rely on indices of change in parent materials Žcom- monly sediments. after deposition. The degrees of rock surface weathering and of soil development are widely used as indicators of age. Relative dating has had some successes, but it should be treated guardedly. In glacier forelands, for example, environmental and sedimentological factors, which can influence commonly used rela- tive-dating variables, may exhibit systematic variations in parallel with, but independent of, surface age, and so invalidate the assumed rates of soil change ŽMcCarroll, 1991; see also Kroonenberg et al., 1990.. The building and interpreting of soil chronosequences are hampered by several well-known problems Že.g., Birkeland, 1990; Gerrard, 1992.. A major difficulty is holding all soil-forming factors constant except time. In particular, climate is highly unlikely to have held steady, even for short periods, and topography and vegetation are unlikely to have remained unchanged for millennia or more. If topography should change, then it may or may not be true that a chronofunction Ža soil property plotted against soil age. reflects a single pedogenic process acting without interruption from geomorphic processes. In some locations, notably in arid environments, soils form chiefly in aeolian sediments and evolve as dust accumulates. At the same time, the dust deposits tend to fill in topographic lows, so reducing the local relief. Soil chronose- quences can be established in such environments. In any environment, mass movements remove and bury pre-existing soils and reset the chronosequence clock to zero. Pedo-  R.J. Huggettr Catena 32 (1998) 155–172 159 genic thresholds create problems as they may produce soil profile characteristics that are very like soil profile characteristics created by climatic change. Only in closely dated chronosequences is it possible to untangle the two sets of effects. Another point to remember is that not all changes in soil processes nor all events in soil history are recorded a soil profile—a chronosequence is only a partial record of the past. And, it may be a biased record, leaving more evidence of resistant soil constituents, such as duricrusts, than of rapidly changing soil properties such as pH and nitrogen content Žcf. Yaalon, 1971.. A practical difficulty lies in finding pedons on dated surfaces that are suitable for chronosequence work. Magnetic susceptibility measurements may be helpful in assessing the uniformity of pedogenesis and thus indicate pedons that might be used in chronosequence construction ŽFine et al., 1992.. Magnetic susceptibility measure- ments may also be used to estimate soil age, at least under certain climates ŽSinger et al., 1992.. A final difficulty is that traditional forms of soil survey Žincluding mapping, profile description, and analysis by horizon., in conjunction with simple transect or random sampling approaches, are inappropriate for the detailed quantitative investigation of soil chronofunctions. Most soil surveys focus on agricultural properties of soils, which are usually uninformative about long-term soil evolution. Seeking the environ- mental factors that control or influence rates of soil development requires a non-tradi- tional approach. One suggestion is to use stratified sampling of the best developed soils within time zones and topographic site-types, followed by soil property analysis on samples collected at specified depths ŽMesser, 1989.. Despite the problems involved in their construction, there is no doubt that soil chronosequences are immensely powerful tools for probing the rate and direction of soil evolution Že.g., Walker and Syers, 1976; Crews et al., 1995; Schaetzl et al., 1994.. Indeed, they are the only way of establishing how pedogenesis operates over centuries and longer periods. Well-dated chronosequences are therefore a boon to pedologists. They are also invaluable to geomorphologists, for, once a soil chronosequence is established, it may be used to investigate other landscape processes. A podzol chronose- quence in the Scottish Highlands has furnished information on saturated hydraulic conductivity changes in each horizon over the last 13,000 years; this information can be used in interpreting changing runoff process and slope stability ŽBrooks and Richards, 1993.. Two alluvial fan chronosequences in the Negev Desert have yielded information on the gravel shattering process in reg soils ŽAmit et al., 1993.. In some cases, soil chronosequences may be used to date land surfaces—known relationships between soil properties and time are used to gauge the age of landforms. This is probably a reasonable practice in cases where geochronometric ages cannot be ascertained. But it is a relatively crude tool—soil property estimates of Quaternary deposits and faulting events give errors of "50%, or even more ŽBirkeland, 1990.. The problem of using soils as a geochronometric tool is also demonstrated by a study of spatial soil variability in the well-dated Cajon Pass chronosequence, southern California ŽHarrison et al., 1990.. It was found that the influence of aeolian silt on pedogenic rates meant that the sequence is not strictly a chronosequence. Furthermore, the evidence for a pedogenic threshold in these soils indicates that the soils do not develop monotonically with time and that soil development cannot be defined by a continuous univariant chronofunction. The discrete nature of the data from a chronosequence and the complex nature of soil 160 R.J. Huggettr Catena 32 (1998) 155–172 development are better analysed by multivariate statistical procedures. Of course, where soil–time relationships are used as a crude dating tool, the dated surfaces cannot then be used to estimate rates of pedogenesis. 3. Studies of soil chronosequences Soil chronosequences record pedogenic change over several time-scales—years to centuries, millennia, hundreds of thousands of years, and even millions of years. Chronosequences evolved during the last millennium or so might be called historical chronosequences to distinguish them from older Holocene and Quaternary chronose- quences. The classic work on soil development on moraines left in the wake of retreating glaciers and on coastal sand dune systems established historical chronose- quences. More recent work on historical chronosequences uses a much wider range of dated surfaces. Landslide scars, old pastures, mined areas, and burnt landscape patches have yielded information on various aspects of short-term pedogenesis. During the Holocene and Quaternary, geomorphic change has produced a selection of datable surfaces that form a basis for chronosequence construction. Lava flows, a range of fluvial landforms Žalluvial fans, terraces, and floodplains., and old coastal sand dunes have been of great service to soil chronosequence builders. This section takes selected examples to illustrate the full sweep of soil chronosequence investigations carried out today Žsee also Table 1.. 3.1. Glacial moraines Glacier retreat in historical times has led to many fine soil chronosequences evolving in moraines. A classic study is the 220-year chronosequence Žand related vegetation succession. in Glacier Bay, National Park, Alaska ŽCrocker and Major, 1955; Olson, 1958.. Neoglacial moraine ridges in southern Norway have provided several suitable sites for soil chronosequence investigations Že.g., Mellor, 1987; Matthews, 1992.. 3.2. Sand dunes Coastal sand dunes commonly comprise a set of dune ridges separated by slacks. The dunes often lie roughly parallel to the coast, the youngest lying nearest to the sea. They thus provide a sequence of dated surfaces suitable for chronosequence work. Indeed, a sand dune sequence on the Lancashire coast, England, was used in a classic study of a soil Žand vegetation. time sequence ŽSalisbury, 1925.. Many recent soil chronosequence studies also employ sand dune systems. For instance, a soil chronosequence evolved in late Holocene dunes along Lake Huron, south-western Ontario, Canada, displayed progressive features of soil weathering ŽVandenBygaart and Protz, 1995.. The plainest indications of progression were an increasing solum thickness and a darkening of chroma in B horizons. Total organic matter in the profiles increased logarithmically with soil age, the fastest change occurring in the first 2000 years or so. The depth to, and thickness of, the B horizons, and the depth of CaCO 3 leaching, all increased linearly  R.J. Huggettr Catena 32 (1998) 155–172 161 with increasing soil age. Soil pH decreased linearly with soil age for the first 2900 years, but after that time, spatial differences in vegetation appear to have influenced the pH and thus pedogenesis. A variation on the sand-dune chronosequence theme comes from Late Pleistocene sand dunes at Tomago, on the east coast of Australia, that have been rehabilitated after sand mining ŽProsser and Roseby, 1995.. The rehabilitation involved spreading stock- piled A1 horizon material over mixed A2 and B horizon materials from the original humus podzols. A four-stage chronosequence of 1, 5, 11 and 17 years since soil rehabilitation was established. Eighty-one percent of the leaching occurred between 1 and 5 years after soil rehabilitation, at a mass loss rate of 3.0 t hay1 yry1 . Such rapid leaching led to thicker A2 horizons than before mining, which fact was attributed to physical destruction of the indurated B1 horizon and its homogenization with A2 and B2 material to form a permeable mass containing easily translocated forms of iron and aluminium. 3.3. Landslide scars Landslides strip the soil leaving patches of regolith within which pedogenesis starts anew. Where landslide scars can be dated, a soil chronosequence can be established. For example, a soil chronosequence is formed in landslide scars, ranging in age from 1 to 55 years, in the Luquillo Experimental Forest, Puerto Rico ŽZarin and Johnson, 1995.. The chronosequence shows that the base saturation index in surface mineral soil Ž0–10 cm. increases during primary succession; that the base saturation index and major nutrient cation concentrations are extremely low on new landslide scars; that, when initial conditions are oligotrophic, both nutrient cation pools and the base saturation index can increase in the surface mineral soil during early pedogenesis; and that the production and decomposition of soil organic matter is the dominant process controlling capture, retention, and within-ecosystem cycling of nutrient cations in the forests. 3.4. Old pasture Pasture that has been abandoned for varying periods provides a ready-made chronose- quence for rapidly changing soil properties, such as carbon and nitrogen levels. An example of this is a study of net nitrogen mineralization and net nitrification along a tropical forest-to-pasture chronosequence in the western Brazilian Amazon Basin state of Rondonia ŽPiccolo et al., 1994.. Present forest and pasture of three different ages—3, 9, and 20 years—were used. The natural abundance of 15 N in soil profiles along two chronosequences was investigated. One chronosequence consisted of forest and 3-, 5- and 20-year-old pasture, the other of forest and 8- and 20-year-old pasture. Pasture surface soils were 1–3 per thousand depleted in 15 N compared with forest soils. The d15 N values in 20-year-old pastures were consistent with greater cumulative inputs of 15 N-depleted atmospherically-derived nitrogen, fixed by free-living bacteria associated with planted pasture grasses in older pastures, or differential plant utilization of soil inorganic nitrogen pools with different d15 N values.  162 R.J. Huggettr Catena 32 (1998) 155–172 Table 1 Some recent soil chronosequence studies Site characteristics Location Main soil changes studied Reference Years to centuries Forest and old pasture Rondonia, western Brasilia Soil nitrogen Piccolo et al. Ž1994. Current and old pasture Sevilleta National Wildlife Refuge, New Soil carbon and respiration Kieft Ž1994. Mexico, United States Clear-cut mixed-conifer stands Sierra Nevada, northern California, Soil nitrogen mineralization Frazer et al. Ž1990. United States Mown grasslands Žunfertilized for 2, ´ Anloebrdiepje, The Netherlands Nitrogen mineralization Olff et al. Ž1994. 6,9, and 45 years. Forelands of Athabasca glacier Canada Soil nitrogen Kohls et al. Ž1994. Sand dunes Westhoek Nature Reserve, Belgium Horizon thickness, litter content, humus Ampe and Langohr Ž1993. content Holocene Andesitic ocean beach-ridge sedi- Costa Rica Weathering and neoformation of miner- Nieuwenhuyse et al. Ž1994. ments als Beach-ridge plain Moruya Heads, New South Wales, Aus- Horizon thickness, chroma, pH, depth to Bowman Ž1989. tralia leaching front, acid-extractable alumi- nium, calcium, iron, magnesium, manga- nese, and sodium Alluvial fans Negev Desert, Israel Gravel shattering by salts Amit et al. Ž1993. River terraces Western Cairngorms, Scotland Exchangeable calcium, magnesium, Bain et al. Ž1993. sodium, and potassium; base saturation; clay type Quaternary Alluvial terraces Appalachian Highlands, south-eastern Clay content, rubification index, iron Leigh Ž1996. United States oxide content, elemental ratios Alluvial terraces Cowlitz River, south-west Washington, Profile morphology and thickness; Dethier Ž1988. United States dithionite-extractable iron, pedogenic clay Alluvial surfaces Inner coastal plain, central Virginia, Soil type, duripans, incipient plinthite, Howard et al. Ž1993. United States ferricrete, weatherable minerals R.J. Huggettr Catena 32 (1998) 155–172 Marine terraces California, United States Dithionite- and oxalate- extractable iron Aniku and Singer Ž1990. and aluminium Alluvium Rock Creek, south-central Montana, Clay mineralogy Reheis Ž1990. United Sates Coastal dunes Cooloola and North Stradbroke Island, Podzol morphology, organic matter Thompson Ž1992.; Skjemstad et al. southern Queensland, Australia composition, elemental composition, 1992. ŽC, Si, Fe, Mn, P, Ti, Zr. Volcanic rocks Hawaii Soil phosphorus Crews et al. Ž1995. Volcanic rocks Hawaii Nitrogen Kitayama et al. Ž1995. Volcanic rocks Islands in south-western Pacific Ferralitization; silica–alumina ratio Zamotayev and Targulian Ž1994. 163 164 R.J. Huggettr Catena 32 (1998) 155–172 3.5. Burnt patches Fire creates burnt patches that are open to vegetation colonization and renewed surface-soil evolution. Repeated fires within an area produce patches of varying age, so creating a chronosequence. For example, the effect of recent fire history and succes- sional stage were evaluated in soils from a steep semiarid shrubland chronosequence in south-east Spain ŽCarreira et al., 1994.. A single fire significantly increased soil mineral nitrogen availability and net nitrification. An increasing fire frequency in the last few decades was associated with a sharp decrease in surface soil organic matter and total nitrogen concentrations and pools, and with changes in the long-term nitrogen dynamic patterns. Another study investigated forest-floor regeneration after fire in Catalonia, Spain. ŽFerran and Vallejo, 1992.. The area considered was a holm-oak Ž Quercus ilex . forest frequently suffering wildfires. Bulk samples from the L, F, and H layers were taken in five burned plots, aged from 0 to 35 years. They were analysed for dry standing weight, and organic carbon and nitrogen content. Within 2 years of a fire, plant cover reached almost 100%. Shrubs and herbs dominated the cover for 20 years, after which time holm-oaks became dominant. The litter decay coefficients were relatively high, allowing for a rapid structuration and formation of the L and F layers after the fire. Indeed, 95% of the maximum steady standing weight accumulated in 8–9 years. 3.6. Old mining areas There are instances where past mining activity has left a series of land surfaces of different ages, in each of which soil genesis has proceeded. An example, is the chronosequences produced by phosphate mining on the island of Nauru ŽManner and Morrison, 1991.. This island, which at 21 km2 is the world’s smallest republic, lies halfway between Australia and Hawaii. A central plateau is rich in phosphate deposits. Mined areas of phosphate ranged from less than 1 year old to 55 years old. Soil carbon and soil nitrogen increased from 0.41–0.48 to 4.56 and from 0.03–0.04 to 0.32%, respectively. The final figures are comparable to concentrations in nearby undisturbed soils. Soils evolved faster in unconsolidated sands and limestone rubble in mining pit bottoms than they did on dolomitic limestone pinnacle surfaces. 3.7. LaÕa flows Many volcanoes obligingly produce lava flows in discrete episodes throughout their active phases. The dates of these flows are often known for historical times, or can be dated with a high degree of precision. Lava flows around a particular volcano thus enable a reliable chronosequence to be constructed. An example is the weathering of volcanic pumices on Mont Pelee,´ Martinique, in the Caribbean ŽQuantin et al., 1991.. The chronosequence comprises three main deposits, dated at 48, 650, and 1670 years, covering the mountain massif. The initial pumices are essentially glassy, rich in silicon and aluminium, calcium and sodium. Within the period covered by the chronosequence, the weathered horizon reached a depth of 30–50 cm, and the median size of particles decreased from approximately 2 to 0.2 mm.  R.J. Huggettr Catena 32 (1998) 155–172 165 3.8. FluÕial landforms (fans, floodplains, and terraces) Several fluvial landforms tend to form episodically and to produce surfaces of differing age. These landforms provide fertile hunting grounds for seekers of soil chronosequences. Investigations of alluvial fans, floodplains, and river terraces have all proved especially rewarding in the search for suitable time sequence. Flights of river terraces provide many soil chronosequences. In the Reefton area of New Zealand, a sequence of dated river terraces has produced a chronosequence with soils ranging from a Dystric Fluvisol on the youngest surface, through Dystric Cam- bisols on middle terraces, to Gleyic Podzols on the higher terraces ŽDawson et al., 1991.. Silicon, titanium, and vanadium Žextracted by hydrofluoric–perchloric acid digestion. have accumulated in the soils over 130,000 years Žon a massrvolume basis., as would be expected given their slow loss during weathering. With the exception of titanium, all the elements have become depleted by over 50% of the original mass within 100,000 years. Four soil chronosequences developed in gravelly alluvial fans in the southern Great Basin, United States, were examined to quantify and help understand soil development during the Quaternary ŽHarden et al., 1991.. The four soils are formed in granite–gneiss, granite and basalt, limestone, and siliceous volcanic rocks. The ages of the soils were approximated from several radiometric and experimental techniques, and rates were assessed using a conservative mathematical approach. Holocene soils in all areas appear to have developed at similar rates, and Pleistocene soils at two of the sites may differ by only a factor of two to four. A preferred model for the age curves over long time intervals is not linear, but may be exponential or parabolic, with rates decreasing with increasing age. It was deduced from these preliminary results that the geographical variation in soil development rates within the southern Great Basin–Mojave region may be much less significant than temporal variation. The reasons for temporal variation in rates are linked to climatic change and related changes in water and dust, erosional history, and internally driven chemical and physical processes. 3.9. Marine terraces Quaternary marine terraces may be suitable for constructing soil chronosequences. In northern California, United States, a chronosequence was built from six soil profiles formed in unconsolidated, sandy-marine, and aeolian parent material deposited on bedrock marine platforms that range from a few years old to 240,000 years old ŽMerritts et al., 1992.. Soil evolution has been dominated by a depletion of silicon, calcium, magnesium, potassium, and sodium; by an enrichment of phosphorus in surface soil horizons; by a relative immobility of iron and aluminium; and by a transformation of iron, silicon, and aluminium in the parent material to secondary clay minerals and sesquioxides. Net mass losses of bases and silicon have been generally uniform with depth, and in some cases have approached 100%. However, the loss rate of each element is markedly different, so that the rank order according to relative abundance shifts with time. 166 R.J. Huggettr Catena 32 (1998) 155–172 4. Theories of pedogenesis and chronosequence studies Empirical investigations in science normally rest upon a theoretical base. New observations sometimes call for a change in the theory. This interplay of theory and observation is how science works. Commonly, the big theories Žmegatheories. become entrenched in the lore of a discipline and are enormously difficult to budge—they become ruling theories or paradigms. In pedology, soil-formation theory is a well-estab- lished paradigm that arose in the closing decades of the nineteenth century. It originally postulated that soil genesis is a product of essentially downwards acting processes that lead to two sets of interrelated layers—the A horizons and the B horizons, which together constitute the solum. Eluviation washes solutes and fine-grained materials out of the A horizons and deposits them in subjacent, illuvial B horizons. Continued eluviation and illuviation produce coarse-textured residual A and E horizons over heavier-textured B horizons. Under some conditions, organic matter accumulates as distinct O horizons that lie on top of the uppermost A horizon. Later, it was realized that some soil processes mix soil materials and in doing so tend to destroy soil horizons. Soil-formation theory was then modified to include the effects of horizon creation Žhorizonation. and horizon destruction Žhaploidization. ŽHole, 1961.. According to soil-formation theory, the nature and rate of pedogenic processes are regulated by ‘factors of soil formation’—climate, organisms, relief, parent material, and time ŽJenny, 1941.. This factorial–functional approach to soil genesis, which had a far-reaching impact in many environmental sciences, was the ruling theory until recently Žsee Johnson and Hole, 1994.. However, soil-formation theory is now challenged by an evolutionary view of pedogenesis. In its traditional version, soil-formation theory assumes that a soil forms or develops progressively under the influence of the environmental state factors. The developmental sequence continues until the soil is in equilibrium with prevailing environmental conditions. Once equilibrium is achieved, the soil will be a mature Žor normal., zonal soil Že.g., a podzol or a chernozem. and will change no more Žor at least until interrupted by geomorphic processes.. This idea was borrowed by Marbut from Davis’s ‘geographi- cal cycle’ concept ŽDavis, 1909., in which the landscape develops progressively from ‘youth’ to ‘maturity’ Žand ultimately ‘senility’. Žsee Jenny, 1941.. The same notion found expression in the developmental view of plant succession as proposed by Cowles Ž1899. and Clements Ž1916., in which new surfaces will be colonized in a definite sequence until a climatic climax vegetation is created that endures indefinitely. Two recent developments have undermined the hegemony of the developmental view, both in pedology and in vegetation science. First, empirical work on environmen- tal change shows that inconstancy, not constancy, of environmental conditions is the norm ŽHuggett, 1997.. Considering this fact, it is improbable that a developmental soil sequence will ever run its full course under a constant environment. Second, it has been found profitable to regard all ecospheric systems Žincluding soils. as dissipative structures forced away from equilibrium states by driving variables. Such structures are dynamical and non-linear. In dynamical systems, there is a turnover of materials through ongoing additions, loses, transfers, and transformations Žcf. Simonson, 1959, 1978.. In addition, state often changes owing to adjustments between inputs,  R.J. Huggettr Catena 32 (1998) 155–172 167 outputs, and storage. Net loses, net gains, and steady states of soil nitrogen exemplify such state changes. Two remarkable properties arise from the non-linearity of dynamical soil systems—depending on internal conditions and external forcing, soils may exhibit stable, periodic, or chaotic behaviour; and they may display self-organizing character- istics. Self-organizing characteristics involve change spatially from a uniform Žundif- ferentiated. state to an increasingly non-uniform Ždifferentiated or segregated. state, largely in response to dynamical instabilities ŽPhillips, in press.. Many soil Žand geomorphic. systems have two modes of behaviour: first, a stable, non-chaotic, non- self-organizing mode; and second, an unstable, chaotic, self-organizing mode ŽPhillips, 1995b.. Chemical weathering transforming rock into saprolite epitomizes stable mode behaviour. Overall, the weathering process is effectively irreversible, monotonically progressing to a single end state. Progressive pedogenesis typifies an unstable mode behaviour—an undifferentiated regolith Žthe product of weathering. is transformed to an increasingly differentiated and organized soil profile. Now, a soil in a stable, non-chaotic mode evolves towards Žor stays in. a single, predetermined, steady-state condition. No matter what the starting conditions, and regardless of any disturbances, the outcome of pedogenesis is always the same. This is mode of pedogenesis envisaged in the classic, developmental view of soils. However, a soil in an unstable, chaotic mode may evolve along any of several evolutionary pathways and at any particular time may exist in several possible states. In this mode, the state at any time is unique and depends upon initial conditions and historical happenstance Žsmall disturbances may be significant., and the soil changes through a series of largely unpredictable states within the bounds set by the local attractor Že.g., Phillips, 1995a, 1997.. This mode of pedogenesis is captured in the evolutionary view of soils. It is supported by numerical models of such processes as humus carbon dynamics ŽRyzhova, 1996. and soil thickness change ŽArlinghaus et al., 1992.. Environmental inconstancy and non-linear dynamics mean that, although some soils and soil processes may stick to the predestined developmental mode, others adopt the far less predictable evolutionary mode ŽHuggett, 1995, 1997; cf. Johnson and Watson- Stegner, 1987.. This line of thinking has led to the developmental view of pedogenesis, as interpreted from the classic chronosequences, being questioned ŽJohnson and Watson-Stegner, 1987; Johnson et al., 1990.. It was anticipated by Jenny Ž1980. Žp. 241. on his reading about the bold challenge to the climatic climax concept in ecology by Drury and Nisbet Ž1973.. ŽJenny’s state-factor approach to soil genesis applies both to the developmental and the evolutionary view of soils, but in the evolutionary view it is associated with a variety of evolutionary pathways or multidirectional succession.. Soil chronosequences are investigative pedological tools. As such, they may be used to test hypotheses of pedogenesis, and seem ideally suited to evaluate the developmental vs. the evolutionary view of soils. It should be possible using chronosequences to see if pedogenesis is unidirectional or whether it multidirectional Žthat is, displays a degree of evolutionary divergence.. It may be wondered why, given that many chronosequences have been established, no signs of multidirectional evolutionary pathways have been reported until recently. However, the interpretation of chronosequences is to an extent preconditioned by theoretical predilections. It is often the case that a ruling theory is so ingrained in researchers’ minds that they unwittingly make ‘the facts fit the theory’. 168 R.J. Huggettr Catena 32 (1998) 155–172 Thus, while the developmental view prevailed, researchers generally sought evidence for unidirectional pedogenic changes leading to a single steady state. It did not occur to anyone to look for multidirectional changes or multiple steady states, although some botanists have never been keen on the idea of a climatic climax vegetation, favouring instead an individualistic interpretation of succession as prosecuted by Gleason Ž1926.. Some recent work on chronosequences supports an evolutionary view of pedogenesis by showing that unidirectional progression towards a single soil type is just one of several evolutionary pathways. A chronosequence formed in terraces of the Almar River, Spain, involving Xerorthents ŽHolocene., Haploxeralfs ŽUpper Pleistocene., and Palexeralfs ŽMiddle Pleistocene. all formed in granite, displayed several trends in chemical and physical properties and in mineralogical components ŽDorronsoro and Alonso, 1994; see also Dorronsoro, 1994.. The horizon development indices and the soil development indices relate to soil age, the rate of ‘development’ declining most strongly in the oldest soils. But interestingly, in the great majority of cases, the properties and development indices continue to evolve through the chronosequence without reaching a steady state. Similarly, in an alluvial soil chronosequence from the inner Coastal Plain of central Virginia, where ages range from 60,000 to 1,600,000 years, not all soil properties showed unidirectional development, a steady state of pedon development was not observed, and the transition from one phase to the next was marked by a change in rate, or sometimes a reversal in the direction, of soil evolution ŽHoward et al., 1993.. A new investigation of the classic chronosequence at Glacier Bay, Alaska, compared the inferred primary vegetation succession with reconstructions of stand development based on tree-ring records from 850 trees at ten sites of different age ŽFastie, 1995.. It revealed three spatially and temporally segregated pathways of vegetation change. The conclusion was that communities of different age at Glacier Bay do not constitute a single chronosequence. The differences among the pathways influence soil evolution. In particular, the long-term accumulation of soil nitrogen becomes spatially differentiated according to the chance location of nitrogen-fixing Sitka alder Ž Alnus sinuata. shrubs. Moreover, where alder shrubs compete with black cottonwood Ž Populus trichocarpa. trees, the black cottonwoods tend to oust the alder and so retard the long-term accumulation of nitrogen at those sites. By altering litter and soil chemistry, black cottonwood could also affect succession to conifers. Other work in Glacier Bay shows similar spatial differentiation of soil properties by tree species ŽBormann and Sidle, 1990.. The O horizon at a Sitka alder site contained greater concentrations of most extractable macronutrients and micronutrients than at a Sitka spruce Ž Picea sitchensis . site. The chemical differences probably result from more rapid soil weathering, vegeta- tive uptake, and cycling back to the O horizon under alder. The soils under Sitka spruce experienced rapid podzolization. 5. Conclusions Soil chronosequences are valuable tools for pedological research. Classic chronose- quence studies were made on sand dunes and glacial moraines. The number of well-dated chronosequences, most of them post-incisive, is growing rapidly. Sequence  R.J. Huggettr Catena 32 (1998) 155–172 169 lengths have been extended to millions of years. The original work on soil Žand vegetation. chronosequences seemed to support the traditional developmental theory of pedogenesis Žsoil formation theory.. A new evolutionary view of pedogenesis, which parallels a new evolutionary view of vegetation change, is supported by some modern soil chronosequence investigations. Soil chronosequences remain potent instruments for probing the pedological past, and for helping to test pedological theories. References Amit, R., Gerson, R., Yaalon, D.H., 1993. Stages and rate of the gravel shattering process by salts in desert reg soils. Geoderma 57, 295–324. Ampe, C., Langohr, R., 1993. Distribution and dynamics of shrub roots in recent coastal dune valley ecosystems of Belgium. Geoderma 56, 37–55. Aniku, J.R.F., Singer, M.J., 1990. Pedogenic iron oxide trends in a marine terrace chronosequence. Soil Sci. Soc. Am. J. 54, 147–152. Arlinghaus, S.L., Nystuen, J.D., Woldenberg, M.J., 1992. An application of graphical analysis to semidesert soils. Geogr. Rev. 82, 244–252. Bain, D.C., Mellor, A., Robertson-Rintout, M.S.E., Buckland, S.T., 1993. Variations in weathering processes and rates with time in a chronosequence of soils from Glen Fleshie, Scotland. Geoderma 57, 275–293. Birkeland, P.W., 1990. Soil–geomorphic research—a selective overview. Geomorphology 3, 207–224. Birkeland, P.W., 1992. Quaternary soil chronosequences in various environments—extremely arid to humid tropical. In: Martini, I.P., Chesworth, W. ŽEds.., Weathering, Soils and Paleosols ŽDevelopments in Earth Surface Processes, 2.. Elsevier, Amsterdam, pp. 261–281. Bockheim, J.G., 1980. Solution and use of chronofunctions in studying soil development. Geoderma 24, 71–85. Bormann, B.T., Sidle, R.C., 1990. Changes in productivity and distribution of nutrients in a chronosequence at Glacier Bay National Park, Alaska. J. Ecol. 78, 561–578. Bowman, G.M., 1989. Podzol development in a Holocene chronosequence: I. Moruya Heads, New South Wales. Aust. J. Soil Res. 27, 607–628. Brooks, S.M., Richards, K.S., 1993. Establishing the role of pedogenesis in changing soil hydraulic properties. Earth Surf. Processes Landforms 18, 573–578. Burges, A., 1960. Time and size as factors in ecology. J. Ecol. 48, 273–285. Carreira, J.A., Niell, F.X., Lajtha, K., 1994. Soil nitrogen availability and nitrification in Mediterranean shrublands of varying fire history and successional stage. Biogeochemistry 26, 189–209. Clements, F.E., 1916. Plant Succession: An Analysis of the Development of Vegetation. Carnegie Institute of Washington, Publication No. 242. Carnegie Institute of Washington, Washington, DC. Cowles, H.C., 1899. The ecological relations of the vegetation on the sand dunes of northeastern Minnesota. Botanical Gazette 27, 95–117, 167–202, 281–308, 361–391. Crews, T.E., Kitayama, K., Fownes, J.H., Riley, R.H., Herbert, D.A., Mueller-Dombois, D., Vitousek, P.M., 1995. Changes in soil phosphorus fractions and ecosystem dynamics across a long chronosequence in Hawaii. Ecology 76, 1407–1424. Crocker, R.L., Major, J., 1955. Soil development in relation to vegetation and surface age at Glacier Bay, Alaska. J. Ecol. 43, 427–448. Davis, W.M., 1909. Geographical Essays. Ginn, Boston, MA. Dawson, B.S.W., Fergusson, J.E., Campbell, A.S., Cutler, E.J.B., 1991. Depletion of first-row transition metals in a chronosequence of soils in the Reefton area of New Zealand. Geoderma 48, 271–296. Dethier, D.P., 1988. The soil chronosequence along the Cowlitz River, Washington. US Geological Survey Bulletin 1590-F, 47 pp. Dorronsoro, C., 1994. Micromorphological index for the evaluation of soil evolution in central Spain. Geoderma 61, 237–250. 170 R.J. Huggettr Catena 32 (1998) 155–172 Dorronsoro, C., Alonso, P., 1994. Chronosequence in Almar River fluvial-terrace soil. Soil Sci. Soc. Am. J. 58, 910–925. Drury, W.H., Nisbet, I.C.T., 1973. Succession. J. Arnold Arboretum 54, 331–368. Fastie, C.L., 1995. Causes and ecosystem consequences of multiple pathways of primary succession at Glacier Bay, Alaska. Ecology 76, 1899–1916. Ferran, A., Vallejo, V.R., 1992. Litter dynamics in post-fire successional forests of Quercus ilex. Vegetation 99–100, 239–246. Fine, P., Singer, M.J., Verosub, K.L., 1992. Use of magnetic-susceptibility measurements in assessing soil uniformity in chronosequence studies. Soil Sci. Soc. Am. J. 56, 1195–1199. Frazer, D.W., McColl, J.G., Powers, R.F., 1990. Soil nitrogen mineralization in a clearcutting chronosequence in a northern California conifer forest. Soil Sci. Soc. Am. J. 54, 1145–1152. Gerrard, A.J., 1992. Soil Geomorphology: An Integration of Pedology and Geomorphology. Chapman & Hall, London, 260 pp. Gleason, H.A., 1926. The individualistic concept of the plant association. Bull. Torrey Bot. Club 53, 7–26. Harden, J.W., 1982. A quantitative index of soil development from field descriptions: examples from a chronosequence in central California. Geoderma 28, 1–28. Harden, J.W., Taylor, E.M., Hill, C., Mark, R.K., McFadden, E.D., Reheis, M.C., Sowers, J.M., Wells, S.G., 1991. Rates of soil development from four soil chronosequences in the southern Great Basin. Quaternary Res. 35, 383–399. Harrison, J.B.J., McFadden, L.D., Weldon, R.J. III, 1990. Spatial soil variability in the Cajon Pass chronosequence: implications for the use of soils as a geochronological tool. Geomorphology 3, 399–416. Hole, F.D., 1961. A classification of pedoturbation and some other processes and factors of soil formation in relation to isotropism and anisotropism. Soil Sci. 91, 375–377. Howard, J.L., Amos, D.F., Daniels, W.L., 1993. Alluvial soil chronosequence in the inner Coastal Plain, central Virginia. Quaternary Res. 39, 201–213. Huggett, R.J., 1995. Geoecology: An Evolutionary Approach. Routledge, London, 320 pp. Huggett, R.J., 1997. Environmental Change: The Evolving Ecosphere. Routledge, London, pp. 378 pp. Jenny, H., 1941. Factors of Soil Formation: A System of Quantitative Pedology. McGraw-Hill, New York, 281 pp. Jenny, H., 1980. The Soil Resource: Origin and Behaviour. Ecological Studies, Vol. 37. Springer, New York. Johnson, D.L., Hole, F.D., 1994. Soil formation theory: a summary of its principal impacts on geography, geomorphology, soil–geomorphology, Quaternary geology and paleopedology. In: Amundson, R. ŽEd.., Factors of Soil Formation: A Fiftieth Anniversary Retrospective. Soil Science Society of America Special Publication 33. Soil Science Society of America, Madison, WI, pp. 111–126. Johnson, D.L., Watson-Stegner, D., 1987. Evolution model of pedogenesis. Soil Sci. 143, 349–366. Johnson, D.L., Keller, E.A., Rockwell, T.K., 1990. Dynamic pedogenesis: new views on some key concepts, and a model for interpreting Quaternary soils. Quaternary Res. 33, 306–319. Karlstrom, E.T., Osborn, G., 1992. Genesis of buried paleosols and soils in Holocene and late Pleistocene tills, Bugaboo Glacier area, British Columbia, Canada. Arctic Alpine Res. 24, 108–123. Kieft, T.L., 1994. Grazing and plant canopy effects on semiarid soil microbial biomass and respiration. Biology and Fertility of Soils 18, 155–162. Kitayama, K., Mueller-Dombois, D., Vitousek, P.M., 1995. Primary succession of Hawaiian montane rain forest on a chronosequence of eight lava flows. J. Vegetation Sci. 6, 211–222. Kohls, S.J., van Kessel, C., Baker, D.D., Grigal, D.S., Lawrence, D.B., 1994. Assessment of N2 fixation and N cycling by dryas along a chronosequence within the forelands of the Athabasca Glacier, Canada. Soil Biology and Biochemistry 26, 623–632. Kroonenberg, S.B., Hoorn, M.C., Moura, M.L., Veldkamp, A., 1990. Variability in bulk geochemistry of ´ fluvial terrace sands: consequences for the study of weathering chronosequences. Pedologie 40, 19–31. Leigh, D.S., 1996. Soil chronosequence of Brasstown Creek, Blue Ridge Mountains, USA. Catena 26, 99–114. Manner, H.I., Morrison, R.J., 1991. A temporal sequence Žchronosequence. of soil carbon and nitrogen development after phosphate mining on Nauru Island. Pacific Sci. 45, 400–404. Matthews, J.A., 1992. The Ecology of Recently-Deglaciated Terrain: A Geoecological Approach to Glacier Forelands and Primary Succession. Cambridge University Press, Cambridge, 386 pp.  R.J. Huggettr Catena 32 (1998) 155–172 171 McCarroll, D., 1991. Relative-age dating of inorganic deposits: the need for a more critical approach. Holocene 1, 174–180. Mellor, A., 1987. A pedogenic investigation of some soil chronosequences on neoglacial moraine ridges, southern Norway: examination of soil chemical data using principal components analysis. Catena 14, 369–381. Merritts, D.J., Chadwick, O.A., Hendricks, D.M., Brimhall, G.H., Lewis, C.J., 1992. The mass balance of soil evolution on late Quaternary marine terraces, northern California. Geol. Soc. Am. Bull. 104, 1456–1470. Messer, A.C., 1989. An alternative approach to the study of pedogenic chronosequences. Norsk Geografisk Tidsskrift 43, 221–229. Mohindra, R., Parkash, B., 1990. Clay mineralogy of the Gandak megafan and adjoining Middle Gangetic ´ 43, 203–212. Plains, India. Bull. Sci. Geol. Niewenhuyse, A., Jongmans, A.G., Van Breeman, N., 1994. Mineralogy of a Holocene chronosequence on andesitic beach sediments in Costa Rica. Soil Science Society of America Journal 58, 485–494. Olff, H., Berendse, F., De Visser, W., 1994. Changes in nitrogen mineralization, tissue nutrient concentrations and biomass compartmentation after cessation of fertilizer application to mown grassland. J. Ecol. 82, 611–620. Olson, J.S., 1958. Rates of succession and soil changes on Southern Lake, Michigan, sand dunes. Bot. Gazette 119, 125–170. Phillips, J.D., 1995a. Time lags and emergent stability in morphogenicrpedogenic system models. Ecol. Modelling 78, 267–276. Phillips, J.D., 1995b. Self-organization and landscape evolution. Prog. Phys. Geogr. 19, 309–321. Phillips, J.D., 1997. Simplicity and the reinvention of equifinality. Geogr. Anal. 29, 1–15. Phillips, J.D., in press. On the relation between complex systems and the factorial model of soil formation. Geoderma. Piccolo, M.C., Neill, C., Cerri, C.C., 1994. Natural abundance of 15 N in soils along forest-to-pasture chronosequences in the western Brazilian Amazon Basin. Oecologia 99, 112–117. Prosser, I.P., Roseby, S.J., 1995. A chronosequence of rapid leaching of mixed Podzol soil materials following sand mining. Geoderma 64, 297–308. ´ Quantin, P., Balesdent, J., Bouleau, A., Delaune, M., Feller, C., 1991. Premiers stades d’alteration de ponces volcaniques en climat tropical humide ŽMontagne Pelee ´ Martinique.. Geoderma 50, 125–148. Reheis, M.C., 1990. Influence of climate and eolian dust on the major-element chemistry and clay mineralogy of soils in the northern Bighorn basin, USA. Catena 17, 219–248. Retallack, G.J., 1990. Soils of the Past: An Introduction to Paleopedology. Unwin Hyman, Boston. Retallack, G.J., 1992. Paleozoic paleosols. In: Martini, I.P., Chesworth, W. ŽEds.., Weathering, Soils and Paleosols, Developments in Earth Surface Processes 2. Elsevier, Amsterdam, pp. 543–564. Rode, A.A., 1961. The soil forming process and soil evolution. Israel Program for Scientific Translations, Jerusalem. Ryzhova, I.M., 1996. Analysis of the feedback effects of ecosystems produced by changes in carbon-cycling parameters using mathematical models. Eurasian Soil Sci. 28, 42–52. Salisbury, E.J., 1925. Note on the edaphic succession in some dune soils with special reference to the time factor. J. Ecol. 13, 322–328. Schaetzl, R.J., Barrett, L.R., Winkler, J.A., 1994. Choosing models for soil chronofunctions and fitting them to data. Eur. J. Soil Sci. 45, 219–232. Simonson, R.W., 1959. Outline of a generalized theory of soil genesis. Proc. Soil Sci. Soc. Am. 23, 152–156. Simonson, R.W., 1978. A multiple-process model of soil genesis. In: Mahaney, W.C. ŽEd.., Quaternary Soils. Geo Abstracts, Norwich, pp. 1–25. Singer, M.J., Fine, P., Verosub, K.L., Chadwick, O.A., 1992. Time dependence of magnetic susceptibility of soil chronosequences on the California coast. Quaternary Res. 37, 323–332. Skjemstad, J.O., Fitzpatrick, R.W., Zarcinas, B.A., Thompson, C.H., 1992. Genesis of podzols on coastal dunes in Southern Queensland. II. Geochemistry and forms of element as deduced from various soil extraction procedures. Australian Journal of Soil Research 30, 615–644. Stevens, P.R., Walker, T.W., 1970. The chronosequence concept and soil formation. Q. Rev. Biol. 45, 333–350. 172 R.J. Huggettr Catena 32 (1998) 155–172 Thompson, C.H., 1992. Genesis of podzols on coastal dunes in Southern Queensland. I. Field relationships and profile morphology. Australian Journal of Soil Research 30, 593–613. VandenBygaart, A.J., Protz, R., 1995. Soil genesis on a chronosequence, Pinery Provincial Park, Ontario. Can. J. Soil Sci. 75, 63–72. Vincent, K.R., Bull, W.B., Chadwick, O.A., 1994. Construction of a soil chronosequence using the thickness of pedogenic carbonate coatings. J. Geol. Educ. 42, 316–324. Vreeken, W.J., 1975. Principal kinds of chronosequences and their significance in soil history. J. Soil Sci. 26, 378–394. Walker, T.W., 1966. The factor time in soil formation. New Zealand Soil News 6, 153–167. Walker, T.W., Syers, J.K., 1976. The fate of phosphorus during pedogenesis. Geoderma 15, 1–19. Yaalon, D.H., 1971. Soil-forming processes in time and space. In: Yaalon, D.H. ŽEd.., Paleopedology—Origin, Nature and Dating of Paleosols. International Society of Soil Science and Israel Universities Press, Jerusalem, Israel, pp. 29–39. Yaalon, D.H., 1975. Conceptual models in pedogenesis: can soil forming factors be solved?. Geoderma 14, 189–205. Yaalon, D.H., 1983. Climate, time and soil development. In: Wilding, L.P., Smeck, N.E., Hall, G.F. ŽEds.., Pedogenesis and Soil Taxonomy: I. Concepts and Interactions. Elsevier, Amsterdam, pp. 223–251. Zamotayev, I.V., Targulian, V.O., 1994. Development of ferralitization in a chronosequence of Andosols. Eurasian Soil Sci. 26, 1–8, ŽTranslated from Pochvovedeniye, 5 Ž1994.: 25–29.. Zarin, D.J., Johnson, A.H., 1995. Base saturation, nutrient cation, and organic matter increases during early pedogenesis on landslide scars in the Luquillo Experimental Forest, Puerto Rico. Geoderma 65, 317–330.