Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene age in Central Mexico

Quaternary International xxx (2014) 1e15

Contents lists available at ScienceDirect

Quaternary International journal homepage: www.elsevier.com/locate/quaint

Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene age in Central Mexico ~ a-Ramírez a, Lorenzo Va zquez-Selem b, Christina Siebe a, * Victor Pen a b

noma de M Instituto de Geología, Universidad Nacional Auto exico, Cd. Universitaria, M exico DF, CP 04510, Mexico noma de M Instituto de Geografía, Universidad Nacional Auto exico, Cd. Universitaria, M exico DF, CP 04510, Mexico

a r t i c l e i n f o

a b s t r a c t

Article history: Available online xxx

The Transmexican Volcanic Belt and its many volcanic fields of different ages offer good opportunities to study soil development on volcanic tephra of intermediate to basaltic composition. We studied a soil chronosequence within the Sierra Chichinautzin volcanic field (SCVF), south of the basin of Mexico, and aimed to establish the rates of pedogenic processes. This field has been active for around 50,000 years, and produced 221 cinder cones with their respective lava flows. We selected 11 sites located on 14C dated lava flows of ages between 1800 and 30,500 BP, at 3100e3200 m above sea level, covered with pineefir forest, with an ustic soil moisture regime and an isomesic soil temperature regime. We also included a younger site (1000 BP) and three older sites (>100,000 years), two at 3100 masl and one at 2600 masl, from nearby volcanic fields to widen the time frame of the chronosequence. Soil profile samples were analysed for total organic carbon as well as for mineral neoformations related to clay contents, selective chemical extractions, and X-ray diffraction analyses. Within the SCVF the total soil thickness, carbon accumulation and Al, Si and Fe extracted with acid ammonium oxalate increased linearly with age on surfaces up to 10,000 years old at rates of 19 cm ky
1, 4.1 kg C m
2 ky
1, 4.6 kg Al, 2.7 kg Si and 2 kg Fe m
2 ky
1, respectively. Crystalline clay and iron oxide formation reach a maximum at the oldest site located at 2600 masl of 1650 kg clay m
2 and 70 kg Fed m
2. Their increase is linear up to ages of 10,000 years at rates of 22 kg clay m
2 ky
1, and 2.8 kg Fed m
2 ky
1. Thereafter the rates are much slower. Allophane and allophane-like minerals dominate in the clay fraction at all sites of the chronosequence, and small amounts of halloysite can be identified in soils older than 6200 years, while kaolinite was only identified at the three oldest sites (>100,000 years). The linear increases of all indicators of pedogenesis in the first 10,000 years are presumably driven by recurrent tephra deposition during this time frame. Thereafter, erosion and colluviation processes seem to disturb pedogenesis, changing its rates and redistributing its products in the landscape. © 2014 Elsevier Ltd and INQUA. All rights reserved.

Keywords: Soil genesis Chronosequence Soil organic matter accumulation Clay minerals Volcanic ash Andosols

1. Introduction Soil forming processes have very slow rates in relation to human lives. Therefore our current knowledge on their rates has been dominantly investigated by selecting study objects affected by the same or at least very similar soil forming conditions over different time spans, from few hundreds to tens of thousands and up to millions of years, i.e. by quantifying indicators of pedogenic processes along so called soil chronosequences (Stevens and Walker,

* Corresponding author. E-mail address: [email protected] (C. Siebe).

1970; Yaalon, 1975; Huggett, 1998). To conform a soil chronosequence all study sites should ideally be located on stable landform positions, have the same parent materials, and similar climatic conditions and thus vegetation cover over time (Bockheim, 1980; Jenny, 1980; Crews et al., 1995; Schaetzl and Anderson, 2005). Additionally the parent material should offer a clear reference of its age, particularly of the age since it has been exposed to weathering at the earth's surface. However, it is well known that environmental conditions are not stable at millennial time scales, and that vegetation changes in response to climate (Heine, 1975; Harden, 1982; Kitayama and Mueller Dombois, 1995; Kitayama et al., 1995; Vitousek et al., 1997). Also, in volcanic landscapes, tephra deposits are seldom homogeneous over larger distances, as coarser particles

http://dx.doi.org/10.1016/j.quaint.2014.11.032 1040-6182/© 2014 Elsevier Ltd and INQUA. All rights reserved.

~ a-Ramírez, V., et al., Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene Please cite this article in press as: Pen age in Central Mexico, Quaternary International (2014), http://dx.doi.org/10.1016/j.quaint.2014.11.032

2

~ a-Ramírez et al. / Quaternary International xxx (2014) 1e15 V. Pen

are deposited close to the source, while smaller size particles can travel much farther (Schmincke, 2004). Recurrent volcanic activity will add fresh material to adjacent locations in variable amounts depending on the meteorological conditions (e.g. Jackson et al., 2005). In several landscapes the inputs of airborne allochthonous materials have been evidenced to alter soil properties significantly (McFadden et al., 1987; Herrmann et al., 1996; Yaalon, 1997). Also, volcanic ash redeposited by wind has been recognized as a source €hling, for loess-like soils (Jackson et al., 2005; Iriondo and Kro 2007). Nevertheless, soil chronosequence studies in which successive stages of one or more pedogenetic processes over several timescales are recorded, have allowed to calculate rates and direction of pedogenetic change during the Quaternary (McFadden and Weldon, 1987; Karlstrom, 1988; Scarciglia et al., 2014) in very different landscapes, as glacial moraines (Hall, 1999), landslide scars (Krasilnikov and Targulian, 2007), old mining areas (RamosArroyo and Siebe, 2007), fluvial landforms such as fans, floodplains and terraces (Howard et al., 1993), marine terraces (Sauer et al., 2010, 2012), and lava flows and volcanic ash deposits of differing ages (Crocker and Major, 1955; Harden et al., 1991; Manner and Morrison, 1991; Miehlich, 1991; Merritts et al., 1992; Crews et al., 1995; Van den Bygaart and Protz, 1995; Zarin and Johnson, 1995; Jahn and Stahr, 1996; Nieuwenhuyse et al., 2000). Several authors report changes of specific indicators of pedogenic processes measured at sites of different age and in which they assume that all other soil forming factors have suffered similar changes. In chronosequence studies, the rate and direction of pedogenetic change can be calculated by correlating soil ages with distinct properties, and by adjusting for example a linear or logarithmic equation to the data (Levine and Ciolkosz, 1983). These chronofunctions are very useful for testing theories of pedogenesis, and to determine the necessary time for certain pedogenetic features or horizons to form (Birkeland, 1990, 1992; Crews et al., 1995; Chadwick and Chorover, 2001; Holzschuh, 2004). Weathering processes have also been studied in the laboratory, by modifying temperature and pressure conditions to force the advance of pedogenesis (Schnoor, 1990; Swodoba-Colberg and Drever, 1993; Sverdrup and Warfvinge, 1995; White and Brabtley, 2003). However the rates calculated in these experiments do not coincide with those inferred from field chronosequences. This discrepancy is attributed to the difficulty of estimating a reactive surface area of field soil minerals, and also to macropore water flow in aggregated natural soils, which alters the reaction contact times between the solid phase and the soil solution (Nahon, 1991; Lasaga, 1998; Chadwick and Chorover, 2001). For this reason, chronosequence studies are still widely used to infer rates of pedogenic processes. However, considering that soil forming factors do not remain constant in time, the data analysis considers a multidirectional rather than a unidirectional driving force of pedogenesis (Phillips, 1993). This recognizes that soil development occurs in episodes during which the soil forming factors are relatively constant, and which are then taken over by a following phase in which the rate of a process changes due to a change in one or several soil forming factor conditions. In some cases, the process can slow down and the expression of an indicator diminishes in consequence. Also, it is now well recognized that the rates of pedogenesis tend to decline as soil development increases, due to the self-limiting nature of some processes, such as the depletion of weatherable minerals, or the decline in weathering rates at the weathering front as regolith thickness increases. In active volcanic fields, soil development is affected by recurrent tephra fall as new eruptions occur at the same or in adjacent volcanoes (Bertrand and Fagel, 2008). As loose ash deposits have a large specific surface and their mineral components are easily

weathered, this can lead to soil aggradation. However, if the ash fallouts have a larger thickness, former soils are buried beneath them, and pedogenesis is interrupted. Often, the ash fallout severely affects vegetation, and subsequent rain events lead to soil erosion. These differences in soil chronosequences have been recognized by Vreeken (1975) and Huggett (1998), who differentiate soil chronosequences as “pre-incisive” or “post-incisive”, depending on whether the starting point of soil formation coincides at all studied sites, or it started at different moments, and according to the time overlap during which all studied sites were subject to pedogenesis (i.e. “time transgressive chronosequences with and without historical overlap”). Many chronosequence studies have been performed in volcanic landscapes, under different moisture regimes, from xeric, i.e. arid and semiarid climates (Jahn and Stahr, 1996; Vaughan, 2008), to udic and perudic, i.e. humid climates (Crews et al., 1995; Nieuwenhuyse et al., 2000; Egli et al., 2008). However, few studies document rates of pedogenic processes under ustic soil moisture regimes (i.e. seasonal climates). The Transmexican Volcanic Belt (TMVB) crosses Central Mexico from the Pacific Coast to the Gulf of Mexico. It offers excellent opportunities to study genesis of volcanic ash soils, as volcanic activity has been more or less constant throughout the Quaternary. One example is the Sierra Chichinautzin volcanic field (SCVF), located south of the Basin of Mexico, where volcanic activity has been recurrent in time intervals of <1700 years over the last 50,000 years (Siebe et al., 2004). It is composed of more than 221 monogenetic cones, which were active for a few years and produced lava flows of dacitic to basaltic composition, as well as ash fallouts, many of which originated during the latest part of the eruptions and covered the lava flows. petl (5450 masl), located Similarly, the strato-volcano Popocate in the Sierra Nevada in the southeast part of the Basin of Mexico has had three major episodes of activity during the last 25,000 years (Siebe and Macías, 2006), offering excellent sites for chronosequence studies. This was acknowledged by Miehlich (1991), who did a very detailed study on soil formation in the Sierra Nevada covering a time span from 1000 to 10,000 BP at altitudes between 2500 and 3000 masl, with soil moisture regimes ranging from ustic/isomesic (2500e3000 m), to udic isomesic/isofrigid (3000e4000 m) and ustic/isofrigid (>4000 m). The objective of this study was to establish rates of pedogenic processes, such as the accumulation of humified organic matter and the neoformation of secondary minerals at 3000 m altitude under coniferous forests. Our aim was to distinguish particularly the rate of neoformation of short-range order minerals, humus-Al and Fe complexes on the one hand, and the formation of crystalline clay minerals and iron oxides on the other hand. We also aimed to investigate whether these indicators of pedogenic processes could provide hints on climate change during the last 100,000 years at this altitude, and whether they are affected by changes in the mineralogical composition of the parent materials. In addition, we investigated if the recurrent inputs of ash fallout deposits can be identified in the field and through various laboratory analyses and to what extent they might affect the overall pedogenic development in relation to the other soil forming factors. 2. Materials and methods 2.1. Study site description and selection of study sites The studied sites are located in the surrounding volcanic mountains of the basin of Mexico, within the central part of the ~ aTransmexican Volcanic Belt (TMVB) (Table 1 and Fig. 1, Pen Ramírez et al., 2009). All sites are between 3000 and 3200 m

~ a-Ramírez, V., et al., Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene Please cite this article in press as: Pen age in Central Mexico, Quaternary International (2014), http://dx.doi.org/10.1016/j.quaint.2014.11.032

~ a-Ramírez et al. / Quaternary International xxx (2014) 1e15 V. Pen

3

petl volcano within the Sierra Fig. 1. Study site locations within the Transmexican volcanic belt, the basin of Mexico and the Sierra Chichinautzin volcanic field (11 sites), Popocate Nevada (1 site) and the Sierra de Monte AltoeMonte Bajo (3 sites).

above sea level (masl) except one, located at 2600 m, and they are either covered by pine or fir forest. We selected sites where the forest cover was relatively well preserved, although all of them showed some disturbance by forest fires (as revealed by charcoal fragments), which have affected the area recurrently. To construct the main part of the chronosequence, we chose 11 sites within the Sierra Chichinautzin volcanic field (SCVF), south of the basin of Mexico. The selection of five lava flows of different age (Chichinautzin, Guespalapa, Tlaloc, Cuauhzin and Pelado) was based on the existence of several 14C dates (Table 2), which determine their ages with confidence (Siebe et al., 2004, 2005). Reported 14 C dates were obtained from charcoal fragments found beneath the lava flows. We selected one or two of the reported dates for each lava flow, which according to the authors are the most reliable. When two dates were considered reliable, we calculated a mean age of the lava flow (see Table 2). petl lava flow was selected, as it is evidently older The Malacate than the Pelado volcano. We defined its age based on Bloomfield (1975), who estimated the ages of 41 volcanic cones in the western part of the Chichinautzin volcanic field using morphological features and a few radiocarbon dates. Following this approach, we petl lava an age similar to the one of El assigned to the Malacate Molcajete volcano dated by Bloomfield (1975) in 30,500 ± 1160 BP (bulk soil buried by the lava flow) (Table 2). In addition, we selected two soils formed also on volcanic fallout deposits, located at other volcanic fields within the valley of petl volcano (Sierra Nevada, Mexico: a younger soil from Popocate

southwest of the basin of Mexico), formed on a dark grey ash fall of andesitic composition which lies on top of a pumice deposit dated 1000 BP by charcoal fragments found within the deposit (Heine, 1975; Miehlich, 1984; Siebe et al., 1996; Panfil et al., 1999); and an apparently much older soil at La Catedral volcano, in the Sierra de Monte AltoeMonte Bajo, northwest of the basin of Mexico. The slopes of La Catedral volcano are formed by lavas and pyroclastic flow deposits dated by K/Ar between 3.71 ± 0.40 and 2.90 ± 0.4 Ma (Osete et al., 2000). A major eruption of the nearby San Miguel volcano dated by K/Ar to 170 ka (Mooser et al., 1986), most probably mantled with tephra the Sierra de Monte AltoeMonte Bajo. Besides the latter K/Ar age, we assign a minimum age of 100,000 years to La Catedral hillslopes and soils based on a comparison of its geomorphic features with those from other dated areas of the n, Blatter et al., TMBV (e.g. shield volcanoes of western Michoaca 2001). We selected three sites within La Catedral volcanic complex; one under fir forest (Abies religiosa (Kunth) Schltdl. & Cham.) at 3200 m; one under pine forest (Pinus montezumae Lamb.) at 3100 m; and the third one also under pine forest (Pinus hartwegii Lindl.), but located at 2600 m. We included this latter site, as we could not find a soil that evidently has evolved beyond the andic phase and is dominated by crystalline clay minerals at a site above 3000 m, neither within the basin of Mexico nor in the western part n). of the TMVB (state of Michoaca At all sites, the parent material for soil formation consists of either volcanic fallout deposits or reworked volcanic ash. Small

~ a-Ramírez, V., et al., Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene Please cite this article in press as: Pen age in Central Mexico, Quaternary International (2014), http://dx.doi.org/10.1016/j.quaint.2014.11.032

~ a-Ramírez et al. / Quaternary International xxx (2014) 1e15 V. Pen

4

Table 1 General characteristics of the sampled sites at the central part of the Transmexican Volcanic Belt. Site Sierra Nevada petl P Popocate

Slope orientation

Rounded average age of the land surfacea [years BP]

Dominant tree species

Altitude [masl]

Coordinates UTM

Evidence of disturbance

NE

1000

Pinus montezumae

3100

X:545524 Y:2107185

Cattle grazing and understory fires

Sierra Chichinautzin Chichinautzin P SW

1835

Pinus montezumae

3100

Quarrying of lava blocks

Guespalapa P1

SW

3800

Pinus montezumae

3100

Guespalapa P2

S

3800

Pinus montezumae

3100

loc P Tla

SW

6200

Pinus montezumae

3100

Cuauhtzin P

SE

8000

Pinus montezumae

3100

Pelado P1

SW

10,000

Pinus montezumae

3100

Pelado P2

SW

10,000

Pinus montezumae

3100

petl P Malacate

SW

30,500

Pinus montezumae

3100

X:482041 Y:2109907 X:481203 Y:2110358 X:481833 Y:2109747 X:493279 Y:2111386 X:490349 Y:2115772 X:475922 Y:2114796 X:475544 Y:2114369 X:472016 Y:2117611

Sierra Monte Alto and Monte Bajo Catedral P NE >100,000

Pinus montezumae

3100

Pinus hartwegii

3100

Catedral Ph

>100,000

N

Sierra Chichinautzin Chichinautzin A SW

1835

Abies religiosa

3200

loc A Tla

SW

6200

Abies religiosa

3200

petl A Malacate

SW

30,500

Abies religiosa

3200

Sierra Monte Alto and Monte Bajo Catedral A NE >100,000

Abies religiosa

3200

a

Cattle grazing and understory fires Cattle grazing and understory fires Cattle grazing and understory fires Understory fires Few evidences of disturbance Few evidences of disturbance Selective logging and understory fires

X:450104 Y:2168173 X:453832 Y:217992

Selective logging and cattle grazing Selective logging and cattle grazing

X:482274 Y:2109903 X:493980 Y:2111353 X:472170 Y:2117954

Quarrying of lava blocks

X:449936 Y:2167350

Selective logging and cattle grazing

Cattle grazing and understory fires Selective logging and understory fires

See Table 2 for the list of dates reported in the literature.

Table 2 Age of the landform at the different sampling locations according to literature reports and the corresponding dating method. Based on the existing reports we calculated an average maximum age for the land surface, which we used to adjust the models. Site

Age of landform [years BP]

Reference

Dated material

Dating method

petl Popocate

855 ± 155 1265 ± 55 1835 ± 55 2835 ± 75 4690 ± 90 6200 ± 85 7360 ± 120 8225 ± 130 9620 ± 160 10,270 ± 190 10,900 ± 280 30,500 ± 1160 170,000

Heine (1975)

Charcoal

non calibrated

14

C

1000

C C

1835 3800

C C

6200 8000

Chichinautzin Guespalapa loc Tla Cuahutzin Pelado

petl Malacate Catedral

Rounded average age used for modeling [years BP]

Siebe et al. (2004) Siebe et al. (2004)

Charcoal Charcoal

non calibrated non calibrated

14

Siebe et al. (2005) Siebe et al. (2005)

Charcoal Charcoal

non calibrated non calibrated

14

Siebe et al. (2004)

Charcoal

non calibrated

14

C

10,000

Bloomfield (1975) Mooser et al. (1986)

Bulk paleosoil Biotite

non calibrated K/Ar

14

C

30,500 >100,000

charcoal fragments distributed homogeneously over a defined loc P, Tl thickness indicated reworking at the following sites: Tla aloc ~ a-Ramírez, 2013). The chemical A, Catedral P, and Catedral A (Pen composition, however, is not homogeneous; the sites at La Catedral and Cuauhtzin are dominantly of dacitic composition, while those of intermediate age of the SCVF and the youngest site at the Sierra petl) are andesitic; the two youngest deposits Nevada (Popocate within the SCVF are basaltic. All sites of the SCVF are oriented to the south, while the site of the Sierra Nevada and La Catedral are oriented to the north. The slope angle at all sites is close to 10 .

14

14

The climate at all sites located between 3000 and 3200 m is €ppen climate temperate, humid with abundant summer rainfall (Ko ~ a-Ramírez et al. classification code: Cw; García (1981), see also Pen (2009)). The mean annual precipitation is 1000e1200 mm and the mean annual temperature is 7e9  C (IMTA, 2000). Frost occurs frequently in winter, which coincides also with the dry season. The soil moisture regime is ustic and the soil temperature regime is isomesic (Soil Survey Staff, 1999). Although paleoenvironmental data from the mountains of central Mexico are scarce, the existing records overall indicate that areas at an elevation around 3000 m have experienced dominantly

~ a-Ramírez, V., et al., Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene Please cite this article in press as: Pen age in Central Mexico, Quaternary International (2014), http://dx.doi.org/10.1016/j.quaint.2014.11.032

~ a-Ramírez et al. / Quaternary International xxx (2014) 1e15 V. Pen

humid conditions during the late Pleistocene, with marked cooling associated to glacier advances of the last glacial cycle. Geomorphological data indicate an extensive glacial advance on the highest zquez-Selem and mountains (>4000 masl) at 205e175 ka (Va Heine, 2011). Our study sites were likely under periglacial conditions with presumably little or no soil formation at that time. A paleopedological record from a near mountain site (Nevado de Toluca) at ~3000 masl suggests that humid climate conditions and forest ecosystems dominated during much of the Late Pleistocene, from the late MIS 5 (ca. 100 ka) to the early MIS 2 (ca. 29 ka), interspersed with short dry phases (Sedov et al., 2001, 2003). The glacial-geomorphological record shows glaciation on the high zquez-Selem mountains from ~21 to 14 ka and from 12 to 10 ka (Va and Heine, 2011). Sites around 3000 masl remained ice-free but likely experienced a cold climate under alpine grassland from 21 to 18 ka, and then were gradually covered by an open pine forest zquezduring the terminal Pleistocene (Lozano-García and Va Selem, 2005). Although low lake levels indicate generally drier conditions at lower elevations during this period (Caballero et al., 2010), mountain sites probably remained relatively moist. Pollen data from a mountain lake at 3100 masl in Sierra Chichinautzin ~ ero et al., 2005) in general indicate temperate con(Almeida-Len ditions with mixed forest (pineeoakealder) throughout the Holocene, with lower temperatures from 10.5 to 5 ka (prevalence of pine), cooler and moister conditions from 2 to 1 ka (fir) and somewhat drier climate since 1 ka (pine). 2.2. Sampling and laboratory analyses Sampling site selection was based on exploratory 1 m-auger observations performed along the slope gradient within previous selected areas that fulfilled the above mentioned requirements with respect to altitude, orientation, slope, and forest cover. Particular care was given to microrelief conditions, in order to choose sites not obviously affected by erosion or colluviation. At each selected site a soil pit was dug, and a soil profile described (FAO, 2006). The detailed soil descriptions are presented ~ a-Ramírez (2013). One disturbed and five undisturbed in Pen (100 cm3 cores) samples were taken from each horizon. An aliquot of the disturbed sample was air dried, sieved <2 mm, and ground with an agate mortar <0.06 mm, while the rest of the sample was preserved field moist at 4  C. Bulk density was determined on the undisturbed samples gravimetrically after drying at 105  C. On air dried, sieved (<2 mm) and ground samples, the following analyses were performed: - Extractable Al, Si, and Fe (Alo, Sio and Feo) in ammonium oxalate at pH 3.25 during 4 h in the dark (Van Reeuwijk, 1992). Elements were quantified by flame atomic absorption spectroscopy (AAS, Perkin Elmer 3110), and Si was quantified immediately after extraction to avoid its reprecipitation. The oxalate extraction dissolves short-range order minerals, such as imogolite, protoimogolite, allophane and allophane-like minerals, as well as Al and Fe humus complexes (Wada, 1980). - Sodium dithionite extractable Al, Si and Fe (Ald, Sid, Fed) (Van Reeuwijk, 1992). These elements were also quantified by AAS, and this extract aims to dissolve crystalline iron and aluminium oxides, in addition to short-range order minerals extracted by ammonium oxalate. - Pyrophosphate extractable Al and Fe (Van Reeuwijk, 1992), (Alp and Fep), in which Al and Fe were quantified by AAS. This extracts aims to dissolve exclusively Al and Fe out of metalehumus complexes (Wada, 1980). The difference AloeAlp is interpreted as Al contained in short-range order minerals (Schwertmann, 1964; Blume and Schwertmann, 1969).

5

- Total carbon and nitrogen were determined in a CNHS auto analyser (Perkin Elmer 2400). Since the soil samples are carbonate free, total carbon equals total organic carbon (TOC). The following analyses were performed in field moist samples that were sieved <2 mm (results are reported on absolute dry mass, correcting for soil moisture content on behalf of sample aliquots dried at 105  C): - Percentage of P fixation capacity (Pfix) was determined after Blakemore et al. (1987), and non-adsorbed P was quantified with a Thermo Spectronic Colorimeter (model Genesys 20) at a wavelength of 466 nm. - pH was measured in 0.01 m CaCl2 in the supernatant of a 1:2.5 (wt:vol) soil suspension with an Aqua Lytic Senso Direct pH24 potentiometer. - Exchangeable base cations were extracted with 1 N ammonium acetate buffered at pH 7, and Ca and Mg quantified by AAS (Perkin Elmer 3100), K and Na by flame emission (Corning). - Exchange acidity (H plus Al) was determined in 1 M KCl extracts by titration with 0.01 N NaOH and with 4% NaF (Hþ), or by AAS (Al3þ). - Soil particle size distribution was determined by a combined sieve (particles 2000e63 mm) and sedimentation procedure (pipette-method: silt and clay size particles). Previous to the analysis soil organic matter was eliminated by H2O2 treatment, and short-range order minerals were dissolved by two subsequent extractions with ammonium oxalate at pH 3.25. Red coloured soils (Catedral Ph, 2600 m), were additionally treated with sodium dithionate to eliminate crystalline iron oxides. Previous to particle separation, samples were dispersed with 50 ml of 0.4 N sodium hexametaphosphate. The reported particle size fractions were calculated on the organic matter and short-range-order-minerals free sample mass. Soil mineralogy was investigated by X-ray diffraction (XRDF) (Shimadzu XRD-6000 and Philips Mod. 1130/96 (generator) and PW 1050/25 (goniometer), both with radiation Cu Ka) in at least one B horizon sample of each profile. For this the clay fraction of each sample was separated by sedimentation in organic matter free sample aliquots. The latter was achieved by pretreating each sample with a 6% sodium hypochloride solution and eliminating the excess salt by permeation through dialysis bags (Siregar et al., 2005). The clay fraction was flocculated by saturating the suspension with MgCl2 and the concentrated suspension was divided into two aliquots: one was shortly dispersed by an ultrasonic treatment and then stepwise dripped on a glass plate allowing the excess water to evaporate until the plate was covered by a homogeneous 1 mm thick clay sheet; the second aliquot was first treated with ammonium oxalate at pH 3.25 in the dark to dissolve short-range order minerals and thereby improve the signals of crystalline minerals, before orienting it on the glass plates. The oriented specimens were then submitted to X-ray diffraction either directly, or after thermic treatment at 100, 150, 450 and 550  C, or after a treatment with formamide solvated with ethylene glycol. 2.3.

14

C dating

In order to corroborate the ages of the studied soils, charcoal fragments found in buried A horizons at some of the sites were collected and dated at the Laboratory of Isotope Geochemistry (conventional dating technique) of the University of Arizona. When charcoal fragments were too small the samples were dated at the AMS facility of the University of Arizona. The non-calibrated 14C age is reported as years before present (BP) in Fig. 2.

~ a-Ramírez, V., et al., Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene Please cite this article in press as: Pen age in Central Mexico, Quaternary International (2014), http://dx.doi.org/10.1016/j.quaint.2014.11.032

~ a-Ramírez et al. / Quaternary International xxx (2014) 1e15 V. Pen

6

Fig. 2. Graphical representation of the described soil profiles, 14C dates (data from this study) and rounded average ages of the landform from the literature (see Table 2), soil classification, altitude and dominating tree species at each site.

2.4. Modelling the rates of pedogenic processes The maximum average age of each landform, i.e. the existing dates of the respective lava flows, was set as the independent variable, while the mass in kg m
2 of the indicator variables in the total soil thickness (m) above the lava flow was considered as the dependent variable. The rates of the respective pedogenic processes were calculated considering only the data of the Sierra Chichinautzin profiles by regression of the data for the first 10,000 years of pedogenesis, since the age of landforms is relatively certain for this time frame (Table 2). For the variables “crystalline clay” and dithionite extractable Al and Si (Ald and Sid), a second linear regression was adjusted to the data for ages 10,000e100,000 years, in order to estimate the rate of pedogenesis for this time frame. The site Catedral Ph, located at 2600 m, was excluded from all models. 3. Results 3.1. Soil descriptions and

14

C dates of buried soils

The field descriptions show that soil depth differs widely along the chronosequence depending on the thickness of the ash fallout deposits and also on their recurrence (Fig. 2). The soil at the petl site developed on a very thick ash fallout youngest Popocate deposit (>100 cm), while at the youngest sites of the SCVF,

Chichinautzin A and P (A ¼ Abies; P ¼ Pinus), and Guespalapa P1, the soils developed on 6e30 cm thick ash deposited on top of basaltic lava. The ash fall deposits at sites older than 2800 years are much thicker, as recurrent volcanic activity has buried former soils loc P and Malacate petl P sites). At most sites, (Guespalapa P2, Tla discontinuities in soil texture and bulk density are indicative of ~ afrequent addition of tephras into existing soil (Fig. 2, and Pen Ramírez, 2013). Charcoal fragments collected in the materials on top of the lava flows (Fig. 2) yield slightly younger ages than those published by Siebe et al. (2004) for the soils buried by the lava flows, thus confirming the ages of the latter. Only the charcoal loc site yielded a dated at the oldest site (Catedral P) and at the Tla considerably younger age than expected (750 ± 40 BP instead of >100,000 BP for Catedral, and 1438 ± 35 instead of 6200 BP at the loc site). At both sites, the soil apparently has formed on Tla reworked materials, suggesting that forest fires (as indicated by the many charcoal fragments found within these layers) disturbed the forest cover and caused erosion and colluviation of former soils. 3.2. Rates of change of selected indicators of pedogenesis The total depth, i.e. the sum of all horizons on top of the lava flow, increases continuously within the SCVF, and reaches almost petl site after 30,500 years. During the first 3 m at the Malacate 10,000 years, the total soil depth increases linearly at rates of 19 cm each 1000 years (Fig. 3a) as shown by the linear regression model

~ a-Ramírez, V., et al., Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene Please cite this article in press as: Pen age in Central Mexico, Quaternary International (2014), http://dx.doi.org/10.1016/j.quaint.2014.11.032

~ a-Ramírez et al. / Quaternary International xxx (2014) 1e15 V. Pen

7

▵

Fig. 3. Soil thickness (a), pH values measured in either Ah ( C - ) or Bw horizons (B , ) (b), amounts of soil organic carbon in the total depth (c), and clay content in the total depth (d) along the soil chronosequence. Continuous lines correspond to the adjusted linear models to the data of sites within the Sierra Chichinautzin volcanic field younger than 10,000 BP. The discontinuous line shows the linear model adjusted to the amount of clay in soils older than 10,000 BP.

adjusted to the data, which explains 68% of the variance. This increase is clearly due to a recurrent addition of new tephras by surrounding volcanic activity and a continuous advance of pedogenic processes within these fine grained unconsolidated materials. However, at the oldest sites of the Sierra Monte AltoeMonte Bajo (Catedral P and Catedral A) the soil is less deep, which can be attributed to a predominance of erosion processes over new tephra depositions at these sites. In this area of the TMVB volcanic activity ceased about 170 ka (Mooser et al., 1986). The pH values of the soils vary between 4.8 and 7.5 along the chronosequence; strongly acid pH values are only observed in the Ah horizons of the youngest soils, which are rich in organic matter, show incipient soil development and only slight weathering of primary minerals. The B horizons have less acidic and even neutral pH values, and only in those of the oldest soils (>100,000 years) the pH tends to values around 6. The latter could either be due to a smaller release of base cations from the parent material (presumably andesitic to dacitic) or to an incipient debasification of the soils (Fig. 3b). The accumulation of humified organic matter is expressed in kg m
2 organic C within the total soil depth along the

chronosequence (Fig. 3c). This was done to include buried Ah horizons and to account for all humification that occurred along the chronosequence in different time intervals since the emplacement of the underlying lava flow. A linear increase of humified organic matter can be observed during the first 8000 years of soil development, when a maximum organic C accumulation of 54 kg m
2 is reached. On older surfaces humus mass decreases to 28 kg m
2 at 3000 m altitude, and to only 12 kg m
2 at 2600 masl at the oldest sites. A linear regression model was adjusted to the first 10,000 years of soil development (Chichinautzin sites) yielding a rate of 4.05 kg m
2 organic C accumulated every 1000 years. The model explains 61% of the variance (Fig. 3c). Total crystalline clay contents increase also linearly in the first 10,000 years, at a rate of 22.4 kg m
2 every 1000 years. On older sites the clay increase rate seems to diminish to 1.5 kg m
2 every 1000 years, although due to the scarcity of data and the increased variability at the oldest sites, this regression coefficient was not statistically significant. The reasons for the smaller rates at older petl site (30,500 years) sites could be the following: at the Malacate a >1 m-thick ash fall deposit yielding a 14C age of 9033 ± 47 BP

~ a-Ramírez, V., et al., Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene Please cite this article in press as: Pen age in Central Mexico, Quaternary International (2014), http://dx.doi.org/10.1016/j.quaint.2014.11.032

8

~ a-Ramírez et al. / Quaternary International xxx (2014) 1e15 V. Pen

covers the original soil. It probably was produced by the nearby Pelado volcano (10,900e9620 BP). This means that clay accumulation corresponds to a time interval of 10,000 rather than 30,500 years. The two sites located at 3000 masl at La Catedral (oldest site) seem to have been affected by erosion and colluviation processes, as revealed by the radiocarbon dates of the charcoal fragments of 750 BP collected at medium depth, which reduced the soil depth to less than 1.4 m, while the soil at 2600 masl is >1.8 m deep and has a five-fold larger clay accumulation of 1679 kg m
2. Here the warmer climate conditions related to a ~400 m lower elevation, as well as deposition of the eroded materials from upslope positions could be the determinants of this large clay accumulation. Neoformation of short-range order minerals in the solum, as indicated by the Alo, Sio and Feo contents, increase linearly at rates of 4.6, 2.7 and 2.0 kg m
2, respectively, every 1000 years during the first 10,000 years within the soils of the SCVF (Fig. 4aec). The fitted models explain 72, 49 and 67% of the variance. The largest Alo, Sio and Feo contents were found after 8000 years (Cuauhtzin site), and they apparently approximate steady state conditions between petl site (30,500 years) 10,000 and 30,500 years. As the Malacate received a large ash fall during the eruption of El Pelado volcano, mineral neoformations at this site have dominantly formed on the 10,000 year old ash. At the oldest sites, the contents decrease markedly, which is consistent with the expected transformation of

short-range order minerals into crystalline secondary minerals with time (Besoain, 1974; Mizota and van Reeuwijk, 1989). Pyrophosphate extractable Al and Fe (Alp and Fep), the indicators of the formation of metalehumus complexes, increase linearly during the first 10,000 years within the SCVF chronosequence (Fig. 4d and e) and the adjusted linear model yields increment rates of 0.3 and 0.1 kg m
2 every 1000 years, respectively. The models explain 57 and 41% of the respective variances. Largest Alp and Fep amounts are shown also at the 8000 year old (Cuauhtzin P) site (4.33 and 1.39 kg m
2). However, the Alp contents decrease towards the 30,500 year old site to around 2 kg m
2, while at the Catedral sites at 3000 m altitude (2.70e4.18 kg m
2), they lie within the same ranges as at the 8000e10,000 year old (Cuauhtzin and Pelado) sites (2.5e4.33 kg m
2). The Fep content decreases at the 30,500 year old site, and keeps diminishing towards the oldest sites (at one of the Catedral sites no Fep could be detected). These findings are coherent with existing knowledge about the depletion of metalehumus complexes as volcanic soils age (Yagasaki et al., 2006; Torn et al., 1997; Tonneijck et al., 2010). The amounts of crystalline pedogenic oxides in the total soil thickness, as indicated by the dithionite extractable Al, Si and Fe contents (Ald, Sid, Fed), increase progressively with age (Fig. 4feh), at higher rates during the first 10,000 years, and at lower rates

Fig. 4. Acid ammonium oxalate Al, Fe and Si amounts (a, b and c), pyrophosphate extractable Al and Fe amounts (d, e) and sodium dithionite extractable Al, Fe and Si amounts (f, g and h), in the total soil depth along the chronosequence. Continuous lines correspond to the adjusted linear models to the data of sites within the Sierra Chichinautzin volcanic field younger than 10,000 BP. Discontinuous line shows the linear model adjusted to the amount of clay in soils older than 10,000 BP.

~ a-Ramírez, V., et al., Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene Please cite this article in press as: Pen age in Central Mexico, Quaternary International (2014), http://dx.doi.org/10.1016/j.quaint.2014.11.032

~ a-Ramírez et al. / Quaternary International xxx (2014) 1e15 V. Pen

thereafter. Adjusted linear models to the data of less than 10,000 years show increment rates of 0.9, 0.23 and 2.8 kg m
2 every 1000 years, respectively, and they explain 65, 29 and 55% of the variance. The largest quantities of Fed are found at the oldest site located at 2600 m altitude, while largest Ald amounts are found at the oldest sites located at 3000 m altitude (La Catedral P), and the largest Sid amounts at the Cuauhtzin sites (8000 years old). The X-ray diffractograms of oriented specimens of clay fractions from most soils that were not previously treated with ammonium oxalate show broad peaks (Fig. 5), which in samples pretreated with acid ammonium oxalate become much better defined. These peaks disappear after heating at 150  C, indicating a predominance of short-range order minerals and organo-mineral Al and Fe complexes (the diffractograms after the different temperature and formamideeethylene glycol treatments are not shown here, but ~ a-Ramírez, 2013). Only the samples from the can be viewed in Pen oldest soil located at 2600 m altitude (Catedral Ph) and the deepest B horizons from 3100 to 3200 m altitude (Catedral P and A sites), did not show this behaviour, and therefore do not contain shortrange order minerals (Fig. 5). Oxalate pretreated samples in which the short-range order minerals were eliminated yielded some better defined peaks, particularly in the lower range of the diffractogram.

9

Halloysite 10 Å was only identified in small amounts in the Bw loc A site (6200 BP). Significant amounts of this horizons of the Tla clay mineral were detected in buried Bw horizons at the Cuauhtzin P site, 5Bw horizon (8000 BP); the Pelado P1 site, 2Bw horizon petl P site, 6Bw horizon and the Mala(10,000 BP); the Malacate petl A site, 3Bw and 4AC horizons (30,500 BP). It was absent in cate all the younger soils, as well as in older soils (Fig. 6). Halloysite 7 Å was present in small amounts at the Tl aloc P, loc A (6200 BP) and Cuauhtzin sites (8000 BP). Several buried Bw Tla horizons did also contain this clay mineral in significant amounts, petl as in the Pelado P1 site, 2Bw horizon (10,000 BP); the Malacate petl A site, 3Bw and 4AC P site 6Bw horizon and Malacate (30,500 BP); and at the oldest site Catedral P it was detected in the 3Bw1 horizon (Fig. 6). Kaolinite was only found at the three soil profiles of the oldest sites (>100,000 years). Also small amounts of smectite could be detected in some horizons of these soils (Fig. 6). Other minerals that could be detected in the clay fraction were opal (Fig. 6), which was present in all B horizons, buried and not buried; plagioclase was particularly present in Bw horizons from the Guespalapa P2 (hopetl A site, while at the rizon: Bw1; 3800 BP) to the Malacate Catedral P and A sites this mineral is present in considerably smaller amounts; sanidine was found exclusively in B horizons of

Fig. 5. X-ray diffractograms from oriented clay specimens before (grey line) and after acid ammonium oxalate treatment. In soils of young ages, short-range order minerals dominate and only after their dissolution by the acid ammonium oxalate treatment, crystalline minerals can be detected; they are more abundant in the oldest soils.

~ a-Ramírez, V., et al., Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene Please cite this article in press as: Pen age in Central Mexico, Quaternary International (2014), http://dx.doi.org/10.1016/j.quaint.2014.11.032

10

~ a-Ramírez et al. / Quaternary International xxx (2014) 1e15 V. Pen

Fig. 6. Dominant clay minerals and minerals identified by X-ray diffraction in oriented clay fraction specimens of selected B horizons along the soil chronosequence. Squares indicate presence of important amounts of the respective clay minerals or minerals, while triangles represent only traces. A more detailed description of the applied sample treatments is provided in the Materials and methods section.

petl P the Tl aloc P and A sites (6200 BP), as well as at the Malacate site (30,500 BP); gibbsite and traces of cristobalite were detected only in Bw horizons and in buried horizons at the oldest sites Catedral P and A (>100,000 years) (Fig. 6). 3.3. Soil classification The youngest soils (<4000 BP) classify as either vitric Leptosols or vitric Andisols, depending on the thickness of the ash fallout and the presence of consolidated rock (lava) near the surface (Fig. 7). In all these soils, the neoformation of short-range order minerals as deduced from the Alo þ 1/2Feo contents is still incipient (<2%), and they still contain large quantities of glass in the sand fraction (data not shown). Only the Guespalapa P2 profile, which developed on

top of a coarse lapilli deposit, shows a cambic horizon, while all the other profiles have AeC-horizon sequences. Soils of ages of 6200 years and up to >100,000 a (sites at La Catedral P and A that are located at 3000 masl), classify predominantly as silandic Andisols. They show a horizon >30 cm thick in which the andic properties criteria are met (Fig. 7), and which additionally have a Sio content >0.6% and an Alp/Alo ratio of <0.5. Only two of the studied soils did not classify as Andisols: one is the petl A site, which has less than 85% of P profile at the Malacate fixation capacity, and classifies therefore as a haplic Phaeozem (siltic); the other profile is the one at 2600 masl at La Catedral (>100,000 years old), which classifies as a cutanic Lixisol. The latter soil is dominated by crystalline clay minerals and no longer has andic properties.

~ a-Ramírez, V., et al., Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene Please cite this article in press as: Pen age in Central Mexico, Quaternary International (2014), http://dx.doi.org/10.1016/j.quaint.2014.11.032

~ a-Ramírez et al. / Quaternary International xxx (2014) 1e15 V. Pen

11

Fig. 7. Depth functions of the different Andosol diagnostic criteria. The first three criteria (Aloþ0.5 Feo), P fixation capacity (P fix) and bulk density allow to distinguish between vitric and andic properties, as defined by the threshold lines shown in each graph, Sio and the ratio Alp/Alo, define silandic and aluandic prefixes, respectively, when the data surpass the marked threshold lines.

~ a-Ramírez, V., et al., Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene Please cite this article in press as: Pen age in Central Mexico, Quaternary International (2014), http://dx.doi.org/10.1016/j.quaint.2014.11.032

12

~ a-Ramírez et al. / Quaternary International xxx (2014) 1e15 V. Pen

The bulk density of most of the soils is very small (<0.9 g cm
3), ptl site except for the horizons of the youngest soil at the Popocate (1000 year old), some of the Bw of the oldest site (Catedral P) and all horizons of the oldest site at 2600 m altitude (Catedral Ph). At these locations bulk densities were larger than >0.9 g cm
3 and up to 1.4 g cm
3 (Fig. 7). 4. Discussion The analysis of rates of pedogenic processes along a soil chronosequence assumes that the studied soils developed from very similar parent materials under the same topographic and climatic conditions and vegetation cover (Harden, 1982). In active volcanic fields soils do not develop from a single volcanic ash deposit, but rather receive recurrent fallout deposits (Bertrand and Fagel, 2008). Field descriptions showed the latter through buried A horizons, differences in soil texture along depth profiles and discontinuities in the amount of coarse fragments (>2 mm) of several of the studied soils. Reworked ash deposits mixed with charcoal indicate erosion and colluviation processes at some sites. All these recurrent events altered the progress of pedogenesis at least short time after petl A, Catedral P their occurrence. At Tl aloc P and Tl aloc A, Malacate and Catedral A no stratification features can be observed, although the land surface was very probable affected by recurrent ash fallouts produced by the neighbour volcanoes. Here, bioturbation by soil organisms as worms and gophers (detailed soil profile de~ a-Ramírez (2013)) seem to have scriptions are presented in Pen almost completely obliterated the stratification features produced by recurrent inputs of allochthonous materials. Soil development is only interrupted for a longer time interval if the deposits are of larger thickness (Buurman et al., 2004), as the petl P site (>1 m). Ash fallouts often one that affected the Malacate burn the existing vegetation cover near the eruptive source; fresh ash fallouts are also easily eroded by rain events or wind, which occur shortly afterwards and transport not only the fresh ash but loc P, Tla loc A, Catedral P, and also the former soils as at sites Tla Catedral A. Sites that are at a larger distance from an eruptive source will receive smaller amounts of fresh materials, which will not affect the existing plant cover. These deposits do not interrupt the general pedogenic process; they only add new material, which is incorporated to the existing soil by bioturbation, and contribute to an accretion of the solum in the medium term, i.e. in the studied chronosequence within the SCVF. The continuous supply of easy weatherable minerals, particularly glass, explains also the constant increases of short-range order neoformations within the still active volcanic field of the Sierra Chichinautzin. Part of this continuous supply may have an aeolian origin, as wind has been recognized as a major source of dust short after an eruption (Arnalds et al., 2013) and as an important geomorphic agent following a recent eruption in central Mexico similar to those of the SCVF (Inbar et al., 1994). Wind-driven redeposition of ash has been identified as the origin of €hling, 2007). loess-like deposits in volcanic regions (Iriondo and Kro In the long term, and particularly when the recurrent volcanic activity ceases, erosion and colluviation processes seem to dominate over soil development, at least at La Catedral sites >3000 masl. This leads to much shallower soils at uphill locations (>3000 masl) and very deep soils at downhill sites (2600 masl). Associated to the respective solum losses or gains are smaller clay amounts (calculated in kg m
2 and total soil depth) in the uphill soils and much larger ones downhill. The smaller amounts of humified organic matter in the uphill soils at La Catedral can partly be explained by erosion. Although the mineralogical composition of the tephras is not constant throughout the studied sites, and not even in the tephra produced by the same volcanic field, this does not seem to alter the

direction or the rate of pedogenesis significantly. In the Chichinautzin volcanic field, the composition has shifted in time from dacitic in the oldest deposits (La Catedral) to intermediate (andesisiticebasaltic) to basalticeandesitic in the youngest deposits (Siebe et al., 2004). Similarly, Nanzyo et al. (2007) did not find significant differences in the mineralogical composition of soils of the same age (Holocene to late Pleistocene) formed on tephras of different mineralogical composition (basaltic to dacitic) under udic soil moisture regimes in Japan. Apparently differences in the mineralogical composition only affect initial rates of secondary mineral formation, but this effect becomes less important in soils of intermediate age, as those studied here. The different mineralogical composition of the parent materials might be the explanation of the smaller amounts of oxalate extractable Fe, Feehumus complexes (Fep) as well as crystalline Feoxide contents in the oldest soils: dacitic tephras have a smaller content of iron-bearing minerals than basaltic ones, and this leads to smaller iron-oxide formation rates at the oldest sites (La Catedral, Fig. 4g). On the other hand, the clear decrease of the Alo and Sio amounts in the oldest soils at the three studied locations could rather indicate that the formation of short-range order Al and Si minerals is a self-terminating process (Yaalon, 1971), once the recurrent deposition of glass contained in allochtonous materials stops as the volcanic activity ceases. Climatic conditions have not been constant through time along the studied chronosequence. In Central Mexico over the late Pleistocene, a more humid climate with dry intervals allowed a pine tree forest to develop at 3000 m (Sedov et al., 2001, 2003). In the periods between 21,000 and 14,000 years and between 12,000 and 10,000 BP, glaciers advanced, the climate became drier and 6e4 C cooler, and the dominant vegetation cover was grassland. The timberline was located around 3000 m and gradually rose to zquez~4000 masl during the Holocene (Lozano-García and Va Selem, 2005), in tandem with changes in vegetation composition at elevations similar to those of our chronosequence (Almeida~ ero et al., 2005). Therefore all the profiles of the chronoseLen quence developed under changing climate-vegetation conditions, during which the primary productivity of the plant cover changed drastically, and also the decomposition rate of the plant debris differed when compared to the actual conditions. During the Last Glacial Maximum a drier and cooler climate than today has been inferred (Caballero et al., 2010). Under these dry-cool conditions pedogenic processes must have slowed down, affecting the oldest petl and La Catedral). Particularly mineral new forsites (Malacate mation and Si leaching likely had slower rates favouring halloysite and cristobalite formation. The presence of these minerals in the soils of >30,000 years could therefore be better explained by periods of drier climate, rather than by crystallization of short-range order minerals due to a shortage of volcanic glass over time, since inputs of glass and other weatherable minerals were constant through the late Pleistocene and Holocene. Results show a constant increase of 4 kg organic C per m2 in soil organic matter every 1000 years during the first (most recent) 10 millennia of the chronosequence. The fitted linear model crosses the y-axis at 1.7 kg C m
2, indicating that during the first 1000 years humus content increased at less than half of this rate. Many authors have considered the accumulation of humified organic matter as a relatively fast process, which reaches a dynamic steady state after a few decades (Zehetner, 2010). In the SCVF chronosequence this process is active for much longer, which is attributable to the soil aggradation produced by the recurrent ash deposition. The latter ~ aimproves site quality and in parallel also biomass production (Pen Ramírez et al., 2009): at the youngest sites, shallow soils with small water holding capacities, small N reserves and excessive drainage petl and conditions allow only an open forest with grass (Popocate

~ a-Ramírez, V., et al., Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene Please cite this article in press as: Pen age in Central Mexico, Quaternary International (2014), http://dx.doi.org/10.1016/j.quaint.2014.11.032

~ a-Ramírez et al. / Quaternary International xxx (2014) 1e15 V. Pen

Chichinautzin sites), but as soil aggradation proceeds, the water holding capacities and the N reserves increase, allowing trees to grow more densely. Additionally, the steady increase of metalehumus complexes as shown by the Alp and Fep contents, stabilize soil organic matter (Tate and Theng, 1980; Wada, 1980; Mizota and van Reeuwijk, 1989; Shoji et al., 1993). Torn et al. (1997) documented the accumulation of up to 60 kg C m
2 in soils of 150,000 years in a soil chronosequence at Hawaii, and attributed this big carbon storage capacity to the large specific surface of the shortrange order mineral new formations (700 m2 g
1) (Torn et al., 1997; Basile-Doelsch et al., 2005; Basile-Doelsch et al., 2007). In our chronosequence, almost the same amount of organic carbon accumulated after only 8000 years (Cuauhtzin: 54 kg C m
2), possibly due to the cooler temperatures at the SCVF. Despite the large production of organic debris by the coniferous vegetation, the soil pH in the surface horizons is well buffered around pH 6 throughout the studied chronosequence, except in the shallow soil of the youngest site, which had a pH of less than 5. Grasses take up important amounts of basic cations, and contribute substantially to their recycling, which in turn buffers the pH around 6. They also take up Si impeding its leaching. Therefore allophane and imogolite formation is favoured under grass cover (Shoji et al., 1993), while pine forests produce litter rich in polyphenols and organic acids, pH values of the soil are more acid and the formation of Al and Feehumus complexes is expected to dominate (Mizota and van Reeuwijk, 1989). In the studied chronosequence these new formations are present in the surface horizons, while in the subsurface horizons the release of Ca and Mg from the tephra rich in plagioclase and mafic minerals buffers the pH between 6 and 7 and thus favours allophane formation in the B horizons. Accordingly, the soils older than 6000 years classified as silandic Andosols. All had andic properties, except at one of the sites at La Catedral. The seasonal climate and its implications for the soil moisture regime do not seem to determine the transformation of short-range order minerals into crystalline ones. As stated before, the latter seem rather to be favoured by a slower release of the dissolution products of the volcanic glass. As even the oldest soils still have important quantities of noncrystalline clays, no clay illuviation features were observed in these soils. However, in the soil at 2600 masl, crystalline clay minerals clearly dominate and an intensive clay illuviation took place. At this lower altitude seasonality is more pronounced, precipitation is about 300 mm lower and temperature 2e4 C higher. This limits the moisture content in the soil for a longer period than at the sites at >3000 masl. Several authors have noted that in the Transmexican Volcanic Belt, the Andosol phase prevails at high elevation for long periods, while below 2700 m Cambisols, Luvisols and Lixisols are the dominating soils (Werner, 1979; Miehlich, 1991). 5. Conclusions In the studied chronosequence, all sites younger than 30,500 years have received recurrent ash depositions over time. However the amount of ash has varied, as well as the effects of the ash inputs: at most sites recurrent but small inputs of ash led to a thickening of the soil, without interrupting pedogenesis, due to bioturbation. Only at one site the new deposit buried the former soil surface with more than 1 m tephra, whereby soil development at this site corresponds to the age of the last deposit. Rates of all pedogenic processes fit linear models during this time span (10,000 years). Once recurrent volcanic activity ceases, erosion and colluviation processes apparently dominate over pedogenesis, leading to decreases in soil thickness. As a result, the total amounts of organic carbon decrease and the rates of crystalline clay mineral and iron oxide neoformations diminish notably. The amounts of short-range

13

order minerals decrease as recurrent inputs of the precursors for their formation cease. Nevertheless, time is still the dominant driving force of pedogenesis; the mineralogical and chemical composition of the tephras seems to play a minor role in determining its direction and intensity. It still remains to be clarified to what extent the presence of halloysite and kaolinite at the soils older than 30,000 years was affected by drier and cooler climate conditions. Acknowledgements This study was financed by the DGAPA-PAPIIT program of the National Autonomous University of Mexico (grant IN-225703). ~ a-Ramírez received a stipend (grant 186228) from the Víctor Pen Consejo Nacional de Ciencia y Tecnología. MSc. Kumiko Shimada and Dr. Lucy Mora Palomino helped with the physical and chemical soil analyses, and Dr. Teresa Pi with the X-ray diffraction analyses. Daniel Pinales elaborated Fig. 1. We also thank two unknown reviewers for their constructive comments. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.quaint.2014.11.032. References ~ ero, L., Hooghiemstra, H., Cleef, A.M., van Geel, B., 2005. Holocene Almeida-Len climatic and environmental change from pollen records of lakes Zempoala and Quila, central Mexican highlands. Review of Palaeobotany and Palynology 136, 63e92. Arnalds, O., Thorarinsdottir, E.F., Thorsson, J., Waldhauserova, P.D., Agustsdottir, A.M., 2013. An extreme wind erosion event of the fresh Eyjafjallajokull 2010 volcanic ash. Scientific Reports 3 (1257), 1e7. Basile-Doelsch, I., Amundson, R., Stone, W.E.E., Masiello, C.A., Bottero, J.Y., Colin, F., Masin, F., Borschneck, D., Meunier, J.D., 2005. Mineralogical control of organic union. European Journal of Soil carbon dynamics in a volcanic ash soil on La Re Science 56, 689e703. Basile-Doelsch, I., Amundson, R., Stone, W.E.E., Borschneck, D., Bottero, J.Y., Moustier, S., Masin, F.D., Colin, F., 2007. Mineral control of carbon pools in a volcanic soil horizon. Geoderma 137, 477e489. Bertrand, S., Fagel, N., 2008. Nature, origin, transport and deposition of Andosol parent material in south-central Chile (36e42 S). Catena 73, 10e22. Blatter, D.L., Carmichael, I.S.E., Deino, A.L., Renne, P.R., 2001. Neogene volcanism at the front of the central Mexican volcanic belt: basaltic andesites to dacites, with contemporaneous shoshonites and high-TiO2 lava. Geological Society of America Bulletin 113, 1324e1342. n de minerales Besoain, E., 1974. Consideraciones generales sobre la formacio  n de secundarios en los andosoles esquema de una secuencia de meteorizacio nicas. Anales de Edafología y Agrobiología 32, 343e357 (in las cenizas volca Spanish). Birkeland, P.W., 1990. Soil geomorphic research a selective overview. Geomorphology 3, 207e224. Birkeland, P.W., 1992. Quaternary soil chronosequences in various environments: extremely arid to humid tropical. In: Martini, I.P., Chestworth, W. (Eds.), Weathering, Soils and Paleosols, Developments in Earth Surface Processes. Elsevier Science Series, Amsterdam, pp. 261e281. Blakemore, L.C., Searle, P.L., Daly, B.K., 1987. Methods for chemical analysis of soils. New Zealand Soil Bureau Scientific Report 80, 103. Bloomfield, K., 1975. A late-Quaternary monogenetic volcano field in central Mexico. Geologische Rundschau 64, 476e497. Blume, H.P., Schwertmann, U., 1969. Genetic evolution of profile distribution of aluminium, iron and manganese oxides. Soil Science Society of American Proceedings 33, 438e444. Bockheim, J.G., 1980. Solution and use of chronofunctions in study soil development. Geoderma 24, 71e85. Buurman, P., Garcia Rodeja, E., Martinez Cortizas, A., van Doesburg, J.D.J., 2004. Stratification of parent material in European volcanic and related soils studied by laser-diffraction grain-sizing and chemical analysis. Catena 56 (1e3), 127e144. zquez-Selem, L., Ortega, B., 2010. Evidencias de Caballero, M., Lozano-García, S., Va tico y ambiental en registros glaciales y en cuencas lacustres del cambio clima xico durante el último m centro de Me aximo glacial. Boletín de la Sociedad gica Mexicana 62 (3), 359e377 (in Spanish, with English Abstract). Geolo Chadwick, O.A., Chorover, J., 2001. The chemistry of pedogenic thresholds. Geoderma 100, 321e353.

~ a-Ramírez, V., et al., Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene Please cite this article in press as: Pen age in Central Mexico, Quaternary International (2014), http://dx.doi.org/10.1016/j.quaint.2014.11.032

14

~ a-Ramírez et al. / Quaternary International xxx (2014) 1e15 V. Pen

Crews, T.E., Kirayama, K., Fownes, J., Riley, R., Herbert, D., Mueller-Dombois, D., Vitousek, P., 1995. Changes in soil phosphorous fractions and ecosystem dynamics across a long chronosequence in Hawaii. Ecology 76 (5), 1407e1424. Crocker, R.L., Major, J., 1955. Soil development in relation to vegetation and surface age at Glacier Bay, Alaska. Journal of Ecology 43, 427e448. €tze, M., Alioth, L., 2008. Clay Egli, M., Nater, M., Mirabella, A., Raimondi, S., Plo minerals oxyhydroxide formation, element leaching and humus development in volcanic soils. Geoderma 143, 110e114. FAO, 2006. Guidelines for Soil Description, fourth ed. Food and Agriculture Organization of the United Nations, p. 97.  n clima tica de Ko €ppen, García, E., 1981. Modificaciones al sistema de clasificacio n offset. Lario adaptado para las condiciones de la República Mexicana, 3ª edicio Editor S.A. pp 252 (in Spanish). Hall, R.D., 1999. Effects of climate change on soils in glacial deposits, Wind River Basin, Wyoming. Quaternary Research 51, 248e261. Harden, J.W., 1982. A quantitative index of soil development from field description: examples from a chronosequence in central California. Geoderma 28, 1e28. 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 Research 35, 383e399. €ren Glazialmorphologie mexikanischer Heine, K., 1975. Studien zur jungquarta Vulkane e mit einem Ausblick auf die Klimaentwicklung. Das Mexiko-Projekt der DFG VII, Wiesbaden, pp. 1e178, Steiner (in German). Herrmann, L., Jahn, R., Stahr, K., 1996. Identification and quantification of dust additions in peri-saharan soils. In: Guerzoni, S., Chester, R. (Eds.), The Impact of Desert Dust across the Mediterranean. Kluwer Academic Publishers, pp. 173e183. Holzschuh, G.M., 2004. Chronosequence of Chemical Weathering of Basaltic Soil in San Francisco Volcanic Field, Arizona (M.S. thesis). Northern Arizona University, USA. Howard, J.L., Amos, D.F., Daniels, W.L., 1993. Alluvial soil chronosequence in the Inner Coastal Plain. Central Virginia. Quaternary Research 39, 201e213. Huggett, R.J., 1998. Soil chronosequences, soil development, and soil evolution: a critical review. Catena 32, 155e172. Inbar, M., Lugo-Hubp, J., Villers-Ruiz, L., 1994. The geomorphological evolution of the Paricutin cone and lava flows, Mexico, 1943e1990. Geomorphology 9, 57e76. Instituto Mexicano de la Tecnología del Agua (IMTA), 2000. Extractor R apido de n Climatolo gica 1920e1998: ERIC II. Instituto Mexicano de la TecInformacio nología del Agua, Juitepec, Morelos (in Spanish). €hling, D.M., 2007. Non-classical types of loess. Sedimentary GeIriondo, M.H., Kro ology 202 (3), 352e368. €nvold, K., Jackson, M.G., Oskarsson, N., Trønnes, R.G., McManus, J.F., Oppo, D.W., Gro Hart, S.R., Sachs, J.P., 2005. Holocene loess deposition in Iceland: evidence for millennial-scale atmosphere-ocean coupling in the North Atlantic. Geology 33, 509e512. Jahn, R., Stahr, K., 1996. Development of soil and site qualities on basic volcanoclastics with special reference to semiarid environment of Lanzarote, Canary gicas 13 (1), 104e112. Islands, Spain. Revista Mexicana de Ciencias Geolo Jenny, H., 1980. The Soil Resource, Origin and Behavior. Springer, New York, p. 377. Karlstrom, E.T., 1988. Rates of soil formation on Black Mesa, northeast Arizona: a chronosequence in late Quaternary alluvium. Physical Geography 9, 301e327. Kitayama, K., Mueller Dombois, D., 1995. Vegetation changes along gradients of long-term soils development in the Hawaiian montane rainforest zone. Vegetatio 120 (1), 1e20. Kitayama, K., Mueller-Dombois, D., Vitousek, P.M., 1995. Primary succession of the Hawaiian montane rainforest on a chronosequence of eight lava flows. Journal of Vegetation Science 6, 211e222. Krasilnikov, P.V., Targulian, V.O., 2007. Soil system and pedogenic processes: selforganization, time scales, and environmental significance. Catena 71 (3), 373e381. Lasaga, A.C., 1998. Kinetic Theory in the Earth Sciences. Princeton University Press, New Jersey, p. 811. Levine, E.R., Ciolkosz, E.J., 1983. Soil development in till of various ages in northeastern Pennsylvania. Quaternary Research 19, 85e99. Lozano-García, S., V azquez-Selem, L., 2005. A high elevation Holocene pollen record from Iztaccíhuatl volcano, central Mexico. The Holocene 15 (3), 329e338. Manner, H.I., Morrison, R.J., 1991. A temporal sequence chronosequence of soil carbon and nitrogen development after phosphate mining on Nauru Island. Pacific Science 45, 400e404. McFadden, L.D., Weldon II, R.J., 1987. Rates and processes of soil development on Quaternary terraces in Cajon Pass, California. Geological Society of America Bulletin 98, 280e293. McFadden, L.D., Wells, S.G., Jercinovich, M.J., 1987. Influences of eolian and pedogenic process on the origin and evolution of desert pavements. Geology 15, 504e508. 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. Geological Society of America Bulletin 104, 1456e1470. Miehlich, G., 1984. Chronosequenzen und anthropogene Ver€ anderungen andesi€den eines randtropischen Gebirges (Sierra Nevada de tischer Vulkanaschebo xico) (Postdoctoral thesis), Hamburg, Germany, p. 402 (in German). Me Miehlich, G., 1991. Chronosequences of Volcanic Ash Soils, 15. Hamburger Bondenkundliche Arbeiten, Hamburg, Germany, p. 207.

Mizota, C., van Reeuwijk, L.P., 1989. Clay mineralogy and Chemistry of Soils Formed in Volcanic Material in Diverse Climatic Regions. In: Soil Monograph 2. International Soil Reference and Information Centre (ISRIC), Wageningen, The Netherlands, p. 185. rrez, C.E., 1986. Características Mooser, F., Tamez, E., Santoyo, E., Holguín, E., Gutie gicas y geote cnicas del Valle de Me xico. Comisio n de Vialidad y Transporte geolo xico D.F., 32 pp. Urbano, Me Nahon, D.B., 1991. Introduction to the Petrology of Soils and Chemical Weathering. Wiley, New York, p. 313. Nanzyo, M., Yamasaki, S., Honna, T., Yamada, I., Shoji, S., Takahashi, T., 2007. Changes in element concentrations during Andosol formation on tephra in Japan. European Journal of Soil Science 58, 465e477. Nieuwenhuyse, A., Verburg, P.S.J., Jongmans, A.G., 2000. Mineralogy of a soil chronosequence on andesitic lava in humid tropical Costa Rica. Geoderma 98, 61e82. Osete, M.L., Ruíz-Martínez, V.C., Caballero, C., Galindo, C., Urrutia-Facugauchi, J., Tarling, D.H., 2000. Southward migration of continental volcanic activity in the Sierra de Las Cruces, Mexico, paleomagnetic and radiometric evidence. Tectonophysics 201e215. Panfil, M.S., Gardner, T.W., Hirth, K.G., 1999. Late Holocene stratigraphy of the petl volcano, central Tetimpa archaeological sites, northeast flank of Popocate Mexico. Geological Society of America Bulletin 111, 204e218. ~ a-Ramírez, V.M., 2013. Dina mica de la calidad de sitio a trave s de una cronoPen secuencia volc anica y sus implicaciones para la productividad forestal (PhD gicas, Instituto de Geología, Universidad thesis). Posgrado en Ciencias Biolo noma de Me xico, 215 p. (in Spanish). Nacional Auto ~ a-Ramírez, V.M., V Pen azquez-Selem, L., Siebe, C., 2009. Soil organic carbon stocks and forest productivity in volcanic ash soils of different age (1835e30,500 years B.P. in Mexico. Geoderma 149, 224e234. Phillips, J., 1993. Progressive and regressive pedogenesis and complex soil evolution. Quaternary Research 40, 169e176. Ramos-Arroyo, Y.R., Siebe, C., 2007. Weathering rates of sulphide minerals and trace element speciation in tailings of the Guanajuato Mining District, Mexico. Catena 71 (3), 497e507. Sauer, D., Wagner, S., Brückner, H., Scarciglia, F., Mastronuzzi, G., Sthar, K., 2010. Soil development on marine terraces near Metaponto (Gulf of Taranto, southern Italy). Quaternary International 222, 48e63. Sauer, D., Finke, P., Sørensen, R., Sperstad, R., Shülli-Maurer, I., Høeg, H., Sthar, K., 2012. Testing a soil development model against southern Norway soil chronosequences. Quaternary International 265, 18e31. Scarciglia, F., Pelle, T., Pulice, L., Robustelli, G., 2014. A comparison of Quaternary soils chronosequence from the Ionian and Tyrrhenian coast of Calabria, southern Italy: rates of soil development and geomorphic dynamics. Quaternary International (in press). Schaetzl, R.J., Anderson, S., 2005. Soils: Genesis and Geomorphology. Cambridge University Press, New York, USA, p. 817. Schmincke, H.U., 2004. Volcanism. Springer Verlag, Berlin-Heidelberg, p. 324. Schnoor, J.L., 1990. Kinetics of chemical weathering: a comparison of laboratory and field weathering rates. In: Stumm, W. (Ed.), Aquatic Chemical Kinetics: Reaction Rates of Processes in Natural Waters. Wiley, New York, pp. 475e504. Schwertmann, U., 1964. The differentiation of iron oxides in soils by extractions €hrung und Bodwith ammonium oxalate solution. Zeitschrift für Pflanzenerna enkunde 105, 194e202 (in German). mez, E., Gonz Sedov, S., Solleiro-Revolledo, E., Gama-Castro, J., Vallejo-Go alezzquez, A., 2001. Buried paleosols of the Nevado de Toluca: an alternative Vela record of Late Quaternary environmental change in central Mexico. Journal of Quaternary Science 16 (4), 375e389. Sedov, S.N., Solleiro-Rebolledo, E., Gama-Castro, J.E., 2003. Andosol to Luvisol evolution in Central Mexico: timing, mechanisms and environmental setting. Catena 54, 495e513. Shoji, S., Nanzyo, M., Dahlgren, R., 1993. Volcanic Ash Soils: Genesis, Properties and Utilization. In: Developments in Soil Science, 21. Elsevier Science Publishers B.V., Amsterdam, The Netherlands, p. 288. Siebe, C., Abrams, M., Macías, J.L., Obenholzner, J., 1996. Repeated volcanic disasters petl, central Mexico: past key to the future? in Prehispanic time at Popocate Geology 24, 399e402. Siebe, C., Rodriguez-Lara, V., Schaaf, P., Abrams, M., 2004. Radiocarbon ages of Holocene Pelado, Guespalapa, and Chichinautzin scoria cones, south of Mexico City: implications for archaeology and future hazards. Bulletin of Volcanology 66, 203e225. Siebe, C., Arana-Salinas, L., Abrams, M., 2005. Geology and radiocarbon ages of loc, Tlacotenco, Cuauhtzin, Hijo del Cuauhtzin, Teuhtli, and Ocusacayo Tla monogenetic volcanoes in the central part of the Sierra Chichinautzin, Mexico. Journal of Volcanology and Geothermal Research 141 (3e4), 225e243. Siebe, C., Macías, J.L., 2006. Volcanic hazards in the Mexico City metropolitan area petl, Nevado de Toluca, and Jocotitl from eruptions at Popocate an stratovolcanoes and ontogenetic scoria cones in the Sierra Chichinautzin Volcanic Field. Penrose Conference Field Trip Guide. Geological Society of America, pp. 253e329. Siregar, A., Kleber, M., Mikutta, R., Jahn, R., 2005. Sodium hypochlorite oxidation reduces soil organic matter concentrations without affecting inorganic soil constituents. European Journal of Soil Science 56, 481e490. Soil Survey Staff, 1999. US Soil Taxonomy. A Basic System of Classification for Making and Interpreting Soil Surveys. Agriculture Handbook N 436. United States Department of Agriculture, Washington D.C.

~ a-Ramírez, V., et al., Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene Please cite this article in press as: Pen age in Central Mexico, Quaternary International (2014), http://dx.doi.org/10.1016/j.quaint.2014.11.032

~ a-Ramírez et al. / Quaternary International xxx (2014) 1e15 V. Pen Stevens, P.R., Walker, T.W., 1970. The chronosequence concept and soil formation. Quaternary Review of Biology 45 (4), 333e350. Sverdrup, H., Warfvinge, P., 1995. Estimating field weathering rates using laboratory kinetics. In: White, A.F., Brantley, S.L. (Eds.), Chemical Weathering Rates of Silicate Minerals, Reviews in Mineralogy and Geochemistry, 31. Mineralogical Society of America, Washington D.C., pp. 485e542. Swodoba-Colberg, N.G., Drever, J.D., 1993. Mineral dissolution rates in plot-scale field and laboratory experiments. Chemical Geology 105, 51e69. Tate, K.R., Theng, B.K.G., 1980. Organic matter and its interactions with inorganic soil constituents. In: Theng, B.K.G. (Ed.), Soils of Variable Charge. New Zealand Society of Soil Science, Soil Bureau, Department of Scientific and Industrial Research, Lower Hutt, New Zealand, pp. 225e252. Tonneijck, F.H., Jansen, B., Nierop, K.G.J., Verstraten, J.M., Sevink, J., De Lange, L., 2010. Towards understanding of carbon stocks and carbon stabilisation in volcanic ash soils in Andean ecosystems of northern Ecuador. European Journal of Soil Science 61, 392e405. Torn, M.S., Trumbore, S.E., Chadwick, O.A., Vitousek, P.M., Hendricks, D.M., 1997. Mineral control of soil organic carbon storage and turnover. Nature 398, 170e173. Van den Bygaart, A.J., Protz, R., 1995. Soil genesis on a chronosequence, Pinery Provincial Park, Ontario. Canadian Journal of Soil Science 75, 63e72. Van Reeuwijk, L.P., 1992. Procedures for Soil Analysis, third ed. International Soil Reference and Information Centre (ISRIC), Wageningen, The Netherlands (Technical Paper No. 9). Vaughan, K.L., 2008. Pedogenesis at Craters of the Moon National Monument and Preserve, Idaho, USA (Ph.D. thesis). University of Idaho, USA, p. 156. V azquez-Selem, L., Heine, K., 2011. Late Quaternary glaciation in Mexico. In: Ehlers, J., Gibbard, P.L., Hughes, P.D. (Eds.), Quaternary Glaciations e Extent and Chronology. A Closer Look, Developments in Quaternary Science, 15. Elsevier, Amsterdam, pp. 849e861.

15

Vitousek, P.M., Chadwick, O.A., Crews, T.E., Fownes, J.H., Hendricks, D.M., Herbert, D., 1997. Soil and ecosystem development across the Hawaiian Islands. GSA Today 7, 1e8. Vreeken, J.W., 1975. Principal kinds of chronosequences and their significance in soil history. Journal of Soil Science 26, 378e394. Wada, K., 1980. Mineralogical characteristics of andisols. In: Theng, B.K.G. (Ed.), Soils with Variable Charge. New Zealand Society of Soil Science, Soil Bureau, Department of Scientific and Industrial Research, Lower Hutt, New Zealand, pp. 87e107. White, A.F., Brabtley, S.L., 2003. The effect of time on the weathering of silicate minerals: why do weathering rates differ in the laboratory and field? Chemical Geology 202, 479e506. Werner, G., 1979. Bodengesellschaften im Zentralen Hochland von Mexico. Catena 6, 331e347. Yaalon, D.H., 1971. Soil forming processes in time and space. In: Yaalon, D.H. (Ed.), Paleopedology. Israel University Press, Jerusalem, pp. 29e39. Yaalon, D.H., 1975. Conceptual models in pedogenesis: can soil forming-functions be solved? Geoderma 14, 189e205. Yaalon, D.H., 1997. Comments on the source, transport and deposition scenario of Saharan dust of southern Europe. Journal of Arid Environments 36, 193e196. Yagasaki, Y., Mulder, J., Okazaki, M., 2006. The role of soil organic matter and shortrange ordered aluminosilicates in controlling the activity of aluminum in soil solutions of volcanic ash soils. Geoderma 137, 40e57. 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, 317e330. Zehetner, F., 2010. Does organic carbon sequestration in volcanic soils offset volcanic CO2 emissions? Quaternary Science Reviews 29, 1313e1316.

~ a-Ramírez, V., et al., Rates of pedogenic processes in volcanic landscapes of late Pleistocene to Holocene Please cite this article in press as: Pen age in Central Mexico, Quaternary International (2014), http://dx.doi.org/10.1016/j.quaint.2014.11.032