Quaternary Geochronology 16 (2013) 129e143

Contents lists available at SciVerse ScienceDirect

Quaternary Geochronology

journal homepage: www.elsevier.com/locate/quageo

Research paper Amino acid racemization in four species of ostracodes: Taxonomic, environmental, and microstructural controls

José E. Ortiz a,*, Trinidad Torres a, Alfredo Pérez-González b a Laboratory of Biomolecular Stratigraphy, E.T.S.I. Minas, Universidad Politécnica de Madrid, C/Ríos Rosas 21, Madrid 28003, Spain b Centro Nacional de Investigación sobre Evolución Humana, Pso Sierra de Atapuerca s/n, 09002 Burgos, Spain article info abstract

Article history: Here we quantified the aspartic acid and glutamic acid racemization rates of the four main ostracode Received 9 September 2011 species (Herpetocypris reptans, Candona neglecta, Ilyocypris gibba and Cyprideis torosa) present in several Received in revised form Iberian Peninsula localities covering a wide chronological range (ca. 1 Ma to present). At low D/L values 27 September 2012 (at Asp D/L < 0.40; and Glu D/L ¼ 0.09e0.18), H. reptans racemized at higher rates than C. neglecta, Accepted 2 November 2012 C. torosa and I. gibba. In contrast, for Asp D/L > 0.4 and Glu D/L > 0.18, H. reptans, C. neglecta and C. torosa Available online 17 November 2012 showed similar racemization rates. I. gibba exhibited the lowest D/L values in old samples (Middle and Lower Pleistocene). We attribute these differences in amino acid racemization rates mainly to variations Keywords: Ostracodes in valve protein composition. We found that the microstructure of the valves of each species (size, Amino acid racemization morphology, and arrangement of crystals) differed, but did not appear to change over time (at least for Microstructure the last ca. 1 Ma). Such differences may also be linked to the type of proteins involved in the respective Iberian Peninsula calcification processes of these organisms. On the basis of our results, and given that other studies have demonstrated that the majority of inter-crystalline proteins are leached early after death (a few centuries or millennia), we propose that the degradation rates of the most resistant inter- and intra-crystalline proteins in each species differ depending on the protein composition of the valves. Although further research is required, we suggest that amino acid racemization in each ostracode species might be related to valve microstructure. Ó 2012 Elsevier B.V. All rights reserved.

1. Introduction bed can be analyzed, thereby enabling easier identification of anomalous results, and estimation of the time-averaging of the Ostracode valves are made of low-magnesium calcite (Kesling, dated event. Recent studies by Bright and Kaufman (2011a,b) and 1951; Sohn, 1958), which is a more stable mineral than aragonite. Reichert et al. (2011) shed light on the processes that affect AAR in The characteristics of these valves make them particularly useful ostracode valves and conclude that D/L values are influenced by for amino acid racemization (AAR) dating purposes (McCoy, 1988; taxonomy, temperature and environmental pH. Oviatt et al., 1999; Kaufman, 2000, 2003; Ortiz et al., 2004, 2009; Given that AAR is a genus-dependent process (Lajoie et al., 1980) Torres et al., 2005; Colman et al., 2006; Owen et al., 2007; Jayko (as well as being time and temperature-dependent), monogeneric et al., 2008; Bright et al., 2010; De Santis et al., 2010; Bright and samples are required to reduce taxonomically controlled differ- Kaufman, 2011a). They show excellent preservation of amino ences in D/L values. However, in some previous studies, various acids (Kaufman and Manley, 1998; Kaufman, 2000; Ortiz et al., ostracode genera have been analyzed together from the same 2002), and the amino acid abundance allows the analysis of horizons/localities in order to establish age, with the assumption a small sample size (even a single ostracode valve z 0.01 mg). The that there are only small differences between D/L values of distinct sample size required for ostracodes (in many cases just one valve) is genera (cf. Oviatt et al., 1999; Kaufman et al., 2001; Kaufman, 2003; much lower than for other organisms, such as mollusks (5 mg; Ortiz et al., 2004, 2009; Torres et al., 2005; De Santis et al., 2010). It Lajoie et al., 1980; Goodfriend and Mitterer, 1988; Wehmiller, 1990, has been demonstrated that for Candona and Limnocythere, which 2000; Goodfriend, 1991; Torres et al., 1997; Ortiz et al., 2002; belong to separate phylogenetic ostracode groups, the D/L values of Penkman et al., 2008). Thus, many ostracode samples from a single valves from the same horizons (Table 1) showed only slight differences (McCoy, 1988; Oviatt et al., 1999; Kaufman, 2000, 2003; * Corresponding author. Tel.: þ34 913366970. Kaufman et al., 2001; Reichert et al., 2011). Nevertheless, Bright and E-mail address: [email protected] (J.E. Ortiz). Kaufman (2011a) reported that the abundance and relative

1871-1014/$ e see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.quageo.2012.11.004 130 J.E. Ortiz et al. / Quaternary Geochronology 16 (2013) 129e143

Table 1 Classification of modern ostracodes used in this study. Compiled using Meisch (2000) and the Integrated Taxonomic Information System. On-line database. www.itis.gov/ index.html. Accessed: July, 15, 2011. Class CRUSTACEA Pennant, 1777

Subclass OSTRACODA Latreille, 1806

Order PODOCOPIDA Muller, 1894

Suborder PODOCOPINA Sars, 1866

Superfamily CYPRIDOIDEA Baird, 1845

Family CYPRIDIDAE Baird, 1845

Subfamily HERPETOCYPRIDINAE Kaufmann, 1900

Genus HERPETOCYPRIS Bradyi and Norman, 1889

Family CANDONIDAE Kaufmann, 1900

Genus CANDONA Baird, 1846

Family ILYOCYPRIDIDAE Kaufmann, 1900

Subfamily ILYOCYPRIDINAE Kaufmann, 1900

Genus ILYOCYPRIS Brady and Norman,

Superfamily CYTHEROIDEA Baird, 1850

Family CYTHERIDEIDAE Sars, 1925

Subfamily CYTHERIDEINAE Sars, 1925

Genus CYPRIDEIS Jones, 1857

proportions of amino acids vary among ostracode genera, even probably recrystallizes to form the calcite layers of the valves when these belong to the same family. The most striking compo- (Rosenfeld, 1982). The secretion of a new valve occurs within a few sitional difference was observed in Heterocypris incongruens (Bright hours of molting (Kesling, 1951; Turpen and Angell, 1971). In all and Kaufman, 2011a), which contained twice the relative propor- arthropods the cuticle consists of two layers (Richards, 1951), tion of aspartic acid (Asp) compared to the other taxa. Bright and namely the epicuticle (the outer part) and the procuticle (the Kaufman (2011b) observed that the racemization rate of Asp is innermost area). higher in Heterocypris valves than in Candona and Ilyocypris ones, The epicuticle is a thin layer that accounts for only 5% of the thereby suggesting a taxonomic effect. Likewise, the same authors whole cuticle (Fig. 1), and its structure is simpler in ostracodes than reported that D/L values differ considerably across family bound- in other arthropods. This layer does not contain chitin, but rather aries in the Family Cypridoidea, thus indicating that these differ- lipids and polyphenols (Jeuniaux et al., 1986). The procuticle ences are linked to taxonomic variability in protein composition. In comprises chitin and is segregated by the epidermic cells before this regard, Kaufman (2003) noted that differences in D/L values are molting takes place (Bate and East, 1972). likely to result from taxon-dependent arrangement of amino acids, From a biochemical, structural, and functional point of view, the or from morphological and structural differences that influence the procuticle shows two distinct layers, namely the endocuticle and differential loss of free and peptide-bound amino acids. exocuticle layers (Fig. 1). In most ostracodes, the latter is thin, Ostracode valves consist of small crystallites of calcium calcified, and pigmented. This layer shows a cross-lattice of chitin carbonate (80e90%, Sohn, 1958) embedded in a chitinous and fibers positioned perpendicular to the surface in the calcareous protein matrix (Bate and East, 1972, 1975; Langer, 1973; Keyser, layer (Bate and East, 1972, 1975; Dépêche, 1982), and also secondary 1982; Rosenfeld, 1979) accounting for 2e15% of the valve weight chitin fibrils without any preferential orientation. The mineralized (Sohn, 1958). Depending on the taxonomy, the microstructure of part of the valve is made up of prisms of calcite disposed in a mosaic the adult valve is expressed in a variety of forms. As in arthropods, of irregular outlines (Langer, 1971, 1973; Dépêche, 1982); however, the main structure of ostracodes consists of a continuous cuticular Jǿrgesen (1970) reported that these prisms are perpendicular to the integument that forms the exoskeleton, which is partly calcified valve surface. According to Dépêche (1982), calcite prisms and (Fig. 1)(Bate and East, 1972, 1975). In ostracodes the integument is fibrils are of the same scale, and fibrils occur both outside and formed by epidermal cells, which periodically secrete crystalline inside calcite crystals (Jǿrgesen, 1970). Chitin layers envelop calcitic layers and a cuticular membrane over the external surface (Bate layers and are highly relevant for the preservation of fossil ostra- and East, 1972, 1975) which comprises mainly chitin and other code valves (Oertli, 1975). proteic material (e.g. arthropodine, resiline, sclerotine and cuticu- The endocuticle is the internal zone of the procuticle and is line), as well as calcite crystals. These materials confer hardness and present in all ostracodes. This layer is calcified and can show high resistance to the valve. A layer of granules secreted by the pigment granules that provide some coloration (Bate and East, epidermis was found to consist of calcite and apatitic calcium 1972, 1975). The endocuticle is made of chitin fibers arranged in orthophosphate (Rosenfeld, 1979, 1982; Bate and Sheppard, 1982; a reticular matrix (feathered-like), although with a certain degree Keyser and Walter, 2004), the main function of which is considered of layering (Bate and Sheppard, 1982). to be the construction of a new calcareous valve during molting. The size, morphology and arrangement of crystals in the This granular layer is present only in living ostracodes, and it microstructure of biomineralized tissues is regulated by proteins, J.E. Ortiz et al. / Quaternary Geochronology 16 (2013) 129e143 131

Fig. 1. Diagramatic section through the anterior ventral part of an ostracode valve (after Kesling, 1951, and Dalingwater and Mutvei, 1990). which are involved in the precipitation of calcium carbonate of Tagus (TAR) drainage basin (Fig. 2), flowing into the Atlantic Ocean carapaces (Albeck et al., 1993; Weiner, 1983, 1984, 1986; Addadi and at Lisbon (Portugal). In this area, distinct aminozones ranging from Weiner, 1985; Aizenberg et al., 1996; Luquet and Marin, 2004; MIS 11 to MIS 1 were established through the AAR values obtained Marin et al., 2008, among others). in H. reptans ostracodes (Torres et al., 2005; Ortiz et al., 2009). Here In brief, ostracode valves show a variety of microstructure we present data for two new localities (GU1.1 and GU1.2; Table 2). patterns and distinct proportions of chitin, protein, and calcite, These sites belong to the same tufa terrace located 5 m above the depending on the family of proteins involved in the formation of current thalweg of the Guadiela River. valves; only calcite crystals, parallel chitinous lamellae together The Blanco River (BR) is a tributary of the Jalón River, which with a layer of crystallites (cf. Keyser and Walter, 2004), or chitinous belongs to the Ebro drainage basin (Fig. 2), which in turn ends in the membranes are included in calcite crystals (Bate and East, 1972). Mediterranean Sea. The low energy conditions and low incision Here we compared the amino acid (Asp and glutamic acid (Glu)) rates produced fluvial terraces (Table 2) of diverse ages, ranging D/L values of the four main ostracode species (Cyprideis torosa, from MIS 7 to MIS 1, at similar elevations and situated between 10 Herpetocypris reptans, Candona neglecta and Ilyocypris gibba)in and 20 m above the current thalweg (Torres et al., 1995). Quaternary continental deposits in the Iberian Peninsula. For this We also sampled two stratigraphic sections of the same fluvial purpose, we selected ostracode valves from 43 localities (Table 2) terrace, consisting of tufa deposits of considerable thickness, in the covering a wide time-span (ca. 1 Ma to Holocene). As the rate at vicinity of Enguídanos (Cuenca Province, Central Spain). This which amino acids racemize depends on various factors, including terrace is linked to the activity of the Cabriel River, a tributary of the taxonomic differences in protein composition, we also performed Júcar River, which ends in the Mediterranean Sea (Fig. 2). These SEM analysis of the valve microstructure of specimens belonging to sections (ENG1 and ENG2) consisted of fluvial sediments, riverine these four species in order to explore relationships between protein tufas and barrage deposits. composition, structural differences, age and D/L values. We also examined whether the taxonomic control reflected in the micro- 2.2. Banyoles Lake structure of the ostracode species influences the racemization rate. Banyoles Lake (BY) (surface area, 1.12 km2; max. depth, 46 m) is 2. Geographical and geological setting located in the northeastern part of the Iberian Peninsula (Fig. 2). This water body is the relict of a much larger Plio-Pleistocene group Ostracode valves were obtained from five major areas (Fig. 2). of karstic lakes, namely the Banyoles-Besalú lacustrine complex Most samples were from tufa deposits deposited in central Spain (Julià, 1980), and it is fed mainly by a large artesian karstic aquifer but belonging to different fluvial systems: Tagus River, Jalón River that developed in Paleogene limestone and gypsum units. A 32 m- and Cabriel River. In addition the following areas were studied: deep core, called La Draga, covering the last 30,000 yr (14C and U/Th Banyoles Lake, the Guadix Baza Basin, the Ambrona archaeological dated) was drilled in the lake (Pérez-Obiol and Julià, 1984). site and the Antas River. 2.3. The Guadix-Baza basin 2.1. Tufa deposits in Central Spain The Guadix-Baza basin (Fig. 2) is an endorheic zone in southern The catchment areas of the fluvial systems studied here are in Spain that covers approximately 4500 km2, with its origin linked to the Iberian Range, which comprises mainly Mesozoic sedimentary the Alpine Orogeny (Soria, 1993). The basin had a centripetal carbonates. Dissolution processes of the carbonate rocks of the depositional model, with alluvial fans at the edges, which gradually drainage areas produce Ca(HCO3)2-rich headwaters, which on passed into a fluvial system that ended in a central system of small reaching the flat basins, the slower flow and development of saline lakes (Torres et al., 2003; Ortiz et al., 2006). macrophytes and algae produced CO2 degassing and consequently Ortiz et al. (2004) provided the chronostratigraphy of a 356 m- the accumulation of tufas (Torres et al., 2005; Ortiz et al., 2009). thick “composite-stratotype-section” of the east domain of the The Henares River (HR) and its tributary the Dulce River (DR), Guadix-Baza basin (CBS). This section ranges from the Plio/Pleis- together with the Cifuentes (CR) and Ruguilla (RR) Rivers, the tocene boundary to the upper part of the Middle Pleistocene. Gárgoles Pond (GP), and the fluvial system of the Guadiela (GU), Likewise, the ages of some palaeontological sites within this basin, Escabas (ES) and Trabaque (TR) Rivers around Priego, belong to the such as Fuente Amarga (FA) (Table 2) were also obtained in Ortiz 132 J.E. Ortiz et al. / Quaternary Geochronology 16 (2013) 129e143

Table 2 Geographical location of the localities. The areas shown in Fig. 2 to which each locality belongs to, are added. Some of them were studied previously but re-sampled here (ES 6.2, TR4.3, TR3.1, CBS-228, CBS-327, CBS-352, FA). HR: Henares River; DR: Dulce River; CR: Cifuentes River; RR: Ruguilla River; GP: Gárgoles de Arriba Pond; TR: Trabaque River; ES: Escabas River; GU: Guadiela River; TAR: Tagus River; BR: Blanco River; SRA: Antas River borehole; ENG: Enguídanos; BY: Banyoles borehole; ABD: Ambrona Doline; ABS: Ambrona archaeological site; CBS: Cúllar-Baza section; and FA: Fuente Amarga. All localities were dated by applying the amino acid racemization method to ostracode valves.

Localities Area Latitude (W) Longitude (N) Elevation (m a.s.l.) Age (ka) References GU1.2 1 219039.800 4033017.600 820 5 1 This paper GP1.4 1 23705800 404504700 875 5 1 Ortiz et al. (2009) RB-80 2 21905400 41110300 980 7 2 This paper RB-81 2 21905400 411101000 980 8 2 This paper RB-82 2 21905200 411102000 960 7 2 This paper DR4.2 1 24404000 405702900 860 7 2 Ortiz et al. (2009) ES6.2 1 22302100 402704000 720 10 5 Torres et al. (2005) GU1.1 1 217012.600 403200.900 845 16 6 This paper BY-4375 4 314024.500 4207037.600 190 20 12 This paper TAR2.3 1 23404100 404104900 730 36 10 Ortiz et al. (2009) BY-21875 4 314024.500 4207037.600 190 36 16 This paper ABS-5 6 229052.300 419039.700 1145 57 15 This paper RB-5 2 220018.300 411008.100 1020 60 11 This paper RB-12 2 219047.700 41110300 1000 65 17 This paper ABS-6 6 229052.300 419039.700 1145 68 20 This paper CR1.1 1 23703500 404602900 890 97 19 Ortiz et al. (2009) ES4.1 1 22004000 402700700 790 99 3 Torres et al. (2005) CR1.11 1 23604700 404205800 785 102 33 Ortiz et al. (2009) ENG2.3 3 136019.400 3940056.200 725 123 29 This paper CR1.5 1 23702700 404305400 810 127 36 Ortiz et al. (2009) ENG1.1 3 136014.700 394203.300 750 129 30 This paper ABD-1 6 229055.800 41905000 1155 130 37 This paper TR4.3 1 21902800 402503200 820 135 15 Torres et al. (2005) ENG2.2 3 136019.400 3940056.200 700 152 22 This paper CR1.8 1 23604000 404301000 775 157 19 Ortiz et al. (2009) SRA2-15 7 14905.200 3712015.700 4.7 171 77 This paper TAR1.1 1 23404100 404101500 760 174 20 Ortiz et al. (2009) CR1.9 top 1 23603300 404204900 781 179 47 Ortiz et al. (2009) SRA1-11 7 149059.300 3713015.100 16.5 208 32 This paper ADB-3 6 22904900 41904900 1155 230 11 This paper ABD-2 6 229057.700 419050.600 1157 246 56 This paper TR4.4 1 22000600 402601800 800 261 14 Torres et al. (2005) TR3.1 1 21904100 402601100 840 273 9 Torres et al. (2005) CBS-352 5 23701000 374602300 920 279 77 Ortiz et al. (2004) CBS-327 5 23602500 374602700 887 389 62 Ortiz et al. (2004) FA 5 23501200 37460700 880 452 72 This paper CBS-310 5 23601700 374603000 880 456 70 This paper CBS-228 5 24405400 373805700 760 863 173 Ortiz et al. (2004)

et al. (2004). Here we present the analysis of ostracode valves of a new sampled horizon (CBS-310, Table 2).

2.4. Ambrona palaeontological site

The Ambrona doline deposits (ABD, Fig. 2) are relicts of former exokarstic forms (dolines) infills located in central Spain. These processes resulted in small palustrine basins (diameter of 3e4m) that show mainly lutite infills. Creek erosion finally captured and opened the infilled dolines, particularly those corresponding to the classical archeological site of Ambrona (ABS), in which abundant Acheulian lithic devices and mammal bones (Equus sp.) were recovered (Howell et al., 1995) and dated at 182.4 18.7 ka through AAR of mammal teeth (Torres et al., submitted for publication). The top of the stratigraphic section of the ABS ends with two lutite beds, which were sampled for this study.

2.5. Antas River

The Antas River (Fig. 2), with a short course, is located in the Almería Province (southeastern Spain). Two borehole cores (15.40- m- and 29.70-m-deep, named SRA1 and SRA2, respectively) were Fig. 2. Geographical location of the localities in this study. 1.1: Tufa terraces of the drilled in its alluvial plain for geotechnical purposes as part of road Tagus River system, 1.2: Tufa terraces of the Blanco River system, 1.3: Tufa terraces of the Cabriel River system, 2: Banyoles Lake; 3: Guadix-Baza Basin; 4: Ambrona site; 5 improvements. We sampled some beds for AAR dating of ostra- Antas River. codes (Table 2). J.E. Ortiz et al. / Quaternary Geochronology 16 (2013) 129e143 133

3. Material and methods alanine (Ala), valine (Val), phenylalanine (Phe), isoleucine (Ile), leucine (Leu), threonine (Thr), arginine (Arg) and tyrosine (Tyr). Valves from C. torosa, C. neglecta, I. gibba and H. reptans species were collected in several localities (Table 2) in the Iberian Peninsula 3.2. Analysis of the ostracode valve microstructure that cover a wide time-span (ca. 1 Ma to Holocene). Some modern specimens were also recovered. We selected ostracodes of various ages (w1 Ma, 450 ka, 200 ka, These species belong to the same Suborder (Table 1), but the 120 ka, 30 ka, 10 ka, and modern) and species (C. torosa, C. neglecta, Cyprideis genus belongs to a distinct Superfamily (Cytheroidea) I. gibba, H. reptans) for analysis by scanning electron microscopy in than the other three, which are included in the same Superfamily order to compare their microstructure with the amino acid content, (Cypridoidea). and to study variations in this structure over time. We also exam- ined the microstructure to test whether differences in the size and 3.1. Amino acid racemization preparation and analysis arrangement of crystals between species are related to racemiza- tion rates. Some data used in this study have already been published Valves were fractured by pressing their center with a needle, (Table 2)(Ortiz et al., 2004, 2009; Torres et al., 2005), although we and the fragments were mounted on stubs for study. Samples were also sampled new localities and re-sampled others. In all cases, covered with an ultrathin coating of gold (electrically conductive samples were taken from beds comprising marls or silts in either material). The valve microstructure was observed under a scanning outcrops or borehole cores. In the first case (outcrops), a 1 m-deep electron microscope (SEM) at the Real Jardín Botánico of Madrid. hole was dug before collecting a 3 kg sample. In the laboratory, the samples were washed and sieved. After drying, the sediment 4. Results remnant (>0.062 mm) was analyzed under a binocular microscope and ostracode valves were picked with the aid of a needle. Ostra- 4.1. Asp and Glu D/L values code valves were carefully cleaned by sonication in distilled deionized (DDI) water and rinsed with DDI water to remove sedi- We analyzed a total of 506 samples of fossil ostracode valves. Of ment. Some valves were also cleaned with a small brush under these, 105 results were rejected (21% of the data) on the basis of the a binocular microscope to eliminate fine debris. In order to remove following two criteria: 1) a cut-off value of 0.8 in the concentration secondary organic molecules adsorbed to the valves, they were of L-Serine (L-Ser) with respect to that of L-Asp; this can determine then submerged in 3% hydrogen peroxide (H2O2) for 2 h following anomalous samples (cf. Kaufman, 2006), as Ser decomposes rapidly Kaufman (2000) and Hearty et al. (2004). This treatment has been and excessive amounts of this amino acid would indicate contam- determined to be the most effective for ostracode valves (Bright and ination of valves by modern amino acids; and 2) Asp and Glu D/L Kaufman, 2011a). We selected translucent specimens only for the values falling outside the 2s range of the group (cf. Hearty et al., analysis to avoid the use of possible contaminated valves (Bright 2004; Kosnik and Kaufman, 2008). Each result and the criterion and Kaufman, 2011a). The number of samples from each locality used for sample rejection are shown in the Appendix. Of these 105 is shown in Tables 3 and 4. Some modern specimens belonging to rejected samples, 39% were rejected on the basis of criterion 1, and the four species (C. torosa, C. neglecta, I. gibba and H. reptans) were 61% on criterion 2. Almost all the latter showed L-Ser/L-Asp values also picked for comparison. close to 0.8, together with Asp and Glu D/L values lower than their Amino acid concentrations and D/L values were quantified using correspondent average values in each locality. We consider that this a high performance liquid chromatograph (HPLC) in the Biomo- observation indicates a slight degree of recent contamination. The lecular Stratigraphy Laboratory, following the sample preparation high permeability characteristic of tufa deposits, which comprised protocol described by Kaufman and Manley (1998) and Kaufman most of the localities studied here, could explain the percentage of (2000). This procedure involves hydrolysis, which was performed samples that appeared to be contaminated. under a N2 atmosphere in 7 mL of 6 M HCl for 20 h at 100 C. The Some previously studied localities (ES 6.2, TR4.3, TR3.1, CBS-228, hydrolyzates were evaporated to dryness in vacuo and then rehy- CBS-327, CBS-352, FA) were re-sampled. The new data is indicated drated in 7 mL 0.01 M HCl with 1.5 mM sodium azide and 0.03 mM L- in the Appendix and therefore the D/L values provided here update homo-arginine (internal standard). those already published, and in all cases they did not vary appre- Samples were injected into an Agilent-1100 HPLC equipped with ciably (Tables 3 and 4). Some of the original values were obtained a fluorescence detector. Excitation and emission wavelengths were using gas-chromatography (GC) (Ortiz et al., 2004; Torres et al., programmed at 335 nm and 445, respectively. A Hypersil BDS C18 2005); however, these values can be compared with the Asp and reverse-phase column (5 mm; 250 4 mm i.d.) was used for the Glu D/L values that we obtained by means of HPLC because simi- separation. larities have been reported in inter-laboratory comparison exer- Derivatization occurred before injection by mixing the sample cises (cf. Wehmiller, 1984; Torres et al.,1997; Wehmiller et al., 2010) (2 mL) with the pre-column derivatization reagent (2.2 mL), which and between several ostracode samples analyzed by GC and HPLC comprised 260 mM isobutyryl-L-cysteine (chiral thiol) and 170 mM in our laboratory (cf. Ortiz et al., 2009, p. 955, see Supplementary o-phthaldialdehyde, dissolved in 1.0 M potassium borate buffer Data-Fig. 1). solution at pH 10.4. Eluent A consisted of 23 mM sodium acetate We selected Asp and Glu because they account for an important with 1.5 mM sodium azide and 1.3 mM EDTA, adjusted to pH 6.00 percentage of the amino acid content of most ostracode valves with 10 M sodium hydroxide and 10% acetic acid. Eluent B was (Table 5, Fig. 3), confirming the results of Kaufman (2000) and HPLC-grade methanol, and eluent C consisted of HPLC-grade Bright and Kaufman (2011a). The mean D/L values of these two acetonitrile. A linear gradient was performed at 1.0 mL/min and amino acids in the four ostracode species examined (C. torosa, 25 C, from 95% eluent A and 5% eluent B upon injection to 76.6% H. reptans, C. neglecta, I. gibba) from the study sites are shown in eluent A, 23% eluent B, and 0.4% eluent C at 31 min. This approach Tables 3 and 4 and their estimated ages appear in Table 2. allowed the separation of Asp, Glu and serine (Ser). For the analysis of modern valves and a subsection of the Pleistocene material 4.1.1. Dating of localities (RB-5, ES4.1, TR4.3, TR4.4, FA, CBS-228), a longer run-time (to Some localities were previously dated (Table 2) by AAR 87 min) was used to separate additional amino acids: glycine (Gly), measurements of ostracode valves (Ortiz et al., 2004, 2009; Torres 134 J.E. Ortiz et al. / Quaternary Geochronology 16 (2013) 129e143

Table 3 Asp D/L values of the valves of four ostracode species from the sampled localities. The numbers in parentheses are the number of samples analyzed. Errors represent the standard deviation of Asp D/L values.

Localities C. torosa I. gibba C. neglecta H. reptans GU1.2 ee0.130 0.032(3) 0.140 0.042(2) GP1.4 ee0.156(1) 0.160 0.009(5) RB-80 ee0.149(1) 0.173 0.012(5) RB-81 ee0.151(1) 0.171 0.001(2) RB-82 ee0.150 0.006 (3) 0.161 0.000(3) DR4.2 ee0.161 0.009(3) 0.179 0.003(4) ES6.2 0.168 0.008(5) e 0.218(1) 0.218 0.014(4) GU1.1 ee0.190 0.010(3) 0.213 0.025(7) BY-4375 0.229 0.004(4) e 0.233 0.011(2) e TAR2.3 e 0.237 0.007(6) 0.239 0.007(6) 0.267 0.018(7) BY-21875 0.249 0.004(5) e 0.270 0.006(4) e ABS-5 e 0.279 0.014(3) 0.287 0.025(3) e RB-5 ee0.239 0.005(10) 0.288 0.010(12) RB-12 ee0.258 0.011(2) 0.299 0.018(4) ABS-6 e 0.275 0.025(4) 0.294 0.027(8) e CR1.1 ee0.289 0.013(6) 0.333 0.012(8) ES4.1 e 0.247 0.006(10) 0.285 0.015(9) 0.338 0.013(10) CR1.11 ee0.296 0.016(5) 0.341 0.012(5) ENG2.3 0.369 0.013(4) e 0.382 0.029(7) 0.375(1) CR1.5 ee0.345 0.037(8) 0.386 0.024(8) ENG1.1 0.334(1) ee0.371 0.009(6) ABD-1 e 0.340(1) 0.325(1) e TR4.3 ee0.346 0.015(7) 0.351 0.024(14) ENG2.2 ee0.341(1) 0.369 0.010(3) CR1.8 ee0.333 0.009(5) 0.398 0.012(6) SRA2-15 0.357 0.012(5) 0.360 0.008(4) e 0.366 0.047(5) TAR1.1 ee0.324 0.021(9) 0.409 0.006(5) CR1.9 top ee0.370 0.013(6) 0.420 0.035(10) SRA1-11 0.410 0.023(4) ee0.422 0.023(2) ABD-3 e 0.365 0.015(2) 0.442(1) 0.437(1) ABD-2 e 0.427 0.031(5) 0.452(1) e TR4.4 e 0.373 0.011(7) 0.448 0.011(7) 0.477 0.030(10) TR3.1 e 0.376(1) e 0.463 0.026(8) CBS-352 0.429 0.023(9) 0.321 0.016(12) 0.373 0.011(8) 0.440 0.033(7) CBS-327 0.467 0.026(14) e 0.455 0.028(9) 0.472 0.033(6) FA 0.562 0.030(8) 0.573 0.026 (7) 0.636 0.021(7) 0.579 0.001(2) CBS-310 0.570 0.021(8) e 0.607 0.018(6) 0.600 0.040(6) CBS-228 0.711 0.029(9) 0.628 0.019(7) 0.699 0.041(7) 0.658 0.017(14)

et al., 2005). The numerical ages of these localities and the new ones presented here were calculated by introducing the Asp and pffiffi 1 þ D=L For Asp : t ¼3:586 þ 19:745 ln (3) Glu D/L values for C. torosa and H. reptans valves into the age 1 D=L calculation algorithms established by Ortiz et al. (2004) for these species in the central and southern part of the Iberian Peninsula. and: These algorithms were achieved by calibration with independently 14 dated samples through C, U/Th, palaeomagnetism, and AAR of pffiffi 1 þ D=L continental gastropods. For Middle and Lower Pleistocene samples For Glu : t ¼3:186 þ 58:972 ln (4) 1 D=L (with D/L Asp > 0.401 and D/L Glu > 0.140), the best fit for Glu, was obtained between the logarithmic transformation of D/L values The estimated age of a single horizon (Table 2) is the average of with respect to time (Equation (1)), whereas for Asp the best fitwas the age estimates obtained for each amino acid D/L value of the achieved with the square root of time (Equation (2)). samples from that bed, after rejection of anomalous values (see Appendix). The age uncertainty quoted in this study is the standard 1 þ D=L deviation of all the age estimates obtained from the Asp and Glu D/L For Glu : t ¼39:59 þ 622:25 ln (1) values of the samples of each locality. 1 D=L 4.1.2. Extent of Asp racemization in valves of ostracode species of pffiffi 1 þ D=L different ages For Asp : t ¼2:666 þ 18:027 ln (2) Plots showing the relationships between mean Asp and Glu D/L 1 D=L values vs. time for the various species are shown in Fig. 4. Herpe- For younger samples (Upper Pleistocene) with D/L Asp < 0.401 tocypris and Candona were the most commonly analyzed genera in and D/L Glu < 0.140, equations (3) and (4) were used. This approach the Iberian localities. is justified because of the “non-linear” behavior of racemization H. reptans valves showed slightly higher Asp D/L values than (Goodfriend and Mitterer, 1988; Goodfriend, 1991), which, in this Candona specimens (0.01e0.02) in younger localities (D/L < 0.25; case (ostracode valves of the Iberian Peninsula ranging from ca. <40 ka), while for intermediate values ranging from 0.25 to 0.40 1 Ma to present) consists of the combination of at least two func- (ca. 40e200 ka) the difference increased to 0.05 or 0.08, with the tions with distinct slopes for Asp and Glu (cf. Ortiz et al., 2004). exceptions of ENG2.3 and TR4.3. In older samples, the differences in J.E. Ortiz et al. / Quaternary Geochronology 16 (2013) 129e143 135

Table 4 Glu D/L values of the valves of four ostracode species from the sampled localities. The numbers in parentheses are the number of samples analyzed. Errors represent the standard deviation of Glu D/L values.

Localities C. torosa I. gibba C. neglecta H. reptans GU1.2 ee0.035 0.004(3) 0.035 0.004(2) GP1.4 ee0.045(1) 0.041 0.004(5) RB-80 ee0.046(1) 0.040 0.002(5) RB-81 ee0.042(1) 0.036 0.007(2) RB-82 ee0.038 0.003(3) 0.036 0.001(3) DR4.2 ee0.045 0.003(3) 0.045 0.008(4) ES6.2 0.044 0.012(5) e 0.048(1) 0.057 0.002(4) GU1.1 ee0.043 0.001(3) 0.047 0.007(7) BY-4375 0.052 0.003(4) e 0.053 0.002(2) e TAR2.3 e 0.060 0.004(6) 0.059 0.005(6) 0.063 0.010(7) BY-21875 0.065 0.009(5) e 0.054 0.007(4) e ABS-5 e 0.085 0.004(3) 0.087 0.015(3) e RB-5 ee0.068 0.010(10) 0.086 0.008(12) RB-12 ee0.080 0.009(2) 0.084 0.005(4) ABS-6 e 0.100 0.011(4) 0.083 0.013(8) e CR1.1 ee0.090 0.010(6) 0.107 0.012(8) ES4.1 e 0.072 0.013(10) 0.083 0.010(9) 0.112 0.010(10) CR1.11 ee0.092 0.004(5) 0.108 0.021(5) ENG2.3 0.133 0.000(4) e 0.131 0.023(7) 0.112(1) CR1.5 ee0.099 0.016(8) 0.112 0.018(8) ENG1.1 0.115(1) ee0.117 0.014(6) ABD-1 e 0.126(1) 0.140(1) e TR4.3 ee0.131 0.005(7) 0.127 0.022(14) ENG2.2 ee0.113(1) 0.136 0.007(3) CR1.8 ee0.101 0.007(5) 0.147 0.014(6) SRA2-15 0.174 0.042(5) 0.194 0.036(4) e 0.224 0.078(5) TAR1.1 ee0.115 0.007(9) 0.135 0.019(5) CR1.9 top ee0.122 0.012(6) 0.156 0.032(10) SRA1-11 0.203 0.027(4) ee0.219 0.022(2) ABD-3 e 0.147 0.031(2) 0.220(1) 0.219(1) ABD-2 e 0.225 0.016(5) 0.304(1) e TR4.4 e 0.141 0.011(7) 0.191 0.018(7) 0.227 0.035(9) TR3.1 e 0.149(1) e 0.201 0.035(8) CBS-352 0.304 0.043(9) 0.166 0.021(12) 0.238 0.015(8) 0.229 0.030(7) CBS-327 0.301 0.048(14) e 0.289 0.044(9) 0.322 0.088(6) FA 0.402 0.036(8) 0.363 0.032(7) 0.462 0.060(7) 0.391 0.019(2) CBS-310 0.377 0.037(8) e 0.364 0.038(6) 0.394 0.090(6) CBS-228 0.614 0.052(9) 0.467 0.055(7) 0.663 0.059(7) 0.604 0.060(14)

Asp D/L values were very small and almost negligible, except in between 0.30 and 0.40), we detected appreciable differences in Asp CBS-352. D/L values of I. gibba valves with respect to those of C. neglecta and Supporting the results of Bright and Kaufman (2011b), Ilyocypris H. reptans species; but such differences were observed to diminish valves showed similar Asp racemization values as Candona valves in older samples belonging to Middle Pleistocene and Lower for Asp D/L < ca. 0.30, with the exception of ES4.1, in which there Pleistocene. was also a significant difference in values of ca. 0.1 with respect to With the exception of the Holocene locality ES6.2, in which Asp those in Herpetocypris valves. In most of the Middle Pleistocene D/L values in C. torosa valves were lower than in those of H. reptans localities, especially those dated at ca. 230e280 ka (with D/L Asp and C. neglecta, the Asp racemization rate appeared to be similar to that found for C. neglecta valves in Upper Pleistocene localities (D/ < e Table 5 L 0.35) and for H. reptans in early Upper Lower, Middle and Amino acid composition (relative percent and total concentration) of valves of Lower Pleistocene localities (D/L > 0.35). modern C. torosa, I. gibba, C. neglecta and H. reptans specimens. Errors represent the In general, C. torosa (Superfamily Cytheroidea) and C. neglecta standard deviation of the percentage for each amino acid content. The numbers in (Superfamily Cypridoidea) valves showed a similar Asp racemiza- parentheses are the number of samples analyzed. tion rate in Holocene and Pleistocene localities (Fig. 5A). In contrast, Amino acid C. torosa (5) I. gibba (6) C. neglecta (4) H. reptans (3) H. reptans valves showed a higher Asp racemization rate than Asp 23.2 2.2 19.4 1.6 17.3 0.6 17.4 0.5 C. torosa and C. neglecta ones for low Asp D/L values (Fig. 5B), Glu 9.6 0.6 12.3 0.5 12.9 0.1 9.8 0.5 although these values were comparable with those of older localities Ser 14.2 1.5 10.6 1.3 13.1 0.4 13.0 0.9 (>130 ka). I. gibba showed a lower racemization rate for Asp than the Ala 6.2 0.7 10.1 1.0 7.7 0.5 9.3 0.3 Gly 22.0 2.7 20.7 1.7 20.2 0.6 24.4 0.4 other species for Middle and Lower Pleistocene localities (Fig. 5A, B). Val 3.7 0.9 6.0 0.8 6.0 0.1 8.1 0.5 Ile 5.8 1.3 3.7 0.4 3.2 0.2 2.4 0.1 4.1.3. Extent of Glu racemization in valves of ostracode species of Leu 3.3 0.8 5.2 0.8 3.9 0.3 2.6 0.1 different ages Phe 3.3 0.3 3.1 0.4 4.3 0.5 1.6 0.1 Thr 4.0 0.3 3.7 0.2 4.1 0.1 3.3 0.1 In contrast to the results for Asp, differences in the Glu D/L Arg 3.3 0.5 3.3 0.4 4.9 0.1 4.9 0.1 values between C. neglecta, C. torosa and H. reptans ostracodes were Tyr 1.4 0.6 1.9 0.3 2.4 0.2 3.2 0.1 negligible at low D/L values (<0.09), with the exception of ES6.2 Total concentration 35.0 10.4 61.8 23.8 72.7 11.6 191.4 63.2 (Fig. 4); in this locality we analyzed only one C. neglecta valve. This (nmol/mg) observation could be explained by the low racemization rate of Glu. 136 J.E. Ortiz et al. / Quaternary Geochronology 16 (2013) 129e143

the Bonneville Basin; Candona specimens exhibit higher values than the ostracode genus Limnocythere for most Late Pleistocenee Holocene samples (those with Asp D/L < 0.25; and Glu D/L < 0.07). However, Kaufman (2003) indicated that, taking into account the standard deviations, racemization rates could be assumed to be the same in these genera. In contrast, we found that similar differences in D/L values were absent in Middle Pleistocene samples, i.e. those with Asp D/L values ranging between 0.5 and 0.6, and Glu D/L values from 0.34 to 0.45.

4.2. Amino acid composition of ostracode valves

The total amino acid concentration for the various ostracode species is presented in Table 5, together with the relative percentage of each amino acid. In general, H. reptans valves showed the highest amino acid content. The amino acid composition varied between the species (Table 5), being specially marked for the most abundant amino acids (Asp, Glu, Ser, Ala, and Gly). C. torosa valves showed the highest Asp percentage (23.2%), followed by I. gibba specimens, whereas C. neglecta and H. reptans representatives showed similar proportions (ca. 17.4%). In contrast, C. torosa and H. reptans valves had similar Glu percentages (ca. 9.6%), which were lower than in those of I. gibba and C. neglecta (ca. 12.5%). The percentage of Ser was similar in C. torosa, C. neglecta and H. reptans valves (13e14%) and lower in those of I. gibba (10.6%). The percentage of Ala was higher in I. gibba and H. reptans (9e10%) than in the other two species, and H. reptans valves registered the highest percentage of Gly (24.4%), followed by C. torosa, I. gibba and C. neglecta specimens, the two latter with similar levels of 20e22%. Fossil ostracodes of the distinct species also showed differences in valve amino acid composition (Fig. 3). The fossilized valves of all four species showed an increased percentage of Asp and Glu contents with respect to modern valves. Of note, Asp content in C. torosa valves reached almost 50% in Middle and Lower Pleisto- cene localities. In contrast, the percentage of Gly decreased in fossilized valves, except for H. reptans, which remained almost the same as in modern valves. The content of other amino acids in the valves of the four species also varied with age (see Supplementary Data-Figs. 2e4).

4.3. Valve microstructure

In order to shed light on the differences observed in the AAR rates of the four ostracode species, we examined the microstructure of their valves by means of SEM. In all cases we observed a thin layer Fig. 3. Percentage of Asp, Gly, and Glu content in valves of fossil representatives of of 1.5e2 mm in the most superficial part of the outer lamella, which C. torosa, I. gibba, C. neglecta and H. reptans from distinct localities. corresponded to the epicuticle (Fig. 6A1, B1, C1, D2). In agreement with Bate and Sheppard (1982) and Sohn and Kornicker (1988),we However, in the Glu D/L range between 0.09 and 0.18 (ca. 40e observed that the epicuticle lacked internally differentiated ultra- 200 ka), H. reptans valves racemized faster than those belonging structure and consisted of a very thin layer above the outer surface to C. neglecta. of the thicker, and more structurally complex, procuticle (Fig. 6A1, As with Asp D/L values, I. gibba valves, like those of C. neglecta, B1, C1, D2). According to Dépêche (1982), the epicuticle decays showed similar Glu racemization rates for Holocene, Upper Pleis- when the valve is fossilized, but SEM examination of all our speci- tocene and lateeMiddle Pleistocene localities (Glu D/L < 0.18); but mens showed that, even in fossil ostracodes, this layer was present for older localities, differences in the racemization rates of this (it is especially visible in the fossil valves shown in Fig. 6A3, B4, C5, amino acid for I. gibba vs. C. neglecta and H. reptans valves were D5), thereby indicating that mineral diagenesis has not strongly notable (Fig. 4). affected the valves. We did not observe a clear differentiation C. torosa valves showed similar Glu racemization rates to those between the exocuticle and endocuticle in any C. torosa valves of C. neglecta and H. reptans for Middle and Lower Pleistocene (Fig. 6B1). but the boundary between these layers was clear in most localities (Fig. 5C, D) although in CBS-352 the mean Glu D/L values specimens of the other species (e.g. Fig. 6A1, C1, D2). were higher (Fig. 4), while I. gibba registered lower Glu D/L values Calcite crystals in I. gibba were arranged in a subparallel pattern for the same sites. both in modern and in fossil valves (Fig. 6A1eA7). Of note, the crystals It is interesting to note that Kaufman (2003) reports differences in this species were larger than in C. torosa (Fig. 6B1eB7) and in Asp and Glu D/L values in ostracode valves of two genera from H. reptans (Fig. 6C1eC7), although the valves were thinner (15e18 mm) J.E. Ortiz et al. / Quaternary Geochronology 16 (2013) 129e143 137

Fig. 4. Comparison of the Asp and Glu D/L values of I. gibba, C. neglecta, C. torosa and H. reptans valves in the sampled localities arranged according to age. than in these latter species (20e25 mm). The internal microstructure 5. Discussion in both modern and fossil I. gibba reproduced the typical external pattern, consisting of nodes and pits (Fig. 6A1eA5). Similar to the The differences observed between these ostracode species may observations of Swain and Kraft (1975), we found that the shell be due to several factors, including environmental conditions. We ornamentation was also reflected in the procuticle. address whether there is a relationship between protein compo- C. torosa valves presented a microstructure of subparallel crys- sition, valve microstructure and Asp and Glu racemization rates. tals that were smaller than in I. gibba (Fig. 6). Although these valves showed a pitted ornamentation, in contrast to I. gibba, the micro- 5.1. Protein composition vs. Asp and Glu D/L values structure only slightly reproduced this pattern (Fig. 6B2eB4, and B6eB7). The composition and structure of proteins influence AAR in In H. reptans, three clearly distinct layers were observed in some organisms; consequently the racemization rates of distinct taxa specimens (Fig. 6C1): the uppermost thin epicuticle, a middle differ (cf. Lajoie et al., 1980). The effect of taxonomy on the rates of lamellar exocuticle, and an innermost endocuticle arranged in AAR is associated with differences in the composition and bond a reticular matrix (feathered-like). However, in most cases, the strengths of the amino acids that comprise the various proteins separation between the endocuticle and exocuticle was not clear present in the organic matrix (Hill, 1965; Hare, 1976; Collins et al., (Fig. 6C3, C4, C6, C7). The valve of this species was thicker than that 1999). According to Wehmiller (1980, 1990), Ile epimerizes more of the others, reaching 25 mm. slowly in foraminifera and mollusk shells with a high Asp content C. neglecta valves showed a cross-lattice structure of parallel and than in shells that contain a low concentration of this amino acid. perpendicular calcite crystals (Fig. 6D1eD6), which were larger This difference may be related to bonding interactions between Asp than in C. torosa and H. reptans (blocky structure). and the shell carbonate (Weiner and Hood, 1975; Weiner, 1979). In all the species, it is noteworthy that the same microstructural It has been hypothesized (Towe, 1980) that amino acids in pattern was observed in modern individuals and in ostracode calcium carbonate biominerals occur both within the crystals (the valves of different ages (1 Ma, 200 ka, 30 ka). intra-crystalline fraction) and between them (inter-crystalline), 138 J.E. Ortiz et al. / Quaternary Geochronology 16 (2013) 129e143

Fig. 5. Comparison of the Asp and Glu D/L values of the valves of four ostracode species in the Iberian Peninsula. A) Asp D/L values of C. neglecta vs. H. reptans, I. gibba, C. torosa;B) Asp D/L values of H. reptans vs. C. neglecta, I. gibba, C. torosa; C) Glu D/L values of C. neglecta vs. H. reptans, I. gibba, C. torosa; D) Glu D/L values of H. reptans vs. C. neglecta, I. gibba, C. torosa. Dashed lines represent the 1:1 relationship.

and these fractions have distinct racemization rates (Sykes et al., We found that the D/L Asp and Glu values of ostracode valves of 1995; Penkman et al., 2007, 2008). While inter-crystalline distinct species (C. neglecta, I. gibba, H. reptans and C. torosa), but of proteins are more susceptible to decomposition or leaching, the the same age, differed. Furthermore, the AAR values of the speci- intra-crystalline fraction has been found to approximate a closed mens studied here did not show any clustering on the basis of their system in various mollusk shells, with the amino acids within phylogenetic group: H. reptans valves consistently showed higher crystals effectively isolated from variable external factors. Inter- Asp D/L values than those of C. neglecta, and C. torosa for Holocene, crystalline proteins in ostracodes are leached out or decomposed Upper Pleistocene and some Middle Pleistocene localities; significantly over a few centuries or thousands years, with only the however, values were similar in older sites (Fig. 5A, B). In contrast, most strongly bound and most resilient inter- and intra-crystalline I. gibba and C. neglecta valves showed similar Asp and Glu D/L proteins remaining (Bright and Kaufman, 2011a,b). As a result of values, but values differed in the oldest localities (Fig. 5). In this these characteristics, these proteins are well protected and highly regard, Bright and Kaufman (2011b) interpreted the high Asp D/L resistant to chemical removal (Bright and Kaufman, 2011a,b), values in Heterocypris valves with respect to those in Candona and thereby likely to be the ones that we measured in this study, Ilyocypris specimens as being produced by taxonomic effects. although it must be considered that in the Ana River section Kaufman (2003) has previously noted that differences in the D/L (Oregon), ostracodes contained consistently and substantially well- values of ostracode valves result from taxon-dependent arrange- preserved proteins for their ages (Bright and Kaufman, 2011b; ment of amino acids, or from morphological and structural Reichert et al., 2011), which were assumed to be produced by differences. environmental conditions. According to Bright and Kaufman Therefore, the mismatch detected between Asp and Glu D/L (2011a), there is a taxonomic control on the effectiveness of values in the valves of the four ostracode species may be linked to bleach at removing proteins, and therefore isolating the intra- the protein composition of their valves, which varies among crystalline fraction. Moreover, Penkman et al. (2008, p. 21) attrib- modern specimens (Table 5) as well as between fossil ones (Fig. 3). uted the poor correlation with age observed in D/L values of Corbicula with respect to other mollusks to differing protein 5.2. Environmental factors composition or shell structure: “the cross-lamellar structure of Corbicula might be more susceptible to incomplete bleaching than The differences observed in the AAR values of the four species the thin-shelled gastropods”. may be linked to environmental conditions. The morphology of Fig. 6. SEM images of the valve microstructure of ostracode species (I. gibba, C. torosa, H. reptans and C. neglecta) of different ages (from ca. 1 Ma to modern). The external surface of the valves is located in the upper part of images. The cuticle layers (1: epicuticle; 2: exocuticle; 3: endocuticle) were marked when possible in selected images for each species. A clear differentiation between the exocuticle and endocuticle was not observed in I. gibba sections. The external pattern of nodes and pits typical of I. gibba valves can be observed in images A1eA5, with the internal structure reproducing this pattern. Calcite crystals in I. gibba (A1eA7) and C. torosa (B1eB7) appear arranged in a subparallel matrix. The microstructure of H. reptans (C1eC7) shows a middle lamellar exocuticle and an innermost endocuticle arranged in a reticular matrix (feathered-like). C. neglecta valves show a cross-lattice structure of parallel and perpendicular calcite crystals (D1eD7). Pore-canals are observed in pictures C3 and D3. 140 J.E. Ortiz et al. / Quaternary Geochronology 16 (2013) 129e143 ostracode valves, including shape, valve thickness and ornamenta- 5.3. Valve microstructure vs. amino acid content and D/L values tion (nodes, spines, pitted, etc.), varies depending on environmental factors such as salinity and/or temperature, with many authors Some authors point out that the valve microstructure of ostra- reporting a positive correlation between valve thickness and orna- codes is linked to their phylogeny. Keyser and Walter (2004) re- mentation and a decrease in salinity of the ambient water (Sandberg, ported that the valve microstructure shows a variety of forms 1964; Carbonnel, 1969; Kilenyi, 1972; Vesper, 1972, 1975; depending on the systematic relationship, generalizing that in Peypouquet, 1977; Garbett and Maddocks, 1979; Bodergat, 1983, Cytheroidea the valve can be almost completely built of calcite 1985; Debenay et al., 1994;andVan Harten, 1996, 2000). High crystals, or in Cypridoidea it comprises parallel chitin protein temperatures may favor the development of ornamentation lamellae, together with a layer of crystallites. We observed that the (Carbonnel, 1975; Peypouquet, 1977; Hartmann, 1982; Ikeya and valve microstructures of Herpetocypris, Candona and Ilyocypris, Ueda, 1988) and, in some cases, nutrients have a considerable which belong to the Superfamily Cypridoidea (and Cyprideis, which influence on this process (Bodergat,1983,1985; Bodergat et al.,1993; is included in the Superfamily Cytheroidea), differ (Fig. 6). These Carbonel and Hoibian, 1988; Ruiz et al., 2006). In brief, environ- variations may be explained by the distinct proteins that constitute mental factors affect the characteristics of ostracode valves. The the valves, i.e. the morphology, arrangement and orientation of valves, in turn, are the structures in which biomineral proteins are calcite crystals is affected by chitin and protein structures located, thus possibly affecting subsequent diagenetic processes; for (cf. Okada, 1982; Addadi and Weiner, 1985; Weiner and Addadi, example, calcite and chitin content has been found to vary 1991; Albeck et al., 1993; Keyser and Walter, 2004), which vary as depending on environmental factors (Keyser and Walter, 2004). a function of the taxonomy. Thus, the microstructure of the valves According to Stankiewicz et al. (1998) and Briggs et al. (1998),the of the four ostracode species (Fig. 6) and their protein composition preservation of arthropod chitin is more strongly related to the (Table 5, Fig. 3)reflect taxonomic control. depositional environment (including temperature, oxygen content, Biominerals are precipitated by means of an organic template, microbial activity, etc.) than age. Bright and Kaufman (2011a) which controls the size and shapes of the crystals, together with the attributed half of the intra-sample (between-valve) variability of texture of the shell (cf. Albeck et al., 1993; Marin and Luquet, 2004). 28% in Asp and Glu content between individual valves within In crustaceans, the organic matrix is responsible for the mineral a modern population of a single taxon (Herpetocypris brevicaudata) polymorph and the morphology of the calcified structures (Hecker to a variety of environmental factors. These included differences in et al., 2003; Luquet and Marin, 2004). In this regard, some kinds of the initial abundance of proteins incorporated into a valve, as well as proteins are involved in the formation and mineralization of the the extent of decomposition after molting (seasonal collection). valves and can modify the structural integrity of calcite crystals Moreover, Bright and Kaufman (2011a) observed differences in the (Albeck et al., 1993). According to Okada (1982), the spacing of the Asp and Glu contents and D/L values of Candona shells from the Ana chitin-protein framework, which varies with the ostracode species, River and other sites of a similar age around the western United affects calcification. Knowledge of the characteristics of the organic States and northern Sonora (Mexico). These differences were matrix for each taxon is required in order to clarify how calcified attributed to the degree of protein preservation in the valves, which mineralization occurs, and how matrix molecules influence the was linked to environmental conditions (geochemistry of the waters, nature of the polymorph obtained. To date, only a few organic mainly the dominant ions). In our study, environmental factors may matrix proteins involved in calcification in crustaceans have been be expected to have affected Asp and Glu D/L values, as conditions characterized (Skinner et al., 1992; Willis, 1999). Little is known would not have been the same in all localities, even during the life of about the types of proteins in ostracode valves, but based on ostracodes in the same locality. However, differences in racemization analogy with mollusks and other crustaceans that secrete values between the four species were, in most cases, higher than the carbonate, there are two main classes of organic matrix in cara- variability of 28% observed by Bright and Kaufman (2011a) between paces: namely the soluble one and the insoluble one. The former is valves of the same species. Furthermore, we found that the differ- rich in acidic amino acids, containing high proportions of Asp, Glu ences between ostracode species tended to consistently follow the and Gly, which play a key role in biomineralization, the precipita- same pattern, i.e. I. gibba valves showed similar D/L values to those of tion of carbonate crystals, and the formation of the carapaces C. neglecta and C. torosa, but these values were lower in older (Weiner, 1979, 1983). In contrast, the insoluble matrix is enriched in localities (with Asp D/L > 0.36 and Glu D/L > 0.15), and these species less-polar amino acids, such as Gly, Ser and Ala, which are believed racemized more slowly than H. reptans.Therefore,wepostulatethat to provide the organic matrix framework for the deposition of other factors affect racemization rates. calcium carbonate (Weiner and Hood, 1975; Weiner and Traub, The differential racemization rates of Asp and Glu between the 1984; Weiner, 1984, 1986; Weiner and Erez, 1984; Robbins and four ostracode species might also be explained by the influence of Brew, 1990; Robbins and Healy-Williams, 1991; Yano et al., 2006). pH. Orem and Kaufman (2011) and Bright and Kaufman (2011b) The soluble matrix affects crystal nucleation, crystal growth and observed an increase in racemization rates of Asp and Glu in inhibition, crystal polymorphism, and atomic lattice orientation a freshwater mollusk and ostracodes (respectively) in response to (Weiner and Addadi, 1991; Wheeler, 1992; Falini et al., 1996; an increase in environmental pH under experimental conditions. Belcher et al., 1996). This matrix holds proteins that control calcite However, we do not attribute the differences in the AAR rates crystal morphology (e.g., Addadi and Weiner, 1985; Fu et al., 2005), between the species in our study to pH. Our interpretation is proteins that stabilize an amorphous calcium carbonate precursor supported by the observation that, when heated at 120 C under to the crystalline calcite phase of the biomineral (Aizenberg et al., the same pH conditions, the valves of two genera (Candona and 1996, 2002; Keyser and Walter, 2004), and proteins that may Limnocythere) that belong to distinct Superfamilies show similar control the polymorphism of a particular mineral phase (Addadi rates and extent of racemization for Asp and Glu (cf. Bright and and Weiner, 1985; Marin et al., 2008). Thus, the family of proteins Kaufman, 2011b). However, certain differences were detected can induce the modification of calcite growth, yielding different between Limnocythere and another species of Candona at low Asp types of crystal forms, sizes and arrangements, which may be the and Glu D/L values. We believe that pH may affect the Asp and Glu result of the distinct rates of crystal growth in various directions. D/L values in valves of a single species between localities, but not Furthermore, it has been proposed that “the nature and organiza- the distinct AAR rates of the ostracode genera for the same locality, tion of functional groups at the surface of the proteins are crucial to as they are subjected to the same environmental conditions. achieve the desired selectivity in polymorph nucleation, as well as J.E. Ortiz et al. / Quaternary Geochronology 16 (2013) 129e143 141 controlling the crystalline nature and morphology of the inorganic 2) Differences in the AAR rates of the four ostracode species are phase” (Lakshminarayanan et al., 2002; p. 5515). In brief, the likely to be due mainly to variations in the protein composition carbonate matrix of the organisms is affected not only by the type of their valves, thus implying that the most resistant inter- and of amino acids present in the proteins, but also by their position in intra-crystalline proteins in each species degrade at different the protein chains. rates, this process potentially varying with age. The ostracode species studied here showed distinct types of 3) Ostracode species belonging to the same Superfamily do not valve microstructure, in spite of three of them belonging to the share the same valve microstructure. C. torosa and I. gibba same Superfamily. We propose that these differences were caused valves showed a lamellar arrangement of calcite crystals by distinct compositions and arrangement of proteins, which also (subparallel pattern), reproducing the external ornamentation, would account for the differences in the Asp and Glu D/L values although crystals in C. torosa valves were smaller than in recorded for the four species. I. gibba, presenting a more packed microstructure. H. reptans Of note, we did not detect any change in valve microstructure valves also showed a lamellar pattern, although with a feath- through time, at least until 1 Ma, for any of the species sampled. The ered-like structure in the endocuticle, while C. neglecta valves same microstructural pattern was observed for modern ostracodes were characterized by a blocky structure. and for specimens from Holocene, and Upper, Middle and Lower 4) Differences in valve microstructures for each species may be Pleistocene deposits (Fig. 6). linked to the proteins involved in their calcification, with In this section we also examine the possible relationship varying percentages of amino acids observed between species. between microstructure and the degree of AAR in ostracode valves. 5) As the valve microstructural pattern does not appear to change According to Weiner (1984), there is a well-defined spatial relation over time (at least for the last ca. 1 Ma), after a certain point in between the mineral crystal and the organic matrix. Bonding degradation, D/L values may be affected mostly by the degra- interactions between different types of proteins and the biomineral dation of intra-crystalline proteins and the most insoluble of carbonate can produce conformational changes in proteins (Weiner the inter-crystalline ones (cf. Bright and Kaufman, 2011a,b) and Hood, 1975; Weiner, 1979), which, in turn, may explain after removal of the inter-crystalline protein fraction. different amino acid racemization/epimerization rates. According 6) The AAR of the valves of the four ostracode species might be to Weiner (1984,p.948)“infrared studies of these proteins related to their microstructure (size and arrangement of crys- (aspartic acid-rich proteins) together with their associated acidic tals), which in turn is determined by the distribution and polysaccharides after extraction from the shell show that both composition of chitin layers and proteins, although additional constituents undergo conformational changes as a result of calcium work is needed. binding, with the proteins adopting the b-sheet conformation”. Thus, we hypothesize that the taxonomic control reflected in the Acknowledgments microstructure of valves of the ostracode species (crystal morphology, size, and pattern), which is attributed to distinct types Funding was obtained through the projects “Paleoclimatological of proteins, might result in the observed differences in AAR rates, at revision of climate evolution in Western Mediterranean region” least in inter-crystalline proteins. C. neglecta valves presented (European Union, CE-FI2W-CT91-0075), “Evidence from Quater- a complex blocky structure together with a lower degree of Asp nary Infills Palaeohydrogeology EQUIP” (European Union, F14W/ racemization in Holocene, Upper Pleistocene and some Middle CT96/0031), “Evolución Paleoclimática de la Mitad Sur de la Pleistocene localities. In contrast, the microstructure of C. torosa Península Ibérica” of ENRESA (National Company for Radioactive valves consisted of densely packed small subparallel crystals, but Waste Management, P-703238) and “Paleoclima” of ENRESA and showed similar racemization rates to C. neglecta, in spite of showing CSN (Spanish Nuclear Safety Council, P-703389. We are indebted to differences in the amino acid composition of their valves, especially Profs. Darrell Kaufman and Matthew Collins, and Jordon Bright and Asp, Glu, Val and Ile in modern individuals (Table 5), and, in an Dra. Kirsty Penkman who helped in the setting up of our HPLC. The appreciable manner, in Asp percentages of fossil ones (Fig. 3). Biomolecular Stratigraphy Laboratory has been partially funded by Although the mechanism by which valve microstructure could ENRESA. We thank Jordon Bright and Eric Oches for their helpful affect AAR is unclear and requires further research, studies between comments on the manuscript. D/L values and microstructural patterns merit further consideration. Editorial handling by: K. Penkman

6. Conclusions Appendix A. Supplementary data

At low D/L values (Asp D/L < 0.40; and Glu D/L ¼ 0.09e0.18), the Supplementary data related to this article can be found at Asp and Glu AAR rates in valves of the ostracode species C. torosa, http://dx.doi.org/10.1016/j.quageo.2012.11.004. I. gibba, C. neglecta, and H. reptans differed. At low Asp D/L values, H. reptans valves racemized faster than those of C. neglecta, C. torosa References and I. gibba, whereas after a certain level of degradation (Asp D/L > 0.40) all species showed similar Asp racemization rates, Addadi, L., Weiner, S., 1985. Interactions between acidic proteins and crystals: except I. gibba, in which this process was slower. C. torosa, stereochemical requirements in biomineralization. Proceedings of the National e C. neglecta, and H. reptans valves showed similar Glu D/L values, Academy of Sciences of the United States of America 82, 4110 4114. Aizenberg, J., Lambert, G., Addadi, L., Weiner, S., 1996. Stabilization of amorphous with the exception of those sampled in some Middle Pleistocene calcium carbonate by specialized macromolecules in biological and synthetic localities (Glu D/L ¼ 0.09e0.18), while I. gibba valves provided lower precipitates. Advanced Materials 8, 222e226. racemization rates for old localities. In summary, we conclude that: Aizenberg, J., Lambert, G., Weiner, S., Addadi, L., 2002. Factors involved in the formation of amorphous and crystalline calcium carbonate: a study of an Ascidian skeleton. Journal of the American Chemical Society 124, 32e39. 1) The valves of C. neglecta, I. gibba, H. reptans and C. torosa show Albeck, S., Aizenberg, J., Addadi, L., Weiner, S., 1993. Interactions of various skeletal different Asp and Glu racemization rates over the course of intracrystalline components with calcite crystals. Journal of the American Chemical Society 115, 11691e11697. their degradation, and do not show any clustering on the basis Bate, R.H., East, B.A., 1972. The structure of the ostracode carappace. Lethaia 5, of phylogenetic group. 177e194. 142 J.E. Ortiz et al. / Quaternary Geochronology 16 (2013) 129e143

Bate, R.H., East, B.A., 1975. The ultrastructure of the ostracode (Crustacea) integu- Quaternary tropical marine sediment cores. Paleoceanography 19, PA4022. ment. Bulletins of American Paleontology 65, 529e547. http://dx.doi.org/10.1029/2004PA001059. Bate, R.H., Sheppard, L.M., 1982. The shell structure of Halocypris inflata (Dana, Hecker, A., Testenière, O., Marin, F., Luquet, G., 2003. Phosphorylation of serine 1849). The secretion process of the ostracod carapace. In: Bate, R.H., residues is fundamental for the calcium-binding ability of Orchestin, a soluble Robinson, E., Sheppard, L.M. (Eds.), Fossil and Recent Ostracods. Ellis Horwood matrix protein from crustacean calcium storage structures. FEBS Letters 535, Limited, Chichester, pp. 25e50. 49e54. Belcher, A.M., Wu, X.H., Christensen, R.J., Hansma, P.K., Stucky, G.D., Morse, D.E., Hill, R.L., 1965. Hydrolysis of proteins. In: Anfinsen Jr., C.B., Anson, M.L., Edsall, J.T., 1996. Control of crystal phase switching and orientation by soluble mollusc- Richards, F.M. (Eds.), Advances in Protein Chemistry, vol. 20. Academic Press, shell proteins. Nature 381, 56e58. New York, NY, pp. 37e107. Bodergat, A.M., 1983. Les ostracodes, témoins de leur environement: Approche Howell, F.G., Butzer, K.W., Freeman, L.G., Klein, R.G., 1995. Observations on the chimique et écologie en milieu lagunaire et océanique. Document des Labo- Acheulean occupation site of Ambrona (Soria Province, Spain), with particular ratoires de Géologie de la Faculte des Sciences de Lyon 88, 246. reference to recent investigations (1980-1983) and the lower occupation. Bodergat, A.M., 1985. Composition chimique des caparaces d’Ostracodes. Para- Jahrbuch des Römisch-Germanischen Zentralmuseum Mainz 38, 33e82. mètres du milieu de vie. In: Oertli, H.J. (Ed.), Atlas des Ostracodes de France, Ikeya, N., Ueda, H., 1988. Morphological variations of Cytheromorpha acupunctata Bulletin des Centres de Recherches et Exploration-Production Elf-Aquitaine, (Brady) in continuous populations at Hamana-ko Bay, Japan. In: Hanai, T., Pau, Mémoire 9, pp. 379e386. Ikeya, N., Ishizaki, K. (Eds.), Evolutionary Biology of Ostracoda. Developments Bodergat, A.M., Carbonnel, G., Rio, M., Keyser, D., 1993. Chemical composition of in Paleontology and Stratigraphy, vol. 11. Kodansha Publications, Tokyo, Leptocythere psammophila (Crustacea: Ostracoda) as influenced by winter pp. 319e340. metabolism and summer supplies. Marine Biology 117, 53e62. Jayko, A.S., Forester, R.M., Kaufman, D.S., Phillips, F.M., Yount, J.C., McGeehin, J., Briggs, D.E.G., Stankiewicz, B.A., Meischner, D., Bierstedt, A., Evershed, R.P., 1998. Mahan, S.A., 2008. Late Pleistocene lakes and wetlands, Panamint Valley, Inyo Taphonomy of arthropod cuticles from Pliocene lake sediments, Willershausen, County, California. In: Reheis, M.C., Hershler, R., Miller, D.M. (Eds.), Late Ceno- Germany. Palaios 13, 386e394. zoic Drainage History of the Southwestern Great Basin and Lower Colorado Bright, J., Kaufman, D.S., 2011a. Amino acid racemization in lacustrine ostracodes, River Region: Geologic and Biologic Perspectives, Geological Society of America part I: effect of oxidizing pre-treatments on amino acid composition. Quater- Special Paper 439, pp. 151e184. nary Geochronology 6, 154e173. Jeuniaux, C., Bussers, J.C., Voss-Foucart, M.F., Poulicek, M., 1986. Chitin production Bright, J., Kaufman, D.S., 2011b. Amino acids in lacustrine ostracodes, part III: effects by animals and natural communities in marine environment. In: of pH and taxonomy on racemization and leaching. Quaternary Geochronology Muzzarelli, R.A.A., Jeuniaux, C., Gooday, G.W. (Eds.), Chitin in Nature and 6, 574e597. Technology. Plenum Press, New York, pp. 512e522. Bright, J., Kaufman, D.S., Forman, S.L., Mead, J.I., Baez, A., 2010. Comparative dating Julià, R., 1980. La conca lacustre de Banyoles-Besalú. Monografies del Centre d’Es- of a Bison-bearing late-Pleistocene deposit, Térapa, Sonora, Mexico. Quaternary tudis Comarcals de Banyoles, Gerona, Spain, 188. Geochronology 5, 631e643. Jǿrgesen, N.O., 1970. Ultrastructure of some ostracods. Bulletin of the Geological Carbonel, P., Hoibian, T., 1988. The impact of organic matter on ostracods from an Society of Denmark 20, 79e92. equatorial deltaic area, the Mahakam Delta e Southeastern Kalimantan. In: Kaufman, D.S., 2000. Amino acid racemization in ostracodes. In: Goodfriend, G.A., Hanai, T., Ikeya, N., Ishizaki, K. (Eds.), Evolutionary Biology of Ostracoda, Its Collins, M.J., Fogel, M.L., Macko, S.A., Wehmiller, J.F. (Eds.), Perspectives in Fundamentals and Applications. Kodansha, Tokyo, pp. 353e366. Amino Acids and Protein Geochemistry. Oxford University Press, New York, Carbonnel, G., 1969. Variations phénotypiques chez une espèce tortonienne du pp. 145e160. genre Elofsonella Pokorny. In: Neale, J.W. (Ed.), The taxonomy, morphology and Kaufman, D.S., 2003. Amino acid paleothermometry of Quaternary ostracodes from ecology of Recent Ostracoda. Oliver & Boyd, Edinburgh, pp. 85e92. the Bonneville Basin, Utah. Quaternary Science Reviews 22, 899e914. Carbonnel, G., 1975. Le facteur lisse chez certains ostracodes Tertiaires: un index Kaufman, D.S., 2006. Temperature sensitivity of aspartic and glutamic acid de paléotempéature. In: Swain, F.M., Kornicker, L.S., Ludin, R.F. (Eds.), racemization in the foraminifera Pulleniatina. Quaternary Geochronology 1, Biology and Paleobiology of Ostracoda, Bulletin American Paleontology 65, 188e207. pp. 285e301. Kaufman, D.S., Manley, W.F., 1998. A new procedure for determining DL amino acid Collins, M.J., Waite, E.R., van Duin, A.C.T., Eglinton, G., 1999. Predicting protein ratios in fossils using reverse phase liquid chromatography. Quaternary decomposition: the case of aspartic-acid racemization kinetics [and discussion]. Geochronology 17, 987e1000. Philosophical Transactions of the Royal Society of London 354B, 51e64. Kaufman, D.S., Forman, S.L., Bright, J., 2001. Age of the Cutler Dam (late Pleistocene) Colman, S.M., Kaufman, D.S., Bright, J., Heil, C., King, J.W., Dean, W.E., Alloformation, Bonneville Basin. Quaternary Research 56, 322e334. Rosenbaum, J.G., Forester, R.M., Bischoff, J.L., Perkins, M., McGeehin, J.P., 2006. Kesling, R.V., 1951. Morphology of ostracod molt stages. Illinois Biological Mono- Age model for a continuous, ca 250-ka Quaternary lacustrine record from Bear graphs 21, 1e126. Lake, Utah-Idaho. Quaternary Science Reviews 25, 2271e2282. Keyser, D., 1982. Development of the sieve pores in Hirschmannia viridis (O. F. Dalingwater, J.E., Mutvei, H., 1990. Arthropod exoskeletons. In: Carter, J.G. (Ed.), Müller, 1785). In: Bate, R.E., Robinson, E., Sheppard, L.M. (Eds.), Fossil and Recent Skeletal Biomineralization: Patterns, Processes and Evolutionary Trends, vol. 1. Ostracods. Ellis Horwood Ltd., Chichester, pp. 51e60. Van Nostrand Reinhold, New York, pp. 83e96. Keyser, D., Walter, D., 2004. Calcification in ostracodes. Revista Española de De Santis, V., Caldara, M., Torres, T., Ortiz, J.E., 2010. Stratigraphic units of the Micropaleontología 36 (1), 1e11. Apulian Tavoliere plain (Southern Italy): chronology, correlation with marine Kilenyi, T., 1972. Transient and balanced genetic polymorphism as an explanation isotope stages and implications regarding vertical movements. Sedimentary of variable noding in the ostracode Cyprideis torosa. Micropaleontology 18 (1), Geology 228, 255e270. 47e63. Debenay, J.-P., Guillou, J.-J., Paypouquet, J.-P., Pages, J., 1994. Encroûtement ferru- Kosnik, M.A., Kaufman, D.S., 2008. Identifying outliers and assessing the accuracy of gineux et dissolution in vivo de la carapace de Cyprideis mandviensis (ostracode) amino acid racemization measurements for geochronology: II. Data screening. dans la ria sursalée de la Casamance (Sénégal). Geobios 27 (6), 677e688. Quaternary Geochronology 3, 328e341. Dépêche, F., 1982. Ultrastructure of the wall of two living ostracods, Herpetocypris Lajoie, K.R., Wehmiller, J.F., Kennedy, G.L., 1980. Inter- and intrageneric trends in chevreuxi (Sars) and Pontocythere elongata (Brady), in comparison with fossil apparent racemization kinetics of amino acids in Quaternary mollusks. In: ostracods from the Middle Jurassic of Normandy. The secretion process of the Hare, P.E., Hoering, T.C., King Jr., K. (Eds.), Biogeochemistry of Amino Acids. ostracod carapace. In: Bate, R.H., Robinson, E., Sheppard, L.M. (Eds.), Fossil and Wiley, New York, pp. 305e340. Recent Ostracods. Ellis Horwood Limited, Chichester, pp. 61e74. Lakshminarayanan, R., Manjunatha Kini, R., Valiyaveettil, S., 2002. Investigation of Falini, G., Albeck, S., Weiner, S., Addadi, L., 1996. Control of aragonite or calcite the role of ansocalcin in the biomineralization in goose eggshell matrix. polymorphism by mollusk shell macromolecules. Science 271, 67e69. Proceedings of the National Academy of Sciences of the United States of Fu, G., Valiyaveettil, S., Wopenka, B., Morse, D.E., 2005. CaCO3 biomineralization: America 99, 5155e5159. acidic 8-kDa proteins isolated from aragonitic Abalone shell nacre can specifi- Langer, W., 1971. Rasterelektronenmikroskopische Beobachtungen über den Fein- cally modify calcite crystal morphology. Biomacromolecules 6, 1289e1298. bau von Ostracod-Schalen. Paläontologische Zeitschrift 45, 181e186. Garbett, E.C., Maddocks, R.F., 1979. Zoogeography of Holocene cytheracean ostra- Langer, W., 1973. Zur Ultrastruktur, Mikromorphologie und Taxonomie des Ostra- codes in the bays of Texas. Journal of Paleontology 53, 841e919. coda-Caparax. Paleontographica Abt A Paleozoology, Stratigraphy 144, 1e54. Goodfriend, G.A., 1991. Patterns of racemization and epimerisation of amino acids in Luquet, G., Marin, F., 2004. Biomineralisations in crustaceans: storage strategies. land snail shells over the course of the Holocene. Geochimica et Cosmochimica Comptes Rendus Palevol 3, 515e534. Acta 55, 293e302. Marin, F., Luquet, G., 2004. Molluscan shell proteins. Comptes Rendus Palevol 3, Goodfriend, G.A., Mitterer, R.M., 1988. Late Quaternary land snails from the north 469e492. coast of Jamaica: local extinctions and climatic change. Palaeogeography, Marin, F., Luquet, G., Marie, B., Medakovic, D., 2008. Molluscan shell proteins: Palaeoclimatology, Palaeoecology 63, 293e312. primary structure, origin, and evolution. Current Topics in Devevelopmental Hare, P.E., 1976. Relative reaction rates and activation energies for some amino acid Biology 80, 209e276. reactions. Carnegie Institution of Washington Yearbook 75, 801e806. McCoy, W.D., 1988. Amino acid racemization in fossil non-marine ostracod shells: Hartmann, G., 1982. Variation in surface ornament of the valves of three ostracode a potential tool for the study of Quaternary stratigraphy, chronology and species from Australia. In: Bate, R.H., Robinson, E., Sheppard, L.M. (Eds.), Fossil palaeotemperature. In: De Deckker, P., Colin, J.P., Peypouquet, J.P. (Eds.), Ostra- and Recent Ostracods. Ellis & Morwood, Chicester, pp. 365e380. coda in the Earth Sciences. Elsevier, Amsterdam & New York, pp. 201e218. Hearty, P.J., O’Leary, M.J., Kaufman, D.S., Page, M., Bright, J., 2004. Amino acid Meisch, C., 2000. Freshwater Ostracoda of Western and Central Europe. In: geochronology of individual foraminifer (Pulleniatina obliquiloculata) tests, Süßwasserfauna von Mitteleuropa, Band 8/3. Spekrum Akademischer Verlag, north Queensland margin, Australia: a new approach to correlating and dating Heidelberg, Berlin, p. 521. J.E. Ortiz et al. / Quaternary Geochronology 16 (2013) 129e143 143

Oertli, H.J., 1975. The conservation of ostracode tests e observations made under Sykes, G.A., Collins, M.J., Walton, D.I., 1995. The significance of a geochemically the scanning electron microscope. In: Swain, F.M. (Ed.), Biology and Paleobi- isolated intracrystalline fraction within biominerals. Organic Geochemistry 23, ology of Ostracoda, Bulletin of American Paleontology 65, pp. 549e575. 1059e1065. Okada, Y., 1982. Structure and cuticle formation of the reticulated carapace of the Torres, T., Canoira, L., Cobo, R., Coello, F.J., García-Alonso, P., García-Cortés, A., ostracode Bicornucythere bisanensis. Lethaia 15, 85e101. Hoyos, M., Llamas, J., Mansilla, H., Soler, V., Valle, M., 1995. Travertinos de río Orem, C., Kaufman, D.S., 2011. Effects of basic pH on amino acid racemization and Blanco (Soria): edad y evolución. Geogaceta 18, 90e92. leaching in freshwater mollusk shell. Quaternary Geochronology 6, 233e245. Torres, T., Llamas, J., Canoira, L., García-Alonso, P., García-Cortés, A., Mansilla, H., Ortiz, J.E., Torres, T., Llamas, F.J., 2002. Cross-calibrations of the racemization rates of 1997. Amino acid chronology of the Lower Pleistocene deposits of Venta Micena leucine and phenylalanine and epimerization rates of isoleucine between (Orce, Granada, Andalusia, Spain). Organic Geochemistry 26, 85e97. ostracodes and gastropods over the Pleistocene in southern Spain. Organic Torres, T., Ortiz, J.E., Soler, V., Reyes, E., Delgado, A., Valle, M., Llamas, J.F., Cobo, R., Geochemistry 33 (7), 691e699. Julià, R., Badal, E., García de la Morena, M.A., García-Martínez, M.J., Fernández- Ortiz, J.E., Torres, T., Delgado, A., Julià, R., Llamas, F.J., Soler, V., Delgado, J., 2004. Gianotti, J., Calvo, J.P., Cortés, A., 2003. Pleistocene lacustrine basin of the east Numerical dating algorithms of amino acid racemization ratios analyzed in domain of Guadix-Baza basin (Granada, Spain): sedimentology, chronostratig- continental ostracodes of the Iberian Peninsula (Spain). Application to Guadix- raphy and palaeoenvironment. In: Valero-Garcés, B.L. (Ed.), Limnogeología en Baza Basin (southern Spain). Quaternary Science Reviews 23, 717e730. España: un tributo a Kerry Kelts. Consejo Superior de Investigaciones Científ- Ortiz, J.E., Torres, T., Delgado, A., Reyes, E., Llamas, J.F., Soler, V., Raya, J., 2006. icas, Madrid, pp. 151e185. Pleistocene paleoenvironmental evolution at continental middle latitudes Torres, T., Ortiz, J.E., García de la Morena, M.A., Llamas, F.J., Goodfriend, G., 2005. inferred from carbon and oxygen stable isotope analysis of ostracodes from the Aminostratigraphy and aminochronology of a tufa system in Central Spain. Guadix-Baza Basin (Granada, SE Spain). Palaeogeography, Palaeoclimatology, Quaternary International 135, 21e33. Palaeoecology 240, 535e561. Torres, T., Ortiz, J.E., Fernández, E., Arroyo-Pardo, E., Grün, R., Pérez-González, A. Ortiz, J.E., Torres, T., Delgado, A., Reyes, E., Díaz-Bautista, A., 2009. A review of the Aspartic acid racemization as a dating tool for dentine: a reality. Quaternary Tagus river tufa deposits (central Spain): age and palaeoenvironmental record. Geochronology (submitted for publication). Quaternary Science Reviews 28, 947e963. Towe, K.M., 1980. Preserved organic ultrastructure: an unreliable indicator for Oviatt, C.G., Thompson, R.S., Kaufman, D.S., Bright, J., Forester, R.M., 1999. Reinter- Paleozoic amino acid biogeochemistry. In: Hare, P.E., Hoering, T.C., King Jr., K. pretation of the Burmester Core, Bonneville Basin, Utah. Quaternary Research (Eds.), Biogeochemistry of Amino Acids. Wiley, New York, pp. 65e74. 52, 180e184. Turpen, J.B., Angell, R.W., 1971. Aspects of molting and calcification in the ostracod Owen, L.A., Bright, J., Finkel, R.C., Jaiswal, M.K., Kaufman, D.S., Mahan, S., Radtke, U., Heterocypris. The Biological Bulletin 140, 331e338. Schneider, J.S., Sharp, W., Singhvi, A.K., Warren, C.N., 2007. Numerical dating of Van Harten, D., 1996. Cyprideis torosa (Ostracoda) revisited. Of salinity, nodes and shell a Late Quaternary spit-shoreline complex at the northern end of Silver Lake size. In: Keen, M.C. (Ed.), Proceedings of the 2nd European Ostracodologist playa, Mojave Desert, California: a comparison of the applicability of radio- Meeting, Glasgow, Great Britain. British Micropaleontology Society, pp. 191e194. carbon, luminescence, terrestrial cosmogenic nuclide, electron spin resonance, Van Harten, D., 2000. Variable noding in Cyprideis torosa (Ostracoda, Crustacea): an U-series and amino acid racemization methods. Quaternary International 166, overview, experimental results and a model from Catastrophe Theory. In: 87e110. Horne, D.J., Martens, K. (Eds.), Evolutionary Biology and Ecology of Ostracoda, Penkman, K.E.H., Preece, R.C., Keen, D.H., Maddy, D., Schreve, D.S., Collins, M.J., 2007. Hydrobiologia 419, pp. 131e139. Testing the aminostratigraphy of fluvial archives: the evidence from intra- Vesper, B., 1972. Zum Problem der Buckelbildung bei Cyprideis torosa (Jones, 1850) crystalline proteins within freshwater shells. Quaternary Science Reviews 26, (Crustacea, Ostracoda, Cytheridae). Mitteilungen Hamburgisches Zoologisches 2958e2969. Museum und Institut 68, 79e94. Penkman, K.E.H., Kaufman, D.S., Maddy, D., Collins, M.J., 2008. Closed-system Vesper, B., 1975. To the problem of noding on Cyprideis torosa (Jones, 1850). In: behaviour of the intra-crystalline fraction of amino acids in mollusk shells. Swain, F.M. (Ed.), Biology and Paleobiology of Ostracoda, Bulletins of American Quaternary Geochronology 3, 2e25. Paleontology 65, pp. 205e216. Pérez-Obiol, R., Julià, R., 1984. Climatic change on the Iberian Peninsula recorded in Wehmiller, J.F., 1984. Interlaboratory comparison of amino acid enantiomeric ratios a 30,000-yr pollen record from Lake Banyoles. Quaternary Research 41, 91e98. in fossil mollusks. Quaternary Research 22, 109e120. Peypouquet, J.P., 1977. Les ostracodes et la connaissance des paleomilieux profonds. Wehmiller, J.F., 1990. Amino acid racemization: applications in chemical taxonomy Application au Cénozoïque de l’Atlantique nord-oriental. Ph.D. thesis., Uni- and chronostratigraphy of Quaternary fossils. In: Carter, J.G. (Ed.), Skeletal versité de Bordeaux, France, p. 443. Biomineralization: Patterns, Processes and Evolutionary Trends, vol. 1. Van Reichert, K.L., Licciardi, J.M., Kaufman, D.S., 2011. Amino acid racemization in Nostrand Reinhold, New York, pp. 583e608. lacustrine ostracodes, part II: paleothermometry in Pleistocene sediments at Wehmiller, J.F., 2000. Aminostratigraphic dating methods in quaternary geology. In: Summer Lake, Oregon. Quaternary Geochronology 6, 174e185. Noller, J.S., Sowers, J.M., Lettis, W.R. (Eds.), Quaternary Geochronology: Methods Richards, A.G., 1951. The Integument of Arthropods. University Press, Minneapolis, and Applications. American Geophysical Union, Washington D.C., pp. 187e222. p. 411. Wehmiller, J.F., Miller, G., De Vogel, S., Kaufman, D.S., Bright, J., Murray-Wallace, C.V., Robbins, L.L., Brew, K., 1990. Proteins from the organic matrix of core-top and fossil Ortiz, J.E., Penkman, K., 2010. Interlaboratory comparison of amino acid D/L planktonic foraminifera. Geochimica et Cosmochimica Acta 54, 2285e2292. values. Geological Society of America Annual Meeting, Denver, Abstracts with Robbins, L.L., Healy-Williams, N., 1991. Toward a classification of planktonic fora- Programs 42(5), p. 86. minifera based on biochemical, geochemical, and morphological criteria. Jour- Wehmiller, J.F., 1980. Intergeneric differences in apparent racemization kinetics in nal of Foraminiferal Research 21, 159e167. molluscs and foraminifera: Implications for models of diageneticracemization. Rosenfeld, A., 1979. Structure and secretion of the carapace in some living ostra- In: Hare, P.E., Hoering, T.C., King Jr., J. (Eds.), Biochemistry of amino acids. John codes. Lethaia 12, 353e379. Wiley, New York, pp. 341e355. Rosenfeld, A., 1982. The secretion process of the ostracod carapace. In: Bate, R.H., Weiner, S., 1979. Aspartic acid-rich proteins: major components of the soluble Robinson, E., Sheppard, L.M. (Eds.), Fossil and Recent Ostracods. Ellis Horwood organic matrix of mollusk shells. Calcified Tissue International 29, 163e167. Limited, Chichester, pp. 12e24. Weiner, S., 1983. Mollusk shell formation: isolation of two organic matrix proteins Ruiz, F., Abad, M., Carbonel, P., Muñoz, J.M., 2006. Polymorphism in recent ostracods associated with calcite deposition in the Bivalve Mytilus californianus. of southwestern Spain: an environmental approach. Geobios 39 (2), 311e317. Biochemistry 22, 4139e4145. Sandberg, P., 1964. Notes on some Tertiary and Recent brackishwater Ostracoda. In: Weiner, S., 1984. Organization of organic matrix components in mineralized tissues. Puri, H.S. (Ed.), Ostracods as Ecological and Paleoecological Indicators, Pub- American Zoologist 24, 945e951. blicazioni della Stazione Zoológica di Napoli 33, pp. 496e514. Weiner, S., 1986. Organization of extracellularly mineralized tissues: a comparative Skinner, D.M., Kumari, S.S., O’Brien, J.J., 1992. Proteins of the crustacean exoskeleton. study of biological crystal growth. Critical Reviews in Biochemistry 20, 365e408. American Zoologist 32, 470e484. Weiner, S., Addadi, L., 1991. Acidic macromolecules of mineralized tissues: the Sohn, I.G., 1958. Chemical constituents of Ostracodes, some applications to pale- controllers of crystal formation. Trends in Biochemical Science 16, 252e256. ontology and paleoecology. Journal of Paleontology 4, 730e736. Weiner, S., Erez, J., 1984. Organic matrix of the shell of the foraminifer Heterostegina Sohn, I.G., Kornicker, L.S., 1988. Ultrastructure of Myodocopid Shells (Ostracoda). depressa. Journal of Foraminiferal Research 14, 206e212. Evolutionary biology of ostracoda: its fundamentals and applications. In: Weiner, S., Hood, L., 1975. Soluble protein of the organic matrix of mollusk shells: Proceedings of the Ninth International Symposium on Ostracoda, Shizuoka, a potential template for shell formation. Science 190, 987e989. Japan. Developments in Palaeontology and Stratigraphy 11, pp. 243e258. Weiner, S., Traub, W., 1984. Macromolecules in mollusc shells and their functions in Soria, J.M., 1993. La sedimentación neógena entre sierra Arana y el río Guadiana biomineralization. Philosophical Transactions of the Royal Society of London Menor (Cordillera Bética Central). Evolución desde un margen continental hasta B304, 425e434. una cuenca intramontañosa. Ph.D. thesis., Granada University, Spain, p. 292. Wheeler, A.P., 1992. Phosphoproteins of oyster (Crassostrea virginica) shell organic Stankiewicz, B.A., Mastalerz, M., Hof, C.H.J., Bierstedt, A., Flannery, M.B., matrix. In: Suga, S., Watabe, N. (Eds.), Hard Tissue Mineralization and Demin- Briggs, D.E.G., Evershed, R.P., 1998. Biodegradation of the chitin-protein eralization. Springer-Verlag, Tokyo, pp. 171e187. complex in crustacean cuticle. Organic Geochemistry 28, 67e76. Willis, J.H., 1999. Cuticular proteins in Insects and Crustaceans. American Zoologist Swain, F.M., Kraft, J., 1975. Biofacies and microstructure of Holocene Ostracoda 39, 600e609. from tidal bays of Delaware. In: Swain, F.M., Kornicker, L.S., Lundin, R.F. (Eds.), Yano, M., Nagai, K., Morimoto, K., Miyamoto, H., 2006. Shematrin: a family of Biology and Paleobiology of Ostracoda, Bulletins of American Paleontology 65, glycine-rich structural proteins in the shell of the pearl oyster Pinctada fucata. pp. 601e621. Comparative Biochemistry and Physiology 144B, 254e262.