Palaeogeography, Palaeoclimatology, Palaeoecology 459 (2016) 86–98

Contents lists available at ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology

j ourna l homepage: www.e lsev ie r .com/ locate /pa laeo

Patterns of drilling predation in relation to stratigraphy, locality and sievesize: Insights from the Eocene molluscan fauna of the Paris Basin

Devapriya Chattopadhyay a,⁎, Martin Zuschin b, Stefano Dominici c, Jennifer A. Sawyer b

a Department of Earth Sciences, Indian Institute of Science Education and Research (IISER) Kolkata, Mohanpur WB-741246, Indiab University of Vienna, Department of Palaeontology, Althanstrasse 14, A-1090 Vienna, Austriac Museo di Storia Naturale, Università degli Studi di Firenze, Florence, Italy

⁎ Corresponding author.E-mail address: devapriya@iiserkol.ac.in (D. Chattopad

http://dx.doi.org/10.1016/j.palaeo.2016.07.0010031-0182/© 2016 Published by Elsevier B.V.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 14 February 2016Received in revised form 30 June 2016Accepted 1 July 2016Available online 2 July 2016

Predatory drilling traces are routinely used to evaluate temporal patterns in animal behaviour. However, fortesting such trends it is important to consider natural andmethodological variability thatmay play a role in shap-ing drilling patterns. The present study attempts to test this notion by evaluating the drilling patterns of themolluscan fauna in relation to stratigraphy, locality, sieve size (4 mm, 2 mm, 1 mm) and taxonomic affinity atclass level. We studied drilling predation using molluscan specimens from three localities (Grignon, La Fermede l'Orme, Fleury-la-Rivière), representing a variety of elementary depositional sequences (EDS) from theMiddleLutetian strata of the Paris Basin.The overall drilling frequency (DF) averaged over the three localities is 17%; bivalves show a significantly higheroverall DF (18%) compared to gastropods (14%) and scaphopods (b0.01%). For all higher taxa and at all localitiesDF is highest in the medium size class (2–4 mm), which we relate to the dominance of preferred prey species inthis category. Despite having similar diversity profiles of molluscan fauna, the three localities show distinctdrilling patterns with highest frequencies in Grignon (overall 25%, bivalves 29%, gastropods 20%), followed byFleury-la-Rivière (overall 18%, bivalves 17%, gastropods 18%) and almost a complete lack of drill holes in LaFerme de l'Orme. This pattern is true for each sieve size of molluscs, bivalves and gastropods. The results aresimilar for common species found in all three localities. EDS2-3 representing a mesotrophic environment duringa regressive phase shows a slightly higher DF compared to oligotrophic EDS4, from the following transgression.Locality emerges as the strongest predictor of DF when evaluated for the relative role of stratigraphy, locality,sieve size and higher taxa. The occurrence of incomplete drill holes is relatively rare and increases with sievesize, implying a possible existence of handling limit of the predatory gastropods. There are very few occurrencesof edge drilling. The combined variation documented in this study is comparable in magnitude with manytemporal shifts documented previously. Therefore, in order to establish any temporal pattern in drilling behaviourit is vital to evaluate and control for the potential gradient of other natural variabilities in predatory behaviour.Species composition and diversity has been used previously to resolve the debate on climate conditions ofthe Eocene Paris basin through a comparison to modern analogues. We have used a similar approach bycomparing predation patterns with Recent localities of subtropical and temperate regions. The available datafrom the Eocene Paris basin, from the warm-temperate northern Adriatic Sea and from the sub-tropical RedSea show very heterogeneous patterns, depending on the drilling metric considered. Considering the strongpatchiness of drilling predation within each of these basins and reported from many other studies, we concludethat a climatic affiliation of the Eocene Paris basin based on drilling metrics is not yet possible.

© 2016 Published by Elsevier B.V.

Keywords:Drilling frequencyPrey-predator interactionPatchinessPrey effectivenessMesh size

hyay).

1. Introduction

Drilling by predatory gastropods, such as naticids and muricids, hascaptured the interest of paleontologists because it preserves biotic inter-actions of the past that can be studied quantitatively. Such data are

87D. Chattopadhyay et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 459 (2016) 86–98

critical in testing evolutionary hypotheses such as coevolution and esca-lation (Kelley and Hansen, 2003). However, for testing such hypothesesit is important to consider various sources of natural and methodologi-cal variability that may play a role in shaping drilling patterns.

Temporal variability in drilling predation is intensely studied (Kelleyand Hansen, 2003 for review). Predation is often considered the mostimportant agent of natural selection and response to predation pressureis often claimed to bring long-term evolutionary change in populations(Vermeij, 1987). Therefore, we may expect to see temporal changein predation intensity. This idea leads to the development ofevolutionary ecological hypotheses like coevolution and escalationand the subsequent effort to document them using drilling frequencydata (Kitchell, 1986, 1990; DeAngelis et al., 1984, 1985; Dudleyand Vermeij, 1978; Kelley and Hansen, 1993, 1996; Kowalewski et al.,1998).

In order to recognize such temporal change, the variation in drillingpattern within a time interval should be much smaller in magnitude incomparison to the changes between subsequent time intervals. Bioticinteraction depends on the composition and density of predator-preycommunities (DeAngelis et al., 1985). As many of the communities arequite patchy in their distribution, wemay expect to see spatial variationin drilling intensity at different scales.

Studies focusing specifically on spatial variation in drilling be-haviour are relatively rare as compared to reports on local drillingpredation. In studying spatial variation in drilling patterns, scalebecomes an important issue. Researchers have recognized variationfrom a few meters to multiple latitudinal bins. Variation along a sev-eral km long transect was recognized in the modern northernAdriatic Sea (Sawyer and Zuschin, 2010) and within and between lo-calities in the Lower andMiddle Miocene of the Vienna basin, Austria(Sawyer and Zuschin, 2011). Using Recent molluscs of the Red Sea,Chattopadhyay et al. (2014a, 2015) reported a change in overalland edge drilling frequencies in various depositional environmentsin one bay of about 75 km2. Using Eocene molluscs of NorthAmerica from multiple depositional environments, Hansen andKelley (1995) demonstrated the variation in naticid drilling preda-tion. Hoffmeister and Kowalewski (2001) documented such varia-tion in drilling pattern for Miocene molluscs between multipledepositional basins of central Europe. Chattopadhyay and Baumiller(2010) reported a case of spatial variation in drilling predation inPleistocene deposits from various basins of the United States. Someresearchers expanded such analyses to a global scale. Vermeij(1980) documented variation in DF for terebrid gastropods by ana-lyzing specimens of Cenozoic occurrences from all over the globe.Often such large-scale spatial variation in DF has been causally linkedto latitudinal variation. Some of the studies demonstrated an in-crease in drilling intensity towards the equator (Visaggi and Kelley,2015) while others found an opposite trend (Hansen and Kelley,1995) or none at all (Marincovich, 1977; Taylor and Taylor, 1977;Dudley and Vermeij, 1978; Vermeij et al., 1989; Allmon et al.,1990). A useful approach to evaluate the extent of spatial variationis to study samples from different localities within one depositionalbasin with distinct environmental characteristics, because ecologicalprocesses operate at local scales and local communities receive spe-cies from a biogeographically delimited metacommunity (Hubell2001). In the fossil record, paleoecological studies in basins with tec-tonically and sedimentologically meaningful frameworks can be per-formed at low taxonomic levels and with highly resolvedstratigraphic control (e.g., Zuschin et al., 2011).

Another important source of variability that could affect the finaldrilling pattern is the taxonomic composition of the prey. As each preypresents a different location on the cost-benefit landscape for a givenpredator (Kitchell et al., 1981; DeAngelis et al., 1985; Chattopadhyayand Baumiller, 2009), we may find different drilling patterns amongdifferent taxa. Such preference for taxa has been documented in Recentas well as fossil ecosystems (Hoffman and Martinell, 1984; Kohn and

Arua, 1999; Sawyer and Zuschin, 2010, 2011; Chattopadhyay andDutta, 2013).

Body size is another source of natural variability. Drilling pattern isoften size dependent primarily because of two factors: 1. Ontogeneticvariation in drilling behaviour of the predator, 2. Cost-benefit relation-ship. Consequently, each size class of the prey may represent a uniquedrilling pattern (e.g., Ansell, 1960; Allmon et al., 1990; Kitchell, 1986;Kelley 1988, 1991; Anderson, 1992; Harper et al., 1998). Therefore,using a specific size class of prey for the study of drilling predationcould present an incomplete result. Moreover, the sampling method(see Kowalewski, 2002) could act as a source of variability that affectsspecimen size; several studies have evaluated this potential source ofbias (Ottens et al., 2012, Hattori et al., 2014, Visaggi and Kelley, 2015).

The natural and methodological variabilities discussed aboveplay important roles in fossil and Recent assemblages. However, apartfrom a few studies (Hoffmeister and Kowalewski, 2001; Ottens et al.,2012; Hattori et al., 2014), rarely has an effort beenmade to disentangletheir relative effects. In this study, we evaluated the individual andrelative effects of stratigraphy, locality, sieve size and higher taxa ondrilling patterns of molluscan fauna. We evaluated these effects ondrilling predation using molluscan specimens from three localities,representing a variety of elementary depositional sequences (EDS)from the Middle Lutetian strata of the Paris basin (Fig. 1).

2. Materials and methods

2.1. Geologic setting

The lists of themacro- andmicrofossils of the middle Lutetian of theParis Basin have been worked out to a large extent (see references inMerle, 2008, Lozouet, 2014). Notwithstanding the basin was locatedaround 30–35°N, sample-level species richness of the molluscan faunamatched values of modern tropical hotspots (Dominici and Zuschin,2016). The thickest part of this interval is composed of unlithifiedfine-grained hybrid arenites (in the sense of Zuffa, 1980), with a verylarge carbonate fraction of biogenic origin (mainly mollusc shells orforaminifer tests), a quartz component and occasional glauconite.Bedding is usually massive, and field correlations among distant out-crops are mainly based on the molluscs (Gely, 1996).

2.2. Stratigraphy

Wehave sampled themiddle Lutetian at three localities: La Fermedel'Orme, Grignon (both near Paris) and Fleury-la-Rivière (near Reims).Strata are roughly horizontal and individual unconformity-boundunits have been correlated between outcrops as far as 160 km (Fig. 1,modified from Dominici and Zuschin (2016), and references therein),following the detailed sedimentologic description of the “Falunière deGrignon” by Jean-Pierre Gély and Didier Merle (in Guernet et al.,2012), interpreted based on available multiproxy data (Huyghe et al.,2012, Guernet et al., 2012, Dominici and Zuschin, 2016). We distin-guished several elementary depositional sequences (EDS) from bottomto top:

EDS1-EDS2 contain bioturbated calcareous sandstone (3–4m), withabundant glauconite and quartz, and sparse shell material. The associa-tion is characterized by Cardium (Orthocardium) subporulosum. Atabular shell bed in the upper part contains sparse gravels and Turritellashell debris. Transgressive gravels are present at Fleury, in associationwith bones of marine vertebrates. EDS 2 sediments at Fleury areinterpreted as lower shoreface deposits. Two samples were analyzedin EDS2, at one locality (Fleury).

EDS 3 contains cross-bedded, locally bioturbated calcareous sand-stones (2 m), rich in quartz and glauconite, interpreted as an uppershoreface deposit, and a highly-bioturbated massive calcareoussandstone (1.5 m), with abundant quartz and glauconite. Loosely-packed shell beds are characterized by Campanile giganteum. The

Fig. 1.Middle Lutetian sedimentary strata at La Ferme-de-L'Orme, Grignon, and Fluery-La-Riviére, with position of bulk samples discussed in the text. Depositional sequences are those ofDominici and Zuschin (2016).

88 D. Chattopadhyay et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 459 (2016) 86–98

upper shell beds,with a species-richmollusc association, are dominatedby Sigmesalia intermedia (Courville et al., 2012). The upper part of EDS 3is interpreted as a lower shoreface deposit. One sample was analyzed inEDS3 (Grignon).

EDS 4 includes a massive calcareous sandstone (2 m) with sparseor densely-packed shell beds, resting on an irregular erosional surface.The associations are characterized by Chama lamellosa, abundantOrbitolites complanatus and high ostracod richness (Guernet et al.,2012), consistent with a lower shoreface paleoenvironment. These sedi-ments are overlain at Grignon by a massive calcareous sandstone(2.5 m), with sparse shell beds dominated by Seraphs sopitus andAvicularium lithocardium. Seraphs dominates the EDS also at La Fermede l'Orme in sediments from a lower shoreface environment. Four sam-ples were analyzed in EDS 4, at two localities (one sample at La Ferme,three samples at Grignon), all from the same paleoenvironment.

Table 1Depositional environment and ecological interpretation of samples used for this study.

Sample Locality EDS Sedimentary environmen

PB_9 Grignon 3 Lower shorefacePB_16 Fleury 2 Lower shorefacePB_27 Fleury 2 Upper shorefacePB_10 Ferme 4 Lower shorefacePB_1 Grignon 4 Lower shorefacePB_3 Grignon 4 Lower shorefacePB_6 Grignon 4 Lower shoreface

EDS 5 includes a massive calcareous sandstone (0.5 m thick) withlaterally continuous shell beds. The association is dominated byPotamides lapidorum, Vicinocerithium echidnoides and Saxolucinasaxorum. This unit, interpreted as an intertidal or shallow subtidal(upper shoreface) deposit, is overlain at La Ferme de l'Orme by amudstone with a freshwater fauna; it is overlain by a calcareoussandstone with a marine fauna (Merle, 2008), representing EDS 6(Fig. 1) (Table 1).

2.3. Environmental interpretation of the samples and localities

All samples were collected in sediments from marine, subtidalbottoms, some at shallower (upper shoreface) other at deeper settings(lower shoreface). Themolluscan assemblages differ significantly. Sam-ples PB9, PB16 and PB27 (EDS 2–3) have a high proportion of turritellid

t Ecological conditions

Oligotrophic, stable conditions in the photic zone, open marineMesotrophic, relatively more unstable conditions, open marineMesotrophic, relatively more unstable conditions, closer to the coastOligotrophic, stable conditions in the photic zone, open marineOligotrophic, stable conditions in the photic zone, open marineOligotrophic, stable conditions in the photic zone, open marineOligotrophic, stable conditions in the photic zone, open marine

Table 2List of species and their drilling frequency. T = Total, G = Grignon, F = Fleury la Riviere, L = Le Ferme de Lorne.

Class Family Species Drilling frequency

T G L F

Bivalve Arcidae Arca (A.) biangula 0.53 0.53 0.00 0.00Arcidae Barbatia (B.) barbatula 0.47 0.47 0.00 0.00Arcidae Barbatia (Rostarca) angusta 0.04 0.06 0.00 0.00Arcidae sp. indet. 1 0.00 0.00 0.00 0.00Aviculariidae Avicularium cymbula 0.00 0.00 0.00 0.00Cardiidae/Cardiinae Cardium (Orthocardium) stilpnaulax 0.29 0.29 0.00 0.00Cardiidae/Cardiinae Loxocardium bouei 0.00 0.00 0.00 0.00Cardiidae/Cardiinae Loxocardium formosum 0.03 0.00 0.00 0.04Cardiidae/Cardiinae Parvicardium plagiomorphum 0.00 0.00 0.00 0.00Cardiidae/Cardiinae Plagiocardium passyi 0.00 0.00 0.00 0.00Carditidae/Carditamerinae Arcturellina elegans 0.42 0.57 0.00 0.00Carditidae/Carditamerinae Arcturellina ornata 0.47 0.47 0.00 0.00Carditidae/Carditamerinae Arcturellina pulchra 0.32 0.50 0.00 0.29Carditidae/Carditamerinae Arcturellina squamosa 0.00 0.00 0.00 0.00Carditidae/Carditesinae Paraglans calcitrapoides 0.33 0.40 0.00 0.00Carditidae/Miodomeridinae Miodomeris (M.) astartoides 1.00 1.00 0.00 0.00Carditidae/Miodomeridinae Miodomeris (M.) modica 0.24 0.24 0.00 0.00Carditidae/Venericardiinae Venericardia imbricata 0.00 0.00 0.00 0.00Carditidae/Venericardiinae Venericardia profunda 0.23 0.00 0.00 0.23Corbulidae Bicorbula gallica 0.00 0.00 0.00 0.00Corbulidae Caryocorbula pixidicula 0.13 0.13 0.00 0.00Corbulidae Notocorbula rugosa 0.34 0.00 0.00 0.36Corbulidae Varicorbula minuta 0.11 0.15 0.00 0.00Crassatellidae Bathytormus lamellosa 0.21 0.24 0.00 0.00Crassatellidae Bathytormus sp. indet. 0.00 0.00 0.00 0.00Crassatellidae Crassatina (Chattonia) laevigata 0.30 0.31 0.00 0.00Crassatellidae Crassatina (Chattonia) triangularis 0.26 0.29 0.00 0.00Donacidae Donax (Tembrockia) nitidus 0.00 0.00 0.00 0.00Glycymeridae Glycymeris pulvinata 0.32 0.33 0.00 0.18Lasaeidae/Erycininae Erycina (E.) calyculata 0.00 0.00 0.00 0.00Lasaeidae/Erycininae Erycina (E.) grignonensis 0.00 0.00 0.00 0.00Lasaeidae/Erycininae Erycina (E.) irregularis 0.00 0.00 0.00 0.00Limarcidae Cnisma nuculata 0.00 0.00 0.00 0.00Limopsidae Limopsis (Pectunculina) granulata 0.33 0.33 0.00 0.00Lucinidae/Divaricellinae Divalinga (Stchepinskya) rigaultiana 0.00 0.00 0.00 0.00Lucinidae/Lucininae Parvilucina (P.) pusilla 0.00 0.00 0.00 0.00Lucinidae/Milthinae “Gibbolucina” brevifulcrata 0.00 0.00 0.00 0.00Lucinidae/Milthinae Gibbolucina callosa 0.00 0.00 0.00 0.00Mactridae Spisula (Austromactra) loustauae 0.67 1.00 0.00 0.00Mactridae Spisula (Austromactra) recondita 0.00 0.00 0.00 0.00Mytilidae Brachidontes (B.) jansseni 0.00 0.00 0.00 0.00Mytilidae Musculus (Skarlatella) seminudus 0.00 0.00 0.00 0.00Noetiidae/Striarcinae Scapularca scapulina 0.57 0.86 0.00 0.00Noetiidae/Striarcinae Striarca (Arcopsis) decipiens 0.55 0.55 0.00 0.00Noetiidae/Trinacriinae Trigonodesma bezanconi 0.11 0.11 0.00 0.00Noetiidae/Trinacriinae Trigonodesma multidentata 0.00 0.00 0.00 0.00Noetiidae/Trinacriinae Trinacria cancellata 0.00 0.00 0.00 0.00Noetiidae/Trinacriinae Trinacria deltoidea 0.00 0.00 0.00 0.00Nuculanidae Nuculana (Saccella) galeottiana 0.00 0.00 0.00 0.00Nuculidae Nucula (N.) parisiensis 0.50 0.00 0.50 0.00Ostreaidae Cubitostrea plicata 0.15 0.17 0.00 0.11Ostreaidae Cubitostrea sp. indet. 0.16 0.00 0.00 0.04Psammobiidae Gari (Psammodonax) caillati 0.12 0.12 0.00 0.00Pteriidae Pteria (Electroma) trigonata 0.00 0.00 0.00 0.00Tellinidae Tellina (Elliptotellina) tellinella 0.00 0.00 0.00 0.00Veneridae/Meretricinae Tivelina analoga 0.17 0.17 0.00 0.00Veneridae/Meretricinae Tivelina baudoni 0.00 0.00 0.00 0.00Veneridae/Meretricinae Tivelina gibbosula 0.00 0.00 0.00 0.00Veneridae/Pitarinae Callista (B.) elegans 0.00 0.00 0.00 0.00Veneridae/Pitarinae Callocardia (Nitidavenus) nitidula 0.14 0.15 0.00 0.13Veneridae/Pitarinae Costacallista laevigata 0.00 0.00 0.00 0.00Veneridae/Pitarinae Pitar (P.) distincta 0.00 0.00 0.00 0.00Veneridae/Pitarinae Pitar (Paradione) ovalina 0.12 0.12 0.00 0.00Veneridae/Tapetinae Katelysia (Textivenus) texta 0.67 0.89 0.00 0.00Veneridae/Tapetinae Venerella secunda 0.30 0.00 0.00 0.30Vermetidae Serpulorbis clathratus 0.05 0.05 0.00 0.00Vermetidae Tenagodus lima 0.06 0.07 0.00 0.00Vermetidae Vermetus (Pyxipoma) praestigiosus 0.00 0.00 0.00 0.00

Gastropod Aclididae Graphis (G.) bouryi 0.00 0.00 0.00 0.00Acteonidae/Acteoninae Acteon (A.) loustaui 0.00 0.00 0.00 0.00Amnicolidae Lapparentia (L.) fischeri 0.08 0.08 0.00 0.00Ampullinidae Gen. et sp. indet. 0.00 0.00 0.00 0.00Calliostomatidae/Solariellinae Solariella (S.) turbinoides 0.00 0.00 0.00 0.00

(continued on next page)

89D. Chattopadhyay et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 459 (2016) 86–98

Table 2 (continued)

Class Family Species Drilling frequency

T G L F

Calyptraeidae Calyptraea (Sigapatella) lamellosa 0.00 0.00 0.00 0.00Cerithiidae Bittium (B.) semigranosum 0.19 0.19 0.00 0.18Cerithiidae Bittium (B.) transenna 0.20 0.00 0.00 0.36Cerithiidae Ptychocerithium inabsolutum 0.00 0.00 0.00 0.00Cerithiidae Serratocerithium denticulatum 0.00 0.00 0.00 0.00Cerithiidae Tenuicerithium goossensi 0.11 0.11 0.00 0.00Conidae/Coninae Hemiconus lineatus 0.00 0.00 0.00 0.00Conidae/Coninae Hemiconus stromboides 0.00 0.00 0.00 0.00Conidae/Conorbinae Cryptoconus sp. indet. 0.00 0.00 0.00 0.00Conidae/Mangeliinae Mangelia (M.) labratula 0.00 0.00 0.00 0.00Costellaridae Conomitra cancellina 0.10 0.10 0.00 0.00Costellaridae Vexillum (Uromitra) terebellum 0.13 0.14 0.00 0.00Costellariidae Conomitra marginata 0.00 0.00 0.00 0.00Cylichnidae Cylichna (Acrostemma) coronata 0.00 0.00 0.00 0.00Cylichnidae Cylichna (C.) bruguierei 0.00 0.00 0.00 0.00Cylichnidae Cylichna (C.) caillati 0.00 0.00 0.00 0.00Diastomatidae ?Diastoma 0.00 0.00 0.00 0.00Fasciolaridae/Fasciolariinae Clavilithes (Rhopalithes) rugosus 0.00 0.00 0.00 0.00Fissurellidae/Emarginulinae/Emarginulini Emarginula costata 0.05 0.05 0.00 0.00Marginellidae Dentimargo dentifera 0.00 0.00 0.00 0.00Marginellidae Gibberula ovulata 0.30 0.33 0.00 0.00Marginellidae Volvarinella chastaingi 0.00 0.00 0.00 0.00Marginellidae Volvarinella contabulata 0.00 0.00 0.00 0.00Mitridae Mitraria (B.) crebricosta 0.00 0.00 0.00 0.00Mitridae Mitraria (M.) plicatella 0.00 0.00 0.00 0.00Naticidae/Naticinae ?Euspira sp. 1 0.00 0.00 0.00 0.00Naticidae/Naticinae Natica epiglottina 0.00 0.00 0.00 0.00Naticidae/Naticinae Natica microglossa 0.00 0.00 0.00 0.00Naticidae/Naticinae Payraudeautia perforata 0.04 0.07 0.00 0.00Neritidae Semineritina mammaria 0.00 0.00 0.00 0.00Nystiellidae Tenuiscala (Cerithiscala) cloezi 0.07 0.07 0.00 0.00Olividae/Ancillariinae Amalda (Baryspira) olivula 0.00 0.00 0.00 0.00Olividae/Ancillariinae Ancillus (Turrancilla) glandinus 0.00 0.00 0.00 0.00Olividae/Olivinae Olivancillaria (Pseudolivella) mitreola 0.13 0.13 0.00 0.00Olividae/Olivinae Olivancillaria (Pseudolivella) parisiensis 0.00 0.00 0.00 0.00Omalaxidae Omalaxis bifrons 0.62 0.93 0.00 0.00Omalaxidae Omalaxis cresnensis 0.00 0.00 0.00 0.00Omalaxidae Omalaxis marginatus 0.51 0.57 0.00 0.44Omalaxidae Omalaxis serratus 0.07 0.00 0.00 0.07Phasianellidae Tricolia (Phasianochilus) turbinoides 0.25 0.25 0.00 0.00Pseudomelaniidae Bayania lactea 0.00 0.00 0.00 0.00Pseudomelaniidae Bayania sp. 2 0.00 0.00 0.00 0.00Pyramidellidae/Odostomiinae Odostomia (O.) modesta 0.00 0.00 0.00 0.00Pyramidellidae/Odostomiinae Odostomia (O.) turbonilloides 0.00 0.00 0.00 0.00Pyramidellidae/Odostomiinae Odostomia sp. indet. 0.00 0.00 0.00 0.00Pyramidellidae/Syrnolinae Syrnola (Pachysyrnola) 0.00 0.00 0.00 0.00Pyramidellidae/Syrnolinae Syrnola (Puposyrnola) acicula 0.00 0.00 0.00 0.00Pyramidellidae/Syrnolinae Tiberia (Cossmannica) emarginata 0.67 0.67 0.00 0.00Pyramidellidae/Turbonillinae/Eulimidini Eulimella (Belonidium) rouaulti 0.00 0.00 0.00 0.00Rostellariidae Hippocrenes sp. 0.00 0.00 0.00 0.00Rostellariidae Rimella fissurella 0.02 0.00 0.00 0.03Tornidae/Circuliinae Circulus (C.) planorbularis 0.00 0.00 0.00 0.00Tornidae/Teinostomatinae Teinostoma (Megatyloma) dubium 0.00 0.00 0.00 0.00Tornidae/Teinostomatinae Teinostoma (T.) striatissimum 0.00 0.00 0.00 0.00Triphoridae Triphora aspera 0.00 0.00 0.00 0.00Triphoridae Triphora inversa 0.00 0.00 0.00 0.00Trochidae/Monodontinae Phorculus sulcatus 0.00 0.00 0.00 0.00Trochidae/Trochinae Jujubinus (Strigosella) lamarckii 0.00 0.00 0.00 0.00Turbinidae/Colloninae Cirsochilus grignonensis 0.67 1.00 0.00 0.00Turbinidae/Colloninae Cirsochilus macrostoma 0.14 0.14 0.00 0.00Turbinidae/Colloninae Cirsochilus striatus 0.00 0.00 0.00 0.00Turbinidae/Colloninae Collonia (C.) marginata 0.57 0.57 0.00 0.00Turbinidae/Colloninae Cyniscella spiruloides 0.00 0.00 0.00 0.00Turridae ?Epalxis sp. 1 0.00 0.00 0.00 0.00Turridae/Crassispirinae Crassispira (Tripia) propeangulosa 0.00 0.00 0.00 0.00Turridae/Crassispirinae Crassispira (Tripia) sp. 2 0.00 0.00 0.00 0.00Turridae/Crassispirinae Crassispira (Tripia) sulcata 0.00 0.00 0.00 0.00Turridae/Crassispirinae Gen. et sp. indet. 0.00 0.00 0.00 0.00Turridae/Turrinae Eopleurotoma insueta 0.00 0.00 0.00 0.00Turridae/Turrinae Eopleurotoma morelleti 0.29 2.00 0.00 0.00Turridae/Turrinae Eopleurotoma plicaria 0.00 0.00 0.00 0.00Turridae/Turrinae Gen. et sp. indet. 0.00 0.00 0.00 0.00Turritellidae Haustator caillati 0.13 0.13 0.00 0.00Turritellidae Haustator elegans 0.38 0.38 0.00 0.00Turritellidae Haustator funiculosus 0.28 0.00 0.00 0.29

90 D. Chattopadhyay et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 459 (2016) 86–98

Table 2 (continued)

Class Family Species Drilling frequency

T G L F

Turritellidae Haustator lamarckii 0.17 0.17 0.00 0.00Turritellidae Haustator mitis 0.18 0.22 0.00 0.18Turritellidae Sigmesalia consobrina 0.48 0.48 0.00 0.00Turritellidae Sigmesalia melanoides 0.50 0.00 0.00 0.50Turritellidae Sigmesalia sp. indet. 0.00 0.00 0.00 0.00Turritellidae Sigmesalia trochoides 0.00 0.00 0.00 0.00Volutidae/Volutilithinae Gen. et sp. indet. 0.00 0.00 0.00 0.00

91D. Chattopadhyay et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 459 (2016) 86–98

gastropods and a relatively low proportion of predatory gastropods.Dominance is higher than in other samples, suggesting that these faunasrepresent settings richer in nutrients, like those presently preferred byturritellids (Allmon 2001). On the contrary, samples PB1, PB6 andPB10 (EDS 4) have the highest rarefied richness, consistent with faunasfrom oligotrophic settings (Table 3 in Dominici and Zuschin, 2016).Their stratigraphic position, corresponding to an interval richer incarbonates typical of clear subtropical waters, confirms this interpreta-tion. Samples are best interpreted as forming a subtidal gradient fromhigher (PB9, PB16, PB27) to lower (PB1, PB3, PB10) nutrient conditions(Table 1). In enclosed basins such as the Paris Basin, not directly facingthe deep sea, nutrients are brought in by rivers, and inputs are seasonal,some seasons being more rainy than others. During a time of generaltransgression, such as the middle Lutetian, subtle sea level changecould trigger faunal variations, with oligotrophic predictable environ-ments at times of transgression (4th order, in coincidence of EDS 4:Dominici and Zuschin, 2016 and references therein) and mesotrophicmore unpredictable settings during regressive phases (EDS 2–3). Theassociation of PB1, PB3 and PB10 with a carbonate-rich interval isconsistent with this interpretation. In summary, locality La Ferme del'Orme consists of one sample (PB10), representative of a vegetated bot-tom, locality Grignon mostly represents oligotrophic, environmentallystable conditions (PB 1, PB 3, PB 6) with one of the samples being indic-ative of nutrient enrichment (PB 9), and locality Fleury-la-Rivière,finally, consists of 2 samples, both indicative of more nutrient-richconditions.

2.4. Material collection and processing

Seven bulk samples, collected from shell beds at three classicmiddleLutetian localities (Grignon, La Ferme de l'Orme, Fleury-la-Rivière) inthe Paris Basin, are used for this study. All samples were processedusing stacked 1 mm, 2 mm and 4 mm sieves. Because of overwhelmingshell densities inmany of the samples, they were split and between onehalf and one sixteenth of the total shell material was used. Drillingpredation analyses are based on proportional data and consequentlythey should not be affected by differences in sample size as long as thenumber of individuals considered is sufficiently high.

Mollusc shells (including identifiable fragments) were picked,sorted, and identified to species-level using Cossmann and Pissarro

Fig. 2. Molluscan drilling frequencies for each size fraction within the Paris Basins. A, Bivalvesdefined by 25th and 75th quantiles; error bars represent 95% confidence intervals.

(1904–1906, 1910–1913). Fragments were excluded from drillingfrequency analyses. For drilling analyses, shells were consideredwhole if they were approximately 90% or more complete (Zuschinet al., 2003).

Analyses were performed on sieve treatments (a proxy for method-ological size filtering). In other words, in the sieve treatment approach,all shells in the 1mm sieve (1–2mm fraction)were counted for drillinganalyses and compared to all of the shells in the 2 mm (2–4 mmfraction) and the 4 mm (N4 mm fraction) sieves.

2.5. Drilling intensities

Each complete shell was examined for predatory drill holes. Thefollowing widely accepted criteria were used to identify such drills asbeing predatory in nature: holes should have smooth sides and be circu-lar in cross section, penetrate from the outside of the shell, perpendicu-larly to the shell surface, and only one valve in articulated bivalves(Carriker and Yochelson, 1968; Rohr, 1991; Baumiller, 1996; Kaplanand Baumiller, 2000; Leighton, 2001). Drilling frequency (DF) wascalculated by dividing the number of shells with complete drill holesby the total number of shells for gastropods (Kowalewski, 2002). Forall the disarticulated valves of bivalves, the frequency of drilling preda-tionwas calculated bydividing the number of bored valves by the half oftotal number of valves in the collection (Kowalewski, 2002).

Drilling frequency (DF) = ND / (N ∗ 0.5)whereND = number of valves with complete drill holeN = total number of valves.The prey effectiveness (PE), however, was calculated by dividing the

total number of incompletely drilled valves by the total number ofdrilled valves (complete and incomplete) present in the collection(Vermeij, 1987).

Prey effectiveness (PE) = NID / (NID + ND)whereNID = number of valves with incomplete drill holeND = number of valves with complete drill hole.We considered a drill hole to be edge-drilled if the boundary of the

hole intersected the commissural line.We calculated per species DF for each locality for various size

fractions. We also identified twelve molluscan species (three bivalvespecies and nine gastropod species) that are present in all threelocalities and compared their DF.

; B, Gastropods. The axis label denotes the lower limit of the size fraction. The boxes are

Fig. 3.Drilling frequency of molluscs among various localities. A. Bivalves, B. Gastropods. The boxes are defined by 25th and 75th quantiles; error bars represent 95% confidence intervals.

92 D. Chattopadhyay et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 459 (2016) 86–98

2.6. Statistical treatment

We compared the DF among size classes and among localities withthe chi-square test. We used Spearman Rank correlation to measurethe correlation of DF with PE. We also estimated the effect of ordinalvariables (stratigraphy, sieve size, locality) in per-species per-stationmethod with simple (analyzing each predictor separately) and withmultiple generalized linear models (GLMs) (analyzing all predictorssimultaneously and evaluating their partial contributions to the totalvariation in DF) (Quinn and Keough, 2002).We used the logit-link func-tion and quasibinomial variance functions suitable for modeling re-sponse variables represented by proportions (McCullagh and Nelder,1989). This model allows the response variables (DF) to range between0 and 1; it also works where errors might not be normally or uniformlydistributed. The “quasi” function adds a parameter to the variance ifdata are under- or over dispersed relative to the variance expectedfrom binomial processes. All statistical analyses were performed in R(R Core Team, 2012).

Fig. 4. Drilling frequency for bivalves among various localities. A. 1–2 mm, B. 2–4 mm.There were not enough specimens in the largest size category (N4 mm) for comparisonbetween localities. The boxes are defined by 25th and 75th quantiles; error barsrepresent 95% confidence intervals.

3. Results

18,861 shells from 116 species were examined, of which 3485were drilled (Table 2). Overall DF is 17% and bivalves show a higheroverall DF (18%) than gastropods (14%) and scaphopods (b0.1%). Ex-cept for scaphopods and gastropods (W = 198, p-value = 0.073), thedifferences are statistically significant (bivalves & scaphopods: W =159, p-value = 0.031; bivalves & gastropods: W = 3026.5, p-value =0.041). Because of the low abundance of scaphopods, the detailed statis-tical analyses are restricted to bivalves and gastropods. The patterns ofDFs across size fractions are different for both classes. Both bivalvesand gastropods from the 2–4 mm fraction are drilled significantly

Fig. 5.Drilling frequency for gastropods among various localities. A. 1–2mm, B. 2–4mm, C.N4 mm. The boxes are defined by 25th and 75th quantiles; error bars represent 95%confidence intervals.

Table 3Summary of the differences in drilling predation between localities.

Combined Bivalve Gastropod

Total Drill Incomplete DF PE Total Drill Incomplete DF PE Total Drill Incomplete DF PE

Grignon 2195 354 40 0.252 0.101 1583 230 33 0.29 0.125 612 124 7 0.20 0.053Le Ferme de Lorne 414 1 0 0.003 0 219 1 0 0.009 0 195 0 0 0 0Fleury la Riviere 1713 213 15 0.178 0.066 1030 87 9 0.169 0.094 683 126 6 0.184 0.045

93D. Chattopadhyay et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 459 (2016) 86–98

more than those from the 1–2 mm and N4 mm fractions, which arestatistically not significantly different from one another (Fig. 2).

We find a significant difference in DF among the three studiedlocalities,with highest values in Grignon (overall 25%, bivalves 29%, gas-tropods 20%), followed by Fleury-la-Rivière (overall 18%, bivalves 17%,gastropods 18%) and almost a complete lack of drillholes in La Fermede l'Orme. This pattern is true for all the size classes ofmolluscs, bivalvesand gastropods (Figs. 3–5) (Table 3).When the same analysis is restrict-ed to the common genera found at all three localities, a similar pattern isevident (Fig.6). All of these differences are statistically significant(Tables 4–6) except for the differences between Grignon and Fleury-la-Rivière for larger size class (N4 mm).

The stratigraphic units, EDS 2 and 3 (mesotrophic), showa slightlyhigher DF compared to EDS 4 (oligotrophic) (Fig. 7); However, thisdifference is not statistically significant (W = 3048, p-value =0.1).The overall DF tends to vary across class, locality, stratigraphicunit and sieve size (Fig. 8). Whereas locality appears to be a significantcontributor in both simple and multiple GLM, stratigraphic unit is onlysignificant in multiple GLM (Table 7).

PE is slightly higher in Fleury-la-Rivière compared to Grignon forgastropods and there is no strong correlation between DF and PE(Fig. 9A, B). Overall PE varies across class, locality, stratigraphic unitand sieve size (Fig. 10). Edge drilling was restricted to single valves ofGlycymeris pulvinata, Gibbolucina (G.) gibbosula, Nucula mixta fromGrignon, 16 valves of Venericardia profunda from Fleury-la-Rivière anda single valve of Nucula parisiensis from La Ferme de l'Orme.

4. Discussion

Drilling predation on molluscan prey from Cenozoic rocks of theParis basin (e.g., Grignon) was first reported by Fischer (1922, 1960,1963, 1965). He identified naticid drill holes in bivalves of differentsizes and documented a decrease in drilling with increasing size classfrom Grignon (Fischer, 1960). The drilling frequency (unadjusted forbivalves) ranges from 27.8% (2–5 mm) to 6% (20–25 mm); there wascomplete absence of successful drilling among bivalves that are largerthan 30 mm. He also studied species-level difference in DF by compar-ing two species of Corbula (Fischer, 1963). He found a significantly

Fig. 6. Drilling frequency among three localities for common species. The boxes aredefined by 25th and 75th quantiles; error bars represent 95% confidence intervals.

higher attack frequency (complete and incomplete drillholes) (49%) inCorbula gallica in comparison to C. rugosa (30%); however, the DF(considering only complete drill holes) of these two species is similar.The drilling pattern from Eocene rocks of the Paris basin was alsostudied by Taylor (1970). His study was based on a collection of 40species (including eight predatory gastropods) from Calcaire Grossier(Lutetian) from Damery and showed an overall DF of 21% with higherDF for bivalves (26%) compared to gastropods (17%). Although hisdata were gained by extracting shells from a large block, using needlepicks, probes and other tools (pers. comm. April 2015), these valuesmatchwell with our results fromGrignon and Fleury-la-Rivière. In addi-tion, we are able to document large variation in DF between size classesand localities.

4.1. Effect of sieve size

The effects of size filtering on studies of ecological interactions maybe quite important in some settings. Due to size-specific selection ofprey by their predators (either directly or indirectly by selecting specificspecies that fall within certain limits), a severe bias may creep inthrough the sievingprotocol (Kowalewski, 2002). Naticids, in particular,are known for a strong size preference while choosing their prey(Edwards and Huebner, 1977; Berry, 1982; Kitchell, 1986; Allmonet al., 1990; Kingsley-Smith et al., 2003a, 2003b; Chattopadhyay andDutta, 2013). Such effects could be especially strong in assemblagesdominated by either large or small predatory drillers. For example,naticid gastropods in the Northern Adriatic Sea tend to be small andattack smaller prey (e.g., Sawyer and Zuschin, 2010). If drilling frequen-cy were evaluated from the Northern Adriatic Sea using larger sievemesh sizes alone, many drilled shells, as well as their drillers, wouldbe discarded and results would erroneously indicate low drillingfrequencies. Such mesh size practices are not uncommon. In a recentstudy evaluating muricid drilling patterns from southern SouthAmerica, Martinelli et al. (2013) only considered the molluscs largerthan 10mm. In another study that evaluated the reliability of encounterfrequency to predict predation frequency, Martinelli et al. (2015), con-sidered molluscan shells larger than 4 mm. They concluded that lackof selectivity demonstrated by the predators in the samples can onlybe applied to the prey size range considered and cannot be generalizedsince drilling gastropods are size selective. Likewise, removal of smallshells in communities dominated by large predators may artificially in-flate DFs by removing small shells that were not drilled by larger pred-ators. Despite this potential dependence of predation frequencies tosize-selection, few studies have examined the effects of size filteringon drilling intensities (see Kowalewski and Hoffmeister, 2003; Visaggiand Kelley, 2015). Our study also demonstrates a significant change inDF in various size classes. For both bivalves and gastropods, themediumsize class has the highest drilling frequency. Such difference in DFbetween size classes could be caused by specific preference of the pred-ator for a) certain ontogenetic stages or b) particular prey species. If it isdriven by ontogenetic selection, then we would expect to see patternsfor all families. Except for Turridae, however, we do not find any signif-icant difference in drilling among size classes for families (Table 1).Therefore, the observed pattern is unlikely due to the preferencetowards an ontogenetic stage. Hence, the difference in DF observedbetween size classes seems to have developed by a preference for

Table 4Summary of χ2 test results showing the differences in drilling frequency between localities at a size class of 1-2 mm. G= Grignon, F = Fleury la Riviere, L = Le Ferme de Lorne. The sig-nificant results are in bold.

Overall Bivalve Gastropod

G vs F χ2 = 0.001P = 0.9

χ2 = 7.86P = 0.005

χ2 = 4.61P = 0.003

G and L χ2 = 20.3P b 0.0001

χ2 = 10.93P b 0.0001

χ2 = 9.33P = 0.002

F and L χ2 = 22.2P b 0.0001

χ2 = 4.01P = 0.04

χ2 = 15.85P b 0.0001

Table 5Summary of χ2 test results showing the differences in drilling frequency between localities at a size class of 2-4 mm. G= Grignon, F = Fleury la Riviere, L = Le Ferme de Lorne. The sig-nificant results are in bold.

Overall Bivalve Gastropod

G vs F χ2 = 3.16P = 0.07

χ2 = 6.79P = 0.009

χ2 = 0.8P = 0.3

G and L χ2 = 42.85P b 0.0001

χ2 = 19.69P b 0.0001

χ2 = 27.17P b 0.0001

F and L χ2 = 32.42P b 0.0001

χ2 = 10.78P = 0.001

χ2 = 22.52P b 0.0001

Table 6Summary of χ2 test results showing the differences in drilling frequency between localities at a size class of N4 mm. G = Grignon, F = Fleury la Riviere, L = Le Ferme de Lorne. Thesignificant results are in bold.

Overall Bivalve Gastropod

G vs F χ2 = 19.86P b 0.0001

χ2 = 10.32P = 0.001

χ2 = 9.58P = 0.001

G and L χ2 = 39.1P b 0.0001

χ2 = 15.13P = 0.0001

χ2 = 36.48P b 0.0001

F and L χ2 = 19.44P b 0.0001

χ2 = 7.85P = 0.005

χ2 = 17.39P b 0.0001

Fig. 7. Plot showing the variation in drilling frequency among two stratigraphic unitsrepresenting mesotrophic (EDS 2 and 3) and oligotrophic (EDS 4) conditions. The boxesare defined by 25th and 75th quantiles; error bars represent 95% confidence intervals.

94 D. Chattopadhyay et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 459 (2016) 86–98

certain species that have theirmaximumabundance in that size class. Inother words, the majority of the preferred mollusc species in this studybelong to the medium size class (2–4 mm). The data on distribution ofpreferred species in various size classes support this claim. The firstfive species of bivalves with the highest overall DF make up 76% in themedium size class; the other two size classes account for 20% (small)and 4% (large) respectively. Some bivalve species with high DF such asStriarca (Arcopsis) decipiens and Arca (A.) biangular exclusively appearin the medium size class. The first five species of gastropods with thehighest overall drilling frequencies make up 48% in the medium sizeclass and the other two size classes account for 27% (small) and 24%(large).

4.2. Effect of spatial variation

Spatial variation in DF found in the present study ranges greatlyfor both bivalves and gastropods (Figs. 5–6). Such variation has been re-ported from fossil (Hansen and Kelley, 1995) and modern environ-ments (Vermeij, 1987; Sawyer and Zuschin, 2010; Visaggi and Kelley,2015).

The reason behind such spatial variation in DF could be manifold.Absence or low abundance of predatory gastropods could result in alow DF. The poleward decrease in predation intensity is often attributedto such decreases in diversity and abundance of predatory groups intemperate zones (Taylor and Taylor, 1977; Vermeij and Dudley, 1982;Vermeij, 1987). However, it is difficult to explain the variation in DFbetween these three localities from the perspective of predatory abun-dance. Six species of naticid gastropods (Amauropsina canaliculata,Natica microglossa, Payraudeautia perforate, Natica epiglottina and twospecies of Euspira) were found and at least two of these species arepresent at each locality. Hence, the extremely low DF observed in LaFerme de l'Orme is not easily explained as this locality has a comparable

proportion of naticid gastropods like Grignon and Fleury-la-Rivière. In-terestingly, this locality has the highest number of naticid individuals(171) among the three localities. The gastropod species richness isalso similar at all three localities. The only naticid gastropod that wasalso common between this study and the report by Taylor (1970) isNatica epiglottina. Conspicuously, this species is absent in La Ferme del'Orme.

Prey availability could also play an important role in dictating thedrilling pattern of an assemblage. Hansen and Kelley (1995) havedemonstrated a strong positive correlation between DF and availabilityof preferred prey. In an experiment with live naticid gastropods,Chattopadhyay et al. (2014b) also demonstrated a significant changein predatory behaviour including a drop in DF with unavailability ofpreferred prey. As these assemblages have only naticids among thedrilling gastropods, the preferred prey item is bivalves (more specifical-ly infaunal bivalves). Interestingly, the proportion of bivalve prey in

Fig. 8. Plot showing the variation in drilling frequency among A. Class, B. Locality, C. Stratigraphic unit and D. Sieve size. The boxes are defined by 25th and 75th quantiles; error barsrepresent 95% confidence intervals.

95D. Chattopadhyay et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 459 (2016) 86–98

these three environments (Grignon: 64%; La Ferme de l'Orme: 45%;Fleury-la-Rivière: 53%) has a positive correlation with DF, showing apossible dependence of predatory attack on availability of prey.However, the difference in DF among twelve molluscan species (threebivalve species and nine gastropod species) common to all threelocalities (Fig. 6) cannot be explained by unavailability of preferredprey species. Moreover, it is important to mention that on multivariateanalysis of a larger set of samples collected in the same sections, foundthat stratigraphic units explained the variability in species-abundancecomposition better than localities (Dominici and Zuschin, 2016).

Some people attributed spatial variation in drilling intensity tovariation in environmental parameters such as temperature, salinityand substrate (Hansen and Kelley, 1995; Hoffmeister and Kowalewski,2001; Visaggi and Kelley, 2015). There are subtle differences indepositional environments among the three localities (Table 1).Thepaleoecological analysis of these three different successions, based ona matrix of abundance data of molluscan species that includes theseven samples here under study, indicates that the Lutetian of LaFerme de l'Orme and Fluery-La Rivière represents shallowerpaleoenvironments, closer to terrestrial inputs, with respect to Grignon.

Table 7The results of GLM on the relationship between DF and various parameters. A. SimpleGLM, B. Multiple GLM.

Estimate Std. error t value Pr(N|t |)

AIntercept −1.1181 0.3271 −3.4184 0.0008Class −0.3723 0.2017 −1.8456 0.0670Intercept −1.0276 0.3154 −3.2578 0.0014Locality −0.3792 0.1719 −2.2057 0.0290Intercept −1.6005 0.2553 −6.2686 0.0000Stratigraphic unit −0.0003 0.0008 −0.4357 0.6637Intercept −1.7022 0.2202 −7.7297 0.0000Sieve size −0.0002 0.0977 −0.0017 0.9987

BIntercept −0.1079 0.5498 −0.1962 0.8447Class −0.3298 0.2033 −1.6223 0.1070Locality −1.8093 0.5231 −3.4590 0.0007Stratigraphic unit 0.0067 0.0020 3.3234 0.0011Sieve size 0.0652 0.1040 0.6271 0.5316

Within the Grignon succession, samples PB1 and PB3 from EDS 4 indi-cate oligotrophic paleoenvironments, typical of very stable tropicalwaters, whereas PB27 and PB16 collected from EDS 2 and 3 at Fleuryrepresent more unpredictable settings in the vast middle LutetianParis gulf (Dominici and Zuschin, 2016). Sample PB10 from EDS 4 atLa Ferme de l'Orme is interpreted as from low-energy subtidal bottoms,possibly shallower and closer to terrestrial inputs with respect toGrignon. The proximity to terrestrial input may have had a significantimpact on predatory behaviour as low salinity is well documented tohave adverse effects on the drilling ability of Recentmuricid gastropods(Manzi, 1970; Garton and Stickle, 1980) and is often used as a predationrefuge (Witman and Grange, 1998). Salinity fluctuations have beenclaimed to affect the efficiency of the accessory boring organ (ABO)and thereby decreasing the effectiveness of drilling behaviour(Hoffmeister and Kowalewski, 2001). However, PB10, according to spe-cies composition and diversity, represents an environment very similarto that of PB1, PB3 and PB6 at Grignon (Dominici and Zuschin, 2016).The conspicuous difference in drilling predation among these three lo-calities, therefore, is hardly entirely controlled by environmentalattributes.

4.3. Other possible source for natural variability

One important aspect to note in explaining natural variation in DF isthe role of taphonomy. A tacit assumption for the biological explanationbehind such variation is that it reflects the original intensity of biogenicinteractions, which may not always be a safe assumption as there aretaphonomic alterations that significantly change DF. The hydrodynamicproperties of shells vary significantly with changes in their externalmorphology (Chattopadhyay et al., 2013a). The presence of drillholesmay further alter the hydrodynamic properties of a shell and couldsignificantly alter the resulting drilling frequency of an assemblageaffected by transportation (Lever et al., 1961; Chattopadhyay et al.,2013b). Roy et al. (1994) found that drilled bivalves are weaker underpointload compression than are undrilled shells. However, Zuschinand Stanton (2001) found that drill holes produced no taphonomicbias during sediment compaction experiments and other studies sup-port this finding (Kelley, 2008; Hagstrom, 1996).In the present study,the chances of such alteration is considered minimal, as we did not

Fig. 9. Plot showing the relationship between drilling frequency and prey-effectiveness for A. Bivalves and B. Gastropods.

Fig. 10. Plot showing the variation in prey effectiveness among A. Class, B. Locality, C. Stratigraphic unit and D. Sieve size. The boxes are defined by 25th and 75th quantiles; error barsrepresent 95% confidence intervals.

96 D. Chattopadhyay et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 459 (2016) 86–98

find any evidence of large-scale transport or destruction of molluscanshells.

5. Implication

Predatory drilling traces are routinely used to evaluate temporalpatterns in animal behaviour. Unlike previous studies on temporalchanges in DF, our study documents the extent of other naturalvariations and emphasizes the need of future research on such variabil-ities of predation proxies. It is important to note that the spatialvariation detected in this study is comparable in magnitude withmany temporal shifts reported previously (Kelley and Hansen, 2003for review). Therefore, studies on the nature and gradient of spatial var-iability in predatory behaviour are essential to establish the credibilityof temporal trends.

In order to shed light on Eocene paleoenvironments, species compo-sition of 12 samples from different localities (including all samples used

Table 8Summary of various drillingmetrics for bivalves of thewarm-temperate northern AdriaticSea, the subtropical northern Red Sea and the Paris basin.

DF PE EDF

Northern Adriatic Sea 23.4 10.3 0.02Northern Red Sea 23 b0.01 1.5Paris basin 18 0.1 b0.01

in the present study) were compared with modern analogues, conclud-ing that subtropical conditions reached as far north as the Paris Basin(Dominici and Zuschin, 2016). A similar approach can be tried forpredation intensity by comparing it with Recent intensities fromtropical and temperate regions. It has been claimed that there shouldbe a difference in drilling intensity between tropical and temperateregions. We therefore compare data on bivalve drilling predation ofthis study with corresponding modern data from the northern Red Sea(Chattopadhyay et al., 2014a, 2015) and northern Adriatic Sea(Sawyer and Zuschin, 2010). This comparison shows that results arevery different, depending on the metric considered (Table 8). Drillingfrequency is much higher in the Adriatic Sea and in the Red Sea thanin the Paris basin. Prey, effectiveness, however, is outstandingly highin the Adriatic Sea, while EDF is by far highest in the Red Sea. Consider-ing the strong patchiness of drilling predation within each of these ba-sins and reported from many other studies, we conclude that aclimatic affiliation of the Eocene Paris basin based on drilling metricsis not appropriate.

Acknowledgements

This studywas supported by project P19013-B17 of the Austrian Sci-ence Fund (FWF). D.C. acknowledges the financial support of IISER Kol-kata (ARF 2014-15) and Ernst Mach Research Grant (Zl.ICM-2014-07215). The comments of Patricia Kelley significantly improved themanuscript.

97D. Chattopadhyay et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 459 (2016) 86–98

References

Allmon, W.D., Nieh, J.C., Norris, R.D., 1990. Drilling and peeling of Turritelline gastropodssince the Late Cretaceous. Palaeontology 33, 595–611.

Allmon, W.D., 2001. Nutrients, temperature, disturbance, and evolution: a model for thelate Cenozoic marine record of the western Atlantic. Palaeogeogr. Palaeoclimatol.Palaeoecol. 166, 9–26.

Anderson, L.C., 1992. Naticid gastropod predation on corbulid bivalves: effects of physicalfactors, morphological features and statistical artifacts. PALAIOS 7, 602–617.

Ansell, A.D., 1960. Observations on predation of Venus striatula (Da Costa) by Naticaalderi (Forbes). J. Molluscan Stud. 34, 157–164.

Baumiller, T.K., 1996. Boreholes in the Middle Devonian blastoid Heteroschisma and theirimplications for gastropod drilling. Palaeogeogr. Palaeoclimatol. Palaeoecol. 123, 434-351.

Berry, A.J., 1982. Predation by Natica maculosa Lamarck (Naticidae: Gastropoda) upon thetrochacean gastropod Umbonium bestiarum (L.) on a Malaysian shore. J. Exp. Mar. Biol.Ecol. 64, 71–89.

Carriker, M.R., Yochelson, E.L., 1968. Recent gastropod boreholes and Ordoviciancylindrical borings. United States Geological Survey Professional Paper 593-B,pp. 1–26.

Chattopadhyay, D., Baumiller, T.K., 2009. An experimental assessment of penetration, ex-cavation and consumption rates of the muricid gastropod Nucella lamellosa.J. Shellfish Res. 28, 1–7.

Chattopadhyay, D., Baumiller, T.K., 2010. Effect of durophagy on drilling predation: a casestudy of Cenozoic molluscs from North America. Hist. Biol. 22, 367–379.

Chattopadhyay, D., Dutta, S., 2013. Prey selection by drilling predators: a case study fromMiocene of Kutch, India. Palaeogeogr. Palaeoclimatol. Palaeoecol. 374, 187–196.

Chattopadhyay, D., Rathie, A., Das, A., 2013a. The effect of morphology on post-mortemtransportation of bivalve and its taphonomic implications. PALAIOS 28, 203–209.

Chattopadhyay, D., Rathie, A., Miller, D., Baumiller, T.K., 2013b. Hydrodynamic effects ofdrill holes on post-mortem transportation of bivalve shells and its taphonomic impli-cations. PALAIOS 28, 875–884.

Chattopadhyay, D., Sarkar, D., Dutta, S., Prasanjit, S.R., 2014b. What controls cannibalismin drilling gastropods: a case study on Natica tigrina. Palaeogeogr. Palaeoclimatol.Palaeoecol. 410, 126–133.

Chattopadhyay, D., Zuschin, M., Tomašových, A., 2014a. Effects of a high-risk environmenton edge-drilling behavior: inference from Recent bivalves from the Red Sea.Paleobiology 40, 34–49.

Chattopadhyay, D., Zuschin, M., Tomasovych, A., 2015. How effective are ecological traitsagainst drilling predation? Insights from Recent bivalve assemblages of the northernRed Sea. Palaeogeogr. Palaeoclimatol. Palaeoecol. 440, 659–670.

Cossmann, M., Pissarro, G., 1904–1906. Iconographie Complète des Coquilles Fossiles deL'Éocène des Environs de Paris. Pélécypodes Tome 1 (Hermann, Paris, 45 pls.).

Cossmann, M., Pissarro, G., 1910–1913. Iconographie Complète des Coquilles Fossiles deL'Éocène des Environs de Paris. Scaphopodes, Gastropodes, Brachiopodes,Céphaloüpodes & Supplément Tome 2 (Hermann, Paris, 65 pls.).

Courville, P., Pacaud, J.-M., Merle, D., Lebrun, P., 2012. Le Lutétien de Damery: géologie,environnements, associations de bivalves et gastéropodes. Fossiles hors-série III57–71.

DeAngelis, D.L., Kitchell, J.A., Post, W.M., Travis, C.C., 1984. A model of naticid gastropodpredator-prey coevolution. In Math. Ecol. 54, 120–136.

DeAngelis, D.L., Kitchell, J.A., Post, W.M., 1985. The influence of naticid predation onevolutionary strategies of bivalve prey: conclusions from a model. Am. Nat. 126,817–842.

Dominici, S., Zuschin, M., 2016. Palaeocommunities, diversity and sea-level change frommiddle Eocene shell beds of the Paris Basin. J. Geol. Soc. http://dx.doi.org/10.1144/jgs2015-150.

Dudley, E.C., Vermeij, G.J., 1978. Predation in time and space; drilling in the gastropodTurritella. Paleobiology 4, 436–441.

Edwards, D.C., Huebner, J.D., 1977. Feeding and growth rates of Polinices duplicatuspreying on Mya arenaria at Barnstable Harbor, Massachusetts. Ecology 58,1218–1236.

Fischer, P.H., 1922. Sur les gastéropode perceurs. J. Conch. 67, 3–56.Fischer, P.H., 1960. Action des gastéropodes perceurs sur un bivalve de l'Etage Lutetien.

J. Conch. 100, 129–131.Fischer, P.H., 1963. Corbules fossiles perforées per des gastéropodes prédateurs. J. Conch.

103, 29–31.Fischer, P.H., 1965. Perforations de fossiles tertiares per des gastéropodes prédateurs.

J. Conch. 105, 66–96.Garton, D., Stickle, W.B., 1980. Effects of salinity and temperature on the predation rate of

Thais haemastoma on Crassostrea virginica spat. Biol. Bull. 158, 49–57.Gély, J.-P., 1996. Le Lutétien du Bassin Parisien: De l’analyse séquentielle haute résolution

à la reconstitution paléogéographique. Bull. Inf. Géol. Bassin Paris 34, 3–27.Guernet, C., Huyghe, D., Lartaud, F., Merle, D., Emmanuel, L., Gély, J.-P., Michel, F., Pilet, O.,

2012. La faune d'ostracodes de la falunière de Grignon (Lutétien du Bassin de Paris):implications stratigraphiques et paléogéographiques. Geodiversitas 34, 909–959.http://dx.doi.org/10.5252/g2012n4a12.

Hagstrom, K.M. 1996. Effects of compaction andwave-induced forces on the preservationand macroevolutionary perception of naticid predator-prey interactions.Unpublished M.S. Thesis. Indiana University.

Hansen, T.A., Kelley, P.H., 1995. Spatial variation of naticid gastropod predation in theEocene of North America. PALAIOS 10, 268–278.

Harper, E.M., Forsythe, G.T., Palmer, T., 1998. Taphonomy and the Mesozoic marine revo-lution; preservation state masks the importance of boring predators. PALAIOS 13,352–360.

Hattori, K.E., Kelley, P.H., Dietl, G.P., Moore, N.O., Simpson, S.L., Zappulla, A.M., Ottens, K.J.,Visaggi, C.C., 2014. Validation of taxon-specific sampling by novice collectors forstudying drilling predation in fossil bivalves. Palaeogeogr. Palaeoclimatol. Palaeoecol.412, 199–207.

Hoffman, A., Martinell, J., 1984. Prey selection by naticid gastropods in the Pliocene ofEmporda (Northeast Spain). Neues Jb. Geol. Paläontol. Monat. 7, 393–399.

Hoffmeister, A.P., Kowalewski, M., 2001. Spatial and environmental variation in the fossilrecord of drilling predation: a case study from the Miocene of Central Europe.PALAIOS 16, 566–579.

Hubbell, S.P., 2001. The unified neutral theory of biodiversity and biogeography (MPB-32). 32. Princeton University Press.

Huyghe, D., Castelltort, S., Mouthereau, F., Cheype, F., Emmanuel, L., 2012. Middle Lutetianclimate in the Paris Basin: implications for a marine hotspot of paleobiodiversity.Facies 58, 587–604.

Kaplan, P., Baumiller, T.K., 2000. Taphonomic inferences on boring habit in theRichmondian Onniella meeki epibole. PALAIOS 15, 499–510.

Kelley, P.H., 1988. Predation by Miocene gastropods of the Chesapeake Group: stereo-typed and predictable. PALAIOS 3, 436–448.

Kelley, P.H., 1991. Apparent cannibalism by Chesapeake Group naticid gastropods: A pre-dictable result of selective predation. J. Paleontol. 65, 75–79.

Kelley, P.H., Hansen, T.A., 1993. Evolution of the naticid gastropod predator-prey system:an evaluation of the hypothesis of escalation. PALAIOS 8, 358–375.

Kelley, P.H., Hansen, T.A., 1996. Naticid gastropod prey selectivity through time and thehypothesis of escalation. PALAIOS 11, 437–445.

Kelley, P.H., 2008. Role of bioerosion in taphonomy: effect of predatory drillholes onpreservation of mollusc shells. Current Developments in Bioerosion. Springer BerlinHeidelberg, pp. 451–470.

Kelley, P.H., Hansen, T.A., 2003. The fossil record of drilling predation on bivalves andGastropods. In: Kelley, P.H., Kowalewski, M., Hansen, T.A. (Eds.), Predator-Prey Inter-actions in the Fossil RecordTopics in Geobiology Series 20. Plenum Press/Kluwer,New York, pp. 113–133.

Kingsley-Smith, P.R., Richardson, C.A., Seed, R., 2003a. Stereotypic and size-selectivepredation in Polinices pulchellus (Gastropoda: Naticidae) Risso 1826. J. Exp. Mar.Biol. Ecol. 295, 173–190.

Kingsley-Smith, P.R., Richardson, C.A., Seed, R., 2003b. Size-related and seasonal patternsof egg collar production in Polinices pulchellus (Gastropoda: Naticidae) Risso 1826.J. Exp. Mar. Biol. Ecol. 295, 191–206.

Kitchell, J.A., 1986. The evolution of predator–prey behavior: naticid gastropods andtheir molluscan prey. In: Nitecki, M., Kitchell, J.A. (Eds.), Evolution of Animal Be-havior: Paleontological and Field Approaches. Oxford University Press, Oxford,pp. 88–110.

Kitchell, J.A., 1990. The reciprocal interaction of organism and effective environment:learning more about ‘and.'. In: Ross, M, Allmon, WB (Eds.), Causes of evolution: a pa-leontological perspective. University of Chicago Press, Chicago, pp. 151–169.

Kitchell, J.A., Boggs, C.H., Kitchell, J.F., Rice, J.A., 1981. Prey selection by naticid gastropods:experimental tests and application to the fossil record. Paleobiology 7, 533–552.

Kohn, A.J., Arua, I., 1999. An early Pleistocene molluscan assemblage from Fiji: Gastropodfaunal composition, paleoecology and biogeography. Palaeogeogr. Palaeoclimatol.Palaeoecol. 146, 99–145.

Kowalewski, M., 2002. The fossil record of predation: an overview of analytical methods.Paleontological Society Special Papers. 8, pp. 3–42.

Kowalewski, M., Dulai, A., Fürsich, F.T., 1998. A fossil record full of holes: the Phanerozoichistory of drilling predation. Geology 26, 1091–1094.

Kowalewski, M., Hoffmeister, A.P., 2003. Sieves and fossils: effects of mesh size onpaleontoligical patterns. PALAIOS 18, 460–469.

Leighton, L.R., 2001. New example of Devonian predatory boreholes and the influence ofbrachiopod spines on predator success. Palaeogeogr. Palaeoclimatol. Palaeoecol. 165,53–69.

Lever, J., Kessler, A., Van Over Beeke, P., Thijssen, R., 1961. Quantitative beach research II,the “hole effect”: a second mode of sorting of lamellibranch valves on sandy beaches.Neth. J. Sea Res. 1, 339–358.

Lozouet, P., 2014. Temporal and latitudinal trends in the biodiversity of European AtlanticCenozoic gastropod (Mollusca) faunas. A base for the history of biogeographic prov-inces. Carnets Geol. 14, 273–314.

Manzi, J.J., 1970. Combined effect of salinity and temperatureon the feeding, reproductiveand survival rates of Eupleura caudata (Say) (Prosobranchia: Muricidae). Biol. Bull.138, 35–46.

Marincovich Jr., L., 1977. Cenozoic Naticidae (Mollusca: Gastropoda) of the northeasternPacific. Bull. Am. Paleontol. 70, 169–497.

Martinelli, J.C., Gordillo, S., Archuby, F., 2013. Muricid drilling predation at high latitudes:insights from southernmost Atlantic. PALAIOS 28, 33–41.

Martinelli, J.C., Kosnik, M.A., Madin, J.S., 2015. Encounter frequency does not predictpredation frequency in tropical dead-shell assemblages. PALAIOS 30, 818–826.

McCullagh, P., Nelder, J.A., 1989. Generalized Linear Models. 2nd ed. Chapman and Hall,London.

Merle, D., 2008. Stratotype Lutétien. Biotope BRGM, Paris (288 pp.).Ottens, K.J., Dietl, G.P., Kelley, P.H., Stanford, S.D., 2012. A comparison of analyses of

drilling predation on fossil bivalves: bulk- vs. taxon-specific sampling and therole of collector experience. Palaeogeogr. Palaeoclimatol. Palaeoecol. 319–320,84–92.

Quinn, G.P., Keough, M.J., 2002. Experimental Design and Data Analysis for Biologists.Cambridge University Press, Cambridge.

R Core Development Team, 2012. R: a language and environment for statistical comput-ing. R Foundation for Statistical Computing, Vienna www.R-project.org.

Rohr, D.M., 1991. Borings in a shell of an Ordovician (Whiterockian) gastropod.J. Paleontol. 65, 687–688.

98 D. Chattopadhyay et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 459 (2016) 86–98

Roy, K., Miller, D.J., Labarbera, M., 1994. Taphonomic bias in analysis of drilling predation:effects of gastropod drill holes on bivalve shell strength. PALAIOS 9, 413–421.

Sawyer, J.A., Zuschin, M., 2011. Drilling predation in mollusks from the Lower and MiddleMiocene of the Central Paratethys. PALAIOS 26, 284–297.

Sawyer, J.A., Zuschin, M., 2010. Intensities of drilling predation of molluscan assemblagesalong a transect through the northern Gulf of Trieste (Adriatic Sea). Palaeogeogr.Palaeoclimatol. Palaeoecol. 285, 152–173.

Taylor, J.D., 1970. Feeding habits of predatory gastropods in a tertiary (Eocene) molluscanassemblage from the Paris basin. Palaeontology 23, 254–260.

Taylor, J.D., Taylor, C.N., 1977. Latitudinal distribution of predatory gastropods on theeastern Atlantic shelf. J. Biogeogr. 4, 73–81.

Vermeij, G.J., 1980. Drilling predation of bivalves in Guam: some paleoecological implica-tions. Malacologia 19, 329–334.

Vermeij, G.J., 1987. Evolution and Escalation: An Ecological History of Life. PrincetonUniversity Press, Princeton, NJ (527 pp.).

Vermeij, G.J., Dudley, E.C., 1982. Shell repair and drilling in some gastropods from theRipley Formation (Upper Cretaceous) of Southeastern U.S.A. Cretac. Res. 3, 397–403.

Vermeij, G.J., Zipser, E., Dudley, E.C., 1989. Successful and unsuccessful drilling predationin Recent pelecypods. Veliger 32, 266–273.

Visaggi, C.C., Kelley, P.H., 2015. Equatorward increase in naticid gastropod drilling preda-tion on infaunal bivalves from Brazil with paleontological implications. Palaeogeogr.Palaeoclimatol. Palaeoecol. 438, 285–299.

Witman, J.D., Grange, K.R., 1998. Links between rain, salinity, and predation in a rockysubtidal community. Ecology 79, 2429–2447.

Zuffa, G.G., 1980. Hybrid arenites: their composition and classification. J. Sediment. Res.50, 21–29.

Zuschin, M., Stanton Jr., R.J., 2001. Experimental measurement of shell strength and itstaphonomic interpretation. Palaios 16, 161–170.

Zuschin, M., Stachowitsch, M., Stanton Jr., R.J., 2003. Patterns and processes of shell frag-mentation in modern and ancient marine environments. Earth Sci. Rev. 63, 33–82.

Zuschin, M., Harzhauser, M., Mandic, O., 2011. Disentangling palaeodiversity signals froma biased sedimentary record: An example from the Early to Middle Miocene of Cen-tral Paratethys Sea. In: McGowan, A., Smith, A.B. (Eds.), Comparing the Geological andFossil Records: Implications for Biodiversity Studies 358. Geological Society, London,Special Publications, pp. 123–139.