A complete Holocene record of trematode–bivalveinfection and implications for the response ofparasitism to climate changeJohn Warren Huntleya,1, Franz T. Fürsichb, Matthias Albertic, Manja Hethkeb, and Chunlian Liud

aDepartment of Geological Sciences, University of Missouri, Columbia, MO 65211; bFachgruppe PaläoUmwelt, GeoZentrum Nordbayern, Friedrich-Alexander-Universität Erlangen-Nürnberg, 91054 Erlangen, Germany; cInstitut für Geowissenschaften, Christian-Albrechts-Universität zu Kiel, 24118 Kiel, Germany;and dSchool of Earth Science and Geological Engineering, Sun Yat-sen University, 510275 Guangzhou, China

Edited by John M. Pandolfi, The University of Queensland, St Lucia, QLD, Australia and accepted by the Editorial Board November 13, 2014 (received forreview August 29, 2014)

Increasing global temperature and sea-level rise have led toconcern about expansions in the distribution and prevalence ofcomplex-lifecycle parasites (CLPs). Indeed, numerous environmen-tal variables can influence the infectivity and reproductive outputof many pathogens. Digenean trematodes are CLPs with interme-diate invertebrate and definitive vertebrate hosts. Global warmingand sea level rise may affect these hosts to varying degrees, andthe effect of increasing temperature on parasite prevalence hasproven to be nonlinear and difficult to predict. Projecting theresponse of parasites to anthropogenic climate change is vital forhuman health, and a longer term perspective (104 y) offered by thesubfossil record is necessary to complement the experimental andhistorical approaches of shorter temporal duration (10−1 to 103 y).We demonstrate, using a high-resolution 9,600-y record of trema-tode parasite traces in bivalve hosts from the Holocene Pearl RiverDelta, that prevalence was significantly higher during the earlieststages of sea level rise, significantly lower during the maximumtransgression, and statistically indistinguishable in the otherstages of sea-level rise and delta progradation. This stratigraphicpaleobiological pattern represents the only long-term high-resolu-tion record of pathogen response to global change, is consistentwith fossil and recent data from other marine basins, and is in-structive regarding the future of disease. We predict an increase intrematode prevalence concurrent with anthropogenic warmingand marine transgression, with negative implications for estuarinemacrobenthos, marine fisheries, and human health.

parasites | disease | global warming | paleoecology | mollusks

Predictions for the increase of global temperature and con-comitant sea-level rise in the coming decades and centuries

have led to concerns about the range expansion or shift of par-asites and pathogens to higher latitudes, resulting in an increasein prevalence and severity of diseases. Indeed, environmentalfactors such as temperature, humidity, and precipitation canstrongly influence the infectivity and reproductive output ofmany pathogens (1, 2); however, the life cycles of parasites canbe quite complex, and accurately predicting how their prevalencewill be influenced by these factors can be difficult. For example,digenean trematodes are complex-lifecycle parasites (CLPs) withmultiple intermediate invertebrate hosts and a definitive verte-brate host. Environmental change may affect each of these hoststo varying degrees, such that a modest increase in temperaturecould result in an increase of parasite reproductive output andinfectivity. However, a more severe increase in temperaturecould be beyond the thermal tolerance of one of the hosts, thusultimately resulting in a decrease in prevalence (3–5). Still otherstudies have produced quantitative models and compared spatialand temporal (seasonal- to historical-scale) aspects of climaticand other environmental variables with outbreak data in an at-tempt to disentangle the controls of disease, and it seems thatthe prediction of a simple range expansion of disease should not

be expected (6). Rather, because taxa have upper and lowertolerances to environmental variables, one should expect a rangeshift rather than expansion; moreover, an expanding range is nota given because other factors (e.g., competition or predation) cantheoretically regulate such expansions (6). Understanding thelikely response of CLPs to anthropogenic climate change is vitalfor human health, but previous efforts have been limited toa shorter temporal scale of observation (10−1 to 103 y). Here, weoffer a complementary longer term perspective (104 y) to un-derstand how CLP prevalence has responded to global change inthe recent geologic past.The fossil record, although sometimes criticized for its in-

complete nature and varying temporal resolution of its strata(time averaging), can be a very reliable chronicle of the history oflife and biotic interactions. Such records are especially reliablewhen fossil datasets are amassed in a high-resolution strati-graphic context wherein temporal trends of diversity or bioticinteractions can be disentangled from shifting environments ofdeposition (7). The subfossil record of the Holocene has provento be particularly useful in providing high temporal resolutionrecords of environmental change and biotic response, bridgingthe temporal divide between traditional ecological and geo-logical studies (8–11). Attempts to quantitatively investigate thefossil record of CLPs are limited in that the preservationalpotential of parasites is typically very low (although they haveoccasionally been preserved) (12) and samples sizes tend to be

Significance

There is growing concern about how parasites will respond toclimate change. Previous studies were based on data fromshort time scales (10−1 to 103 y), but here we present, to ourknowledge, the first analysis of the response of parasites toglobal change on a longer time scale (104 y), utilizing thesubfossil record. A 9,600-y record of clams and flatworm par-asites from the Pearl River Delta exhibits a significant spike inparasite prevalence during the initial phase of sea-level rise.This increase is not related to changes in salinity or intermediatehost availability. Temperature and productivity cannot be quan-tified and tested as driving factors. These results suggest starkimplications for macrobenthos, fisheries, and human health in thecontext of climate change.

Author contributions: J.W.H., F.T.F., M.A., M.H., and C.L. designed research; J.W.H., F.T.F.,M.A., M.H., and C.L. performed research; J.W.H. analyzed data; and J.W.H. wrotethe paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission. J.M.P. is a guest editor invited by the EditorialBoard.1To whom correspondence should be addressed. Email: huntleyj@missouri.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1416747111/-/DCSupplemental.

18150–18155 | PNAS | December 23, 2014 | vol. 111 | no. 51 www.pnas.org/cgi/doi/10.1073/pnas.1416747111

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 8,

202

1

quite small; however, in some cases, it can be possible to greatlyincrease sample size by documenting parasite-induced malfor-mations on hosts that have significantly greater preservation po-tential. Digenean trematodes induce the growth of characteristicpits and blisters on the interior of their bivalve mollusk secondintermediate hosts (Fig. 1). These traces are readily preserved inthe fossil record (10, 13, 14) and indicate that trematodes haveindeed been completing their life cycle by infesting multipleinvertebrate and vertebrate hosts in deep time. Trematodes areresponsible for many diseases of both humans (e.g., schistoso-miasis and swimmer’s itch by freshwater Schistosomatidae) andnonhuman animals, but not all trematodes are known to inducetraces in their hosts. The majority of studies addressing trace-makeridentity seem to suggest that the Gymnophallidae (typically Parva-trema and Gymnophallus) are primarily responsible for the in-duction of pit formation in bivalve hosts (Table S1); therefore, ourresults should serve as a proxy for the response of this diversedigenean family.

Here we take advantage of a high-resolution complete Holo-cene record from the Pearl River Delta (China), which containsan abundant record of trematode-induced malformations inbivalves, (i) to determine how trematode prevalence has respon-ded to eustatic sea-level change (and its environmental correlates)in the recent geologic record and (ii) to evaluate the prospects fortrematode prevalence within the context of anthropogenic climatechange and sea-level rise.

ResultsPRD-10 is a 24-m-long core sample that was drilled in the centerof the Pearl River plain from which were collected 168 samplesthat record complete deposition through the Holocene (15). Tenenvironmental zones within this core were established basedupon molluscan, ostracod, foraminiferal, and lithological dataand reveal, at the coarse scale, a protracted sea-level rise fol-lowed by delta progradation and lobe switching (see Table 1 fordetailed analysis) (15). A detrended correspondence analysis

Fig. 1. Photomicrographs of trematode-induced malformations on the interior of P. amurensis valves. (A) Whole specimen from sample 154 with shallowpits. (B) Partial specimen from sample 154 with deep pits. (C) Partial specimen from sample 157 displaying pits on multiple growth layers. (D) Incipientsteinkern from sample 162 displaying pits preserved as positive relief on lower half of specimen. Note how the pits are preserved quite readily on bivalvesfrom a broad spectrum of taphonomic grade.

Huntley et al. PNAS | December 23, 2014 | vol. 111 | no. 51 | 18151

EART

H,A

TMOSP

HER

IC,

ANDPL

ANET

ARY

SCIENCE

SEC

OLO

GY

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 8,

202

1

(DCA) of molluscan presence/absence data identified a primarygradient related to salinity, and the stratigraphic sequence

of sample DC1 scores reveals a dynamic history of initialtransgression (sequence boundary at base of zone B followed by

Table 1. Environmental zones of the PRD-10 core, ages, sample numbers, environment of deposition (15), and sequence stratigraphicinterpretation

Zone Age, cal y BP Sample no. Environment of depositionSequence stratigraphic

interpretation

H <200 4-1 Floodplain HSTG 2,300–200 30-5 Moderately brackish, minor fluvial and major tidal influences HSTF 3,200–2,300 49-31 Lowermost brief transgression, followed by fluvial-dominated

setting punctuated by transgressions and brackish water,transgressions likely the result of delta lobe switching

HST

E 4,200–3,200 70-50 High energy, freshwater dominant, near river mouth HSTD 6,800–4,200 99-71 Decrease in water depth and salinity, delta progradation HSTC3 >6,800 113-100 Peak transgression, highest depth and salinity values MFZ

TSTC2 ̶ 117-114 Moderately brackish, 3 graded beds indicate reworking and

pulses of transgressionTST

C1 <9,300 147-118 Moderately brackish, low-moderate energy, peak in molluscanrichness, increasing abundance and richness of ostracods

TST

B 9,600–9,300 159-148 Upper estuary TSTFSSB

A >9,600 168-160 Pedogenically altered deltaic sediments HST

cal y BP, calendar years before present; FS, flooding surface; HST, highstand systems tract; MFZ, maximum flooding zone; SB, sequence boundary; TST,transgressive systems tract.

Fig. 2. Detrended correspondence analysis of molluscan presence/absence data. (A) DC1 vs. DC2 scatterplot of scores for samples (gray boxes) and taxa(colored circles: blue, marine tolerant; green, brackish; yellow, freshwater). (B) Temporal sequence of sample DC1 scores (black line, raw data; red line, threepoint moving average). Note the strong decrease in sample DC1 scores between zones C3 and D indicating the transition from the maximum flooding zonewithin the transgressive systems tract (peak of relative sea-level rise) to the highstand systems tract and delta progradation (relative fall in sea level).

18152 | www.pnas.org/cgi/doi/10.1073/pnas.1416747111 Huntley et al.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 8,

202

1

flooding surface), continued transgression (zones C1–C2), max-imum transgression (maximum flooding zone in zone C3), anda forced regression through delta progradation with fluctuationsin salinity and depth controlled by delta lobe switching (zonesD–G) (Fig. 2).We examined Potamocorbula amurensis (n = 3,984), an op-

portunistic species that lives in shallow waters near estuaries,for trematode-induced malformations. Ninety-two out of 137(67.2%) species occurrences of P. amurensis preserved trematode-induced pits, proportional prevalence values ranged from 0.011 to0.397 in species occurrences with n ≥ 10 valves (Fig. S1), andprevalence values displayed statistically significant temporal andenvironmental variation. Median trematode prevalence valueswere significantly higher than expected by random chance [>99%confidence interval (CI)] during the initial onset of flooding (zoneB) that led to upper estuarine conditions at the location of PRD-10 (Fig. 3). The bootstrapped 95% CI for median prevalence andrandomization and bootstrap analysis of mean values produced

similar results (Fig. 3 and Fig. S2). Median and mean trematodeprevalence values were significantly lower than expected by ran-dom chance (>95% CI and >99% CI, respectively) for samplesfrom the maximum flooding zone (zone C3) (Fig. 3 and Fig. S2).The bootstrap analysis produced results consistent with therandomization. The median and mean prevalence values forall other zones are not statistically distinguishable from thatexpected by random chance (Fig. 3 and Fig. S2).

DiscussionWe have demonstrated a strong link between the initiation(∼300 y) of sea-level rise in the Pearl River system and a signif-icant increase in trematode prevalence. Indeed, this Holocenerecord may well serve as an example for what we can expect inthe coming decades and centuries for estuarine environmentssubject to sea-level rise. There has been a growing recognition ofthe influence of eustatic sea-level change (and the resultingpatterns of sediment accumulation) upon biotic patterns pre-served in the fossil record. Indeed, the stratigraphic paleobiologyparadigm provides a rigorous methodology for disentanglingbiotic history from stratigraphic bias (7). A number of ecologicalvariables in estuaries (e.g., relative abundances of taxa, salinity,productivity, and temperature, among others) can covary withsea-level change, and it is possible to rule out some of thesevariables as factors influencing the increase of trematode prev-alence. There is no significant correlation between P. amurensisabundance and trematode prevalence (Spearman’s r = 0.06, P =0.62) (Fig. 4), suggesting that the changing availability of in-termediate host individuals through time did not influenceparasite prevalence. Due to the lack of a comparable shore-bird fossil record in these strata, it is impossible to assess theinfluence of temporal variation in definitive host abundance onparasite prevalence. There is no significant correlation betweenDC1 sample scores and trematode prevalence (Pearson’s r =0.05, P = 0.67) (Fig. 4), suggesting that changes in salinitythrough time did not influence prevalence either. We currentlyhave no proxy for productivity and cannot assess its role ina quantitatively meaningful way. Liu et al. (16) have analyzedstable isotope values of ostracods from a nearby core but havedetermined that δ18O variation is the result of fluctuating salinityvalues; therefore, we are unable to test the contributing roleof temperature to trematode prevalence at this time.By the end of the 21st century, mean sea-surface temperature

is expected to increase 1–3 °C, and the global mean sea level isexpected to rise 0.44–0.74 m (17). Numerous studies have

Fig. 3. Median arcsine-transformed prevalence values by zone for samplesincluding at least 10 valves. (A) Bootstrapped 95% confidence intervals. (B)Results of randomization (10,000 iterations). Dark gray, 95%; light gray, 99%.

Fig. 4. Relationships between arcsine-transformed trematode prevalence values and (A) host abundance and (B) sample DC1 score as a proxy for salinity forsamples with n ≥ 10 valves.

Huntley et al. PNAS | December 23, 2014 | vol. 111 | no. 51 | 18153

EART

H,A

TMOSP

HER

IC,

ANDPL

ANET

ARY

SCIENCE

SEC

OLO

GY

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 8,

202

1

suggested that the ranges of many pathogens will expand or shiftpoleward and potentially increase prevalence as a result of thisanthropogenic climate change (1–6), but lack a longer temporalperspective. The insights offered by our high-resolution Holo-cene data are that trematode-prevalence values in brackish set-tings will be significantly higher during the initial stage of rapidsea-level rise in the coming centuries and will significantly de-crease by the maximum flooding stage. An increase in trematodeprevalence will have significant implications for ecological andhuman health. Increased gymnophallid trematode prevalencecould harm populations of the dominant molluscan macrobenthosby castrating and consuming various organs of their molluscanintermediate hosts and potentially altering their behavior tomake them more susceptible to their own predators (13, 14, 18–20), thereby significantly altering the estuarine trophic web atmultiple levels. Moreover, estuaries serve as nurseries for manycommercially important marine fishes, and their juvenile pop-ulations are particularly sensitive to climatic variation (21). Sig-nificant alterations in the estuarine trophic web induced byincreased trematode prevalence coupled with anthropogenic cli-mate change could function as a double whammy for commercialfisheries.A strictly conservative interpretation of our data would be

limited to gymnophallid trematodes in estuarine environments;however, our results may have implications beyond the estuarybecause increased prevalence during transgression has beenidentified in very different environmental settings. Trematodeprevalence among marine bivalves from Pleistocene and Holo-cene strata of the Po River plain is significantly higher intransgressive systems tracts (TSTs) than in highstand systemstracts (HSTs) (10). Moreover, this pattern persists when com-paring modern bivalve death assemblages from TST-like andHST-like settings in the northern Adriatic (22). Huntley andScarponi (10) suggested that environmental disturbances (e.g.,high sedimentation rate and salinity fluctuations) were lesscommon in TST lower shoreface settings than in comparablefacies of the HST, making for more suitable conditions for ma-rine trematodes. Although the pattern of increased trematodeprevalence during sea-level rise in the Pearl River Delta estua-rine samples is consistent with the pattern identified among theNorthern Adriatic marine samples, Huntley and Scarponi’s (10,22) hypothesized mechanism does not satisfactorily explain thesenew data. Regardless of the mechanism producing the pattern,trematode prevalence is significantly greater during times oftransgression in both marine and estuarine settings than duringregression. It is, therefore, not unreasonable to explore theimplications of increased trematode prevalence in freshwatersettings during transgression. Perhaps we should expect to see anincrease in the prevalence of schistosomiasis accompanying sea-level rise in the coming decades and centuries. Indeed, Schisto-soma japonicum, the trematode genus responsible for schisto-somiasis in China, thrives in freshwater marshes and lakes, whichare suitable environments for their gastropod intermediate hosts(23). There is grave concern that the construction of the ThreeGorges Dam and its 600-km-long reservoir will serve as a vastnew habitat suitable for schistosome-bearing gastropods (24).Similarly, continued flooding of river valleys by sea-level riseshould increase the number of subdivided, yet connected,freshwater habitats in upper estuaries, thus expanding habitatssuitable for schistosomes and their hosts. This problem is ex-acerbated by the high density of humans (definitive hosts) in thePearl River Delta and the Guangdong Province (>100 millioninhabitants) (25). Great strides have been made in reducingschistosomiasis infections in China from ∼12 million people in1949 to less than 1 million in 2004 (24); however, the Holocenefossil record of trematode trace fossils seems to suggest thatthese gains are threatened by continued sea-level rise associatedwith anthropogenic climate change.

MethodsThe PRD-10 corewas drilled in the center of the Pearl River Delta plain (22° 43′22.0′′ N, 113° 14′ 42.0′′ E) to a depth of 24 m spanning the Holocene andpenetrating Pleistocene deposits. The 21 segments of core were cut in half;one half was used for facies analysis and the other half was sampled every10 cm (3-cm-thick samples) for micro- and macrofossil analysis for 168 sam-ples. Seventeen 10-cm-thick samples were taken for radiocarbon analyses.The radiocarbon dates suggest continuous deposition through the Holoceneand depositional rates of ∼0.32 cm/y in the Early and Late Holocene and∼0.19 cm/y in the Middle Holocene. Abundance data were collected forbivalves, gastropods, ostracods, and foraminifera. Paleocommunity anal-yses (nonmetric multidimensional scaling and cluster analyses) coupled withmacrofacies analysis allowed for the establishment of nine zones (A–I)and three subzones (C1–C3), indicating the depositional environments. Here,we consider zones and subzones A–G and refer to them simply as zones.See Alberti et al. (15) for details on the methodologies, results, and inter-pretations of these analyses.

The core was highly fossiliferous, and 18,274 mollusk shells were re-covered. The bivalves Cyrenodonax formosana and P. amurensis were by farthe most abundant bivalve, and indeed molluscan, taxa, constituting 71.8%and 25.7% of bivalve individuals, respectively. C. formosana (9,576 valves)and 3,984 valves of P. amurensis were inspected for pits and blisters in-dicative of trematode infestation (13) using a Zeiss V12 SteREO Discoverystereo microscope, and photomicrographs were taken with a Jenoptik SpeedXT5 digital camera. The oval-shaped pits with raised rims that are charac-teristic of trematode infestation were very common in P. amurensis, but notso common in C. formosana. Rather, C. formosana displayed purple blistersand discoloration ranging in morphology from the size and shape of trem-atode pits found in P. amurensis to linear raised purple streaks on the in-terior of the valves. Such discolorations and blisters have been linked totrematode infestation (26), but they are not as diagnostic as the pits.Therefore, the conservative approach to data interpretation dictates that wenot consider the C. formosana traces further in this study.

P. amurensis was present in most samples of the core, although not atconsistent abundances. Trematode-induced pits were similarly presentthrough much of the core although their abundance was not consistent (Fig.S1). Trematode prevalence was calculated as the proportion of valves withina single-species occurrence that exhibited trematode-induced pits or blisters.Further analyses of trematode prevalence values were restricted to singlespecies occurrences with 10 or more valves. Trematode prevalence valueswere arcsine-transformed and sorted by zone, and 95% confidence intervalswere bootstrapped with SAS 9.3. A randomization was conducted in SAS 9.3to explore the range of median and mean prevalence values expected byrandom chance. Trematode prevalence values were randomly shuffled intobins that mimicked the sample sizes of zones A, B, C1, C2, C3, D, E, and G,and the median or mean prevalence value (randomizations were run sepa-rately for each metric) was calculated for each bin. This process was con-ducted at 10,000 iterations, and 95% and 99% confidence intervals werecalculated (2.5 and 97.5 percentile and 0.5 and 99.5 percentile values fromeach bin’s distribution of 10,000 estimates of central tendency, respectively).Zone F was not included in the randomization because none of its single-species occurrences of P. amurensis contained more than 10 valves. Zone Hwas not included in the bootstrap or randomization simulations because itdid not have any occurrences of P. amurensis.

Detrended correspondence analysis (DCA) was conducted in PAST version2.17c on a matrix of molluscan (bivalve and gastropod) presence/absencedata. Abundance data were calculated as the number of individuals per 100 gof sediment. Taxon singletons were removed from thematrix before analysis.See ref. 15 for this data matrix. Abundance data were converted to presence/absence data to reduce the dominance of the ordination by the two highlydominant taxa C. formosana and P. amurensis. Select taxa were coded forsalinity tolerance based upon descriptions in ref. 15 and the World Registerof Marine Species database (27).

ACKNOWLEDGMENTS. The authors acknowledge discussions with K. De Baets,W. Kiessling, J. D. Schiffbauer, and the comments of the editor (J. Pandolfi)and two anonymous reviewers. J. Janoso assisted with imaging technology.Funding for this project was provided by the Alexander von Humboldt Stiftung(J.W.H.), the University of Missouri Research Council (J.W.H.), National ScienceFoundation of China Grants 40872024 and 40272011 (to C.L.), and DAAD-CSC (Deutscher Akademischer Austauschdienst–China Scholarship Council)(M.A., M.H., and F.T.F.).

18154 | www.pnas.org/cgi/doi/10.1073/pnas.1416747111 Huntley et al.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 8,

202

1

1. Harvell CD, et al. (2002) Climate warming and disease risks for terrestrial and marinebiota. Science 296(5576):2158–2162.

2. Poulin R (2006) Global warming and temperature-mediated increases in cercarialemergence in trematode parasites. Parasitology 132(Pt 1):143–151.

3. Paull SH, Johnson PTJ (2011) High temperature enhances host pathology in a snail–trematode system: Possible consequences of climate change for the emergence ofdisease. Freshw Biol 56:767–778.

4. Studer A, Thieltges DW, Poulin R (2010) Parasites and global warming: Net effects oftemperature on an intertidal host-parasite system. Mar Ecol Prog Ser 415:11–22.

5. Thieltges DW, Rick J (2006) Effect of temperature on emergence, survival and in-fectivity of cercariae of the marine trematode Renicola roscovita (Digenea: Re-nicolidae). Dis Aquat Organ 73(1):63–68.

6. Lafferty KD (2009) The ecology of climate change and infectious diseases. Ecology90(4):888–900.

7. Patzkowsky ME, Holland SM (2012) Stratigraphic Paleobiology: Understanding theDistribution of Fossil Taxa in Time and Space (Univ of Chicago Press, Chicago).

8. Dexter TA, et al. (2014) A continuous multi-millennial record of surficial bivalvemollusk shells from the São Paulo Bight, Brazilian shelf. Quat Res 81:274–283.

9. Kowalewski M (2009) The youngest fossil record and conservation paleobiology:Holocene shells as eco-environmental recorders. Conservation Paleobiology: Usingthe Past to Manage for the Future, eds Dietl G, Flessa K (Yale Univ Printing andPublishing Services, New Haven, CT), pp 1–23.

10. Huntley JW, Scarponi D (2012) Evolutionary and ecological implications of trematodeparasitism of modern and fossil northern Adriatic bivalves. Paleobiology 38:40–51.

11. Goodfriend GA, Gould SJ (1996) Paleontology and chronology of two evolutionarytransitions by hybridization in the Bahamian land snail Cerion. Science 274(5294):1894–1897.

12. Fry GF, Moore JG (1969) Enterobius vermicularis: 10,000-year-old human infection.Science 166(3913):1620.

13. Ruiz GM, Lindberg DR (1989) A fossil record for trematodes: Extent and potentialuses. Lethaia 22:431–438.

14. Ituarte CF, Cremonte F, Deferrari G (2001) Mantle-shell complex reactions elicited bydigenean metacercariae in Gaimardia trapesina (Bivalvia: Gaimardiidae) from theSouthwestern Atlantic Ocean and Magellan Strait. Dis Aquat Organ 48(1):47–56.

15. Alberti M, Hethke M, Fürsich FT, Liu C (2013) Macro- versus microfauna: Resolutionpotential of bivalves, gastropods, foraminifera and ostracods in reconstructing theHolocene palaeoenvironmental evolution of the Pearl River delta, China. Palaeo-biogeogr. Palaeoenv 93:327–353.

16. Liu C, et al. (2013) Late Quaternary palaeoenvironmental changes documented bymicrofaunas and shell stable isotopes in the southern Pearl River Delta plain, SouthChina. J Palaeogeogr 2:344–361.

17. Intergovernmental Panel on Climage Change (2013) Climate Change 2013: The PhysicalScience Basis, eds Stocker TF, et al. (Cambridge Univ Press, Cambridge, UK). Available atwww.ipcc.ch/report/ar5/wg1/. Accessed 2 July 2014.

18. Swennen C (1969) Crawling-tracks of trematode infectedMacoma balthica (L.). Neth JSea Res 4:376–379.

19. Curtis LA (1987) Vertical distribution of an estuarine snail altered by a parasite. Sci-ence 235(4795):1509–1511.

20. Thomas F, Poulin R (1998) Manipulation of a mollusc by a trophically transmitted para-site: Convergent evolution or phylogenetic inheritance? Parasitology 116(Pt 5):431–436.

21. Attrill MJ, Power M (2002) Climatic influence on a marine fish assemblage. Nature417(6886):275–278.

22. Huntley JW, Scarponi D (2015) Geographic variation of parasitic and predatory traceson mollusks in the northern Adriatic Sea, Italy: Implications for stratigraphic paleo-biology of biotic interactions. Paleobiology, in press.

23. Ross AGP, et al. (2001) Schistosomiasis in the People’s Republic of China: Prospects andchallenges for the 21st century. Clin Microbiol Rev 14(2):270–295.

24. McManus DP, et al. (2010) Schistosomiasis in the People’s Republic of China: The eraof the Three Gorges Dam. Clin Microbiol Rev 23(2):442–466.

25. National Bureau of Statistics of China (2011) Communiqué of the National Bureau ofStatistics of People’s Republic of China on Major Figures of the 2010 PopulationCensus (National Bureau of Statistics of China, Beijing). Available at www.stats.gov.cn/english/NewsEvents/201104/t20110429_26450.html. Accessed 2 July 2014.

26. Vázquez NN, Ituarte C, Navone GT, Cremonte F (2006) Parasites of the stout razor clamTagelus plebius (Psammobiidae) from the southwestern Atlantic Ocean. J Shellfish Res25:877–886.

27. WoRMS Editorial Board (2014) World Register of Marine Species (World Register ofMarine Species). Available at www.marinespecies.org. Accessed 18 June 2014.

Huntley et al. PNAS | December 23, 2014 | vol. 111 | no. 51 | 18155

EART

H,A

TMOSP

HER

IC,

ANDPL

ANET

ARY

SCIENCE

SEC

OLO

GY

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 8,

202

1