colonization of a 'lost world': encrustation patterns in modern subtropical brachiopod...

RESEARCH REPORTS 381

Copyright Q 2004, SEPM (Society for Sedimentary Geology) 0883-1351/04/0019-0381/$3.00

Colonization of a ‘Lost World’: Encrustation Patterns inModern Subtropical Brachiopod Assemblages

DAVID L. RODLANDDepartment of Earth Sciences, Southern Connecticut State University, 501 Crescent St., New Haven, CT 06515, Email:

[email protected]

MICHAL KOWALEWSKIDepartment of Geosciences, 4044 Derring Hall, Virginia Tech, Blacksburg, VA 24061

MONICA CARROLLDepartment of Geology, University of Georgia, Athens, GA 30602-2501

MARCELLO G. SIMOESInstituto de Biociencias, Universidade Estadual Paulista, Distrito de Rubiao Junior, CP. 510, 18.610-000,

Botucatu, SP, Brazil

PALAIOS, 2004, V. 19, p. 381–395

The encrustation of Paleozoic rhynchonelliform brachio-pods has been studied for decades, but modern brachiopodshave not received similar scrutiny. The discovery of abun-dant subtropical brachiopods from the Southeast BrazilianBight provides an unprecedented opportunity to assess epi-biont abundance, diversity, and encrustation patterns inmodern brachiopod assemblages. Across the outer shelf, en-crustation frequencies vary among taxa, from mean valuesof 0.45% for Platidia to 9.3% for Argyrotheca. Encrustationfrequencies for Bouchardia increase from 1.6% on the outershelf to 84% on the inner shelf. Larger valves are encrustedmore frequently, and epibionts preferentially colonize valveinteriors. Increased encrustation on the inner shelf may re-flect the greater surface area of larger hosts, longer exposureof dead shells, water-mass characteristics, sedimentationrates, productivity, or other factors that vary with depth. In-ner-shelf brachiopods exhibit encrustation frequencies com-parable to those reported for epifaunal bivalves. The epi-biont fauna is dominated by bryozoans and serpulids, withminor roles played by spirorbids, bivalves, barnacles, fora-minifera, algae, and other taxa. Epibiont abundance ateach site is highly variable, but sites are similar in rank im-portance of epibiont taxa. A different suite of epibionts col-onized Paleozoic brachiopods, but similar patterns of en-crustation have been observed, including preferential settle-ment according to valve morphology. These results providea baseline for evaluating the encrustation of modern bi-valves and ancient brachiopods, and may elucidate themacroevolutionary history of epibionts and their relation-ship to their hosts.

INTRODUCTION

Paleozoic fossil assemblages commonly are dominatedby rhynchonelliform (former Class Articulata, informally‘articulate’ or ‘articulated’) brachiopods, but their modernrepresentatives tend to be restricted to cool- and deep-wa-

ter niches and cryptic habitats (Rudwick, 1970; Jackson etal., 1971; Tunicliffe and Wilson, 1988; James et al., 1992;Richardson, 1997; Wilson, 1998). However, recent workhas shown that the tropical and subtropical shelf of theSoutheast Brazilian Bight harbors large numbers of artic-ulate brachiopods, in places exceeding gastropod and bi-valve abundance combined (Kowalewski et al., 2002). Incontrast to previously documented cool-water settings,this region provides a better analogue to Paleozoic sub-tropical- and tropical-shelf assemblages. From an actuo-paleontological approach, the Southeast Brazilian Bight isa classic pipe dream: a refuge for taxa representing thedominant fauna of a bygone era, as in ‘‘The Lost World’’ ofSir Arthur Conan Doyle’s 1912 story. It provides an oppor-tunity for unprecedented actualistic studies on this long-ranging phylum, including biogeographic, taxonomic, eco-logic, and taphonomic approaches (e.g., Kowalewski et al.,2002; Carroll et al., 2003; Simoes et al., in press).

Brachiopods have been subject to neontological and pa-leontological research from perspectives as disparate asfunctional morphology (Thayer, 1975; Alexander, 1984),predation (Kowalewski et al., 1998; Leighton, 2001), pa-leocommunity ecology (Ziegler et al., 1968; Patzkowskyand Holland, 1999), experimental taphonomy (Emig,1990; Daley, 1993; Torello et al., 2002), macroevolution(Gould and Calloway, 1980), and molecular phylogenetics(Cohen and Gawthrop, 1997). However, studies of encrus-tation of modern brachiopods are practically non-existent,and provide limited quantitative data (e.g., Paine, 1963;Ruggiero, 1996; Fagerstrom, 1996). This stands in con-trast to extensive work on the encrustation of Paleozoicbrachiopods (e.g., Richards, 1972; Kesling et al., 1980;Watkins, 1981; Alvarez and Taylor, 1987; Alexander andScharpf, 1990; Bordeaux and Brett, 1990; Meyer, 1990;Gibson, 1992; Lescinsky, 1997).

Biomineralized epibionts (encrusters) have colonizedhard substrates since the Early Cambrian (Kobluk andJames, 1979; Kobluk, 1981a, 1981b; Brett et al., 1983;Kobluk, 1985), and by the Ordovician, articulate brachio-pods were encrusted frequently (e.g., Alexander and

382 RODLAND ET AL.

FIGURE 1—Map of study area and brachiopod abundance across theshelf of the Southeast Brazilian Bight. (A) Location relative to SouthAmerica. (B) The Southeast Brazilian Bight, with the geographic lo-cation of sites indicated by dots. Brachiopod abundance at each sitevaries according to the legend inset. The box along the coast indicatesthe inner-shelf study area. (C) Close-up inset showing position of in-ner-shelf sites and the abundance of brachiopods at each site. Allmaps adapted from figures generated at http://www.aquarius.geomar.de/omc/makepmap.html using Generic Mapping Tools (Wes-sel and Smith, 1998; http://gmt.soest.hawaii.edu/).

Scharpf, 1990). Encrusters possess a good fossil record,and their unique preservational characteristics have beensubject to a wide array of taphonomic studies (e.g., Lescin-sky, 1993; Parsons-Hubbard et al., 1999; Best and Kid-well, 2000a, 2000b; Lescinsky et al., 2002). Epibiont as-semblages have extraordinarily high spatial fidelity andresolution and the potential for high temporal resolution(e.g., Alexander and Brett, 1990; Lescinsky, 1997), andthus can be used to address paleoecological questions be-yond the scope of soft-substrate benthic faunas. Previousworkers have used epibionts to investigate interactionwithin and between species, larval-settlement patterns,and taphonomic pathways in past and present communi-ties (e.g., Alexander and Scharpf, 1990; Bordeaux andBrett, 1990; Meyer, 1990; Gibson, 1992; Fursich and Osch-mann, 1993; Lescinsky, 1993, 1997; and numerous refer-ences therein). It has suggested that encrustation pat-terns may reflect productivity or nutrient levels throughtime (Vermeij, 1995), but this proposition has yet to beevaluated rigorously.

Despite a great deal of research on the preservation ofmodern bivalves and their encrusters (e.g., Parsons-Hub-bard et al., 1999; Best and Kidwell, 2000a, 2000b; Lescin-sky et al., 2002), these taphonomic patterns may not bereadily applicable to brachiopods. Bivalves and brachio-pod shells exhibit many differences in the mineralogy, per-iostracum, microstructure, shape, and size, each of whichcan influence epibiont colonization patterns. In addition,Paleozoic brachiopod body fossils (and presumably theirencrusters) are more frequently preserved due to theircomposition: low-Mg calcite is less subject to dissolutionthan aragonite, which is a major component of many mol-lusk shells (e.g., Cherns and Wright, 2000; Wright et al.,2003). Therefore, brachiopods may represent a bettermodel system for investigating encrustation through thePhanerozoic. Ideally, encrustation in naturally occurring,modern brachiopod assemblages should be documentedfor comparison with the encrustation of both present-daymollusks and ancient brachiopods.

Finally, an increasing understanding of encrustation onbrachiopods from the present day can be used to addressmacroevolutionary questions. Brachiopods remain alivetoday, but the end-Permian mass extinction wiped outperhaps 95% of their genera (Carlson, 1991), and they areno longer as abundant or diverse as they were in the Pa-leozoic. The transition between Paleozoic and Modernbenthic faunas has been studied from many different per-spectives, but how this transition affected epibiont com-munities remains unclear. Similarly, if encrustationvaries as a function of productivity (Vermeij, 1995), eval-uation of encrustation trends through the Phanerozoiccould test arguments for increasingly productive ecosys-tems (e.g., Bambach, 1993). In any case, the evaluation ofmacroevolutionary patterns in epibiont faunas requires afirm understanding of modern patterns in analogous set-tings. Until recently, these studies have been limited bythe comparatively minor role brachiopods play in modernoceans and a dearth of controlled taphonomic deploymentexperiments utilizing their shells.

This study presents the results of the first large-scale,quantitative investigation of the encrustation of naturallyoccurring modern tropical and subtropical brachiopods.The primary focus of this paper is the documentation andanalysis of spatial variation in encrustation at multiple

levels of resolution, ranging from frequency trends acrossthe entire shelf to variations in diversity and abundanceamong patches of brachiopods. This is the first step for fu-ture comparison of modern encrustation patterns withthose reported from Paleozoic brachiopod-dominated as-semblages. In addition, comparison of epibiont coloniza-tion patterns on modern bivalve and brachiopod shells al-lows evaluation of the effects of host identity on encrusta-tion, and can be used to test the applicability of modernencrustation studies to the fossil record.

STUDY AREA

The study area is part of the Southeast Brazilian BightMarine Province (Fig. 1), influenced by the warm-waterSouth Brazil Current and subject to a humid tropical cli-mate (Knoppers et al., 1999). The shelf is a passive marginextending 90–180 km offshore, characterized by relict car-bonate and terrigenous sediments (da Rocha et al., 1975).Upwelling brings cooler South Atlantic Central Watersonto the outer shelf, below the thermocline at approxi-mately 100 meters water depth (Campos et al., 1995,2000). The distribution of brachiopods appears to be relat-ed to areas of upwelling and carbonate-rich substrates(Kowalewski et al., 2002). Samples were taken from near-

ENCRUSTATION OF MODERN SUBTROPICAL BRACHIOPODS 383

FIGURE 2—Size-frequency distributions for each brachiopod speciescollected from the outer shelf.

coastal areas and across the shelf, over a depth range of;800 m (for additional details on the study area and sam-pling regime, see Kowalewski et al., 2002; Simoes et al., inpress). Only sites that provided abundant brachiopodswere considered in this paper.

Outer-shelf specimens were collected under the scope ofthe Revisee Score Sul Benthos project. These sites weresampled offshore from Parana and Sao Paulo (SP) states(Fig. 1A, B), and include 31 brachiopod-rich sites, rangingfrom 99 meters to 500 meters water depth. Brachiopodswere confined to substrates composed of at least 40% car-bonate by weight (Kowalewski et al., 2002).

On the inner shelf, seven sites were sampled from coast-al areas near Ubatuba, SP. These sites are located in twogeographically distinct regions, referred to here as thenorthern and southern study areas (Fig. 1C). Collectionswere made from known and persistent brachiopod-rich lo-calities across more than 20 kilometers of tropical and sub-tropical coastline, where water depth varied from 6 metersto 30 meters. The northern area consists of Ilha das Cou-ves, Ubatumirim 1, and Ubatumirim 2 study sites, whichwere sampled at 15 meters, 6 meters, and 23 meters waterdepth, respectively. The southern (Ubatuba) area wassampled more extensively, and is represented by four bra-chiopod-dominated sites along a transect trending off-shore from northwest to southeast. Ubatuba Station 9 wascollected at 10 meters water depth, and the sediment ispoorly sorted coarse sand, with 25% carbonate and 7% or-ganic matter by weight. Ubatuba Station 5 was collectedfrom 20 meters water depth, and is also characterized bypoorly sorted coarse sand, with smaller contributions ofcarbonate (17%) and organic matter (1%). In contrast,Ubatuba Station 3, collected from 25 meters, is moderate-ly well sorted, very fine siliciclastic sand (1% carbonate,2% organic matter by weight). Ubatuba Station 1 is thedeepest inner-shelf station sampled (30 meters), with bot-tom sediments composed of moderately sorted coarsesand, including 25% carbonate by weight. These resultsare consistent with previous investigations of the sedi-mentary regime of the Ubatuba region (e.g., Mahiques,1995; Mantelatto and Fransozo, 1999). Further discussionof the depositional setting can be found in Carroll et al.(2003).

MATERIALS AND METHODS

A total of 12,731 specimens was collected from the outershelf, including four brachiopod species: Bouchardia ro-sea, Argyrotheca cf. A. cuneata, Terebratulina sp., and Pla-tidia anomioides (Fig. 2). Three collection methods wereemployed: dredging, Van Veen grab, and box core. Al-though repeated sampling by different methods did not al-ways produce the same results, no systematic biases wereobserved. The longest dimension of each brachiopod shellor fragment was measured using 1-mm-grid graph paperunder a binocular microscope.

The frequency of encrustation was determined sepa-rately for each species as the number of brachiopods bear-ing epibionts relative to all brachiopods collected from thatsite. As a ratio measurement, encrustation frequencymeasures the percentage of the population that was en-crusted (just as gene frequency measures the proportion ofa population bearing a given gene), and does not refer tothe temporal rate at which valves are colonized. This mea-

sure can be used to assess variations among differenthosts, or among hosts of the same species at different sites.When encrusting taxa are solitary, encrustation frequencyalso can provide an estimate of relative abundance amongepibionts, although gregarious taxa are underrepresent-ed. Encrustation frequency is simple to measure (as pres-ence or absence of each taxon on a shell), can be readily in-corporated into any taphonomic or paleoecological study,and is an effective way to determine the relative impor-tance of epibiont taxa. However, it provides no informa-tion on the relative number or biomass of epibionts, or onthe surface area covered by them.

Additional brachiopod samples were collected from thenear-coastal inner shelf, primarily represented by large(.3 mm) specimens of Bouchardia rosea (Fig. 3) Samplingwas conducted by the marine ecology group at UNESP us-ing a Van Veen grab sampler (1/40 m2). This material wassorted from sediment using mesh sizes two millimeters (2mm) across or wider, providing a total of 1029 specimensof Bouchardia rosea. Each specimen was measured to thenearest 0.1 mm using electronic calipers, and the encrust-ing fauna was examined under a binocular microscope.Epibionts were identified at higher taxonomic levels (e.g.,serpulid worms, bryozoans, foraminifera), an approach

384 RODLAND ET AL.

FIGURE 3—Size-frequency distributions for whole valves of Bouchar-dia rosea evaluated for encrustation frequencies. (A) Overall distri-bution for all specimens from both inner and outer shelf. (B) The samesize-frequency distributions, excluding specimens less than 5 mm inlength.

commonly used in modern encrustation studies (e.g., Les-cinsky, 1997; Best and Kidwell, 2000a, 2000b; Lescinskyet al., 2002).

In order to investigate the encrusting fauna more close-ly, and to evaluate the effects of taxonomic resolution onencrustation studies in general, 135 of the 1029 specimensfrom the inner shelf were studied in more detail (15 to 40randomly selected specimens per site). For these speci-mens, epibionts were identified to morphospecies level(when preservation allowed) using a binocular microscope;taxonomic identification has not yet been performed. Theabundance and location of each epibiont morphospeciesand overall taxonomic richness was recorded for eachvalve. Measurements of abundance focus on the number ofepibiont larvae that successfully colonize the substrate,without regard to biomass. Because of this focus on colo-nization and larval arrival, colonial organisms are regard-ed as single individuals, while multiple colonies on a valveare counted as separate individuals.

Valve size and exposure history control the surface areaavailable for colonization and the temporal window inwhich encrustation may occur, so both factors should betaken into account. If colonization only occurs during ashort interval (e.g., Lescinsky, 1997), interaction betweenindividual epibionts may result in competition for limitedresources, such as space and food, and larger valves maybe able to accommodate a greater number of epibionts.The length of time available for valve colonization may de-pend on a variety of factors, including valve size, sedimen-tation and bioturbation rates, storm frequency, and thedurability of a valve (Lescinsky et al., 2002). Consideringvalve identity and size allows for more detailed analysis.In a similar manner, studying the encrustation of valvescollected from the same site minimizes the potential fordifferent exposure histories (although time-averaging isstill likely to occur). Thus, each valve can be considered areplicate sample substrate deployed naturally at each site.Another factor complicating interpretation is the life-stateof the host: the valves of dead hosts provide greater sur-face area (valve interiors) than live hosts. While both liveand dead hosts are considered here, consideration of valveexteriors alone can be used to evaluate encrustation in acomparable manner. Given constant size and substratecomposition along with comparable exposure histories,epibiont abundance (measured as the number of larvalcolonization events) may prove to be a useful measure ofplanktonic productivity at a site in the fossil record.

RESULTS

Out of a total of 13,760 brachiopods examined in thisstudy (12,731 from the outer shelf and 1,029 from the in-ner shelf), 1,106 were encrusted, for an overall encrusta-tion frequency of 8.0%. A wide degree of spatial and envi-ronmental variation occurs in brachiopod encrustationacross the Brazilian shelf. Encrustation frequency rangedfrom zero beyond the slope break, at 500 meters waterdepth, to 100% at a depth of 6 meters on the inner shelf.When all sites are pooled, exterior encrustation frequen-cies for Bouchardia are not directly correlated with depthat a 5 0.05 (r 520.248, p 5 0.266), but they are correlatedwith mean host size, (r 5 0.688, p 5 0.0004), which de-creases with increasing depth (r 520.584, p 5 0.004), asshown in Figure 3.

The encrusting fauna of the outer shelf is composed pri-marily of foraminifera, although a few serpulids and bryo-zoans have been noted. By contrast, calcified polychaeteworm tubes, bryozoa, and foraminifera characterize theinner shelf, with a lesser role played by bivalves, algae,barnacles, and other taxa. Four different morphotypes ofserpulid worms were observed, along with sixteen mor-photypes of bryozoa, three morphotypes of foraminifera,and one variety each of spirorbids, bivalves, algae, andbarnacles, for a total of 27 distinct encrusting morphospe-cies on 4 brachiopod species. Diversity, like encrustationfrequency, increases towards shore.

There are distinct differences in sampling regime anddata collection between inner and outer shelf. Among oth-er factors, many outer-shelf brachiopods were collectedalive and preserved soft tissues, while inner-shelf brachio-pods represent a time-averaged assemblage covering morethan 3000 years (Carroll et al., 2003). Because of these dif-ferences, inner- and outer-shelf encrustation patterns arediscussed separately.

Outer Shelf

Analysis of encrustation frequency by depth, latitude,longitude, and collection method failed to identify trendsin brachiopod encrustation across the outer shelf (Table1). On the other hand, the abundance of brachiopodsshows a spatial pattern, with Bouchardia primarily foundin the southwest, and the other taxa found only at north-eastern sites. Only 1.9% of all brachiopods collected fromthe outer shelf were encrusted. Overall, brachiopod en-crustation frequencies on the outer shelf vary from zero to100% per site, although the highest percentages are lim-ited to small samples and thus may represent a statisticalartifact. After excluding small samples (fewer than 10 bra-chiopods), frequencies per site varied over a much smallerrange, from 0–44%, with the highest recorded encrusta-tion frequency (44%) observed for a sample of 50 speci-mens of Bouchardia.

The frequency of encrustation varies widely among bra-chiopod species (Fig. 4). Argyrotheca was colonized morecommonly by epibionts than other brachiopod species: thepooled encrustation frequency for all specimens of Argy-rotheca was 9.3%, compared to pooled frequencies of 3.9%for Terebratulina, 1.6% for Bouchardia, and 0.4% for Pla-


TABLE 1—Encrustation frequencies for brachiopods from the outer shelf of the Southeast Brazilian Bight, in order of sample depth. Totalsand mean values are calculated separately for each site and for each species; # brachs 5 number of brachiopods; # encr. 5 number ofencrusters; % encr. 5 percentage of encrusters.

Site

Collection Locality

Depth(m)

Latitude(S)

Longi-tude (W)

Argyrotheca

#brachs

#encr.

%encr.

Bouchardia

#brachs

#encr.

%encr.

Platidia

#brachs

#encr.

%encr.

Terebratulina

#brachs

#encr.

%encr.

Total

#brachs

#encr.

%encr.

6678670466696674670366736677(19)67936808

99100101122133133137138141

24.46.35725.14.6024.07.34724.31.0825.39.7024.17.93924.40.74727.46.4928.48.67

45.11.13546.03.0044.42.14244.54.0046.13.2044.35.98344.50.82247.40.4548.00.19

10

4665

39

0.00.00.00.00.06.5

13.80.00.0

911

10360920

50226

165

223

0.00.00.01.0

10.70.00.0

44.01.3

0.00.00.00.00.00.00.00.00.0

3

3

5452

2

0.00.00.00.00.03.70.00.00.0

2211

10660912011750

226

16559

223

0.00.00.00.9

10.74.27.7

44.01.3

6661669967876676668666956676(18)6686(28)66536666

147150151153153153153153155163

24.07.63726.01.2627.27.8324.49.69925.36.98826.17.5124.49.69925.36.98825.43.5024.17.129

45.51.89546.25.2647.24.2244.44.96545.13.57146.41.2344.44.96545.13.57146.02.5044.12.179

688

99

36

1

16

7

0.00.00.01.50.00.0

16.20.00.0

19.4

12695

17

18809

273

150

0.02.1

60.00.00.00.00.0

100.00.60.0

1

3

0.00.00.00.00.00.00.00.00.00.0

22

395

128

30

1

1

8

0.050.00.02.60.00.06.30.00.00.0

21271

51071417

2271

880969

2832

241

507

0.02.2

60.01.90.00.0

10.6100.0

0.610.1

6666(8)6672(14)6681(23)6706664666526698665166656665(7)

163165168184198206241256258258

24.17.12926.27.7525.11.00525.48.6025.43.7825.51.0426.10.8725.53.5824.20.84424.20.844

44.12.17944.30.35144.56.645.44.5045.16.0645.47.3046.20.0145.42.1344.09.91344.09.913

153410

1

17

2 13.30.00.00.00.00.00.00.00.00.0

1

137691

5712

62

3

0.00.00.04.42.90.05.30.00.00.0

11

440 2

0.00.00.00.00.00.00.00.00.50.0

19331111

981

411

1

21.13.09.10.0

100.00.00.00.00.00.0

356921

138711

571

5571

61163

3

2

17.11.44.84.34.20.05.30.00.40.0

6685665066446777

282417485500

25.41.82725.57.3925.45.8026.51.76

45.11.68645.34.2545.11.7746.18.37

0.00.00.00.0

11

0.00.00.00.0

0.00.00.00.0

23

1

0.00.00.00.0

2411

0.00.00.00.0

Totals

Mean

perspeciesby site

409

34.1

38

3.2

9.3

5.9

11390

517.7

183

8.3

1.6

10.6

446

89.2

2

0.4

0.4

0.1

488

24.4

19

1.0

3.9

9.8

12733

385.8

242

7.3

1.9

8.8

tidia. Encrustation frequencies for each species vary sig-nificantly from the frequencies derived by pooling the oth-er taxa: with one degree of freedom, Chi-square valuesrange from 5.23 to 124 for such comparisons, and p valuesrange from 0.22 to , 0.0001. Similar size-frequency distri-butions for all species (Fig. 2) suggest valve size does notaccount for these differences.

Because encrustation frequencies vary widely amongsites, not all individuals of a species have the same expo-sure to colonization. Calculated frequencies of encrusta-tion per site can be used to assess spatial variability in theencrusting fauna, as shown in the ‘Total’ column in Table1, and indicate wide variation in encrustation frequenciesper site. However, variations among sites do not accountfor the observed variation in encrustation frequenciesamong species. Argyrotheca occurs only at sites where Ter-ebratulina is present, but a significantly higher number ofArgyrotheca were encrusted (Fisher’s Exact Test: p 50.0014).

Inner ShelfTotal encrustation frequencies for inner-shelf Bouchar-

dia are considerably higher than values determined from

the outer shelf (Fig. 5), ranging from 75% to 100%, in con-trast to the outer-shelf range between 0% and 44% for allsamples where n $ 10 brachiopods. There is a strong pref-erence for the encrustation of valve interiors of deadshells, and since most outer-shelf brachiopods includedsoft tissues internally, this presents a possible bias. If onlyvalve exteriors are considered, the difference is reduced:1.6% of Bouchardia specimens are encrusted on the outershelf, compared to 27.3% of all Bouchardia from the innershelf. This still records a strong preference for inner-shelfencrustation, however: 532 out of 1029 inner-shelf speci-mens were encrusted on the exterior of the shell, com-pared to 183 out of 11,390 outer-shelf specimens (Fisher’sExact Test: p , 0.0001), although the role of differentialtime-averaging remains unknown. Within the inner shelf,whole-fauna variations in encrustation frequency do notvary consistently between samples as a function of depthor latitude (Table 2). In contrast, when inner-shelf sitesare pooled together, encrustation frequencies show astrong increase as a function of valve size (Fig. 6).

Specific taxonomic groups of epibionts do show somespatial distribution trends, particularly along the south-

386 RODLAND ET AL.

FIGURE 4—Encrustation frequencies across the outer shelf, by spe-cies. Each histogram plots the number of sites where brachiopod en-crustation frequencies fall within a 5% bin, ranging from 0% (no en-crustation at that site) to 96–100% (all, or nearly all, brachiopods areencrusted at that site). Only sites represented by n $ 10 individualsof the plotted species are included.

FIGURE 5—Comparison of encrustation frequencies between inner-and outer-shelf brachiopods. (A) Histograms as in Figure 3. Encrus-tation frequency is plotted by the number of sites that fall into 5% bins,labeled by the maximum frequency for that bin. Only sites representedby n $ 10 brachiopods are included. (B) Scatter plot of encrustationfrequency versus depth for the same sites. (C) Exterior encrustationfrequency versus depth. Comparing data from valve exteriors exclu-sively reduces taphonomic biases introduced by live/dead collectiondifferences between inner- and outer-shelf sites. (D) Mean valvelength per site as a function of depth.

ern Ubatuba transect. The percentage of brachiopods en-crusted by serpulid worms increases steadily northwardalong the Ubatuba transect and through Ilha das Couvesand Ubatumirim. Within the southern transect, serpulidsform a decreasing percentage of the encrusting fauna withdepth. Up to 49 individuals were counted on a single valvefrom the 23-m-depth site at Ubatumirim, making serpu-lids the most gregarious of encrusters present in the studyarea. Bryozoa show a similar trend to serpulids, compris-ing an increasing percentage of the fauna from south tonorth and with decreasing depth through the southerntransect, but their abundance is more variable throughthe northern study area. Barnacles colonize brachiopodsrarely, and only in the northern parts of the study area,where epibiont abundance is highest. Spirorbids may beconfined to areas near shore, as they were found at allsites except for Ubatuba 1 and Ilha das Couves. The dis-tribution of bivalves, foraminifera, and algae is patchy, fol-lowing no apparent trend.

The abundance of epibiont taxa and relative percentageof valves encrusted by each taxon vary among sites, evenat coarse taxonomic scales (Fig. 7). In contrast, the relativeproportion of shells encrusted by each taxon is remarkably

consistent among samples collected from the same site;random subsets of as few as ten brachiopods show little de-viation from the patterns derived from larger sample siz-es. Epibiont abundance patterns are similar to the pat-terns observed by measuring encrustation frequenciesseparately for each epibiont taxon. The only exception isthe intensely encrusted 23-m site at Ubatumirim 2, wherethe gregarious tendencies of serpulids skew the abun-dance pattern. The similarity between encrusting faunasat each site was tested using the Spearman rank correla-tion test, assuming significance levels of a 5 0.05 (Spear-man rank correlation identifies significant correlation inthe rank order abundance of taxa at different sites, withsignificant results indicating a high degree of similarity


TABLE 2—Encrustation frequencies for the inner shelf as a total for all encrusters and broken down by taxonomic group.

Encrusting taxa: SerpulidSpiror-

bidBryo-zoan

Barna-cle Bivalve Algae

Forami-nifera Other Total

Ubatumirim 1DepthUbatumirim 2DepthUbatumirimDepthIlha das CouvesDepth

n 5 166 mn 5 1623 mn 5 120??n 5 1415 m

# Bouchardia encrusted% Bouchardia encrusted# Bouchardia encrusted% Bouchardia encrusted# Bouchardia encrusted% Bouchardia encrusted# Bouchardia encrusted% Bouchardia encrusted

1487.51381.257461.77

50.0

16.254

254335.800

161001487.59175.87

50.0

212.50021.73

21.4

4254

254134.26

42.9

212.516.25

2117.52

14.3

85016.250000

531.252

12.51210.04

28.6

161001593.8

108901178.6

Ubatuba 9DepthUbatuba 5DepthUbatuba 3DepthUbatuba 1Depth

n 5 77610 mn 5 1520 mn 5 4025 mn 5 3230 m

# Bouchardia encrusted% Bouchardia encrusted# Bouchardia encrusted% Bouchardia encrusted# Bouchardia encrusted% Bouchardia encrusted# B. rosea encrusted% B. rosea encrusted

33643.35

33.312307

21.9

222.82

13.37

17.500

46860.37

46.71332.57

21.9

30.4000000

425.400

1332.55

15.6

9712.5000039.4

10113.04

26.71127.57

21.9

9412.17

46.710251237.5

644831386.730752784.4

FIGURE 6—Mean encrustation frequency per size class for all inner-shelf brachiopods, plotted separately for valve interiors, exteriors, andper shell. Each size class is represented by all individuals in 1-mmincrements of anterior-posterior valve length.

between them; see Kidwell, 2001 for more details on thisapproach). Each site shows significant rank-order corre-lation between the abundance of epibionts and the relativefrequency at which each taxon encrusts valve surfaces (p, 0.05).

Epibiont abundance and diversity are more complicatedat finer taxonomic resolution. Taphonomic factors limitedthe ability to distinguish all taxa, but the composition offaunas varies widely among sites. Ubatuba Station 9showed the highest diversity (24 morphospecies), whileUbatuba Station 5 showed the lowest diversity (8 morpho-species). Epibiont distribution is patchy: only one bryozo-an, one serpulid, and one foraminiferan morphospecies oc-cur in all inner-shelf sites. Six taxa were found only at onesite each; the mode was four sites, represented by ninetaxa. No fewer than 11 epibiont morphospecies were dif-ferentiated from sites with 30 or more brachiopods. Thereis a strong association between diversity and abundance ofepibionts (r 5 0.939, p 5 0.0055), but no significant corre-lations occur at a 5 0.05 between depth, abundance, diver-sity, and the number of brachiopods sampled at each site.

Using Spearman correlation to test for similarities in

rank order of morphospecies abundance between sites,there is a significant correlation (p , 0.05) between all buttwo sites (Table 3). The only exception (p 5 0.113) is thepairwise comparison between Ubatumirim 2, character-ized by the highest epibiont abundance, and Ubatuba 1,which has the lowest epibiont abundance. One problemwith Spearman rank correlation is its susceptibility to mu-tual absences, but the degree to which these influence thedata can be tested by excluding taxa that are absent frommultiple sites. Table 3 presents results of such an analy-sis, excluding taxa that are absent from more than threesites. This analysis loses resolving power as original dataare excluded, but the results are similar to the uncorrectedcorrelation: all sites within the southern study area stillgroup together, while Ubatumirim 1 is correlated withboth Ubatumirim 2 and Ubatuba 9.

INTERPRETATIONS AND FUTURE RESEARCHDIRECTIONS

The results of this study can be compared to encrusta-tion studies on articulate brachiopods in the fossil record(e.g., Kesling et al., 1980; Wilson, 1982; Nield, 1986; Alva-rez and Taylor, 1987; Alexander and Scharpf, 1990; Bor-deaux and Brett, 1990; Cuffey et al., 1995; Lescinsky,1995, 1997), as well as studies on modern mollusks,whether based on field sampling (e.g., Best and Kidwell,2000a, b) or experimental deployment studies (e.g., Par-sons-Hubbard et al., 1999; Lescinsky et al., 2002). Thesecomparisons are necessary to develop a better under-standing of long-term temporal patterns of encrustationthrough the Phanerozoic. If fundamental differences existin encrustation between brachiopods and bivalves, thehistory of encrustation must take into account the long-noted decline of brachiopods since the Paleozoic (Gouldand Calloway, 1980, and references therein). Similarly,careful investigation may reveal changes in substrate se-lectivity by epibionts (e.g., taxonomic trends, rugophilic,photophobic, or cryptic settlement behavior) through geo-logic time. These questions must be addressed before mea-sures of encrustation frequency can be used as proxies forenvironmental parameters in the fossil record.

388 RODLAND ET AL.

FIGURE 7—Epibiont associations on the inner shelf, by site. The area of each pie chart corresponds to the mean number of epibionts pervalve at each site, ranging from 1.81 at Ubatuba 1 (30 m) to 16.1 at Ubatumirim 2 (23 m). (A) Pie charts indicating the relative number ofencrusted brachiopod surfaces (interior and exterior surfaces counted separately) colonized by each epibiont taxon at each site. (B) Pie chartsindicating the actual abundance of epibiont taxa at each site.

Variations Among Sites

Brachiopod assemblages and the frequencies and pat-terns of encrustation are markedly different between in-ner- and outer-shelf localities. A 69-meter-depth gap ex-ists between the shallowest outer-shelf site that yielded

brachiopods (99 m) and the deepest inner-shelf site sam-pled (30 m). Inner-shelf brachiopod faunas are over-whelmingly dominated by Bouchardia (Kowalewski et al.,2002) with a similar size range to those on the outer shelf,but a larger proportion reach larger size classes (Fig. 3).However, due to the sampling regime, only shells longer


TABLE 3—Spearman correlation coefficients and p values for rank abundances of epibionts in inner-shelf assemblages. Correlation coefficientsare rounded to the second decimal place, while p values are rounded to the first significant digit. Pairwise comparisons of sites resulting in pvalues below 0.05 are interpreted as significant correlations, and reflect recruitment of epibiont larvae from the same pool of potential colonists.

Site Ubatuba 9 Ubatuba 5 Ubatuba 3 Ubatuba 1 Ubatumirim 1Ubatumi-

rim 2 Comments

Ubatuba 9Ubatuba 5Ubatuba 3Ubatuba 1Ubatumirim 1Ubatumirim 2

p 5 0.002p 5 0.0002p 5 0.005p 5 0.001p 5 0.005

0.55

p 5 0.001p 5 0.005p 5 0.02p 5 0.009

0.650.59

p 5 0.0001p 5 0.03p 5 0.01

0.520.520.66

p 5 0.04p 5 0.1

0.580.440.410.39

p , 0.0001

0.520.490.470.310.77

Results of Spearmancorrelation of morpho-species abundancewithout correction formutual absences oftaxa

Ubatuba 9Ubatuba 5Ubatuba 3Ubatuba 1Ubatumirim 1Ubatumirim 2

p 5 0.04p 5 0.03p 5 0.002p 5 0.003p 5 0.07

0.48

p 5 0.02p 5 0.03p 5 0.3p 5 0.4

0.510.53

p 5 0.001p 5 0.4p 5 0.6

0.680.510.70

p 5 0.2p 5 0.9

0.660.260.230.33

p , 0.0001

0.430.190.120.030.80

Results of Spearmancorrelation of morpho-species abundance,excluding taxa absentfrom more than threesites

than 3 mm were evaluated for encrustation from the innershelf (Table 2). In addition, for reasons that remain un-clear, outer-shelf sites were dominated by live-collectedspecimens (.90%), while dead specimens comprised theoverwhelming majority of inner-shelf specimens. Thesefactors complicate direct comparison of inner- and outer-shelf data, and the effects of exposure history and host sizeon encrustation require evaluation.

On a site-by-site basis, mean shell length and encrusta-tion frequency increase towards shallow waters (r 5 0.688,p 5 0.0004 for exterior encrustation; Fig. 5), and inner-shelf data indicate that larger shells are encrusted muchmore frequently than smaller shells, particularly for shellexteriors (Fig. 6). This suggests that much of the differ-ence in inner- and outer-shelf encrustation could be due toshell size alone. Whether this relationship is causal orsimply correlative remains uncertain. Larger shells aremore likely to be colonized because they make larger tar-gets for epibiont larvae settling at random, but epibiontabundance and host size could be responding positively toother environmental factors, such as food supply.

Valve interiors are encrusted more frequently than ex-teriors, and thus most encrustation occurs after the deathof the host. As a result, the preponderance of dead hosts onthe inner shelf could bias the record towards higher en-crustation frequencies. Because of this, it is necessary toexamine encrustation separately between valve interiorsand exteriors. All interior encrustation must occur afterthe death of the host; therefore, taphonomic studies em-phasizing post-mortem processes may focus on interior al-teration exclusively (e.g., Best and Kidwell, 2000a, b). Ex-terior encrustation can occur during the life of the host aswell as after death, and may reflect commensal relation-ships between host and epibiont (e.g., Fagerstrom, 1996;Lescinsky 1997). Even when exterior encrustation fre-quencies are considered alone, nearly three times as manyinner-shelf specimens of Bouchardia were encrusted thanouter-shelf Bouchardia. This is all the more remarkablebecause outer-shelf specimens outnumber inner-shelfspecimens by more than an order of magnitude. Becauseencrustation of shells occurs rapidly, within months oryears (e.g., Parsons-Hubbard et al., 1999), and brachiopodshells may decay rapidly (e.g., Emig, 1990), fossil brachio-pod encrusting assemblages may differ little from those

found on live hosts (Lescinsky, 1997). This is supported bywork demonstrating that the ages of dated shells do notcorrelate with encrustation or other taphonomic scores(Carroll et al., 2003). While differences in duration of ex-posure could bias these data, it is possible that the differ-ence in exterior encrustation between outer shelf and in-ner shelf reflects a real ecological phenomenon ratherthan a taphonomic artifact.

At finer spatial scales the picture is more complicated,as brachiopod abundance, frequency, and intensity of en-crustation, along with the composition of encrusting fau-nas, do not show clear depth trends within either inner-shelf (Fig. 8) or outer-shelf environments. For instance,while epibiont abundance decreases with depth along theUbatuba transect on the inner shelf, the 23-meter-depthUbatumirim site records the highest abundance observedin this study. In addition, Spearman rank correlation sug-gests that the epibionts found at each site on the innershelf came from no more than two initial pools of potentialcolonists, characterizing the northern and southern studyareas. Ubatumirim 1 correlates with Ubatuba 9 as well aswith Ubatumirim 2; however, the epibiont fauna at thissite may result from mixing between both areas.

Depth alone is probably not a primary control on en-crustation frequency, epibiont abundance, or diversity,but is correlated to the controlling variables instead. Be-sides exposure history and valve size, an array of physicaland biological environmental factors co-vary with depthand are likely to influence encrustation. Water depth canbe directly related to water-mass characteristics such astemperature and salinity, which affect the growth ratesand survival of marine invertebrates. Additional physicalinfluences often related to water depth include sedimentcomposition and sedimentation rates. Even a thin veneerof sediment can deter encrustation, and could result inlower encrustation frequencies below storm wave base(Parsons-Hubbard et al., 1999; Lescinsky et al., 2002), al-though this may not be true on live hosts. The variationsin encrustation of brachiopod taxa from the same sites in-dicate that the constituents of the benthic fauna and thephysical characteristics of their shells also must be takeninto account.

It has been suggested that various measures of encrus-tation should increase with increasing productivity (Ver-

390 RODLAND ET AL.

FIGURE 8—Scatter plots showing the relationships between water depth, epibiont abundance (total number of individuals or colonies), diversity(epibiont morphospecies richness), number of brachiopod shells, and the frequency of encrustation for each inner-shelf site where epibiontswere counted at morphospecies resolution. (A) Increase in epibiont diversity with increasing epibiont abundance. (B) Decrease in epibiontdiversity with increasing water depth. (C) Encrustation frequency versus epibiont abundance. (D) Decrease in epibiont abundance with increas-ing water depth. (E) Epibiont abundance versus number of brachiopods sampled. (F) Epibiont diversity versus number of brachiopods sampled.

meij, 1995; Lescinsky et al., 2002). While this hypothesis isattractive, it remains unclear precisely which measures ofproductivity are related to encrustation, or how encrusta-tion should be measured to interpret productivity. In ad-dition, other influences (e.g., sedimentation rates) alsomust be taken into consideration and complicate the inter-pretation of productivity from epibiont data (Lescinsky etal., 2002). While the data available here are not sufficientto test the hypothesis that encrustation can be a proxy for

productivity, it is worth noting the oceanographic and eco-logical context of the study area. The Ubatuba region gen-erally is considered to be oligotrophic or oligo-mesotrophicin terms of nutrient availability, with nutrients providedfrom atmospheric sources, runoff, and upwelling of SouthAtlantic Central Water (SACW) onto the shallow shelf(Braga and Muller, 1998). Natural eutrophication occursduring summer months due to the upwelling of SACWinto the euphotic zone and consequent nutrient regenera-


tion (Aidar et al., 1993), which certainly would be favor-able for phytoplankton. Similarly, zooplankton density ishighest on the shallow shelf, above 50 meters water depth(Vega Perez, 1993). While the linkages between primaryproductivity and encrustation are tentative and requirefurther evaluation, these results are consistent with thehypothesis that epibiont abundance and encrustation fre-quency should increase with increasing nutrient availabil-ity.

Variations Among Taxa

The data collected support the suggestion that encrus-tation frequencies vary among brachiopod species (despitesimilar size-frequency distributions and sample sizes),and are consistent with evidence from the fossil record(e.g., Richards, 1972; Alexander and Scharpf, 1990; Bor-deaux and Brett, 1990; Lescinsky, 1997). However, thecauses for these interspecific variations bear further in-vestigation. Different pictures emerge depending onwhether encrustation frequencies are computed as anoverall value for the species or as mean frequencies persite. Frequencies calculated per species assume that eachindividual brachiopod within a species has the samechance of being encrusted as any other member of the spe-cies. This ignores variation in encrustation between sites,and assumes that variation between species is the resultof substrate selectivity among settling larvae of encrust-ing taxa.

Notably, the highest per-species encrustation frequencyon the outer shelf was calculated for Argyrotheca, whichhas pronounced radial ribbing. The second highest per-species encrustation frequency was reached by Terebratu-lina, with fine radial ribs, and the lowest frequencies wereattained by the smooth-shelled Bouchardia and the spi-nule-bearing Platidia. This suggests preferential coloni-zation of the exteriors of ribbed taxa, due perhaps to theincreased surface area of the shell or, alternatively, to apreference for cryptic habits provided by concave surfaces,such as grooves. Given the similarities in size-frequencydistribution and in cross-sectional area among these bra-chiopod species (Fig. 2), the latter explanation seems morelikely. Other factors may also play a role, such as faciesdifferences, punctae density, chemical deterrents, and (asnoted previously) host size and exposure history, but forshells of equivalent size collected from the same settings,shell topography appears to be the strongest candidate forexplaining these data. Paleozoic spirorbids show distinctpreference for the spaces between brachiopod ribs (Alex-ander and Scharpf, 1990), and are almost entirely con-fined to concave valve interiors of specimens observed inthis study. Cryptic settlement patterns are commonamong modern epibionts, but appear to have been lesscommon on Paleozoic brachiopod hosts (e.g., Watkins,1981; Alexander and Scharpf, 1990). It remains uncertainwhether this is a taphonomic bias resulting from post-mortem disarticulation, or an actual trend towards cryptichabitation over geologic time.

Paleozoic brachiopods also show extensive variation inmean encrustation frequency among species, even aftercorrection for surface area (Alexander and Scharpf, 1990;Bordeaux and Brett, 1990). Mean encrustation frequen-cies calculated for each of ten brachiopod species collected

from Upper Ordovician strata in southeastern Indiana(Alexander and Scharpf, 1990) range from a low of 0%(Thaerodonta clarksvillensis) to a high of 55% (Rafinesqui-na alternata), although facies differences may play a role.Both Ordovician and Devonian encrusting taxa show ageneral settlement preference for ribbed brachiopods, andavoidance of brachiopods bearing spines and punctae (Al-exander and Scharpf, 1990; Bordeaux and Brett, 1990).This tendency continues on the outer shelf of the modernSouth Atlantic: the most coarsely ribbed brachiopod spe-cies (Argyrotheca) is most frequently encrusted, while theleast-encrusted taxon (Platidia) possesses spinules.

It is worth emphasizing that shell ornamentation is notthe dominant control on encrustation, either in the mod-ern oceans or in the Paleozoic. For instance, the smooth-shelled brachiopod Composita was the one of the most fre-quently encrusted brachiopods from the Carboniferous ofNorth America (Lescinsky, 1997). Similarly, Bouchardiahas a smooth shell, but is the most frequently encrustedbrachiopod on the Southeast Brazilian Bight. However,this is biased by size and environment: outer-shelf Bou-chardia are encrusted at lower frequency than ribbed bra-chiopod taxa in the same size range. This juxtaposition il-lustrates the need for size control in evaluating modernand fossil encrustation.

These substrate-specific variations between inner- andouter-shelf encrustation may be complicated further bythe role of punctae in epibiont deterrence. Each of the spe-cies identified from the Southeast Brazilian Bight possesspunctae, as do all terebratulids, and therefore, variationsin the density of punctae may influence encrustation fre-quencies. Moreover, some brachiopods secrete mucosac-charides through their punctae, and these substances de-ter colonization by epibionts during the life of the individ-ual (Curry, 1983). Because the overwhelming majority ofbrachiopods from the outer shelf were collected alive, dif-ferences in encrustation frequency also may reflect differ-ent degrees of epibiont deterrence by chemical means.However, other than estimating the density of punctae, itmay be difficult to assess the role of chemical deterrenceon encrustation in fossil brachiopod assemblages

Comparison With Paleozoic Brachiopods

When the epibiont fauna is examined at fine taxonomicresolution, a substantial proportion of diversity is foundamong the rare taxa, which is typical in Cenozoic soft-bot-tom, benthic-marine assemblages (e.g., Bambach, 1977;Powell and Kowalewski, 2002). These observations con-firm the patchy nature of encrustation, and, in particular,demonstrate the same lack of community coherency seenin many ecological and paleoecological studies (Springerand Miller, 1990; Bennington and Bambach, 1996; Lescin-sky, 1997; Patzkowsky and Holland, 1999; Olszewski andPatzkowsky, 2001).

Analyses of species richness trends in epibiont assem-blages through geologic time are complicated by a varietyof factors, such as changing epibiont and host identity,lack of facies control, and variable sample size and taxo-nomic resolution. Some studies underestimate epibiont di-versity because they are conducted at coarse taxonomicresolution (e.g., Best and Kidwell, 2000a, b; Lescinsky etal., 2002). In contrast, oversplitting of taxa potentially can

392 RODLAND ET AL.

FIGURE 9—Comparison of epibiont diversity per site between Car-boniferous Composita and Recent Bouchardia. The mean number ofepibiont species per brachiopod shell is plotted separately for eachsite. Data for Composita taken from Lescinsky, 1997.

create ecologically improbable scenarios: for example, 14separate species of Hederella were reported on Paraspiri-fer from the Silica shale, accounting for more than a thirdof epibiont species richness (Kesling et al., 1980). Theseproblems present significant obstacles to determining ad-equate estimates of diversity of epibionts through geologictime.

A more reliable estimate of epibiont diversity may bethe mean number of encrusting species per brachiopod.While this measure still depends on adequate distinctionof species, it is less dependent on the number of hosts stud-ied than total richness, and it measures ecological com-plexity (Lescinsky, 1997). In a sense, each valve can beconsidered as a separate island that can be colonized byepibiont larvae, each of which has limited space and foodresources available. Multiple valves provide replicatesamples for colonization, and can be used to estimateniche occupation rather than the diversity of the pool ofpotential epibionts. The mean number of morphospeciesper encrusted valve, therefore, provides a degree of ecolog-ical standardization between samples in a given localityand of equivalent age, at least for hosts of equivalent size.

Unfortunately, few studies have provided data suitablefor this sort of analysis, but a great deal of information isavailable from Mississippian and Pennsylvanian encrust-ing communities in North America. Encrusted Mississip-pian brachiopods host, on average, 1.05 to 3.43 species perspecimen; the highest value is for Anthracospirifer whilethe lowest is for Reticulariina (Lescinsky, 1997). Pennsyl-vanian specimens of Composita host anywhere from 1.06to 3.30 species per encrusted shell (Lescinsky, 1997).These encrusted Composita should be comparable to theresults of this study: a single genus was studied by a singleauthor using a consistent methodology, using multiplesamples representing tropical or subtropical environ-ments in an icehouse world.

For this comparison, the mean number of epibiont spe-cies per host species was calculated for all sites, andweighted by epibiont abundance at each site (i.e., themean number of encrusting species per encrusted brachio-pod was multiplied by the number of encrusted brachio-pods at that site, the total for all sites was calculated, andthen divided by the total number of encrusted brachio-pods). This value is estimated only for encrusted brachio-pods because brachiopods that were not encrusted provideno useful information on epibiont diversity patterns. Amean value of 1.99 epibiont species was calculated perhost (median 5 1.93) on 695 specimens of CarboniferousComposita. By comparison, inner-shelf Bouchardia hostsprovide mean values per site ranging between 1.63 and4.69 species per encrusted valve, with a mean of 3.21 epi-biont morphospecies per host (median 5 3.30) for 142specimens. A Wilcoxon 2-sample test indicates that thisdifference is marginally significant (Z 5 1.95, p 5 0.05).Factors including size and facies differences could under-lie this result, and the data ranges overlap. Unfortunately,valve size data was not reported for Composita, so its roleis difficult to evaluate.

Higher species richness has been reported in Devonianepibiont assemblages (Kesling et al., 1980; Gibson, 1992),and it is possible that these settings preserved a greaternumber of epibiont taxa per shell. More detailed compari-sons of abundance and diversity in epibiont assemblages

through the Paleozoic are needed to identify potentialmacroevolutionary trends. Mean and maximum species-richness values per shell are higher for modern subtropi-cal brachiopods from the southeast Brazilian shelf thanthey were in comparable Pennsylvanian settings from themidcontinent of North America (Fig. 9). Whether this re-flects evolutionary changes in the epibiont fauna, oceano-graphic and environmental controls, differences in theability of epibionts to colonize these hosts, or differences inthe hosts themselves, remains uncertain.

Encrusting fauna have changed dramatically since thePaleozoic, because many once-dominant taxa (cornulitids,edrioasteroids, stenolaemate bryozoans, and tabulate andrugose corals) are extinct (Palmer, 1982). Their role as en-crusters of brachiopods has been supplanted by serpulidworms and cheilostome bryozoans. Spirorbids remain asmall but important part of the fauna today, and displayedsimilar, if highly variable, relative abundance in Devonianand Mississippian encrusting communities (Bordeaux andBrett, 1990; Lescinsky, 1997). Encrusting communitiesexperienced significant turnover through the late Paleo-zoic, a situation aggravated by the end-Permian mass ex-tinction (Lescinsky, 1996). Spirorbids are the only en-crusting taxa identified in Triassic faunas from the surviv-al interval immediately postdating the extinction (report-ed in Paull and Paull, 1994), and more diverse Mesozoicencrusting communities were not established until theLate Triassic (Taylor and Michalik, 1991). The new suiteof taxa may reflect the results of differential survivalthrough mass extinction, or they may track ecologicalchanges through time, such as increasing biomass andmorphological complexity (e.g., cheilostome versus steno-laemate bryozoans, Sepkoski et al., 2000) in epibiont taxa.Despite these changes in epibiont faunas and whatevertheir causes, these results suggest that meaningful com-parisons can be made between modern and ancient bra-chiopod encrustation patterns in open-marine, warm-wa-ter settings.

Comparison with Modern Bivalves

Modern articulate brachiopods are encrusted by manyof the same taxa that colonize bivalve mollusks in modernmarine settings. Bivalves from a tropical, mixed siliciclas-tic and carbonate subtidal embayment in Panama are en-


crusted primarily by bryozoans and calcareous tube-dwell-ing worms (both serpulids and spirorbids), with minorroles played by algae, foraminifera, mollusks, and corals(Best and Kidwell, 2000a). Encrusting taxa in the JavaSea also vary by site, dominated by bryozoans, barnacles,bivalves, and sponges in a eutrophic setting, but by bryo-zoans, serpulids, and algae at a mesotrophic site (Lescin-sky et al., 2002). The observed similarity of dominant en-crusting taxa among brachiopods and bivalves in threeseparate tropical seas suggests that host identity is not aprimary control on the identity of encrusting taxa at high-er taxonomic levels. Despite many differences in perios-tracum and shell microstructure and mineralogy, metab-olism, and filtration rates, the potential for infaunalmodes of life among bivalves, and other characteristics be-tween bivalves and brachiopods, these seem to present notaxonomic barriers for encrusting taxa. However, sub-strate specialization may occur (and seems more likely) atfiner levels of taxonomic resolution.

Encrustation frequencies among modern bivalves alsovary widely depending upon environment of deposition.Best and Kidwell (2000a) reported mean encrustation fre-quencies for tropical bivalves ranging from 0% in soft,muddy bottoms to a high of 45% in hard substrates (patchreefs and Halimeda sands). However, they also show awide range of variation within each setting, with patch-reef encrustation frequencies ranging from under 30% toover 70% (Best and Kidwell, 2000a).

Encrustation frequencies determined for Panamanianbivalves (Best and Kidwell, 2000a) are somewhat lowerthan the inner-shelf encrustation frequencies observed inthis study. This is not entirely unexpected—many bi-valves are infaunal, and thus, encrustation of the valve ex-terior is unlikely during the life of the animal. Further-more, the valves of infaunal bivalves must be exhumed be-fore encrusters can colonize them. Higher encrustationfrequencies were observed among epifaunal bivalves rela-tive to infaunal bivalves in all environments studied byBest and Kidwell (2000b), with frequencies exceeding 80%for some epifaunal bivalve assemblages. An additional, al-though minor, bias exists in the form of data collection: thePanama study was limited to valve interiors, while thisstudy examined all epibionts colonizing a host shell. Be-cause some hosts are colonized on the exterior exclusively,total encrustation frequencies in this study will be higher.When the interior encrustation of Bouchardia is evaluatedalone, inner-shelf specimens are encrusted at a pooled fre-quency of 80.8%.

Bivalves generally may exhibit lower frequencies of en-crustation than articulate brachiopods, because the latterare exclusively epifaunal and have greater exposure to en-crusting organisms. However, these results suggest thatstrong similarities exist in encrustation frequency and inthe composition of the fauna encrusting bivalves and bra-chiopods in present-day tropical-shelf assemblages. Thebrachiopod data reported here suggest that differences inencrustation at higher taxonomic levels may be maskedentirely by variations among genera or across environ-ments. This implies that the composition of the host valvemay be less important than other factors, such as waterdepth, valve surface texture, and primary productivity, indetermining the frequency of encrustation and the com-position of the encrusting fauna. However, the encrusta-

tion of co-occurring brachiopods and bivalves from thesame localities requires further evaluation, as do patternsof epibiont abundance and species richness per shell.

CONCLUSIONS

This study provides the first large-scale, quantitativestudy of encrustation of modern brachiopods from a tropi-cal and subtropical passive-margin shelf. While many Pa-leozoic brachiopod assemblages were deposited in epeiricseaways, rimmed margins, and foreland basins, this set-ting provides the best modern analogue for many ancientfaunas dominated by rhynchonelliform brachiopods. Theresults demonstrate the extent of spatial variation andemphasize the difficulties involved in interpreting fossil-encrustation data in terms of controlling factors. Unlikemany taphonomic-deployment experiments, a collection-based study cannot control all the variables that might in-fluence encrustation (e.g., host size, shell architecture, wa-ter depth, productivity), although many of these factorscan be measured as they vary from site to site.

While encrustation frequencies drop off markedly fromthe inner shelf to the outer shelf, they remain variablewithin both settings, and do not appear to correlate direct-ly to depth. Instead, the patterns observed appear to be re-lated to other factors that vary with depth, especially host-shell size and exposure history. Larger shells are encrust-ed more frequently than small shells, and mean shell sizeper site decreases with increasing depth. However, it isnot clear whether the increase in encrustation is simplybecause larger shells are bigger targets for settling epi-biont larvae, or if both encrustation frequency and hostsize are positively related to another factor, such as pro-ductivity.

The diversity, abundance, and taxonomic composition ofencrusting taxa are likely to be better proxies for produc-tivity than encrustation frequency. Barnacles appear lim-ited to areas of higher encrustation intensity on the innershelf, while foraminifera predominate on the outer shelf,where encrustation frequencies are low. However, the per-centage of shells encrusted by each epibiont taxon appearsto be adequate for finding the relative proportion of taxathat do not show gregarious settlement patterns.

Similar encrustation patterns are observed among Pa-leozoic and modern brachiopods, including variations inencrustation frequencies among species, preferential set-tlement of epibionts on ribbed species, and their avoidanceof spines or spinules, despite changes in the encrustingfauna. Whether the host substrate is bivalve or brachio-pod, modern shallow-water tropical and subtropical epi-biont assemblages appear to be dominated by serpulidworms and cheilostome bryozoans. The Mesozoic and Ce-nozoic record of brachiopod encrustation remains unex-plored, but patterns of encrustation established early inthe Paleozoic still can be observed in the modern ocean.Comparison of modern and ancient brachiopod encrusta-tion and evaluation of the controls on encrustation amongdifferent host taxa will provide a better understanding ofthe macroevolutionary history of epibiont faunas and theirrelationship with their hosts.

394 RODLAND ET AL.

ACKNOWLEDGEMENTS

Funding for this project was provided by the NSF grantEAR-0125149 and the Sao Paulo Science Foundation (FA-PESP) grant 00/12659–7. Outer-shelf brachiopod sampleswere collected in the scope of the Score Sul Benthos ocean-ographic project (http.//www.cnpq.br/programa/revizee.htm), a Revizee project funded by CNPq (the FederalAgency of Sciences, Brazil). Personal thanks are extendedto the coordinator of the Score Sul part of the project,A.C.Z. Amaral (Universidade Estadual de Campinas), forproviding brachiopod specimens and data on samplingmethods and sampled sites, and A. Fransozo (Universida-de Estadual de Sao Paulo, Botucatu campus) for providingmany specimens from the coastal region of Ubatuba Bayand for assistance during boat trips in the study area. Ad-ditional thanks go out to R. Bambach, K. Eriksson, J. F.Read, S. Scheckler, A. Hoffmeister, and S. Walker for help-ful comments on this project, to K. Parsons-Hubbard, T.Olszewski, H. Lescinsky, and one anonymous reviewer forthought-provoking and useful reviews, and to S. Barbour-Wood for assistance in brachiopod wrangling.

REFERENCES

AIDAR, E., GAETA, S.A., GIANESELLA-GALVAO, S.M.F., KUTNER,M.B.B., and TEIXEIRA, C., 1993, Ecossitema costeiro subtropical:nutrientes dissolvidos, fitoplancton e clorofila—a e suas relacoescom as condicioes oceanograficas na regiao de Ubatuba, SP: Pub-licacao Especial do Instituto Oceanografico Sao Paulo v. 10, p. 9–43.

ALEXANDER, R.R., 1984, Comparative hydrodynamic stability of bra-chiopod shells on current scoured arenaceous substrates: Lethaia,v. 17, p. 17–32.

ALEXANDER, R.R. and BRETT, C.E., 1990, Symposium on Paleozoicepibionts; introduction: Historical Biology v. 4, p. 151–153.

ALEXANDER, R.R. and SCHARPF, C.D., 1990, Epibionts on Late Ordo-vician brachiopods from southeastern Indiana: Historical Biology,v. 4, p. 179–202.

ALVAREZ, F., and TAYLOR, P.D., 1987, Epizoan ecology and interac-tions in the Devonian of Spain: Palaeogeography, Palaeoclimatol-ogy, Palaeoecology, v. 61, p. 17–31.

BAMBACH, R.K., 1977, Species richness in marine benthic habitatsthrough the Phanerozoic: Paleobiology, v. 3, p. 152–167.

BAMBACH, R.K., 1993, Seafood through time; changes in biomass, en-ergetics, and productivity in the marine ecosystem: Paleobiology,v. 19, p. 372–397.

BENNINGTON, J.B., and BAMBACH, R.K., 1996, Statistical testing forpaleocommunity recurrence: Are similar fossil assemblages everthe same?: Palaeogeography, Palaeoclimatology, Palaeoecology,v.127, p. 107–133.

BEST, M.M.R., and KIDWELL, S.M., 2000a, Bivalve taphonomy in trop-ical mixed siliciclastic-carbonate settings: I. Environmental vari-ation in shell condition: Paleobiology, v. 26, p. 80–102.

BEST, M.M.R., and KIDWELL, S.M., 2000b, Bivalve taphonomy in trop-ical mixed siliciclastic-carbonate settings: II. Effect of bivalve lifehabits and shell types: Paleobiology, v. 26, p. 103–115.

BORDEAUX, Y.L., and C.E.BRETT, 1990, Substrate specific associa-tions of epibionts on Middle Devonian brachiopods: implicationsfor paleoecology: Historical Biology, v. 4, p. 203–220.

BRAGA, E.S., and MULLER, T.J., 1998, Observation of regeneration ofnitrate, phosphate and silicate during upwelling off Ubatuba,Bra-zil, 23 degree S: Continental Shelf Research, v. 18, p. 915–922.

BRETT, C.E., LIDDELL, W.D., and DERSTLER, K.L., 1983, Late Cambri-an hard substrate communities from Montana/Wyoming: the old-est known hardground encrusters: Lethaia, v. 16, p. 281–289.

CAMPOS, E.J.D., GONCLAVES, J.E., and IKEDA, Y., 1995, Water masscharacteristics and geostrophic circulation in the South BrazilBight: Summer of 1991: Journal of Geophysical Research, v. 100,p. 18537–18550.

CAMPOS, E.J.D., VELHOT, D., and DA SILVERIRA, I.C.A., 2000, Shelfbreak upwelling driven by Brazil Current cyclonic meanders: Geo-physical Research Letter, v. 27, p. 751–754.

CARLSON, S.J., 1991, A phylogenetic perspective on articulate bra-chiopod diversity and the Permo–Triassic extinction: in Dudley,E., ed., The Unity of Evolutionary Biology, Proceedings of the 4th

International Congress of Systematic and Evolutionary Biology:Dioscorides Press, Portland, p. 119–142.

CARROLL, M., KOWALEWSKI, M., SIMOES, M.G., and GOODFRIEND,G.A., 2003, Quantitative estimates of time-averaging in brachio-pod shell accumulations from a modern, tropical shelf: Paleobiol-ogy, v. 29, pg. 381–402.

CHERNS, L. and WRIGHT, V.P., 2000, Missing molluscs as evidence oflarge scale, early skeletal dissolution in a Silurian sea: Geology, v.28, p. 791–794.

COHEN, B.L., and GAWTHROP, A.B., 1997, The brachiopod genome: inKaesler, R.L., ed., Treatise on Invertebrate Paleontology, Part H,v. 1, Revised, Brachiopoda: Geological Society of America and Uni-versity of Kansas, Boulder, Colorado and Lawrence, Kansas, p.189–211.

CUFFEY, C.A., ROBB, A.J., III, LEMBCKE, J.T., and CUFFEY, R.J., 1995,Epizoic bryozoans and corals as indicators of life and post-mortemorientations of the Devonian brachiopod Meristella: Lethaia, v. 28,p. 139–153.

CURRY, G.B., 1983, Microborings in Recent brachiopods and the func-tions of caeca: Lethaia, v. 16, p. 119–127.

DA ROCHA, J., MILLIMAN, J.D., SANTANA, C.I., and VICALVI, M.A.,1975, Southern Brazil: in Milliman, J.D. and Summerhayes, C.P.,eds., Upper Continental Margin Sedimentation off Brazil: Contri-butions to Sedimentology No. 4: Schweizerbart, Stuttgart, p. 117–150.

DALEY, G.M., 1993, Passive deterioration of shelly material: a studyof the recent eastern Pacific articulate brachiopod Terebrataliatransversa Sowerby: PALAIOS, v. 8, p. 226–232.

DOYLE, A.C., 1912, The Lost World: The Strand, v. 43–44.EMIG, C.C., 1990, Examples of post-mortality alteration in Recent

brachiopod shells and (paleo)ecological consequences: Marine Bi-ology, v. 104, p. 233–238.

FAGERSTROM, J.A., 1996, Paleozoic brachiopod symbioses: testing thelimits of modern analogues in paleoecology: Geological Society ofAmerica Bulletin, v. 108, p. 1393–1403.

FURSICH, F.T., and OSCHMANN, W., 1993, Storm shell beds as tools inbasin analysis: the Jurassic of Kachchh, western India: Journal ofGeological Society of London, v. 150, p. 169–185.

GIBSON, M.A., 1992, Some epibiont–host and epibiont-epibiont inter-actions from the Birdsong Shale Member of the Lower DevonianRoss Formation (west-central Tennessee, U.S.A.): Historical Biol-ogy, v. 6, p. 113–132.

GOULD, S.J., and CALLOWAY, C.B., 1980, Clams and brachiopods;ships that pass in the night: Paleobiology, v. 6, p. 383–396.

JACKSON, J.B.C., GOREAU, T.F., and HARTMAN, W.D., 1971, Recentbrachiopod-coralline sponge communities and their paleoecologi-cal significance: Science, v. 173, p. 623–625.

JAMES, M.A., ANSELL, D., COLLINS, M.J., CURRY, G.B., PECK, L.S.,and RHODES, M.C., 1992, Biology of living brachiopods: Advancesin Marine Biology, v. 28, p. 175–387.

KESLING, R.V., HOARE, R.D., and SPARKS, D.K., 1980, Epibionts of theMiddle Devonian brachiopod Paraspirifer bownnockeri, their rela-tionships to one another and to their host: Journal of Paleontology,v. 54, p. 1141–1154.

KNOPPERS, B., EKAU, W., and FIGUEIREDO, A.G., 1999, The coast andshelf of east and northeast Brazil and material transport:Geo-Ma-rine Letters, v. 19, p. 171–178.

KOBLUK, D.R., 1981a, The record of early cavity-dwelling (coelobion-tic) organisms in the Paleozoic: Canadian Journal of Earth Sci-ence, v. 18, p. 181–190.

KOBLUK, D.R., 1981b, Earliest cavity-dwelling organisms (coelo-bionts), Lower Cambrian Poleta Formation, Nevada: CanadianJournal of Earth Science, v. 18, p. 669–679.

KOBLUK, D.R., 1985, Biota preserved within cavities in CambrianEpiphyton mounds, upper Shady Dolomite, southwestern Virgin-ia: Journal of Paleontology, v. 59, p. 1158–1172.

KOBLUK, D.R. and JAMES, N.P., 1979, Cavity-dwelling organisms in


Lower Cambrian patch reefs from southern Labrador: Lethaia, v.12, p. 193–218.

KOWALEWSKI, M., DULAI, A., and FURSICH, F.T., 1998, A fossil recordfull of holes: the Phanerozoic history of drilling predation: Geolo-gy, v. 26, p. 1091–1094.

KOWALEWSKI, M., SIMOES, M.G., CARROLL, M., and RODLAND, D.L.,2002, Abundant brachiopods on a tropical, upwelling-influencedshelf (southeast Brazilian Bight, South Atlantic): PALAIOS, v. 17,p. 277–286.

LEIGHTON, L., 2001, New example of Devonian predatory drillholesand the influence of brachiopod spines on predator success: Pa-laeogeography, Palaeoclimatology, Palaeoecology, v. 165 p. 53–69

LESCINSKY, H.L., 1993, Taphonomy and paleoecology of epibionts onthe scallops Chlamys hastata (Sowerby, 1843) and Chlamys rubi-da (Hinds, 1845): PALAIOS, v. 8, p. 267–277.

LESCINSKY, H.L., 1995, The life orientation of concavo-convex bra-chiopods: overturning the paradigm: Paleobiology, v. 21, p. 520–551.

LESCINSKY, H.L., 1996, 100 million years of community evolution:gradual replacement and ecological trends in Paleozoic shell-en-crusting communities: in Repetski, J. E., ed., Sixth North Ameri-can Paleontological Convention Abstracts of Papers, Paleontolog-ical Society Special Publication No. 8, p. 235.

LESCINSKY, H.L., 1997, Epibiont communities: recruitment and com-petition on North American Carboniferous brachiopods: Journalof Paleontology, v. 71, p. 34–53.

LESCINSKY, H.L., EDINGER, E., and RISK, M.J., 2002, Mollusc shell en-crustation and bioerosion rates in a modern epeiric sea: taphono-my experiments in the Java Sea, Indonesia: PALAIOS, v. 17, p.171–191.

DE MAHIQUES, M.M., 1995, Sedimentary dynamics of the bays offUbatuba, state of Sao Paulo: Boletim do Instituto Oceanografico,Sao Paulo, v. 43, p. 111–122.

MANTELATTO, F.L.M., and FRANSOZO, A., 1999, Characterization ofthe physical and chemical parameters of Ubatuba Bay, northerncoast of Sao Paulo State, Brazil: Revista Brasileira de Biologia, v.59, p. 23–31.

MEYER, D.L., 1990, Population paleoecology and comparative taphon-omy of two edrioasteroid (Echinodermata) pavements: Upper Or-dovician of Kentucky and Ohio: Historical Biology, v. 4, p. 155–178.

NIELD, E.W., 1986, Liljevallia gotlandica: encrustation patterns inthe earliest cemented articulate brachiopod, and their implica-tions for its larval behaviour: Palaeogeography, Palaeoclimatolo-gy, Palaeoecology, v. 56, p. 277–290.

OLSZEWSKI, T.D., and PATZKOWSKY, M.E., 2001, Measuring recur-rence of marine biotic gradients: a case study from the Pennsyl-vanian–Permian Midcontinent: PALAIOS, v. 16, p. 444–460.

PAINE, R.T., 1963, Ecology of the brachiopod Glottidia pyramidata:Ecological Monographs, v. 33, p. 255–280.

PALMER, T.J., 1982, Cambrian to Cretaceous changes in hardgroundcommunities: Lethaia, v. 15, p. 309–323.

PARSONS-HUBBARD, K.M., CALLENDER, W.R., POWELL, E.N., BRETT,C.E., WALKER, S.E., RAYMOND, A.L., and STAFF, G.M., 1999, Ratesof burial and disturbance of experimentally-deployed molluscs:implications for preservation potential: PALAIOS, v. 14, p. 337–351.

PATZKOWSKY, M.E., and HOLLAND, S.M., 1999, Biofacies replacementin a sequence stratigraphic framework: Middle and Upper Ordo-vician of the Nashville Dome, Tennessee, USA: PALAIOS, v. 14, p.301–323.

PAULL, R.K., and PAULL, R.A., 1994, Shallow marine sedimentary fa-cies in the Earliest Triassic (Griesbachian) Cordilleran Miogeoc-line, USA: Sedimentary Geology, v. 93, p. 181–191.

POWELL, M.G., and KOWALEWSKI, M., 2002, Increase in evenness andsampled alpha diversity through the Phanerozoic: comparison of

early Paleozoic and Cenozoic marine fossil assemblages: Geology,v. 30, p. 331–334.

RICHARDS, R.P., 1972, Autecology of Richmondian brachiopods (LateOrdovician of Indiana and Ohio): Journal of Paleontology, v. 46, p.29–32.

RICHARDSON, J.R., 1997, Biogeography of articulate brachiopods: inKaesler, R.L., ed., Treatise on Invertebrate Paleontology, Part H,v. 1, Revised, Brachiopoda: Geological Society of America and Uni-versity of Kansas, Boulder and Lawrence, p. H463–472.

RUDWICK, M.J.S., 1970, Living and Fossil Brachiopods: Hutchinsonand Co., London, 199 p.

RUGGIERO, E.T., 1996, Notes on living brachiopod ecology in a sub-marine cave off the Campania coast, Italy: in Copper, P., and Jin,J., eds., Brachiopods: Proceedings of the Third International Bra-chiopod Congress, A.A. Balkema, Rotterdam, p. 227–231.

SEPKOSKI, J.J., MCKINNEY, F.K., and LIDGARD, S., 2000, Competitivedisplacement among post-Paleozoic cyclostome and cheilostomebryozoans: Paleobiology, v. 26, p. 7–18.

SIMOES, M.G., KOWALEWSKI, M, MELLO, L.H.C., RODLAND, D.L., andCARROLL, M., in press, Present-day terebratulid brachiopods fromthe southern Brazilian shelf: paleontological and biogeographicimplications: Palaeontology.

SPRINGER, D.A., and MILLER, A.I., 1990, Levels of spatial variability:the ‘‘community’’ problem: in Miller, W., ed., PaleocommunityTemporal Dynamics: the Longterm Development of MultispeciesAssemblies; Paleontological Society Special Publication v. 5, p.13–30.

TAYLOR, P.D., and MICHALIK, J., 1991, Cyclostome bryozoans fromthe Late Triassic (Rhaetian) of the West Carpathians, Czechoslo-vakia: Neues Jahrbuch fuer Geologie und Palaeontologie v. 182, p.285–302.

THAYER, C.W., 1975, Morphologic adaptations of benthic inverte-brates to soft substrates: Journal of Marine Research, v. 33, p.117–189.

TORELLO, F.F., SIMOES, M.G., and PASSOS, J.R.S., 2002, The tapho-nomic tumbling barrel—a methodological review to understandpreservational biases in the fossil record: First International Pa-laeontological Congress, 2002, Sydney, The Geological Society ofAustralia, Abstracts, McPherson’s Printing Group, Sydney, v. 68,p. 285–286.

TUNICLIFFE, V., and WILSON, K., 1988, Brachiopod populations: dis-tribution in fjords of British Columbia (Canada) and tolerance oflow oxygen concentration: Marine Ecology Progress Series, v. 47,p. 117–128.

VEGA PEREZ, L.A., 1993, Estudo do zooplancton da regiao de Ubatuba,Estado de Sao Paulo: Publicacao Especial do Instituto Oceano-grafico Sao Paulo, no. 10, p. 65–84.

VERMEIJ, G.J., 1995, Economics, volcanoes, and Phanerozoic revolu-tions: Paleobiology, v. 21, p. 125–152.

WATKINS, R., 1981, Epizoan ecology in the type Ludlow Series (UpperSilurian), England: Journal of Paleontology, v. 55, p. 29–32.

WESSEL, P., and SMITH, W.H.F., 1998, New, improved version of theGeneric Mapping Tools released: American Geophysical Union,EOS Transactions, v. 79, p. 579.

WILSON, M.A., 1982, Origin of brachiopod-bryozoan assemblages inan Upper Carboniferous limestone: importance of physical andecological controls: Lethaia, v. 15, p. 263–273.

WILSON, M.A., 1998, Succession in a Jurassic marine cavity commu-nity and the evolution of cryptic marine faunas: Geology, v. 26, p.379–381.

WRIGHT, P., CHERNS, L., and HODGES, P., 2003, Missing molluscs:field testing taphonomic loss in the Mesozoic through early large-scale aragonite dissolution: Geology, v. 31, p. 211–214.

ZIEGLER, A.M., COCKS, L.R.M., and BAMBACH, R.K., 1968, The com-position and structure of lower Silurian marine communities: Le-thaia, v. 1, p. 1–27.

ACCEPTED JANUARY 30, 2004

colonization of a 'lost world': encrustation patterns in modern subtropical brachiopod...

Documents