Substrate Specificity in the Devonian Tabulate Coral Pleurodictyurn

Substrate specificity in the Devonian tabulate coral Pleurodictyurn

CARLTONE. BRETT AND JOHN F. COTTRELL

Brett, Carlton E. & Cottrell, John F. 19820815: Substrate specificity in the Devonian tabulate coral LETHAIA pleurodicryum. Leihain, "01.15. pp 247-262. o110 ~ssN0024-1164. The tabulate coral Pleurodiciyum nmericonum Roemer has been cited as an example of a host-specific organism occurring exclusively on the shells of gastropods, particularly Paloeozvsopkuro homihonioe (Hall). Examination of over 16W specimens of P. omericonum, from the Middle Devonian Hamilton Group of wcrtcrn New York, reveals additional cornplcriticr which require reinterpretation. While substrate selectivity for Palaeoiygopleuro shells is evident in all 42 subsamples, a variety of other substrates were also utilized by Pleurodictyum including corals, brachiopods, other molluscs and pebbles. Recent rcleractinian corals inhabiting soft bottoms show similar substrate preference, selecting for the tubes of live serpulids, or gastropod shells (invariably with a secondary sipunculid host), but also occasionally settling on unoccupied shells or pebbles. Shell surfaces of P. harniltoniae, preserved as external molds on the Pieurodiciyum epitheca, exhibit cnciustation by worm tubes and bryozoans as well as boring5 and mechanical shell damage, suggesting that these were not ihc shells of live gastropods. However, the invariant aperture-downward orientation and thc high degrcc of selectivity of P. arneri- canurn strongly suggest that the shells were occupied by secondary hosts. O Subsrrarespecifi~iry,cornmen- mlirm, tobulore coral, garfropod, sipunculid, Devonian, Hamilton Group, New York. Corlron E. Brert ond John F. Cornell, Depcrimsnr of Geologicoi Sciences, the Univrrsiry of Rochester, Rochesrer, New York 14627; 15th March, 1981.

Larvae of many marine invertebrates require substrate specificity in Pleurodictyum; however, hard, stable substrates for successful settlement it also reveals additional complexities which re- on the sea floor (Thorson 1950; Gray 1974). quire reinterpretation of the relationship. Many species exhibit nonrandom colonization of particular suhstrate types, including the skeletons of other living host organisms (Scheltema Materials and methods 1974; Knight-Jones 1951; Crisp & Meadows 1967; and others). Substrate specificity can also Pleurodictyum is a small tabulate coral with a be demonstrated among fossil organisms in roughly-. discoidal corallum composed of relative- which the original attachment site is incorporated ly large, flaring, prismatic corallites and hearing into the adult skeleton or into the skeleton of a wncentricallv wrinkled basal eoitheca (Hill & another organism (Clarke 1908, 1921; Teichert Stumm 1957). The corallum exhibits a high de- 1945; Thayer 1974; Brett 1978; MacNamara gree of size and morphologic variability, ranging 1978; Conway Morris 1981). The Devonian tabu- from large hemispherical, flat-based forms, late coral Pleurodictyum has previously been which are generally the most abuhdant morpho- cited as an ancient organism showing suhstrate type, to conical and biconical forms, and irregu- specificity (Ager 1963). As early as 1908, John lar or aberrant morphotypes (Fig. 1). All forms M. Clarke noted the frequent association of this are intergradational and prohabiy represent eco- coral with particular -gastropods and inferred the phenotypic variants of a single species (Ross existence of a commensal relationship, in which 1953). Pleurodictyum americanum occurs ahnn- the coral larvae selected live snail hosts on which dantly in the Hamilton Group of New York to settle. While this suggestion has been much State, in a wide variety of lithologies. The corals quoted, the relationship has never been thor- are most abundant and largest in offshore medi- oughly documented and the hypothesis remains um to dark gray shales of western New York untested. Our detailed study of approximately (e.g. Wanakah and Windom members), but also 1600 specimens of Pleurodictyum americanum occur rarely in coarser clastic facies in the Hamil- Roemer from the Hamilton Group of western ton Group of eastern New York (e.g. King Fer- New York State supports Clarke's suggestion of ry). Pleurodictyum frequently occurs in sparsely 248 Carlton E. Bren and John F. Comell

Pleurodictyurn occur in the Jaycox Member of the Ludlowville Formation and in the Kashong Shale Member of the Moscow Formation (Baird 1979). The highest stratigraphic occurrences of Pleurodictyum in western New York are in two richly fossiliferous zones of the Windom Shale in the upper portion of the Moscow Formation (Brett 1974; Baird & Brett, in press). A total of 42 large samples of Pleurodictyum were obtained from various horizons at 31 localities in western and west-central New York State (Fig. 3; Appendix). By far the largest samples were derived from the Wanakah Shale Pleurodic- tyum beds and the basal Windom Shale coral beds, both in Erie County. Specimens were prepared by boiling in Quaternary-0 for several hours and then treated with an ultrasonic clean- er. For most specimens in the soh shales of western New York this treatment served to re- move the matrix completely from basal surfaces of the corals and their attachment scars. After cleaning, latex casts were prepared of certain samples that showed attachment scars of Pleuro- dictyum. In three sites, where Pleurodictyum Fig. 1. Morphologic vanation in the corallum of Pleurodictyum specimens were particularly common, large cf. americonum Roemer. OA. Morphotype 1: planar base with blocks of matrix rock were collected and disag- nearly planar 'upper' orallite surface, BMS E 24955. OB. gregated in the lab. All contained fossils and Morphotype 2: planar base with convex upper surface, BMS E 24954. OC. Morphotype 3: convex base with planar upper their preservations were recorded to determine suriace, BMS E 24956. OD. Morphotype 4: convex base with the bulk composition of the Pleurodictyum-bear- convex upper surface, BMS E 24957. OE, F. Morphotype 5: ing fossil assemblages. All corals included in the specimens exhibiting irregular coralla, BMS E 24958. E 24959. present study are believed to represent a single B others x from Loc. 5A, all from Loc. 5B. AU views 1.2. species, Pleurodictyum americanum; however, further detailed taxonomic studies of the genus Pleurodictyum are needed. fossiliferous mudstones which lack rugose corals. It evidently was capable of inhabiting relatively soft mud bottoms which otherwise supported Substrates of Pleurodictyum only small brachiopods and mollusks. Indeed, Pleurodictyum has been cited as an example of The basal epitheca of Pleurodictyum commonly an organism adapted for soft substrates (Thayer incorporates the remains or impressions of ob- 1975). jects which provided initial attachment sites for Pleurodictyum is largely confined to several the corals (Fig. 4). Thus, in the majority of cases, thin but persistent horizons (Fig. 2; Cooper 1930, it is possible to identify the initial settlement 1957). The lowest horizon occurs at the base of substrate. A wide variety of hard substrates was the Ludlowville Formation in the transition be- utilized by Pleurodictyum including both skeletal tween the Centerfield Limestone and the Led- and inorganic materials such as phosphate peb- yard Shale (McCollum 1980). A very widespread bles and hiatus concretions (Fig. 4L, M). Over 60 unit termed the 'Pleurodictyum beds' by Grahau different skeletal substrates have been identified (1899) occurs at the base of the Wanakah Shale in our study (Tables 1 and 2). These include from Lake Erie to the central Finger Lakes re- mgose and auloporid corals (Fig. 4F), brachio- gion. Two or three other levels exist in the higher pods (Fig. 4C), and a variety of mollusks includ- units of the Wanakah and the laterally equivalent ing cephalopods, bivalves, and gastropods (Figs. King Feny Shale in the Seneca-Cayuga Lake 4A, B, E, G-I, K). Thus, statements which im- region (Baird 1981). Scattered specimens of ply that Pleurodictyum americanum occurs exclu- JO uo!$ualuoa aql ~ioddnsisal s!ql JO sqnsax .sa%e[q .sa~dmesy1nq pue alerpqns aql qaoq u! a8qq -masse a!io!q Ieg!u! aq, u!q!m spodo1lse8 asaql -massa l!ssoj aqllo siaqmam laqio lsu!r?8e av!uoJ JO aauepunqe Leu!ploelixa ue )aaUai d[dm!s -[!uvy vlnaldo8i(zoav~vd$0 aauepunqe aq) arud ~q%!rusa)eilsqns se v.inaldo8i(zoavlvd JO uoyod -moa 01 alqel Laua8u!luoa z x z e 8u!sn 'saldmes -old q%!q aql 'rahamoH .i(lyy!aads aleqsqns lsa2 aaiq) aq) $0 qaea 103 elep asaq uo pamiopad 4ns elep mo laadsai s!q) UI .alerasqns [elalays sem aauapuadapu! jo pa] ,X v .a%e~qmasseqlo!q JO sauads a[Zu!s uommoa Isom aql s! vrnaldo8 paleraosse aql jo % E ueq, ssa1 sasudmoa a! pay -i(zoav1vd 'saldmes lsom UI .Lalle~eu8!epueue~ -mexa suamuads un4xpodnald aq, jo %og ueql aql $0 aIeqs qeyemfi aql u! % 18 01 bun03 aux aiom u! alensqns Iuamqaene ue se srnaao vlnald $0 aleqs mopu!M a41 u! % LZ moq sa%uelvlnald -o2i(zoav~vdseaiaqm IE~Ismoqs 'huno3 aug -02zoav~vduo aauarrnaao 30 a8emaarad aq) 'uogeas yooig Bu~rd~ae apqS qeyeuefi lam01 'uozuoq a!qde@lens al8u!s e moq suam~aads aql 30 paq uni(rxpo.n~aldaqlloj elep xulem aq) unti03!polnald arolu lo 0s jo saldmes UI (,u!e%e pue aleilsqns aqljo uosuedmos e 'aldmexa rod aauo aep!)emauoxoT aqi q$!m sapads s!ql Bu!pn[a '(E alqe~)urnAj3!polnald $0 Sa)ellsqns mamqael -u! zuelrem Lem qalqm qauoaoloid aqljo s[!e)ap -,e 40 4!iolem aq) sasudmoa I! aiaqm sa!l!Ieao[ a~asaidel[eloa urnAj2!polnald lo aseq aql uo awes aql le sa8elqmasse pssoj pa~asaidIEIOI aq, splom leil4eN 'uo!,sanb u! s! 'r~61'ral[!d~ pue jo uo!podoid rou!m e Sluo io~slunosae vlnaldo8 a%paip[~'su!llox Lq aepuualdo%koae[edaq) ol -Lzoavlvd asea qaea u~ .paq [el03 aIeqS mopu!M 'av!uoll!wvy vurauoxo7 dllauuoj 'sapads s!ql JO aql moq auo io~pue spaq uni0qpornald aleqs auamu%pseaq) 'Spnls luasaid aq) JO aauanbasuoa qeyeuefi aql moq sa~dluesom) loj auop sem s!ql e sv) .av!uojpwvy vma~do2i(zoav~vdpodoilse8 :Laqesol pue uozuoq ames aq) morj saldmes yInq aql uo inaao saldmes rojem [le u! suampads jo u! exel 30 suo!llodord ql& paledmoa alam sald uogiodoid a%rqLra~ e )eql alerisuomap sa!pn)s -mes ie[na!lred u! exel alensqns unAj~!po~na1d ino 'pueq iaqlo aq, uo .roria u! Lpeap are (£961 JO sno!lrodord 'i(~![!q!ssod s!ql au!mexa o~ 1a8v :so61 ayreg) sI[aqs podonse8 uo Llan~s 250 Carlion E. Bren and John F. CottreN LETHA~A15 (1982)

-.. Erie CO. I .- . -. - .- . - . -. !Ontario ! I .. 7 ! a9 co. .6 ! ..! 11. . . 5 !.. !.. 1 Yates Co,

I I Living -

...... - . .. . - . - ., 8. . -. ,' 1, .,

- - - krn 0 10

Fig. 3. Index map for localities in western and west central New York State. Numbered localities are dercribcd in Appendix. Modified from Baird 1979, Fig. 1B. substrate selectivity by Pleurodictyum for P. ha- quencies of various skeletonized taxa in the origi- miltoniae in each case (e.g. for Wanakah Shale, nal community (i.e. random settling of coral lar- sample 1, test statistic = 314; X21,0.01= 6.64). vae onto skeletons on the sea floor). Then, the The comparison of substrate and matrix data is discrepancy with the matrix data would be due to complicated by taphonomic processes, in particu- loss of aragonitic shells; Comparing the ratio of lar, the differential ~reservabilitvof calcitic vs. Palaeozyaopleura.-. to readily. preservable. calcitic aragonitic fossils. Some taphonomic loss of Pa- fossils (e.g. brachiopods) in the substrate vs. ma- l~~ozy~y~pl~rrreand other irqnnltlc tosstls due ru rr~r,:tnlples 11 is evl~lcnrth:

Fis. 4. Attachment substrates of Pieurodicrywn americonum Roemer; A, B, D, E, G, H and K arelatex peels iromimpressions on basal epitheca of Pieurodicryum coralla. OA. Gastropod, Mourionio capiilario (Conrad). Loc. 6, BMS E 24960, x 4. OB, E. Bivalve, Nucula sp. Coralla attached to articulated shells in two different orientations. In B. attachment is to umbonal region, and in E to the commissural area of gaping valves. Loc 58, BMS E 24961, E 24964. x 4, x 2.3. OC. Brachiopod, Mediospirifer oudncuiu (Conrad), Loc. 3, BMS E24962, x 1.2. OD. Undetermined substrate, possibly lithistid sponge. Loc. SB, BMS E24963, x 4. OF. Rugose coral, Srpreoiarmorecrum (Hall). Loc. 3, BMS E24965, x 1.6. OG. Gastropod, bellerophontid. Loc. 5B, BMS E 24966, x3.7 OH. Molluse(?) Coieolur sp., Loc. SB, BMS E 24967, xZ.5 01. External mold of orthoconic nautiloid on Pieurodicryum epitheca. Note impression of a septum with riphunde on broken end of phragmacone. Loc. 5B, BMS E 24968, x 3.1. OJ. Crinoidcolumn embedded in bare of Pleurodicryum. Loc. 10, BMS E 24969, x 1.6. OK. Gastropod, Poiaeozygopieuro homiltonioe (Hall). Loc. 9, BMS 24970, x 2.2. OL, M. Picurodvryum on bored and epizoan-encrusted hiatus concretions. L is view of corallum from above, M is a cross-section through Pieurodicryum and concretion. Loc. 23A, BMS E 24971, E 24972, X 1.0. LETHAIA 15 (1982) Substrate of PLEURODICTYUM251 252 Carlton E. Brett and John F. Cottreli LEl'HAIA 15 (1982)

Toble I. Pleurodiclyurn substrates. Key to locality numben in Appendix.

Substrate Wanakah Jaycox King Kashong Windom Ferry

Locality numbers: 1-7 a10 11-13 14-15 1hU 13B 188 22-23 8B,9 11 14 18C.21 3.58.78

Gastropods Poloeozy~opleurn bcllerophontids others Bivalves Poleoneilo Cypticordella nuculoids others Other mollusn Coleolur nautiloid fragments ~OIIUSCfragments Corals Amplexipkyllum SLereolmma Pleurodictyum auloporids "gore othen Brachiopods Tropidoleprur Mediospitifer chonetids spiriferidr strophomenids others Trilobites Bryozoans Crin0idr Other organia Inorganics phosphate pebbles hiatus concretions Undetermined substrate present no substrate

Totals: 510 10 71 70 905

this is an unrealistically high value. In processing tyum specimens were collected with matrix rock, bulk samoles we were careful to extract all fossils several show the oresence of smeared-out molds regardless of their preservation. These samples of Palaeozygopleura in the attachment scar (Fig. demonstrate that there was not complete loss of SA, B). Evidently, some recognizable molds aragonitic fossils. Rather, severely compressed, were preserved, and provide a record of these plastically deformed external molds of mollusc aragonitic shells. These observations suggest that shells occur abundantly within the shale matrix. the demonstrable discrepancy between substrate Molds of Palaeozygopleura are readily recogniza- abundances and matrix abundances is a real phe- ble because of their elongate shape and distinc- nomenon. In turn, this indicates a high degree of tive ornamentation (Fig. 5). They occur in all selectivity for Palaeozygopleura shells by settling sizes, suggesting that taphonomic losses, if pre- larvae of Pleurodictyum. Thus, we characterize sent, were not size-selective. Where Pleurodic- Pleurodictyum as a semi-specific epizoan organ- LETHAIA 15 (1982) Substrate of PLEURODICTYUM253

Table 2. Substrates of attachment for Pleurodic~umomeri- cnnum expressed as percentage of total identified substrates.

Substrate Erie Generec Gencrcs &an Cayuga county county valley dalgua Valley valley

WanakahKine Ferry localities: Paloeozygopleura 55.3 46.7 77.3 81.5 71.2 Other molluscs 24.5 20.0 18.1 1.9 15.4 Corals 3.1 - - 1.9 5.8 Brachiopods 9.2 26.7 - 13.0 3.9 Other organics 4.0 6.7 4.6 1.9 3.9 Kashong localities: Paineoiygopleurn - 60.0 83.3 30.0 - Othcr molluscs - 20.0 6.6 20.0 - Corals - - - - Brachiopodr - 20.0 6.6 40.0 Other organics - - 3.3 10.0 Windom localities: Polneorygopleuro 27.7 Other molluscs 22.1 Corals 27.9 Brachiopodr 20.8 Other organics 4.6

Table 3. Comparison of taronomic composition (numbers of specimens and corresponding percentages) of Pieurodicryum substrates with that of the total preserved assemblages in the Wanakah Shale (Soring.. - Brook Station, tributari to Buffalo Creek) and the Windom Shale (Carenovia Creek).

Tara Wanakah Shale Windom Shale ----substrate 'rota1 Substrate Toul n en %n %n %

Pnloeoiygopleura 78 61.4 15 2.4 120 26.7 1 0.3 Othergastropods 2 1.6 9 1.5 13 2.9 1 0.3 Bivalves 11 8.7 14 2.3 27 6.0 I5 3.8 Nautiloids 0 0 6 1.0 22 4.9 3 0.8 Othermollurcs 20 15.7 4 0.6 42 9.3 0 0 Corals 1 0.7 6 10 123 Ti3 110 28.1 Brachiopods 10 7.9 487 78.9 84 18.7 131 33.5 Bryozoanr 2 1.6 12 1.9 1 0.2 2 0.5 Trilobites 3 2.4 21 3.4 13 2.9 88 22.5 Crinoidr 0 0 43 7.0 5 1.1 40 10.2 Undetemined 11 53 No substrate 4 91

ism. It could, and frequently did, colonize virtu- Fis. 5. Presewation modes of Poineozvgopieura homdroniae. ally any hard substrate hut it shows a OA. Compressed compoitte internallexternal shalc mold re- nonrandom preference for Palaeozygopleura. mainine intact on undersurface of ~ieurodictyumspecimen. Modern marine larvae are known to become Lac. 17, BMS E 24991, x 1.6. OB. uncompressed iteinkern decreasingly selective for substrate in later devel- (calcareo~~shale) on underside of Pleurodicwum. Loc. 22, BMS E 24992, x3.6 OC. Highly compressed ('smeared') 'pmenta1 larval stages (Knight-Jones 1953)2 internalierternal molds of Paloeory~opleuro.Note smooth sur- which provides a probable explanation for semi- h~,,,eth ,scken,ide, (arrow points to area of preserved lime selectivity of host substrates. We assume that the ornamentation). Loc. 16, BMS E 24993, x 1.8. 254 Carlton E. Brett and John F. Cottrell LETHA~A15 (1982) same may have applied to Pleurodictyum. Observations of Recent empty turreted gastro- Hence, while the Pleurodictyum larvae probably pod shells, although slightly different in morpho- had a preference for Palaeozygopleura shells, if logy from Palaeozygopleura, suggest that the such shells were unavailable the larvae did ulti- most stable orientation is aperture upward. Even mately settle on any other available substrate. in a strictly random distribution, at least some Finally, P. hamiltoniae is probably not the only portion of the shells would be expected to occur species which was selected as a substrate by in this orientation. Indeed. accumulations of Pa- Pleurodictyum. Certain other shells also occur in laeozygopleura on bedding planes in the Hamil- greater than expected frequency (based on bulk ton Group do show both aperture-upward and sample data). Most notable is a smooth, tube- aperture-downward orientations. Observations like, annulated shell, tentatively identified as Co- of 56 pyritized steinkerns of P. hamiltoniae from leolus sp. (Fig. 4H). Coleoll*~is generally quite bedding planes of the Ledyard Shale at 11 Mile rare in the matrix rocks. Although this tawon Creek, Genesee County, New York, show an accounts for only a small proportion of Pleuro- aperture-downlaperture-upratio of 1:1.95. Simi- dictyum substrates, it is present in many of the lar observations of 53 'smeared' shale molds samples. This suggests possible selectivity, al- from the upper Wanakah Shale at Deep Run thouehv insufficient samole sizes are available for Gully, Canandaigua Lake, yield a ratio of 1:1.3. statistical testing. Evidently the aperture-upward configuration was, if anything, more common in most accumulations of Palaeozygopleura shells. Thus, our data strongly suggest that Pleurodictyum did not Biostratinomy of substrate fossils settle randomly on empty Palaeozygopleura shells. Rather, the shells were preferentially ori- Orientation ented aperture downward, i.e. in apparent life The orientation of shells utilized as substrates by orientation. Pleurodicryum provides additional insights into Observations were also made on the site of the relatioushio between this coral and its sub- attachment of Pleurodictyum on the Palaeozygo- strates. Aside from Palaeozygopleura, most skel- pleura shells, to determine whether or not the etal remains were encrusted by Pleurodictyum as coral had a preference for a particular region of they lay on the sea floor in a variety of post- the shell. Each shell was subdivided into three mortem orientations (compare Fig. 4B and E). regions corresponding to the apex, central shell, However, in the case of Palaeozy~opleura, the and the body whorl. That portion of the shell orientations of the shells at the timi of encrusta- (i.e. apex, body whorl or central region) lying at tion were remarkablv consistent. Wherever the the approximate center of the epitheca of each body whorl of the gastropod has been incorpo- Pleurodictyum colony was judged to be the posi- rated into the Pleurodictyum epitheca, the shell tion of initial attachment of the founder corallite is observed to he oriented in an aperture-down- of that colony (Fig. 61, J). Counts revealed a ward position. In no case have we noted the relatively high frequency of settlement on the aperture of Palaeozygopieura preserved as a body whorl relative to the center or apex (55.3 % mold on the base of Pleurodictyum. vs. 24.3 % vs. 20.4%). Initially this suggested a

Fit. 6. Biostratinomy of shells utilized by Pleurodicryum as attachment substrates. Specimens AX, E, F and H are later peels from external molds of Poloeozy~opleurohomiltonioe (Hall) on the basal epitheca of Pleurodicryurn cf. nrnericnnum Roemer. UA. Pre-monem breakage and repair in Poloeozygopleuro. Note crescentic reflection of growth lines in body whorl representing a 'healed' break in the apenural lip which occurred during growth of the gastropod. Note also the encrusting discoid foraminifers (?) and serpulids. Loc 6, BMS E 24973, x3.6. OB. C. Solution (?) pitting in Polncoiygopleurn shells. Pits appear centered on possible sponge botings; note the 'bullseye' pattern due to succesrive penetration of shell laminae. In B, the shell shows well dcfined ornament, whereas in C, the exterior of the shell has been desrroyed. Loc. 5B, BMS E 24974, E 24975, x 4.1. OD. Analogous 'bullseye' solution pits on shell of Recent freshwater gastropod Helkorno rp., x4.1. OE, H. Post-mortem rhell breakage in Polaeozygopleuro, both specimens show broken margins on the body whorls and E displays a broken apex. Loc. 6, BMS E 24976, E 24977, x 3.6, x4.1. UP. Shell 'smoothed' by complete removal of ornamented outermost layer (remnanls of this layer remain along sutural shoulden). Loc. 3. BMS E 24978, x 4.1. OG. Basal epitheca of Pleurodicrywn incorporating a shell of Pnlaeorygopleura which has been completely ensheathed with encrusting trepostome bryozoan (Poloeschora?) prior to overgrowth by the Pleurodicwum. Note the bryozoan sheath has been crushed following dissolution of the gastropod shell. Loc. 11, BMS E 24979, x2.1 01, J. Basal surfacer of Pleurodicryum coralla showing varying positions of colony center with respect to Poloeozygopleura shells. In I, the orallum is centered on the body whorl of the gastropod shell. In 1, the colony center is near the apex of the shell. Loc. 8B. BMS E 24980, E 24981, x 1.1. LETHAIA 15 (1982) Substrate of PLEURODICTYUM255 256 Carlton E. Brett and John F. Come11 LETHAIA 15 (1982) preference for settlement of the coral onto the paired breaks in the shell, and corrosion or disso- body whorl: however, it is also evident that the lution pitting. In most samples, a significant pro- body whorl presented a considerably. larger. target portion of the shell casts show breakage. These are~forscttlemcntthxn d~dthz apcal reglonof;he ;nclude broken and :h~ppe~lspice, as well :a\ hhcll. Rv cumD;!rlne freauenc\ uf ;lttachment in :I hr~lc*throueh the sllell\ cFia nE, 11) Jistin:- .".2 ,- . . particular section with the relative proportion of tive type of corrosion pitting, characterized by a ;hell surface area represented by that section, it concentric or 'bullseye' pattern (Fig. 6C) appar- can be shown that Pleurodicivum was not. in fact. ently reflects breaching of successive shell layers selective with respect to position on the around an initial hole through the periostracum. Palaeozygopleura shell. Using a x2 test, the null Very similar pitting, of unknown origin, has been hypothesis - that the proportion of corallite at- found to occur in Recent freshwater gastropods tachments in a given region of the shell is not (Fig. 6D). Complete removal of the outer shell significantly different from the proportion pre- layers results in some cases of 'smoothed' shell dicted by the relative surface area of that shell surfaces (Fig. 6F). portion - cannot be rejected at all normally ac- The substrate shells were alsb tv~icallv,. . bored cepted levels of probability (test statistic= 4.15; and encrusted by epizoic organisms prior to the x22,0.05=5.99, x22,0.1=4.60). final overgrowth by Pleurodictyum (Fig. 7). In some cases the gastropod shells were completely ensheathed by a layer of encrusting bryozoan. In Condition PALAEO~YGOPLEURAshellsprior of Fig. 6G the bryozoan layer has collapsed due to to encrustation by PLEURODIC~UM removal, by dissolution, of the aragonitic gastro- Latex peels of Palaeozygopleura shell molds re- nod shell and subseauent com~action.Most Pa- veal fine details of shell surface including evi- laeozygopleura and other aragonitic shells were dence for encrustation, shell breakage, and cor- encrusted by serpulid worms (Fig. 7A, D). A few rosion. Both pre-mortem and post-mortem shell also exhibit Spirorbis, ctenostome bryozoans, damage are in evidence. Pre-mortem shell dam- and possible sponges (Fig. 7G, H, I). Several age consists of readily discernible shell repairs. molds also showed the presence of portions of These are recognizable as crescentic lines which the articulate brachiopod Schuchertella, seeming- cut across the vertical shell ornamentation, ly attached to the Palaeozygopleura shell prior to marking fractures of the former shell apertural its encrustation by a Pleurodictyum; this supports margin. The broken apertures were 'healed' by previous suggesti-ons that these brachiopods were later accretion of shell material, often accompa- cemented bv the interarea reeion. Manv of the nied by distortion of growth lines and ornamen- Palaeozygopleura shells also exhibit minute cir- tation (Fig. 6A). Crescentic breaks (repaired and cular pits, most likely the borings of clionid unrepaired) occur frequently in Recent snail ap- sponges (Fig. 7D, E). ertures and they often result from the action of The abundance of damaged and encrusted Pa- predators such as crustaceans (Vermeil 1977). laeozygopleura shells suggests that these were Discovery of similar breakage and 'healing' in not occupied by living snails at the time of en- Devonian gastropod molds represents the oldest crustation by Pleurodictyum. Most gastropods, record of such breaks (G. J. Vermeii,. personal. includiny Palaeozy~opleura, as demonstrated communication, 1980). This also provides evi- herein, are capable-of healing breakage in the dence for (attempted) predation on gastropods in a~erturalreeion " of the shell:. vet, numerous Pa- the Devonian. laeozygopleura specimens show unhealed break- Probable post-mortem damage includes unre- age. High-spired snails such as Palaeozygopleura

Fig. 7. Epifaunal and endolithic organisms on shells utilized as substrates by Ple~~odicryurn.With the exception of H and I, all specimens are later casts. OA. Basal 'U-tube' of Hverer on Coleolu sp., Loc. 5B. BMS E 24982, x 4.1. OB. Minute pits and sinuous channels (sponge borings?) and serpulids on a fragment of mollusc shell. Loc 5B, BMS E 24983, x 4.1. OC.Encrusting bryozoan (mold of zooecial apertures) on the shell of Poloeozygopleurn. Loc. 6, BMS E 24984, x 3.8. OD. Serpulid tuber, sponge(?) borings and the brachiopod Schuchenello d. nrctosnioto (Hall), on body whod of Pnlneorygopleuro. Loc. 5B. BMS E 24985, x4.1. OE. Poloeorygopleura shell 'riddled' with minute sponge(?) pi-. Loc. 6, BMS E 24986, x3.1. OF. Stolons of ctenostome blyowans on Paioeozygopleuro. Loc. 5B. BMS E 24987, x 3.6. O G. Polaeozygopleuo encrusted by several serpulid and Spirorbir worm tubes and blyomans. Loc. 9 BMS E 24570, x 3.1. 0 H. Base of Pleurodicryum on a fragment of brachiopod valve encrusted (on interior of shell) with the etenostome bryoroan Ascodicryon. Loc 5B, BMS E 24988, r 3.9. 01. Valve of bivalve Prerinopecten rp. cut by largechannel-like annelid (?) boring. Loe. 2, BMS €24989, x 2.0. 01. Undetermined encrustation on body whorl of Poloeozygopleura, possibly algal or sponge. Loc. 5B, BMS E 24990, x 3.7. LETHAIA 15 (1982) Substrate of PLEUROOIC~UM257 258 Carhon E. Breff and John F. CottreN LETHAIA 15 (1982)

Table 4. Known associations of encrusting corals and gastropod shells.

Coral Attachment substrate Age Sediment Remarks

Heteropsammio gastropod Recent muddy rand Coral settles on gastropod shell occupied by rec~ ondary host, Aspidosiphon (sipunculid) (Schindewolf 1959) Heteeocyothlrr gastropod Recent muddy rand Coral settles on gastropod shell occupied by set- ondary hort, Aspidosiphon (sipunculid) (Schindewolf 1959) Psommoserir gastropod Recent Coral settles on gastropod shell occvpied by secondary host, Aspidosiphon (sipunculid) (Schindewolf 1959) Caryopkyllio Dinupo (rerpulid) Recent tine sand Although Ditrupo appean to be the primary rite selected for, the coral also settles on shells of Aponhnir and Dentaliurn which are occupied by the secondary host, Pkascolfon (sipunculid) (Wil- son 1976) Syrnbiongia Terebro (gastropod) Miocene Small coral completely encrusts gastropod shell; ~o~spicuousmbe indicates possible secondary sipunculid hort (Wells 1972)

are almost invariably shell-draggers; i.e. they (Table 4). In all known cases, the corals are small carry the shell subhorizontally (Linsley 1977, forms which live predominantly on mud or mud- 1979; P. W. Signor, personal communication). dy sand substrates. All are somewhat selective This might inhibit the growth of epizoans on the for particular snail shell types, but as far as shell. Also, the spiral growth of Palaeozygo- known, none settle on living gastropods. Rather, pleura shells would result in.epizoan encrusters they invariably occur on the shells of dead snails (including Pleurodictyum) being periodically ro- which are secondarily occupied by sipunculid tated downward into the mud of the sea floor. worm hosts. Evidently, such corals are capable These lines of evidence suggest that the gastro- of inhabiting muddy areas inhospitable to most pod shells were not occupied by live snails at the other species because of 'towing' by the sipuncu- time of attachment by Pleurodicryum. lids which maintains the polyps above the sediment-water interface during critical early growth stages (Gill & Coates 1977). The most detailed study of a Recent coral- Discussion substrate relationship was that made by J. B. Data presented herein suggest a major paradox. Wilson (1976) for the living coral Caryophyllia On the one hand, the high degree of substrate smithii. This species occurs on both hard rock selectivity by Pleurodictyum for Palaeozygo- substrates and on offshore muddy sands where it pleura and the consistent aperture-downward frequently encrusts the tubes of the polychaete orientations suggest that Palaeozygopleura shells Ditrupa. On sands, this coral also selects particu- were.occupied by a living organism. On the other lar types of gastropod shells; however, again, hand, unrepaired shell damage and encrustation they are never the shells of live snails, but those argue against occupation of these shells by living occupied by a specific species of sipunculid snails. Any theory of the relationship involved worm. Like Pleurodictvum., , this coral is semise- must explain these facts and also take into ac- lective; whereas it favors Ditrupa tubes or secon- count the semiselective nature of the Pleurodic- darily occupied gastropod shells, it may also oc- tyum encrustation. cur on dead bivalves and pebbles or cobbles. We sought an answer to this paradox in mod- Desoite its verv distant taxonomic relations hi^ ern coral analogs. Several living scleractinian with Pleurodictyum, we believe that Caryophyl- corals are known-to encrust shells of a lia provides an excellent functional analog for the oarticular soecies (Bouvier 1895: Schindewolf Devonian coral relationshin described herein. i959; ~oreau& ~onge1968; ~eisbord1971; In summary, our studies demonstrate that Wells 1972; Gill & Coates 1977). These include Pleurodictyum was a semiselective epifaunal or- both solitary and colonial abermatypic corals ganism. It could and did inhabit a wide variety of angaalas aql vodar osle (~~61)qlrea ;s laquea .roopzas dppnm aql anoqe [lam suo!l!sod ie naqs lsoq aqi .mo~sp!lnaund!s ,i!uuaq, aq& .sp!~ dolanap 03 saNolos aql pamolle s!q) 'u!eBe :(0861

-naund!s~ aql~~ jo saauegqns au!raowa 01 puodsar qsoiupm) sp!ouua aBre1 u!euaa jo suurnloa aeiue~leroa aql 'dlq!ssod aim6 .slerm dqpaisrua aql uo dlan~snpxalsourfe Zu~uaao'nualajard -ua 'urn1 u! 'are qs~qmsllaqs pa!dnaaoar asaql ajeqsqns re[lm!s e si!q!qxa 'sal!loyruv 'snua4 s! 11 .(adeqs r!acp jo asneaaq d[qeqo~d)sm~om pro3 ue!uonaa palefar e leql alqelou s! $1 '(8~61 p!1nannd1s $0 sayads 6q pal~qequ!i([an!)aalas are raqael!a~ :SL61 radeq~)(laaga 'Zraqaa!,) aseq sllaqs pus dldura u!eiraa ieql sreadde 'peaauI [ea!uoa e pue (13a33a ,aoqs mous,) eare Buueaq .spodorlseZ an![ morj ![nm!ls [eu!maqa 01 puods re^ 'paorq e apnlau! asau roo^ eas gos aqi -ar IOU op aeNe[ leroa asaql 'dpuap!n9 .anoqe uo lroddns roj suo!pidepe Lrepuoaas padolanap paiou xopered aq, 01 uo!lnlosar 'uanordun [[!is pue aierlsqns TeNel aql lano papuedxa duolos 3! 'alqeuosear e ap!nord sllaqs [!nus uo d1an!palas wnhqpo~naldaql 'qlmor8 [euuou jo asrnoa aliias qa!qm sleroa iuaaax uo e)ep 'd[leuy aql UI '(~SL-SSL:LL~Iuosyaer .p) aaepalu! rai 'aep!u!laqag.q aql $0 sraqluam -em-luan~pas aql anoqe pau!elu!em aiam qalqm Buoure uomuros arah sisoq an!l ql!m sd~qsuog salerlsqns apqour uo aeuel u!eiraa 30 luama[) -e[ar lesuammos leql1saZ8ns suo!leNasqo asaqL -las Zuunsse UI an!$depe sem ms!ueqaam an~iaal .uogeisnraua 30 amp aql ie Zuvg uaaq aneq -as s!q) leql aaZZns aM .an!uojl!wvy v.ina~doSLz 01 parraju! daql qa!qm 'sahIen!q Zu!~oq ue!uonaa -oaulvd jo s[[aqs aql paronej aaNe1 Zutluas 'iana jo sllaqs uo o!doj!suv.i~o!u~layn~ 30 aauarnuso -moq 'sloopas jsom uo 'sa%erlsqnsalqels aZrq

.LEx OOOSZ 3 SYV8 '9 3rrl .mnproJ ~ndj?p~~naldaql jo ase~lns~apunaqi jo ise3 saml e r! uam!aads lool lo^ wndj?!po~nalde 103 alellnqnr e Ee pahlar qqqm llaqr olnaldo2Lzoanlnd 01 paq3eize aqw nza3!H padsqs-n jo uo!llod lesea 'am s1 x '666~23 SNB 'ufiouyon i[l!te~o? ,saia?!Hjo saqnl snonu!s 'pal~q-sqappus pnm a~moqrelle~o~ undj3!pomald qaninorql suo!i~ar-sso13 -330 .OZr "66~~SNB 'PZ 307 (mane) raja=!H jo aqm padeqsn pangu!-luam!par BU!MO~Euola~ays wndj(r~ipo~na1d30 plom leu~a~u~80 .O.E x '966Z 3 SYVB'6 301 -(~fione)sZu!uado I~I~JI!~om, aroN -rara3!~80 aqm pa!j!qe$ ayl jo uo!llod e 8u~oqslunllelOJ ~"dj~tp~maldlo neps ,add" 10 ma!A 'VO 'elleIoJ uni0~!podnaldu! raqnl sa,a~!~-8 -S!J 260 Carlton E. Brett and John F. CottreN

Tnble 5. Co-accurrence of Hicetes wiUl Pleurodictym for dif- americanum is evidenced by two circular tuhe ferent substrates at four stratigraphic horizons, expressed as the apertures on the upper surface of the coral colo- percentage of Pleurodicryvrn specimens containing Hiceru. ny and/or by a 'U' shapeddepression at the site of the attachment scar. Hicetes must have settled Substrate Wanakah King Ferry Kanhong Windom on the substrate at nearly the same time as the Pnlneorygqpleurn 81.8 41.2 91.2 89.4 Pleurodictyum larvae to produce a scar near the Bivalves 58.1 50.0 50.0 84.6 corallum attachment site (Fie. 8). Previous au- Other molluscs 83.6 42.8 83.3 84.5 thors (Gerth 1952; ~chindewolfi959) suggested Corals 83.3 -' - 67.6 that the Hicetes worm occupied empty gastropod Brachiopods 43.2 - 54.5 77.5 shells prior to their encrustation by Pleurodic- Bryoroans 66.7 - - 83.3 Trilobites 71.4 - - 70.6 tyum, in analogy with a Recent coral-sipunculid Crinoids 66.7 - - 57.1 association (Heteropsammia-Aspidosiphon; Other organics - - - Bouvier 1895). ~iweve;,such an association can Inorganics - 33.3 - - Undetermined 50.9 26.3 75.0 74.4 be ruled out for Pleurodictvum which occur on NOsubstrate 88.9 - - 83.9 suhstrates such as phosphate pebbles yet, never- Total 74.4 31.9 77.8 79.2 theless, have Hicetes tubes. We suggest that the Number of Hicetes larvae were capable of detecting and set- specimens 510 71 70 905 tling on Pleurodiciyum 'spats'. That the 'worms' were also capable of settling on established colo- nies is evident from specimens in which tubes Feustel 1965). However, the shell is not moved penetrate only part way into the colony. downward into the substrate nor is it actively In any case, our material proves a very close growing. Hence, epibionts (including corals) association of Hicetes with ~feurodictyum~(~able have the advantage of a mobile substrate without 5). More than 75 % of all Pleurodictvum soeci- the disadvantages associated with live gastro- mens examined were found to possess a single pods. Thus, in contrast to the interpretation of centrally located Hicetes tube. In certain samples Clarke, we do not believe that the interaction as many as 90% of Pleurodictyum specimens between Pleurodiciyum and Palaeozygopleura were infested. This association does not appear involved settlement onto a live gastropod, but to be related to the type of suhstrate utilized by rather we suggest that Pleurodiciyum selectively Pleurodiciyum. slight differences were observed encrusted shells (of Palaeozygopleura and per- in the .orooortion . of Hicetes-infested soecimens haps other taxa, e.g., Coleolus) which were occu- occurring among Pleurodiciyum specimens on pied by secondary hosts, possibly sipunculid various substrates, but these are largely artificial. worms. For example, subsamples of Pleurodictyum at- Marek & Galle (1976) discuss the semiselec- tached to brachiopods and corals tend to exhibit

tive encrustation of hyolithids, gastropods and fewer apparent cases of Hicetes -partly ~ because nautiloids by the tabulate (alveolitid?) coral these calcitic substrates remain intact so that the

Hvostraeulum.u in the Lower Devonian of Bohe- basal 'scar' can not be examined for traces of the mia and Moravia. Because of their consistent Hicetes tuhe. Thus, we judge the Pleurodictyum- growth orientations these corals were judged to Hicetes association to he independent of any have settled on living and slightly active hosts; Pleurodiciyum-substrate association. the possibility of secondary occupancy of the shells was not considered. Acknowledgernenb. - We are particularly indebted to Harvey A. Ellis of West Falls, N. Y., for loan of an exc:ceptionally large sample of Pleurodicryum from the Windom Shale. Gordon C. Pleurodictyum-Hicetes association Baird, Carl K. Seyferl and Charles Klein also furnished collections of specimens from various localities. Field research, par- The association of Pleurodiciyum with a spiral tially funded by the Rochester Academy of Science, war car- tube was first noted by Clarke (1908, 1921) and later studied in detail by Gerth (1952) and Schindewolf (1959). Hicetes innrxus Clarke is viewed this paper and offered numerous su&estion; for its inferred to have been a commensal tubicolous improvement. Robert M. Eaton carefully prepared the photo- worn which formed an embedment structure graphs and Margrit Gardner typed various drafts of the manus- cript. To all, we express our sincere appreciation. Specimens (see Bromley 1970) within the corallum of illustrated in this paper are deposited in the collections of the Pleurodictyum. The occurrence of Hicetes in P. Buffalo Museum of Science (BMS). LETHAIA 15 (1962) Substrate of PLEURODIC~VUM261

References aspectsof settling behavior of Spirorbis. Mor & Biol Ajroc. U.K. I. 30, 201-222. Ager, D. V. 1963: Principles of Poleoecology. 371 pp. Knight-Jones, E. W. 1953: Decreased discrimination during McGraw-Hill, New York. settling after prolonged planktonic life in larvae of Spirorbis Baird.. G. C. 1979: Sedimentarv relationshios of Ponland Point borealis (Serpulidae). Mar Biol. Asroc. OK.1.32, 337-345. and associated Middle Devonian rocks: in central and west- Linsley, R. M. 1977: Some 'laws' of gastropod shell form. ern New York. N.Y. Sf. Mu. Bull. 433. 24 pp. Paleobiology 3(2J, 196206. Baird, G. C. 1981: Submarine erosion on agentle paleoslope: a Linrley, R. M. 1979: Gastropods of the Devonian. In House, study of two discontinuities in the New York Devonian. M. R., Scrutton, C. 7. & Bassett, M. G. (edr.): Devonian Letkoio 14, 105-122. Systcm, Special Papers in Paleontolo~y3, 249-254. Paleont. Baird, G. C. & Brett, C. E. 1983, in press: Re*onal variation Assoc., London. and paleontology of two coral beds in the Middle Devonian [McCoUum, L. B. 1980: Distribution of marine fossil assem- Hamilton G,roup of western New York. I. Poleonlo1 57. blages in a Middle Devonian stratified basin, lower Ludlow- Barthel, K. W. & Barth, W. 1972: Paleoecologic specimens viUe Formation, New York. Unpubl. Ph. D. Diss., SUNY from the Devonian of Bolivia. Neuer Ib. Geol Pol., Mo- Binghamton, Binghamton, N.Y.] natrh. 10, 57S581. Mclntosh, G. C. 1980: Dependent commensal association be- Bouvier, E.-L. 1895: Le commensalisme chez certainr polyps tween Devonian crinoids and two species of tabulate corals. madrCporaires. Ann. Sti. Not Zoo1 Polgont 20, 1-32. GeoL Soc. Am. Absnacis with Progrom 12f7). 481. Brett. C. E. 1974: Biontratigraphy and paleoecology of the MacNamara, K. J. 1978: Symbiosis between gastropods and Windom Shale Member, (Moscow Formation) in Erie Coun- bryozoans in the late Ordovician of Cumbria, England. Lsth- ty, New York. NY Sr Geol Asroc. Guidebook, 4th Ann. ,"in 11, 25-40. Mrg., G-I-G-10. Marek, L. & Galle, A. 1976: Thc tabulate coral Hyosfrogulum, Brett, C. E. 1978: Host~rpecificpit-forming epizoans on Siluri- an epizoan with bearing on hyolithid ecology and syrtemat- an crinoids. Letkaio 11, 217-232. ics. Le~koio9, 5144. Bromley, R. G. 1970: Barings as trace forrilr and Entobin Miiller, A. H. 1979: Fossilization (taphonomy), In Robiron, R. cretoceo Portlock, as an example. In Crimes, T. P. & Har- A. & Teichert, C. (eds.): Treolise on Invertebrate Poleontol- per, J.C. (eds.): Trace Fossilr, 49-90. See1 House Press, OD, A. Introduction, A1-A78 Geol. Soc. America and -- Univ. Press of Kansas, Boulder, Colorado, and Lawrence, Clarke, J. M. 1908: The beginnings of dependent life. N.Y. St. Kansas. Mur. Bull. 121, 14h149. Rollins, H. B., Eldredge, N. & Spiller, 1. 1W1: Gastropods Clarke, J. M. 1921: Organic dependence and disease: their and Monoplacophora of the Solsville (Middle Devonian, origin and significance. NY St Mw. Bull 221, 222, 1-113. Marcellus Formation) in the Chenango Valley, New York Conwav Morris. S. 1981: Parasites and the fossil record. Para- State. Am. Mus. Nor. Hist. Bull. 144, 131-170. sitology 82, 489-509. Ross, M. H. 1953: The Favoritidae of the Hamilton Group Cooper, G. A. 1930: Stratigraphy of the Hamilton Group of (Middle Devonian) of New York. Buffalo Soc Not. Sci. New York. Am. J. Sci. 5rh ser, 19, 11G134, 21&236. Buil 21(2), 37-89. Cooper, G. A. 1957: Paleoecology of the Middle Devonian of Seheltema, R. S. 1974: Biological interactions determining lar- eastern and central United States. In Ladd. H. S. (ed.): val settlement of marine inverfebratcs. Tholossia Ju~oslovico Treatise on Marine Ecology and Paleoecology. Geol. Soc. 10, 26S296. Am. Mem. 67, 249-278. Schindewolf, 0. H. 1959: Wllimer und Korallen a$ Synbken, Crisp, D. I. & Meadows, P. S. 1967: The chemical basis of zur Kenntnir der Systeme Aspidosipkon Heleroprommin, gregariousness in barnades. Proc. R. Soc. Lond, (61 156, und HicetesiPIeurodiciyum. Abh. Akad Wirr. Moinz, Mar. 5KL520. Not. K1. 1958, 6, 259-328. Feustel, H. 1965: Anatomische Unfersuchungen zum Problem Seilacher, A. 1978: Paleoecology of primary and secondary soft der Aspidosiphon-Heterocyuthw-Symbios. Ve'erhondl. bottom dwellen. Geol Soc. Am. Absnncrr whk Programs Deutrch. 2001. Gesell 29, 131-143. 10(7J, 489. Genh, H. 1952: Die von Sipunculiden bewohnten lebenden Teichert, C. 1945: Parasitic worms in Permian brachiopod and und jungtertiiiren Korallen und der wurmfbrmige Kbrper pelecypod shells in Western Australia. Am. J. Sci. 243, 197- von Pleurodiciyum. Poloont Zschr 25(3), 119-126. ?W. Gill, G. A. & Coates, A. G. 1977 Mobility, growth patterns Thayer, C. W. 1974: Substrate rpccificity of Devonian epiroa. and substrate in some fossil and Recent corals. Lethoia 10, 3. Pnkonfol. 48(5), 881-894. 119-134. Thayer, C. W. 1975: Morphologjc adaptations of benthic inver~ Goreau, T. F. & Yonge, C. M. 1968: Coral community on tebrates to soft substrata. I. Mar. Rcr. 33, 117-189. muddy sand. Narure217:5127, 421423. Ihonon, G. 1950: Reproduction and larval ccology of marine Grabau, A. W. 189M899: Geology and paleontology of bottom invertebrates. Wol. Rev. 25, M5. Eighteen-mile creek and the Lakeshore sections of Erie Vermeij, G. J. 1977: The Mesozoic marine revolution: evi- County, New York Pt. 1, Geology; PI. 2, Paleontology. dence from snails, predators and grazers. Poleobrology 3(3), Buffalo Soc. Nor. Sci. Bull. 6. 403 pp. 24-258. Gray, J. S. 1974 Animal sediment relationships. Oceonogr Weisbord, N. E. 1971: Corals from the Chipola and Jackson Mar Bid. Ann. 12, 223-261. Bluff Formations of Florida. Florida Bur. Geol. Bull. 53, 1- Hill, D. & Stumm, E. C. 1957: Tnbulara. In Moore, R. C. 105. (ed.): Treatise on Invertebrate Paleonfology. F Coelenteroto, Wells, J. W. 1972: Notes on the fauna of the Chipola Formation FWF477. Geol. Soc. America and Univ Kansas hess., - V Symbinnyia, a new rhizangiid coral. Tulam Slud. Geol. Boulder, Colorado, and Lawrence, Kansas. ~oleo"t.10(lj, s28. Jackson, I. B. C. 1977: Competition on marine hard suhtrata: Wilson, I. B. 1976: Attachment of the coral Caryophyliio the adaptive sigzificance of solitary and colonial strategies. smirhi; S. & B. to tubes of the polychaete Dihupo arietino Am. Not. 111 (9801, 74S767. (Miiller) and other substrates. J. Mar. Blol Assoc. U.K. Knight-Jones, E. W. 1951: Gregariousness and some other 262 Carlton E. Brett and John F. ComeN LETHAIA 15 (1982)

13A. Geneseo Quadrangle - Banks of Jaycox Creek in lower Appendix: Locality register part of Jaycor Run and Wheeler Gully 0.5 miles west of (AU quadrangles 1:24,WO) N.Y. Route 39 and 2 miles nonh northeast of Geneseo, Livingston Co., N.Y. (Wanakah Shale, Pieurodicryurn 1. Eden Quadrangle - Erie lakerhore bluffs 0.H.5 miles beds). south of Wanakah waterworks Wanakah, Erie Co., N.Y. 13B. Geneseo Quadrangle - Jaycox Creek in upper part of (Wanakah Shale, Pleurodicfyum beds). Jaycox Run, 0.3 miles east of lot. 13A (Jaycox Shale). 2. Buffalo Southeast Quadrangle - Shale pit immediately 14. Crn,mJlliur Irkr O~~Jr.~n~lc- (h.,lc hmk. .d]clrr.llll north of unnamed wert-flowing creek, and 0.3 miles trim hnph vlllr III \Ic~.~rlllClt~lh b uz>t of \\c~IL.>kr southwest of WKBW radio towers along Big Tree Road, 11034 e,I 0, ..II <1,,t,r,o (i . Blasdell, Erie Co., N.Y. (Wanakah Shale, Pleurodic- N.Y. (Kashong Shale). ryum beds). 15. Rushville Quadrangle - Banks of small unnamed west- 3. Buffalo Southeast Quadrangle - Large abandoned shale flowing creek 0.3 miles north of US. Routc 20, 3 miles pit of Penn Dlnie Cement Co., 0.3 miles northwest of southwest of Hopewell and2 miles west of Aloquin, intersection of Big Tree and Bay View roads, Blasdcll, Ontario Co.. N.Y. (Wanakah Shale, Pleurodicryum Eric Co., N.Y. (Lower Windom Shale, 'coral beq. bedr). 4. Buffalo Southcast Quadrangle - Banks of small north- 16. Canandaigua Lake Quadrangle - Banks of Deep Run flowing tributary of south branch of Smoke Creek, 0.3 Gully 1.2 miles south of Kipp Road east side of Canan- miles north of Mile Strip Road and 0.5 miles west of daigua Lake, Gorham, Ontario Co., N.Y. (Wanakah Abbott Road, Orchard Park, Erie Co., N.Y. (Wanakah Shale, Pleurodicryurn beds). Shale, Pleurodicryurn bedr). 17. Canandaigua Lake Quadrangle - Banks of Gage Gully 0.30.7 miles southeast of East Lake Road, east side of Canandaigua Lake, Gorham, Ontario Co., N.Y. (Ka- shaog Shale).

5B Orchard Park Quadrangle - Banks of Caenovia Creek 18A. Geneva South Quadrangle - Shale banks above upper north and south of Nonhmp Road, Spring Brook, Erie falls in Kashong Creek 4.2 miles west of N.Y. Route 14, Co., N.Y. (Windom Shale, 'coral bed'). south of Geneva, Ontarlo Co., N.Y. (Kashong Shale). 6. Orchard Park Quadrangle - Bed of north-flowing tribu- 188. Geneva South Quadrangle Lower (high) falls and adja- tary of Buffalo Creek, immediately north of Bullis Road cent shale banks along Karhong Creek about 0.3 miles and 0.7 miles east of intersection with Pound Road, east of loc. 18A. (Jaycon Shale). Spring Brook Station, Erie Co.,N.Y. (Wanakah Shale, Pleurodioyum beds). 18C Geneva South Quadrangle - Shale in creek bed and banks of Kashong Creek about 0.3 miles west of N.Y. 7,\ I::<* Aurora i)u;Jrlnple - Inu lank\ .>I Htmllil. Cw~k Route 14 and 0.6 miles east of loc 188. (Wanakah hcluce~Rnul,n inJ (imrdlc Rol.1 Llmr k.ilr ( I . U Y Shale, Pleurodicryurn beds). (W~n.tL,>hSl~.slc., 1'1et~r~d~~~;un~I><,I>!. 19. Romulus Quadrangle - Small roadside quarry along 7B. East Aurora Quadrangle - Banks of Buffalo Creek 0.1- N.Y. Route 336, 3.5 miles east of MacDougall, Seneca 0.2 miles south of Bulk Road old bridge, Elm, Erie Co.,N.Y. (Wanakah Shale, Pleurodicryum beds). Co.,N.Y. (Windom Shale, 'coral bed'). 20. Dresden Quadrangle- Small manmmde atalong lower 8A. Alexander Quadrangle - Shale bank along west branch reaches of stream that marks nonh border of Sampson of Murder Creek, 0.6 miles south of Shanack Road and Naval Facility, east side Seneca Lake, Seneca Co., N.Y. 0.6 miles nonh of U.S. 20, Darien, Generee Co., N.Y. (Wanakah Shale, Pleurodicryum beds). (Wanakah Shale, Pleurodicryum beds). 21. Dresden ,Quadrangle -Shale banks along Indian Creek 8B. Alexander Quadrangle - Bed of west branch of Murder between N.Y. Route 96A and the east shore of Seneca Creek, 04miles south of US. 20, Darien, Genesee Co., Lake near Willard, Ontario Co.,N.Y. (Kashong Shale). N.Y. (Kashong Shale). 22. Ovid Quadrangle - Shale banks along Big Hollow Creek, 0.3 miles north of Swick Road and about 2 miles north of Hayt Corr~en,Seneca Co., N.Y. (King Ferry Shale; upper hiatus concretion zone). shong Shale). 23A. Ovid Quadrangle - Shale bank along Mack Creek 2 10. Batavia South Quadrangle - Raulroad at along Eri- miles east of Hayt Cornerr and 0.5 miles west of N.Y. Lackawanna Railroad tracks, just east of Francis Road Route 89, Seneca Co., N.Y. (King Ferry Shale; upper overpass and 1.7 miles north of U.S. 20, Bethany, Gene- hiatus concretion zone). see Co., N.Y. (Wanakah Shale, Pleurodictyum bedr). Z3B. Ovid Quadrangle - Shale banks along Bloomer Creek 11. Leicester Quadrangle - Shale dumps and milroad cut at 0.5 miles south of loc. 23A (King Ferry Shale; upper international Salt Company mine, 0.35 miles south of hiatus concretion zone). N.Y. Route 63, Retrof, Livingston Co., N.Y. (Kashong Shale). 24. Gilboa Quadrangle - Roadcut on unnamed road off Route 30, 3 miles southeast of North Blenheim, Scho- 12. Leicester Quadrangle - Shale banks along Bidwell Creek harie Co., N.Y. ('Portland Point' equivalent). 0.3 miles northwest of Retrof Road, Retsof, Livingston Co.,N.Y. (Wsnakah Shale, Pleurodictylrm beds).