15643

8/12/2019 15643

1/12

A m od ern analogue for the trace fossilGyrolithes:burrows of thethalassinidean sh rim pAxianassa australisPETER C. DWO RSCHAK AND SERGIO DEA. RODRIGUES

LETHAIADwors chak, P.C. Rodrigue s,S. de A. 1997 04 15: A modern analogue for the trace fossilGyrolithes:burrows of the thalassinidean shrimpAxianassa australis. Lethaia,Vol. 30, pp. 41-52. Oslo. ISSN 0024-1 164.

The tidal flats at Praia do AraGa, Brazil have m uddy siliciclastic sedim ents on the surface and alayer of heavily packed shells down to 30-40 cm depth. The most obvious elementof theinfauna is the thalassinidean sh rimpAxinnassa australis.Several animals were captured with ayabby pum p. Burrow open ings were characterized by a low mound (1-2 cm high and 10-30 cmin diamete r at the base) with one or two simple holes nearby (20-70 cm away). Counts alongtwo transects showed a m ean density ofAxianassa burrow openingsof 1.4 m-2 (range: 0-7),mounds ranged in density from0 to 3 m- (mean 1.25). Three nearly com plete (a nd severalincom plete) resin casts showed a unique b urrow shape, with spiral shafts leading to wide hori-zontal galleries from which several evenly proportioned corkscrew-shaped spirals branchedoff,

leading to further horiz ontal galleries at greater sediment depths. Burrows had up to15 suchspirals and a total length of over8 m. The total bur row dep th was between 106 and 130 cm. Therole of the spirals and th e similarity ofAxianassaburrow s to the trace fossilGyrolithesare dis-cussed. halassinidea, AXIANASSApiral burrows, tracefossils GYROLITHES,razil.

Peter C. Dworschak [[email protected]], 3. Zoologische Abteilung,Naturhistorisches Museitni, Burgring7, A-1014 Wie n, Austria, arid Instirutfur Zoalogie derUniversitiit Wie n, Althanstrajle 14, A-10 90 Wie n, Austria; Se'rgio deA. Rodrigues, Departamento deEcologia Geral, Instituto de BiociPncias, and Centro de Biologia Marinha, Universidade de SdoPaulo, C.Postal 11461, BR-05422-970SdoPaulo, SdoPaulo,Brazil; 5th March, 1996; revised26thSepteni ber, 1996.

Crustaceans are among the most imp ortant burrowers inintertidal and shallow subtidal sediments.A number oftrace fossils have been attributed to crustaceans, mainlystoma topods, crabs, lobsters and m ud-s hrim ps (e.g., Freyet al. 1984; Pem berton et al. 1984; Bromley 1990). Ofthese (seven ichnogene ra), three are of particular interest:OphiomorphaLundgren, characterized by burrows w ith aknobby o uter wall lining;ThalassinoidesEhrenberg, char-acterized by burrows with a three-dimensional networkof sm ooth-walled cylindrical comp onents of variable di-

ameter with Y- to T-shaped branches; and GyrolithesSa-porta characterized by burrows w ith a regular spiral form.Gyrolitheshas a world-wide distribution, ranging frommarine deposits of Jurassic through Miocene (G ernant1972; Bromley Frey 1974; Hantzsch el 1975; Chris-tiansen Cur ran 1995). A recent analogue ofOphiomor-pha can be found in the burrows of the thalassinideanshrimp Callichirus major,which is characteristic of beach-es along the Atlantic coast of the Americas. Burrows ofThalassinoidesform are made in modern environmentsby a number of marine organisms, including cerianthidanemones, enteropneusts, and fish, but most im portantly

by decapod crustaceans, primarily thalassinidean shrimps(Myrow 1995). Modern spiral burrows similar toGyro-litheshave only been reported for capitellid polychaetes(Powell 1977). Hantzschel(19 34,1935) argued thatGyro-litheshad most probably been m ade by polychaetes, butthe majority of palaeontologists have interpreted thesespirals to be the bu rrows of decapods (e.g., Kilpper 1962;Ger nant 1972; Bromley Frey 1974; May oral 1986;Christiansen Curran 1995).

Among the burrow ing decapods, the Thalassinidea are

of greatest impo rtanc e; this group lives almost exclusivelyin burrows. Many intertidal and shallow-subtidal sedi-ment substrates are inhabited by large numbers of mem -bers of the species-rich families Upogebiidae and Callia-nassidae. The genu sAxianassahas been placed alternatelyin the fam ilies Axianassidae (established by Schm itt 1924)and Laomediidae (Rodrigues Shimizu 1992) and, untilrecently, was thought to be quite rare. Up to 1990, onlyfive specimens representing two species had beenrecorded. The num ber of specimens was increased to ninewith the description of three more species by KensleyHeard (1990). The description ofAxianassa australisfrom

8/12/2019 15643

2/12

42 Peter C. Dworschak andSergiode A . Rodrigues LETHAIA 30 (199 7)

Brazil by Rodrigues Shimizu (1992 ) was based on sixspecimens and increased the number of species to six.

The present work reports on the results ofa study of theburrowing crustaceans of intertidal and shallow subtidalhabitats around Sao Sebastiao, Sao Paulo, Brazil, whichrevealed a relatively large population ofAxiannssa oirstra-

lisat the tidal flat at AraGa. Yabbv -pum p sam pling yieldeda total of nine specimens. Resin castingof the burrowsresulted in several incomplete and three nearly completeburrow replicas, which all showed a unique shape closelyresembling Gyrolitlzes.This paper describes these unusua lburrow s and considers the possible m ethodsof construc-tion a nd the possible function s of the spirals.

Study sitePraia do Arap is situated south of the town of Sao

Sebastiao, Sao Paulo, Brazil (Fig. 1) . Here there is a bayabout 600 m wide and 350 m deep, measured from theline between Pon ta d o AraCa in the south and th e cornerof the pier in the no rth. There are two rocky outcrop s inthe middle of thebay. From the m ore southern of the tworocks, a bank m easuring 100 m long and ca. 80 cm abovelow water level and consisting mainlyof mollusc shellsstretches to the inner sideof the bay. The in ner edgesofthe bay and the south end of the bank support a sparsegrowth of mangrove trees. The present investigationswere carried out in a n area nort h of the bank , which wasexposed duringlow tide (Fig. 2 ) . The tides are semidiur-nal, and the mea n tidal range is 1.5 m .

The sediment at the surface issoft, whereas th e deepersediment layers show compacted shell accumulations.The sediment depth in which the shell accumulationsbegan ranged between 10 cm near the shell bank a nd30cni in the middle of the tidal flat. At the siteof the resincasts and the sediment core the acc umula tion started at 17cm depth.

The surface sediment (0-11 cm deep in the core) con-sists of silty, very tine san d with 17.7% silt an d 8.7% clay,an organic content (wet oxidation) of7.9 , and a car-bonate contentof 9.8Oh. At 22-33 cm depth , the sedime ntconsists of 12.2% particles larger than4 mm , mainly ofmollusc shells and shell fragmen ts (Fig.3). The remainingsediment is a medium sand with only5.5% silt, an organi ccontent of 6.6% and a c arbonate conten t of 21.6%.

Most of the shells were of bivalves, mainly th e veneridAnomalocardin brasiliana(Gme lin) and the lucinidPha-codespectinatirs (Gmelin) .A few shells were fro m gas tro-pods, mainly Cerztlziirrn tzus carttmSay and Biilln ?striataBrugiere.

Other importan t infaunal organisms living in this tidalflat were the stomatopod Lysiosquilla scabricaudo(Lam arck), he polychaetesDiopatrn cf. cupren (Bosc) andClzaetopterus variopedatus (Renier), an unidentified

Tropic of Capricorn

Sao Paulo ~

0

llha de

Atlantic Ocean

100 m

Canal deS l o SebastiBc

Fig. 1. Map showing the study si te.LWL: low-water line on April 28th.1994;dotte d areas: mangroves;A-A' a n d B-B': transects. Modified afterRodrigues (196 6) and Rodrigues Rocha(1993)

Fig. 2. View no rth on Praia do AraGa.

nereid and a terebellom orph, the bivalvesAnomalocardiabrasiliana (Gmelin), Tagelits plebeitis (Lightfoot) andPhacoides pectinatus (Gmelin) (= Lucinn jamaicensis),asipunculid (?SipunculusnudusL.), an echiurid, a new spe-cies of nemertean (Wolfgang Senz, personal comm unica-tion ), he tanaidaceanK a h p s e u d e ssp., and an as yet uni-dentified speciesof alpheid s hrim p (which probably livesin the burrows of Axiannssa;see Felder et al. 1995). Near

8/12/2019 15643

3/12

LETHAIA 30 (1997) A modern analoguefor GYROLITHES3

60

50

40

g 30.

20

10

0 Surface

22-33 cm depth

-2 1 1 2 4 5 6 7 8

Grain size (phi)

Fig. 3 . Grain-size distribution of sedim ents from the tidal flat at Praia doAraga as percent w eight retainedon sieve. White columns: sedime nt atthe surface (0-11 cm); black columns: sediment at a depth between22and 33 cm.

the rocks, a specimen of the thalassinidean shrim pVpoge-bia omissaGomes Correa was captured.

Material and m ethods

Axianassa australis (Fig. 4A) was described recently byRodrigues Shimizu (1992). It was collected from man -grove mud in Valenqa, Bahia, Brazil, and was first cap-tured at the tidal flat at AraGa in 1985. Anim al density wasestimated by counting holes and m ounds in a 1 m2 framealong two transects, one across the bay, the other fro m themangrove trees to the low water line nearly at right anglesto the first (Fig. 1).

Animals were captured with a yabby pump similar tothat described by Rodrigues(1966). Preserved specimenswere deposited in the N aturhistorisches Museum, Wien(NHMW) and in the Mustum National dHistoireNaturelle, Paris (MNHM). Sediment samples wereobtained using the yabby pum p as a corer and were ana-lysed using the methods outlined by Buchanan (1984).

Resin casting was don ein situ using an epoxy resin (seePervesler Dworschak 1985). Initial attempts to castmou nds resulted in only short casts of vertical shafts (5-20 cm deep).In subsequent attem pts, suction was exertedat the m ounds with a small yabby pum p and resin pouredinto those holes by visibly responding to the suction. Aslow-setting mixture of the resin (Araldite GY257:HY830:HY850 in pro portio ns by weight of 20:9:3, whichneeds 4-6 days hardening time, instead of 25:7:8 with ahardening time of 2 days) yielded the most completecasts. For reference, three casts were deposited in theNaturhistorisches Museum, Wien (NHMW).

The termino logy used for description of the spirals fol-lows that of Gernant (1972) and Bromley Frey (1974).

Results

Appearanceof openings and density

At the se diment surface, burrow openings ofA. australisare characterized by low mou nds, 0.5-1.5 cm high an d 3-27 cm in diameter at their base (Fig. 4B, D). Thesemo unds are grey to black in colour, in contrast to the sur-rounding brown, ripple-marked sediment surface. Insome cases, faecal pellets were fou nd am ong the sed imentforming the mo und (Fig.4C).When the burrow openingswere submerged, suspended sediment blown out of theopening by the pum ping activity of the shrimp could beseen. This pump ing activity was also observed during lowtide, when the receding water forme d typical rill marksonthe mounds (Fig. 4B, D) and murky water filled thetroughs of the ripple marks (Fig. 4C,D). Most of themo unds had a small opening on top (0.5-1.4 cm in diam-

eter) while some had the shape of a crater (2-3 cm indiameter). Near to the m ound , at a distance between 10and 49 cm, a se cond ope ning was visible, which was eithera simple hole (Fig. 4B) or a funnel 2-3 cm deep, me asur-ing 1.2-3.6 cm in diameter a t the surface and narro wingto 0.5-0.9 cm (Fig. 4C) . These open ings became visiblewhen suction with the yabby pump was exerted at themound opening.

The density of burrow openings attributed toA. austra-lis ranged between 0 and 8 m-2 (mean s 1.35 and 1.68 forthe first an d second transect, respectively). Moun dsoccurred a t a density of between0 and 5 m-2 (me ans 1.22

and 1.27), and simple holes at a density of between0 and4 rn- (m eans 1.17 and 1.5).

In m ost cases, only one specimen was captured whenapplying suction to a mound and then to the correspond-ing second opening. In one case a male-female pair wasobtained. Females outnumbered males, and all femalesexcept one were ovigerous. The measurements of all ninespecimens ofA. australis are summarized in Table 1.

Burrowshape and dimensions

Burrows of A. australis follow a general pattern (Figs.5 ,6). From both the hole and the m ound site (marked withh an d m in Figs. 5A-D an d 6A-B), a vertical shaf t (7-12m m in diameter, circular in cross section) leads to a depthbetween 13 and 30 cm (Table2). At this depth, it eitherleads after a short w horl (Fig. 5A, C) into a w ide tunnel orcontinues as a wide spiral shaft (Fig. 6A, B), sometimeswith a series of chambers at dep ths between 30 and 60 cm ,which opens in to a wide tunnel. This wide tunnel occursat depths between 13 and 60 cm from which one or tworegular, vertical spirals branch off. These spirals leadeither to other (less wide) tunnels m ade a t greater sedi-ment depth, o r up to the surface (Fig. 5C, F). Several other

8/12/2019 15643

4/12

LETHAIA 30 (1997)

Fig. -1. 14 pecimen of 4 r i i ~ r 1 n s mfemale, X H l I \ V 8452 1 111 side view; scale is 1 cm. Photo A. Schumacher.OB-D. Burrow openingsof A. australis;f r a m e k 1 0 x 1 0 i m .

94(J41594042 i940425910425

940425

94042;910426

940126940376~

T d d c I . ,2ledaurementh1 i n n i m ot A . Y I ~ Z U ~ I S L Ii i rs t rd i j specimens collected at Praiado . \ r a p with a vabbypump; \ex, f: female. fo: wigerous female, ni: male; TL: otal length; CL: arapace length;CW: carapacewidth: 51: bide of maior ihel iped. r: right, 1: left; PLma: length of nidior cheliped;P\Vma: width of majoriheliped;PLmi: length of min or cheliped; PL\'mi: widthof minor cheliped; hluseum in which specimenshdvr been deposited f XH\l\ \ ' : Ndturhistorisshesl l u s e u m in \\'ien; XISHX: h luseuniNation al d'HistoireSaturelle P a r i \ ~ .

l h t e Sex T L C L C\ \ i PLnia P\Vma PLmi P\\'mi Xluseuni~~~~ ~~ ~~~~~~~~ ~~~ ~~~~~~~ ~~~ ~ ~ ~ ~ ~~~~

__ _ _2 5 N H h l W 84523 0 l l N H N - T h 1292

N H h W 84702 5 SHhlW 8455

2 5 XINHN-Th 1292

2 NHhlI2 ' 8469NHhlLV 8453

2 9 S H h l M ' 8 4 5 6

7- i NHhliZ.' 8454

spirals are linked by another seriesof tunnels. In the caseof cast 940430 ( Fig. s E ) , w o parallel spirals bran choff thetunnel at 13 cni depth, and another pairof parallel spiralswas present in greater sedime nt depth s (Fig.SF) . Both leftand right tur ning spirals can be present ina single burrow(see Table 2 ) .

The tunnels are either straight or curved, and occurwith or w ithout side branches. In cross-sections the t un -nels are circular or elliptical whe n the roof is mor e archedthan th e bottom or triangular to pear-shaped (see cross-sections in Figs.5A, C and 6 A ) . The length of these tun-nels ranges from 10 to 40 cm. The irregular shapeof the

8/12/2019 15643

5/12

LETHAIA 30 (1997) A niodern analogueforGYROLITHES5

Fig.5. Burrows of A. mistralis.OA. Cast 940427/1(NHMW 13723) in side view. OB. Same cast viewedfrom above. UC. Cast 940430 ( NHMW 13724)in side view. OD. Same cast viewed obliquelyfrom above.EIE. Detail o f C showing upper parallel spirals, the left one turn ing left down , the right oneturning right down. OF. Detail of C showing lower parallel spirals, both turn ingdown to the left; note rougherroofscompared to the upper spirals. m:mound site;h: hole ( funn el) site. Arrow shows position where an a nima l was entom bed. Vertical scale(For resin casts) is 10 cm; horizontalscale (forcross-sections in A andC ) is 2 cm.

tunnels can be seen particularly when the casts are viewedfrom above (Figs.5B, D, 6B). The dip angle of the tunnelsranges between0' and 30'.

The burrows in the regular 'corkscrew spirals' have anelliptical cross-section,and the roof is more arched thanthe bottom. The outsideof these burrows is more acute

than the rounded inner side and sometimes forms a keelrunning along the outsideof the spirals (seeFig. 5E).Thedip angle of the burrow within these spirals rangesfrom20 to 40 .

No complete animals were entombed by the resin. In940427/1 two chelipeds (PLma= 3.3 mm; P h i = 2.5

8/12/2019 15643

6/12

46 Peter C. Dworschakand Sergio de A. Rodriglies LETHAIA 30 (1997)

l i ible2. Measurementsof Aximassn nustrnlisburrow casts.1 : weight; TL: total length;Tlst: depth of first ( wid e) spiral;TU: depth of U, TG: total depth;\: burrow volume icalculated from de nsityof resin 1; left: numb er of spirals turning d own to the left; right: num berof spirals turning do wn to the right;A: distance of openings; HE: horizontal extension; BH: burrow height;BIY: burrow width; SH: spiral height; CD : outer coil diameter;L length of thespiral, i.e. the vertical extensionof on e entire spiral between separate tunnels and side branches.. .. -

9.10427/1 9 4 0 4 2 i i 2 94042713 94042714 9404 28 940430 940501_______ _ _ _ . .~ ~ _

w igl 186 177 150 92 179 1605 1572TL Ism ] 182 119 148 94 180 1020 809Tls t [mi] I3 I3 18 32 28 12 36

TG [cm] 34 62 62 67 79 106 130TL icm] 13 33 81

A jcmj 31 40 20HE [ i m ] 40 44 37 30 44 70 80\ jcmi] 135.8 129.2 109.3 67.2 130.7 1171.5 1147.5left 1 9 2right 1 1 2 6 5remarks ihelde incomplete incomplete incomplete incomplete chela

10-1813-37

11-27 10-2914-30 19-36

SpIrlikHH [ ~ n n i ] 6-1 1.3 11 8-14 10-15HIT [ m m ] b 11.5 14 10-18 15-23

CII) , n i n ] 12-27 2.1 18-24 28-37 38-60L ~ c m ] h-10 7 8-25 7-30

SH [ m m ] 13-27 28 21-27 17-40 16-35

~~~ ~~~ ~ ~ ~~

nim) and part of the cephalothorax were caught in theresin (see arrows in Fig. 5A, B). In cast 940430/1 onlyone cheliped (pre suma bly the m ajor; PLma= 7 m m )was entombed (see arrows in Fig. 5C, D). From themeasurements of the chelipeds the total lengthof theinhabitants was calculated to be between 21 and 23 m min the small cast (940427 /1) and ca.50 m m in the largecast ( 94043011 .

The dimensions of the three complete burrows andfour incomplete burrows are summarized in Table 2.hleasurements of the partly entombed animals indicatethat the burr ow dim ensions are related t o animal size. Themean burrow width of the spirals shows a more regularcross-section than th e tunnels. This param eter was there-fore chosen as the independent variable in calculatingregressions of the me an spiral radius, mean spiral height,and mean coil diameter in relation to the mean burrowwidth (Fig.7 ) .

The inner coil diameter ranged from 0 to 10 mm.Because of the com plex shape an d varying diam eter of theburrows, no exact calculationsof the bu rrow surface havebeen made. Very rough estimates indicate a surfaceofmore than 0.5 m2 for the largest burrow.

Properties of the burrow wall

The burrows are unlined, an d the wallsof the tunnels andspirals are irregular a nd rou gh. In several places im prin tsof shells and shell fragm ents are visible, and in som e cases

entire shells have been e mbedd ed i n the resin (Fig.6C, D) .The floors of the tunne ls an d spirals are smooth . The gra-dation from tunn els to spirals often has paired ridges orfurrows (round ed,0.5 m m high/deep,5 m m wide) on thefloor runni ng in th e direction of the burrow segm ent overa distanceof 10 cm (Fig. 6D ).

During excavation of the casts it was observed that th eburro w wall of the spirals was a light brown colour (ind i-cating that it was oxidized), comp ared to th e surround ingblack t o grey colour characteristicof reduced sediment.

Burrowing behaviour

In a laboratory aquarium,A . australism ade only feeble at-tempts to construct a new burrow. O ver a periodof fourdays the specimen created only a slight depression at thesediment surface.

Discussion

Recent and fossil spiral burrows

This is the first reportof a symm etrical corkscrew spiralin an extant speciesof burrowing shrimp. All burrowsofthalassinideans can at least occasionally have burrowcomponents with a more-or-less spiral shape (e.g.,Dw orsch ak 1983; Bromley 1990). A spiral is even presentin the Y-shaped b urrows o f the suspension-feeding Up o-

8/12/2019 15643

7/12

LETHAIA30 (1997) A modern analogue or GYR OLIT HES7

Fig. 6. Resin cast 950501 (NH MW 13725) of anA. australis burrow. OA. Side view with sup erimpo sed outlines of cross-sections. OB. Viewed from above.OC. Detail of spiral showing embedded ga stropod shell; note keel on sides of the tunne ls (ar row) .OD. Detail of tunnel with furrows on the floor (arrow

up) and packed bivalve shells within spiral whorl (arro w down ). m: burrow o pening showing mo und; h: bu rrow op ening showing hole; burrow belowupper shafts incompletely filled an d partly reconstru cted with wire. Vertical scale (for resin casts) is 10 cm; horizonta l scale (fo r cross-sections inA) is2 cm.

gebiidae (Dworschak 1983). However, in such cases thespiral is generally wide with bran ching blind tu nnels (e.g.,in Glypturus acanthochirus;Dworschak Ott 1993) andoften has of a series of chambers linked by a sha ft. This lat-ter pattern is seen in the burrow sof the callianassidCalli-anassa bouvieri(Dworschak Pervesler 1988) and in thelaomediidJaxea nocturna(Pervesler Dworschak 1985).

The spiral burrows of Axianassa are very similar inshape and dimensions to the trace fossilsGyrolithes(Fig.8). The shape an d dimensions of these trace fossils, how-ever, vary considerably. The burrow cross-section hasbeen described as either circular (Kilpper 1962; Gernant1972; Christiansen Curran 1995) or elliptical (Man s-field 1927, 1930; May oral 1986; Follm i Grim m 1990;Grim m Follmi 1994). An elliptical cross-sec tion hasbeen interpreted to be du e to early compaction (Stenzeletal. 1957). How ever, the e lliptical cross-sec tionof the bur-rows of Axianassa is most probably due to the naturalburrowing actions of the animals and not to non-bio-logical forces such as com paction.

Fig. 7 shows three dimensions of selected fossil andrecent spiral burrows plotted in relation to the burrowwidth of the spirals. All burrow widths in the spirals ofAxianassa burrows from Brazil fall within the rangeofthose reported for Gyrolithes.A spiral cast taken in Belize(probably of an Axianassaspecies) was found to have amuch wider burrow. Spiral height was found to be a linearfunction of burrow width inAxianassaburrows (line b inFig. 7). The sp iral height of fossil burr ows is smaller thanin Axianassaburrows of comparable burrow w idth in thecase of sm all specimens, but larger in spiral burrows ofgreater burro w w idth (line c' in Fig. 7). The coil diameterof spirals in Axianassaburrows was not fo und to be line-arly related to burrow width, but showed logarithmicrelation with coil diameter in spirals showing a smallerincrease with increasing burrow width (line c in Fig. 7).Nearly all fossil spiral burrows have a larger coil diame terthan the spirals ofAxianassaburrows of comparable bur-row width. This is mainly due to the fact that som e of theGyrolitheshave a much wider inner coil diameter- e.g.,

8/12/2019 15643

8/12

48 Peter C. Dworscknkatid Sergiode A . Rodriglies LETHAIA 30 (1997)

6 0

5 0

EE 4 0

n

cn3 0

re2 0

1 0

0

0i

C

70

a

3 1 7

0 1 0 1 5 2 0 2 5 3 0 3 5

B W mm)F I X .7. Dimensions of ni dr rn and fossi l spi ra l bur ro w.A s i m i o u a bu rro ws filled symbols, solid lines;Notornastirs lobatus:crossed symbols;Gyrolithes:open symbols. dashed line.\. hLean burrow width (BM ');mean spiral radius(SR, triangles,a: SR = 0.3847BW + 3.6053, 6 = 0.6133); mean spiral height( SH, uadrats, b: SH = 0.7373BJV + 13.804,F = 0.9614; b': SH = 29.675 InBM'- 53.74, = 0.706); mean coil diameter (CD , circles, c: CD= 29.85 lnBW

5 I . j i 1 , ? = 0.8193; c : CD = 14.289 InB\\' + 4.4614, F = 0.2723). Sources:1 4 : his study: 1: cast 9405 01; 2: cast 940430;3: cast 940427; 4: cast 940428.. burrow cast, probablyof Asir~riczssarom Carrie Bow Cay.Belize i Dworschak, unpublished ).6: Notoniastuslobatus, intertidal, North Carolina (Powell1974, Fig. 2 ) .7: Gyrditht-s dnvre ~rzi,'pyer Cretaceoub. Belgium(Bromley ; Frey 1974, Fig. 1).8: Gyrolithes,Oligocene, Germany (vo n Ammon 1900cxGernant 1972,Table 1 I. 9: G1mlitke.ir i i c ~ r ~ l [ 7 r i ~ f i ~ - i ~ . ii iocme, St. X i a r y b Formation , Maryland (Mansfield 1927, PI. 3:2).10:Gyrolithesclarki,Miocene,hlonterey Group, Californiai Xlanstield 1930, PI. 2: l) .11: G)ndithrs rwximia Cretaceous, Mexico (Mansfield 1930, PI. 1:3]. 12:Gyrolithes,Eocene,Stone City Beds. Texah (Stenzelf t a/. 1957 ex Gernant 1972, Table 1 1 13: Gyrolithrscf. r?iarylafidictrs,Miocene, Braunkohlenformation, Germany (Kilp-per 1962). 14: G i w l i t h t yRliocene , Borneo (Keii 1965, Fig. 29/31, 15, 16:Gyrolithesn~orylnridicirsGernant 1972, Table 1 andPI. 1:l) . 17: Gyrolithes,hlio irne. Brazil bernan deseC Assis 1980 . 18: G).rolit/irsd d i ,Pliocene, Spain (M ayoral 1986, Table1 ) . 19: Gyrolithes,Oligocene-Miocene, San Gre-gory Formation.Haia California Follrni8; Gri mni 1990, Fig.2B 1 20: G)vdi thcsfriarylmidictrs,Miocene, St. Mary's Form ation, Maryland (ChristiansenK Curran 19Y5 dnd Allen Curran . personal communication, 1995 2 1: C; .rolit/irs, Oligocene-Miocene, San Gregorio Formation, Baja CaliforniaI G r i m m 8; Follnii 1994. Figs.hB and 10) .Dimensions given as mean valueof ranges men tioned in text or measured from figures in original publications.

up to 3 cm in Gyrolitkes dnvreuxi(Bromley Frey 1974)- han the spirals of Axinnnssaburrows, where the innercoil diameter is zero in m ost cases.

Recent spiral burrows of similar size have also beenreported by Powell (1977)for a polychaete (Notoinnstussp.) from tidal flats in North Carolina.A much smallerspiral burrow ( burro w diameter3 m m, coil diameter 1.6c m ) oriented obliquely or horizontally was observedforanother capitellid polychaete, Heteromastits Iatericus,from the No rth Sea near Helgoland (Hertw eck Reineck1966; Reineck rt 11.1967). Members of the polychaetefamily Paraonidae(Pnroonisfirlgens, Aricidea fragilis)alsomake spiralled, corkscrew-like structures in the deeperparts of their burrows (Roder 1971; Ruppert Fox 1988).In Pnrnonis frrlgeiisthese spirals lie in a horizontal plane(Rod er 1971). According to Ruppert Fox (1988), the

enteropneust Saccoglossus kowalevskiifrom tidal flats inNor th C arolina constructs helical U-shaped burrow s.

Spiral burrows have been observed in a numberofbrachyuran crabs(for a review, see Van nini, 1980). Som eadult male ocypodid crabs, e.g., Ocypode saratan (Al-Kholy 1958; Magnus 1961; Linsenmair 1965, 1967) and0 cerutophtlialma (Farrow 19711, dig spiral burrowsassociated with sand pyra mids ('reproductive burrows').Other ocypodid crabs, e.g., Mucropkthalmus definitus(Verwey 9301,Macropktkalmussp. (Farrow 1971),Ucaannulipes and Macrophthalmus pawi manus(Braithwaite

Talbot 1972), and Uca beebei (Christy Schober 1994),also dig spiral burrows. These crab burrows, however,have an incomplete single, wide whorl ( up to60 cm coildiameter in Ocypode)with a circular cross-section, and amuch larger burrow diameter (u pto 10 cm) when com-

8/12/2019 15643

9/12

LETHAIA 30 (1997) A modern analoguefor GYROLITHES9

Fig. 8. Trace fossil Gyrolithes marylandicus, Little Cove Point M emberofSt. Marys Formation (Miocene)of Maryland,USA. Photo H.A. Curran.

pared with the fossil spiral burrows described under theichnogenus Gyrolithes.

Of the above-mentioned m odern spiral burrows, onlythat of Notomastuswould qualify asGyrolithes.

Intergradational morphologies among ossil andrecent burrows

Intergradation among the trace fossilsGyrolithes, Op hio-morpha, Spongeliomorphaand Thalassinoideshas been re-ported a num ber of times by palaeontologists (see Fursich1973; Bromley Frey 1974; May oral 1986; Bromley1990). It has been explained either by the presence ofmorphologically different burrow parts (in a burrow pro-duced by a particular species) or by the co-occurrence ofseveral different burrowing species in a particular sedi-mentary setting. Studies on recent burrowing crusta-ceans, particularly thalassinidean shrimp, indicate thatthe burrows have a wide diversity, both in their m orphol-ogy and in their function (see Nickell Atkinson 1995).For exam ple, the burrows ofA. australisshow three dis-

tinct burro w parts: vertical shafts, branched horizontaltunnels, and regular corkscrew spirals. On the basisofcomparative morphology alone, the corkscrew spiralswould be referred to the ichnogenusGyrolithes,while theirregular branched horizontal tunnels would be referredto the ichnogenus Thalassinoides(Frey et al. 1984). Fur-

thermore, in addition toA.

australis, there are a numberof other burrow ing organisms foun d in this tidal flat. Inthe 1960s, several species of callianassid shrim p were col-lected at this location (Rodrigues 1966, 1971), includingC. major which was also collected in 1994,200 m south ofthe tidal flat. The burrows ofC. major are one of the re-cent ana logues for the trace fossilOphiomorpha(Weimer

Hoyt 1964; Frey et al. 1978). Stom atopods(Squillasp.)have also been observed to cons truct burrows with agglu-tinated walls and a knobby exterior which are similar toOphiomorpha (Vaugelas 1991). Alachosquilla floridensis(Mann ing) is comm on in the subtidal of Baia do Segredo(Coelho 1995) and Praia de Barequeqaba (which are bothbays south of Praia do Araqa) and constructs a burrowwith a n agglutinated wall (Dw orschak, unpublished). A n-other stomatopod,Lysiosquilla scabricauda,is commonlyfoun d in the tidal flat in Brazil, whereA. australislives. Inthe m angrove channel of Twin Cays (Belize),where a spi-ral burrow most probably made byAxianassawas found,there were also burrows m ade by alpheid shrimps and bythe callianassid Glypturus acanthochirus(DworschakOtt 1993).

Construction of the spirals

In addition to speculation on the identity of the organ-isms responsible for constructing spiral burrows, theorieshave been presented on how such burrows may havebeen constructed. Toots (1963) discussed several possiblemechanisms and considered this burrow shape to be theresult of asymmetric digging movements. Another m odelof construction of Gyrolitheswas presented by M ayoral(1986)) who considered the paired striae on the outsideof the whorls to be scratch marks of a presumed crabschelipeds.

A num ber of works dealing with the burrowing behav-

iour of decapods have been reviewed by Atkinson Tay-lor (1988). Thalassinideans construct burrows mainly byexcavation. Sediment is loosened with the first two pairsof pereiopods and is carried in a basket formed by thesetae of these first two pairs of appendages (MacGinitie1930; Sanko lli 1963; Devine 1966; Rodrigu es 1966; LeGall1969; Thom pson, 1972; Ot tet al. 1976; Rodrigue s Hod11990). The sediment rem oved dur ing the construction ofa new burr ow is first brought to the surface. Later, the sed-iment is transported t o other parts of the burrow , whereit is placed in blind tunnels or inco rporated into the b ur-row wall. A different m odeof sediment transpo rt has beenobserved in the laomediid Jaxea nocturnaby Nickell

8/12/2019 15643

10/12

50 Peter C. Dworschak and Sergio deA. Rodrigues LETHAIA 30 (1997 )

Atkinson (1 995 ), which bulldozes sedim ent between itschelipeds. This method of sediment tran spor t by carryingsediment on the chelae and pushing with the outside ofthe chelae has been observed to be typical of burrowingalpheid s hrim p (M agn us 1967; review in K arplus, 1987).Bulldozing may be responsible for the ridges and furrow s

seen on th e burrow floora t transitions from the spiral sec-tion to tunnels in the burr ow castsof Axianassa.

In the burrowing ocypodid crabs, the direction of thespiral is related to the h andedness of the anim al. The bur -row spiral runs clockwise or counterclockwise, accordingto whether the small chela is the left or the right one(Linsenmair 1967; Farrow 1971). Ocypodids loosen sed-iment with the walking legs and carry it between themin or cheliped and the second an d third walking leg ofthe same side (V annin i 1980). These anim als always digsidewards and always enter the bur row with their m inorcheliped first.

Handedness may also bea factor in the construction ofspiral burrows in the thalassinideans. InAxianassa,how-ever, there is not a great difference in size between themajor an d mi nor cheliped (see Table 1. In addition, bothleft- and right-turning spirals are found in the same bur-row (see Table 2 .Callianassids havea more pronounceddifference between their major and m inor chelipeds, butboth left- and right-turning spirals have been observed ina single burrow inhabited by a single animal, e.g., inCal-lianassa bouvieri (Dworschak Pervesler 1988;Dworschak, unpu blished).

The trace fossil Dairnonhelix, known as the Devilscorkscrew, have spirals of a much larger dimension thanGyrolithes. Howe ver, the Devils corkscrew has be enattributed to the work of terrestrial beavers(Palaeocas-tor) ,which made the burrows by a series of right- or left-handed incisor strokes (M arti n 1994).

Recently, Follmi Grim m (1990) and Grim m Follmi( 1994) attem pted to interpret the occurrence ofGyrolithesand Thalassinoides in laminated sediments associatedwith event beds. They proposed the theory of doomedpioneers, whereby thalassinidean shrimps were carriedwithin gravity flows to a deeper but more oxygen-poorenvironment, where they reworked extensively the lam i-nated sediments with no o ther sign of bioturb ation. Theseautho rs speculated on the origins of spiral burrows: (a)the burrows are m ade by a decapod amputee (which lostits large cheliped durin g transp ort), ( b ) variations in thecoiling direction of spiral burrows may reflect the norm alsexual dim orph ism of differential cheliped enlargement,and (c) the trans port of an enrolled crustacean may havedisturbed the statocyst balance system, making the animaldizzy, which resulted in it digging a spiral burrow. Noneof these conditions would provide a satisfactory explana-tion for the means of construction of the spiral burrow ofAxianassa described in th e present work.

Fiinctional niorphologyof lANAssA burrows

The burrows of Axianassa australis, with their uniquecorkscrew sp irals, cann ot readily be classified using exist-ing models, such as the one proposed by Gr ifi s Sucha-nek (1991). They definedsix burrow types, based o n the

absence or presence of(1) sediment moun ds, 2) seagrassin chambers or the burrow lining and 3 ) a simple Ushaped burrow design, and a ttribute the types to one ofthe three general trophic modes (deposit feeding, driftcatching and filter/suspension feeding). Using the alter-native approa ch propo sed by Nickel1 Atkinson (1995),features such as surface mound, deep burrow, andsub-circular tunnel cross section found inAxianassaburrows are indicative of sediment processing for feedingor for burro w expansion and repair. This would infer thatthese animals are deposit feeders utilizing sub-surfacesediments.

The compact sediment and t he relatively weak abilityfor reburrowing observed in aquaria indicate that at leastthe deeper parts of Axianassa burrows are quite perm a-nent. The occurrence of mounds with freshly expelledsediment and the ventilation activity observed duringboth water coverage at high tide an d exposure at low tideindicate that sediment is removed fro m the deeper layers,either for burrow expansion or for the repair of collapsedparts of the upper burrow.

A close fit between anim al size and b urro w diam eter isindicative of suspension feeding, as in upogebiids(Dworschak 1983) and suspension feeding callianassids,e.g., Callichirus rnajor(Rodrigues 1966; Rodrigues Hod11991). A close fit is necessary for effective ventilation ofthe burrow, for both respiratory and feeding purposes(Dworschak 1981, 1987). Deposit-feeding callianassidsand axiids such asAxiopsis serratifrons(Dworschak Ott1993) and laomediids do not fit tightly in their burrows.They have a large relative burr ow diam eter an d may turnaro un d anywhere in it. The extreme case isJaxea nocturna(Pervesler Dworschak 1985),where anim als with a totallength of 5 cm and a bo dy diameter of less than 1 cm diga burrow 3-4 cm wide and 1.5-2 cm high. Here , gentlysloping bu rrow segments (with d ip angles of between0and 30) conne ct with even larger (6-8 cm wide,2-4 cmhigh) chambers.

Deposit feeding (rather than suspension feeding) isimplied by the poo r fit ofA. australis to its burrow dimen-sions (see Tables 1 and 2). This poor fit would makemovem ent in vertical shafts, and especially bulldozing ofsediment, impossible. The corkscrew spirals fou nd in theburrows of A. australis may therefore serve to allow theanimals to bur row to greater depths, with gently slopingburrow floor in order to exploit deeper sediment layersrich in organic matter.

Thalassinidean burrows with large chambersin differ-ent sediment depths (such as those ofJaxea nocturna)

8/12/2019 15643

11/12

LETHAIA 30 (1997) A modern analogue fo rGYROLITHES1

suggest tha t a nimals selectively seek sedime nt layers richin organic matter (Pervesler Dworschak 1985). Atightly layered lattice (such as is foun d in the bu rrows ofCallianassa subterranea) is considered to represent themaximal exploitation of a given volume of sediment bythe animal, with the depth of the lattices indicating the

thickness of the layers rich in organic matter (Nickel1Atkinson 1995).The regular spirals in the burrowso f A x i -nassa offer an optimum of accessible surface area in agiven volume of sediment. This indicates that organicmatter is more or less evenly distributed over the entiresedime nt layer in which the spirals occur.

It should be noted that the above explanations for thefunction of the spirals are tentative and n ot intended toprovide a complete answer. Direct observations of theburrowing a nd feeding behaviour ofA. australis, togetherwith a study of the spatial distribution of organic matterin the sediment, will undoubtedly help to clarify the func-

tion of the unu sual spiral burrows.Recently, Darryl Felder (personal comm unication, S ep-tember, 1996) has hypothesized an anti-predator func-tion for spirals in the burrow s ofAxianassasp. that he castin sou th F lorida. Experimental work is being designed byhim to test this.

Environmentalsignificanceof A I A N A S S A burrowsLittle is known about the biology and ecology ofAxianas-sa. Five of the know n six species have been rep orted to oc-cur from the intertidal zone, with only one species,A. are-naria, living in deeper water (38 m; Kensley Heard

1990). The sediment in which theAxianassawere foundranged from intertidal stones on rocky and gravellybeaches (forA. minieri)to sa nd or sandy silt (forA. amai-censisand A. arenaria), to muddy sand or mud (forA. in-termedia and A. australis) (Kensley Hea rd 1990; Rod-riguez Shim izu 1992).Axianassa australis has beenreported to occur at, or near to, mangroves (RodriguesShimizu 1992; this study), a nd several specimens of unde-scribed species of Axianassa have been collected in theCaribbean and the Gulf of Mexico from m angrove envi-ronments (Felder et al. 1995; Darryl Felder, personalcommunication, 1994).A cast similar to those ofA. aus-

tralis was made in a mangrove channel in Belize(Dworschak, unpublished). The genusAxianassathere-fore seems to occur mainly in poorly oxygenated sedi-ments in protected environments.

Acknowledgements.-Financial supp ort was given to PCD within theframe of the Convenio between the University of Vienna and the U ni-versity of SBoPaulo. Part of the resin was provided by courtesy of Ciba-Geigy Br a d W e are grateful to the entire staff of CEBIMar for their hos-pitality during our stay inSHo SebastiHo. Our thanks are due to Ms.Vania R. Coelho and M r. Paulo Pezzuto for help du ring field work, toDrs. Allen Curran an d Darryl Felder for unpublish ed information , andto Dr. Jorg A. Ott for critical reading of the m anuscrip t.Dr. Neil Cum-berlidge helped to impro ve the English. We also than k Drs. R. Jam esA.Atkinson andH. Allen Curran for their constructive comm ents.

ReferencesAl-Kho ly, A.A. 1958: Larval stages of four br achy uran Cru stacea.Pub-

lications of the Marine Biological Station, Al Ghardaqa, Egypt,10239-246.

Ammon, L. von 1900: Ueber das Vorkommen von Steinschrauben(Daemonhelix)in der Oligocanen Molasse Oberbayerns.Geognos-tischeJahreshefte13,55-69.

Atkinson , R.J.A. Taylor, A.C. 1988: Physiolog ical ecology of bur row -ing decapods. In Fincham, A.A. Rainbow, P.S. (eds.):Aspects ofDe -capod Crustacean Biology,201-226. Symposia of th e Zoological Societyof London59. C larendon , Oxford.

Braithwaite, C.J.R. Talbot, M.R. 1972: Crustacean burrows in the Sey-chelles, Indian Ocean.Palaeogeography Palaeoclima tology Palaeoecol-

Bromley, R.G. 199 0TraceFossils.Biology and Tap hon omy . Special Topicsin Paleontology3.280 pp. Unwin Hyman, London.

Brom ley, R.G. Frey, R.W. 1974: Redescriptio n of the trace fossilGyrolithesand taxonomic evaluation ofThalassinoides, Ophiom or-pha and Spongelimorpha. Bulletin of the Geological Society ofDen-mark 23, 311-335.

Buchanan, J.B. 1984: Chapter 3. S edimen t analysis.In Holme, N.A.McIntyre, A.D. (eds.):Methods for the S tudy of Marine Benthos. 1BP-

Handbook 16.2ndEd. 41 45 . Blackwell, Oxford.Christiansen , J.C. Cu rra n, H.A. 1995: The trace fossilGyrolithes:un-winding the spiral enigma in the St. Marys Formation (Miocene)ofMaryland. Geological Society of America, Abstracts with Programs,Northeastern Section27, 35-36.

Christy, J.H. Schober, U.M. 1994: A test for resource-defence matingin the fiddler crab Uca beebei. An ima l Beh aviour48,795-802.

[Co elho , V.R. 1995: Ecologia deUpogebiaomissa Gomes CorrCa, 1968(Decapoda: Thalassinidea: Upogebiidae). Dissertaqao apresentada aoDepar tamento de Ecologia Geral do Instituto d e BiociCncias da Uni-versidade de S o Paulo como parte dos requisitos para obtenqao dotitulo de M estre de Ecologia. 78 pp.]

Devine, C.E. 1966: Ecology ofCallianassa filholiMilne Edwards 1878(Crustacea, Thalassinidea).Transactions of the Royal Society of NewZealand, Zoology8,93-110.

Dworschak, P.C. 1981: The pum ping ratesof the burrowing shrimpUp-ogebia pusilla(Petagna ) (Decapoda: Thalassinidea).Jour nal of Exper-imen tal Marine Biology and Ecology52,25-35.

Dwors chak, P.C. 1983: Th e biology ofUpogebiapusilla(Petagn a) (Deca-poda, Thalassinidea).I. The burrows. P.S.Z.N.I.: Marine Ecology4,19-43.

Dworschak, P.C. 1987: Feeding behaviour ofUpogebia pusillaand Calli-anassa tyrrhena (Crustacea, Decapoda, Thalassinidea).1nvestigacid.nPesquera, Barcelona51,421-429.

Dworschak, P.C. Ott, J.A. 1993: Decapod burrows in mang rove-cha n-nel an d back-reef en vironm ents at the Atlantic Barrier Reef, Belize.lchnos2, 277-290.

Dwo rschak, P.C. Pervesler, P. 1988: Burrow s ofCallianassa bouvieriNobili 1904 from Safaga (Egypt, Red Sea) with some remarks o n thebiology of the species.Senckenbergiana ma ritima20, 1-17.

Farrow, G.E. 1971: Back-reef an d lagoonal en vironm ents of Aldabra At-oll distinguished b y their crustacean burrows.Symposia of theZoolog-ical Society of London28,455-500.

Felder, D.L., Ma nnin g, R.B. Lema itre, R. 1995: Ecologica l and bioge-ographical implications of distrib utions in mangrove-associated tha-lassinidean and alpheoid sh rimp (Decapoda).American Zoologist35,14A [Abstrac t].

Fernandes, A.C.S. Assis, J.F.P. 1980: Sobre a oco rrh cia d e ichnofos-seis na Formaqao Pirabas (Mioceno Inferior) no Estado do Para.Anais da Academia Brasiliera de Ciincias52,327-334.

Follmi, K.B. Grimm , K.A. 1990: Doom ed pioneers: Gravity-flow dep-osition and bioturbation in marine oxygen-deficient environments.Geology18 1069-1072.

Frey, R.W., Curr an, H.A. Pemb erton, S.G. 19 84 Tracemaking activi-ties of crabs and their environmental significance: the ichnogenusPsilonichnus. Journal ofPaleontology58, 333-350.

O ~ Y1,265-285.

8/12/2019 15643

12/12

52 Peter C. Dworschak andSergio de A . Rodrigties LETHAIA 30 (1997)

Frey, R.W., Howard, I.D. Ei Pryor, \V.A. 1978: OphiornoThu:its mor-phologic, taxonomic, and environmantal significance.Paltvgeogrn-p h y Paleoclirnatolog~~nleoecologj,73,199-229.

Fursiih, F.T. 1973:A revision ofth e trace fossilsSpotigeliornorpha,O p h -oniorpho and Thdnssinoides.Veues Johrbuc-h irr Geologieund Palnoti-tologic,Monntshefie1-7, 719-735.

Gerna nt, R.E. 1972: The p aleoenv ironme ntal significance ofGyrolithes(Lebensspur). orrrnnl ofPaleontology46, 735-741.

Grim m, K.A. Follmi. K.B. 1994: Doomed pioneers: allochtho nouscrustacean tracemakers in anaerobic basinal strata, Oligo-MioceneSan Gregorio Formation, Baja CaliforniaSur, hfedco. Palnios 9,313-334.

Hantzschel, W, 1934: Schraubenforniige und spiralige Grabgange inTuronen Sandsteinen des Zittauer Gebirges.Seric-keriberginrtn 16.313-324.

Hantzschel.W . 1935:Serioh m o n i c n n s p . und ihre Deutung.Setirk-enberginnu 17. 105-108.

Hantzschel, \