^3

BULLETIN OF MARINE SCIENCE, 28(1); 1-31, 1978

POLYPIDE MORPHOLOGY AND FEEDING BEHAVIOR IN MARINE ECTOPROCTS

Judith E. Winston

ABSTRACT Observations on living colonies of 56 species of marine bryozoans from Florida and Panama have shown that these organisms possess a variety of morphological and be- havioral adaptations related to feeding activities. In the species studied mean lophophore diameter varies from 187 to 1,012 fivn, mouth size from 15 to 91 ium, and tentacle number from 8 to 31. Polypide morphology, particularly introvert length and lophophore sym- metry, also varies from species to species, and this variation is linked to behavioral strat- egy. With respect to individual polypide behavior species can range from passive filterers (e.g. Crisia elongala), to tentacle feeders (e.g. Pasylhea lulipifera), to cage-captors (e.g. Bugiila neriiina), to particle jugglers (e.g. Siindanelta sibogae). Colony-wide patterns of morphology and behavior also reflect various methods of dealing with water currents. These range from weakly integrated patterns characterized by separation and a high degree of functional independence of individual polypides to highly integrated patterns in which both skeletal morphology and polypide orientation serve to enhance and/or chan- nel feeding currents. Similar strategies have evolved in all three orders of marine bryo- zoans, apparently in response to the common problem of changing the characteristics of water flow in the immediate vicinity of the colony so that food particles may be extracted.

Behavior is not a word one commonly as- early 19th century. Farre (1837) was able sociates with colonial marine organisms, to make detailed and accurate observations particularly those as sessile and apparently on the structure of the polypide in several inactive as ectoprocts. Seeing the calcified species, noting size and activity of the ten- crusts or seaweed-like festoons of their colo- tacles, distribution of cilia and "bristles," the nies among the rocks of breakwaters or on form of various portions of the gut, the fate the undersurfaces of platy corals, it is hard of food particles and the action of swallow- to imagine writing of any of them as, "re- ing. Hincks (1880) included in his survey markable among an active tribe for the of the British marine Polyzoa much informa- vivacity of its movements," or, "A colony in tion indicating his understanding of individ- full health and vigor affords a rare display ual morphology and behavior, including the of delicate structure, vivacious movement presence of sensory cells on the tentacles, and graceful form" (Hincks, 1880). Yet to though most of this information is buried in 19th century naturalists, enchanted by the species descriptions. Yet the comparative newly discovered miniature world of the aspects of polypide morphology have not tide-pool, the ectoprocts were indeed active received attention, and, indeed, examination creatures. of later 19th and 20th century literature In most species feeding activities of in- shows a gradual loss of information on poly- dividual zooids are limited to movements of pide structure and function. the polypide: protrusion, retraction, ex- Although the 19th century knowledge of pansion and bending of the lophophore, and individual (polypide) behavior was never associated actions of the ciliated tentacles. codified, it is apparent from species descrip- The basic features of the anatomy of the tions and illustrations (Grant, 1827; Lis- ectoproct polypide were understood by the ter, 1834; Farre, 1837; Hincks, 1880, and BULLETIN OF MARINE SCIENCE, VOL. 28, NO. 1, 1978

Others) that individual behavior has been related to parameters of morphology, ecol- observed. Even the process of feeding was ogy and taxonomy. This paper covers qual- studied (Grant, 1827; Farre, 1837; Hincks, itative observations on morphology and 1880), and, while the language in which it behavior. The quantitative relationships be- was described is overly metaphorical for tween them will be explored in a future present-day scientific tastes, the observa- publication (Winston, in prep.). tions themselves are generally correct. These authors noted both that the current MATERIALS AND METHODS generated by the cilia of the tentacles acts Ectoprocts were collected in various sub- to "create a very maelstrom in the water, tropical and tropical habitats; in Florida, which sweeps the passing animalcule or the from estuarine grassbeds, fouling panels, floating food particle toward the central breakwaters, and intertidal beachrock ledges, mouth" (Hincks, 1880, p. xiv), and that the in Panama, from coral reef and intertidal tentacles themselves might play an impor- and subtidal rocky areas. No attempt was tant role in feeding: "The tentacula are made to collect all the ectoprocts from any exquisitely sensible, and we frequently ob- one habitat, rather the emphasis was on col- serve them either singly or all at once, strik- lecting as many different kinds of colonies ing in their extremities to the centre of the and as many representatives of the different bell-shaped cavity, when any minute floating groups, cheilostomes, cyclostomes, and body comes in contact with them" (Grant, ctenostomes, as possible. 1827, p. 114). Animals were taken to the laboratory While individual polypide behavior was where they were observed and measured in understood by those early workers, the phe- seawater under the dissecting microscope. nomena of colony-wide patterns in feeding Most observations were made as soon as behavior do not seem to have been dis- possible after collection, as the behavior of covered until much more recently. Banta, many species changes markedly within a few McKinney, and Zimmer (1974) showed the hours after being collected. Some species presence of excurrent chimneys, based on could be kept in aquaria without harm, and polypide morphology and orientation in these were maintained and fed with Dunal- Membranipora membranácea and specu- ¡ella or Isochrysis. lated on the possible function of monticules Because there was so little information and other skeletal modifications in fossil and available on the ways in which polypides be- living ectoprocts in providing effective water have, the first few weeks of the project were circulation throughout colonies. spent in making detailed obsei-vations on a In reviewing the feeding biology of ma- variety of species in order to learn what be- rine ectoprocts (Winston, in press) I be- havioral characters occurred and how they came aware of the limited knowledge, the could be combined in a data sheet that could controversies, and the possibilities, regard- be used with other species. Table 1 sum- ing polypide morphology and feeding be- marizes the characteristics for each of 56 havior in these animals. The research re- species of marine ectoprocts from Florida ported here was started in order to learn (1 ) and Panama studied. The analysis for each the range and variation of morphology pos- species included characters of polypide mor- sible in the gymnolaemate polypide, with phology: shape of the lophophore, shape of specific reference to variation in lophophore the mouth, length of the introvert; individual shape, and in the positioning of possible sen- behavior: tentacle flicking activity, other sory structures, (2) what different types of types of individual actions, particle rejection behavior patterns were possible for individ- mechanisms; and colony level behavior: ual polypides and for colonies, and (3) in relative amount of time polypides remained what ways these behavior patterns could be expanded, level of behavioral integration, WINSTON: BRYOZOAN MORPHOtOGY AND FEEDING BEHAVIOR

^ S ^ ^ S ^ 5: ^ ^ 00 00 CO C/) GO

.•. (N f^ • • ^H r-í n oi r^i (N cj HO CQ

as OOCO 00 ooc/^t/íoooo CO CO C/5 CO GO CO CO C/-1 CO 0 CO O 2S -J S2S, 2S > > > > > > HJ J :¿ > < X w oí < o < > u o o < o o ü u < O Ü o s O < <: PH

¿•30 î^ciE LÍ ra '-^ï .E í-í S < SSKW< < < XSKSS 2Ï

S t/5 HJ HJ H4 2 s s s S J

H H H i- [5 C¿ Pí W UJ Q< W U W Pi W c¿¡ c^ O >^ Ü o «; 0.5 : (^1 M- o\ 00 o r^ o 01 t^ "^ m r*-) ^ &> Pä O ¿ a> ü H H H H H H H H H H H H H H a H 2 0.0 a cq u W UJ W UJ tij w < W W w 0 0 0 0 0 0 CO UJ Q < < >^ o o (U o í 00 o 00 ^ GO r^ "^ -^ ^D r^l • " n • r-l 2: _o o TT • O *^ • ^JO r- r-J O 00 C^l V-i r^ •. • • ^ r^i 01 rJ ÇX SH2o^ • in ^£5 r^ fNj ^ "^ ri •^ -^ o e C ¿ 1) p M >Ai O r^ c^i ^ •. r^ : [^ r- _^ r^ r-- •1 o o o-i 'o 00 •0 r^ >/l 0 i^ r^ r^ 00 n r*^ CM vn ^O r^ r- ^ d •^ ^ r~t MD r^ •' ^

ri rJ

, i _^ '^

lit ~^"'-^o r- ^ '^ •::: r-

^ -i; tj o o o ^ X X CQ CQ 0:3 CQ ^ cS 0 0 Ou 0 0

r-i •^ 'O "sO t^ GO •• rJ r^ '^ »^ ^¿ r-^ 00 o< Ö BULLETIN OF MARINE SCIENCE, VOL. 28, NO. 1, 1978

c c o ^ ^ ^ Oil

o >-.2

fio- _ T:f fvj M5 (N

(A o o 2 o o o C/5 CO 00 OO CO 00 O ''^ 2 s -I > ^

o

•=¿"2-; ü o o o o ü" tu tu 5

S S Ä KW S < < < :s Ä s <

j Z

j (^ H c¿ Qá c¿ s« o p¿ tó f¿ oí Pí W O c4 Oá Pí Pí o Pí o Píí .c< o< oí oí

r^ SO »/~i »/^ ''BSP -H ^^ r-i es ^^ ÍN (N

H o i-^ ^ í^ H H H H H H H H H H H H H H H tu c^ W W U-1 Ö tü O w m m rt w w O w O w O m CO CO CO

oo oo

Su O oo ON r-1 >í^ r^, r^ ^o oo a\ O OO o so r- c\ so o --. SH5 = r-1 SO (N v-i r-i ^- c-i xf

^ 2 " E •-£> o oo >0 SD G\ r- r^ as ¿'S, o ra ; ^ o (N Os OJ ri oo ri ^ ^ (N et ^

-2 C^ -< to i> •5 g 3 è c <-^ ¿ 'S Í3 C2 S, _2 .S ^ ^ g r^ O •t3 ^ • 'Si 3 Is c O 3 •^ •f O io ^ 5-1 o o 5 r^ S g Ci 3 S o .^ 5 o ï g 5 Sj o Ci. 2 S o o o ^ o ^ 1 ta. ^ s: c -S -S c -S Ci. o. 5 5 5 o G G G 5 C O í ¿ Í Í 5 S § § I¿ ^== Cu ^ ^ »T) (N r^ •^ >n ^o r^ 0¿ a\ Ö ri en ^ H !N r) (N ri ri (^1 OJ

= § •? s "^ : 5 S o w '^'^S^^^S-^ S ^ i: 2 s ; -t X •; I " " \ XI "O

og'.S " ° ra lo "^ r^, r^ »i^ r^ r^i - c ^ y ,

lyj p. Cl o H ce O o ^ O O > en CÛOJ c O ^ II < o o II o ^ rii w II c II O o -* .. c H H ^ V t/i >. O c > o Cí) II o< < o ^ ex o a> II 5= C¿ M •5 D. < -S o "o i M a e¿ o".> o O o o O < u. .2 s s

: 5 til "II J ÄS :s < :s 5 S < £ •" >, a 11- .t: fe.S o " > I Jí " ïi II s s s J J Z h-l c/) J o# >» -•:: g-S- : £ -^ S é E S

: '' r- cij •- c tí a 2 H H tu LU pj Ci Ci C¿ O íií c^ « J " "H i I " 5 " 'S =^ :• i Î "^ o S c o > g Ï ^- > M o-U , = ~S 11 • Es t; ! ^

ses. H H H H H H f- H H (- H H H Í- O O O U4 ttj rj UJ w O ¡n uj Lu SJJ !- ¿ « "-^ W 1 a« O o S-^IIHJH r o --o S

j o '-^ CJ, cJ/ p, cu "O

i - 3,11 1! a ^ p o O 5 " c - xj != < o <= _ E "

c I ^ h-- -O - 'D-O t3 Ü ^ oo ÍÍ-1 CO o n c\ t-A r^i oo o oo s H -- •o m ''^ r^i (N vi • '> "^ o ^ *^ 0$ '^ ri "S r-i 1^ ON '-0 ^O o\ rn OJ ! s o Ï ü i M .^ -^ í'í ^> O C\ • ^ IV-, ^o ; ^ ä "i ^ a £ (^ »v^ ^ r^ ri y:? s li •EQ .|:"-a g E "»MO ; o 73 o II c ° " i = s am S-g 11 i o II o • -S s •-

Cl. 5 o O LU M c CL =: o 0¿ -5 ^ .§-s ¡2 c Si 5 -o c OK ._ II til í ti "? •- ^ NOD. « K f^ K h ;^ à

; S ^ J u. -= • BULLETIN OF MARINE SCIENCE, VOL. 28, NO. 1, 1978

separated from the peritoneum by a layer of connective tissue containing collagen fi- bers, and with a lumen filled with coelomic fluid (Lutaud, 1955; Smith, 1973; Gordon, 1974). By various means, in ctenostomes, cheilostomes and cyclostomes, muscle ac- tion upon flexible portions of the zooid wall yields a change in hydrostatic pressure re- sulting in protrusion or retraction of the lophophore. When the lophophore is re- tracted it is enclosed in a tentacle sheath, proximal distal ^ which when expanded functions as an in- trovert. Figure 1. Diagram of a gymnolaemate polypide The gut in bryozoans (reviewed by Win- •showing orientation and main features (Refer to text for explanation of abbreviations). ston, in press) is U-shaped, beginning with a ciliated pharynx (ph), an esophageal or cardial region (car), in some ctenostomes and strength of colony currents. Measure- modified into a gizzard. The stomach con- ments were also made of several skeletal sists of a sac-like caecum (cae) and a ciliated characters: length and width of zooecium, pyloric region (py) where food is processed length and width of orifice, as well as the in a whirling mass. This is divided by a most important polypide characters: num- constriction from the rectal region (r) where ber of tentacles, lophophore diameter, ten- digested food remains are formed into fecal tacle length, mouth diameter (greatest pellets (fp) to be excreted through the anus length), and tentacle width so that a quan- (an). The shape of the gut also varies from titative analysis of the possible relationships species to species and group to group (Silén, between skeletal, morphological, and be- 1944), and seems primarily a function of havioral characters could be carried out the shape of the zooid, though little com- (Winston, in prep.). The observations on parative study has been made. morphology and behavior were documented with slides of polypide morphology and 16- Lophophore Shape and mm microcinematography of polypide ac- Tentacle Orientation tivity. The gymnolaemate and stenolaemate lophophore is generally thought of as cir- POLYPIDE MORPHOLOGY cular (in cross-section or oral view) in con- The most important parameters describ- trast with the horse-shoe or spiral shaped ing polypide morphology are illustrated by lophophores of phylactolaemate ectoprocts, Figure 1. Figure lA shows the arrangement phoronids, and brachiopods. Yet, while a of the lophophore as viewed from above. simple circular lophophore is common in The tentacles are arranged in a circle around these groups within the restrictions imposed the mouth (m), with the anus (an) outside by their small size, the morphology of the and below the funnel formed by the tentacles. lophophore has undergone modifications In side view (Fig. IB) the complete poly- and elaborations with respect to size, num- pide (lophophore, tentacles, sheath-intro- ber of tentacles, lophophore shape and ten- vert, digestive tract, and muscular and ner- tacle positioning. vous tissue, including the nerve ganglion (n) associated with it) can be seen. Equi-tentacled Lophophores The tentacles comprising the lophophore Lophophores in the form of circular fun- are hollow tubes, each with an epidermis nels with all tentacles of equal length are WINSTON: BRYOZOAN MORPHOLOGY AND FEEDING BEHAVIOR common in all three groups of marine ecto- procts: ctenostomes, cheilostomes, and cyclostomes. Most species with small lopho- phores are of this simplest type, though even these can be quite different in ap- pearance, depending on whether the ten- tacles are held moderately close together (Fig. 2A), or spread far apart (Fig. 2B). Species with large lophophores may also be of this type. In some large species the nu- merous long tentacles, held very straight, give the lophophore a crowded narrow ap- pearance (Fig. 2C). In other species the tentacles appear more flexible and bend outward at the free ends giving the lopho- phore a bell-like shape (Fig. 2D).

Obliquely Truncate Lophophores It was noted by Farre (1837) that some polypides of Alcyonidium gelalinosum have tentacles much shorter on one side of the lophophore than on the other. Hincks (1880) also noted a dissimilarity of ten- Figure 2. Examples of equi-tentacled lopho- tacle length in Eleclra pilosa: "Very com- phores: A. Crisia sp., B. Anguinella palmata, C. Beania inlermedia, D. Watersipora siibovoidea. monly those on one side are inferior in height to those on the other, and the ten- tacular bell is obliquely truncate above." unison to produce a very strong current flow Recently Gordon (1974) has pointed out through the colony, the shape of the lopho- this same phenomenon in Cryptosula pal- pore appears suited to increasing current lasiana, but no one has attempted to explore production. its functional or taxonomic significance. As In other colonies, polypides may have can be seen from Figure 3 the anal (distal) lophophores equi-tentacled or obliquely tentacles are the longest and the abanal truncate depending on their position with (proximal) tentacles are the shortest, with respect to incurrent cells. In colonies in the intermediate tentacles showing a smooth which the polypides form fixed clusters, gradation in height. Because of the flexi- either of the skeletally or non-skeletally re- bility of the introvert, however, the polypide flected type (see behavior section), the poly- may be twisted on expansion so that the pides in the center of the incurrent cell are longest tentacles may appear to be oriented equi-tentacled while those toward the out- in various other directions. side of the cell have obliquely truncate The present study has shown that oblique- lophophores with the longer (anal) ten- ly truncate lophophores occur in two pat- tacles being toward the outer margin of the terns (Table 1). Species with erect branch- incurrent cell, while the excurrent chimneys ing colonies such as Bugula stolonifera or (Banta, McKinney, and Zimmer, 1974) are Caulibugula ármala, commonly have all fringed with polypides having extremely long polypides of the colony with obliquely trun- anal tentacles (Fig. 3B). In many of these cate lophophores (Fig. 3A). In such spe- species the longest tentacles may be bent in cies, in which the individual polypide has at the tips, while the abanal tentacles bend little flexibility, but all polypides act in out, giving the lophophore a decided scoop BULLETIN OF MARINE SCIENCE, VOL. 28, NO. L 1978

some role in increasing the efficiency of cur- rent channelling. It also indicates that there may be even more elaborate patterns that have not yet been observed.

Bent-tentacled Lophophores One of the most important observations made by Hincks (1880) was that in the lophophores of several species of cteno- stome ectoprocts two tentacles were consis- tently bent away from each other (campy- lonemidan). Unfortunately, Hincks used the campylonemidan lophophore structu.re as the basis for his classification of the cteno- stomes, and when this classification was dis- carded on various grounds, his observations on campylonemidan symmetry were also discarded. Later authors (KraepeHn, 1887; Loppens, 1908) questioned its existence and Hyman (1959) did not even include the term. But, although possession of a campylo- nemidan lophophore did not provide a good Figure 3. Examples of obliquely truncate and basis for ctenostome classification, Hincks' scalloped lophophores: A. Bitgiila slolonifera or observations of the peculiar lophophore Caiilibugiila pearsei, B. Alycoiiidiiiin polypyli/m, structure of some members of this group, C. Cellepororia albiroslris. D. Tremaiooecia tur- ril a. Vclkeria uva, Valkeiia trémula, and Vic- torella pavida are valid, although he be- lieved that the two everted tentacles were shape (Fig. 3C). In such species, viewed anal while in all the specimens I have exam- from above, the tentacle tips form an oval ined they are abanal. Like the other lopho- or triangle (apex anal, base abanal), ]'ather phore shapes, the significance of the cam- than a circle, so that the lophophore ap- pylonemidan lophophore appears more proaches a bilateral symmetry. functional than taxonomic. In the cteno- stomes it occurs at least in Victoreila (subor- Scalloped Lophophores der Carnosa), Valkeria and Aeverrillia (Fig. 4A) (Suborder Stolenjfera), all delicate Other elaborations of lophophore shape creeping forms. In the cheilostomes, it is are possible. Several species have been ob- well developed in at least one species of the served (e.g. CeUepoiaria albiroslris) in which Sertellidae (Fig. 4C), Reteporellina evel- the tentacles of the obliquely truncate lopho- inae, a group characterized by a rigid phore do not grade evenly toward the ven- tral side, but by variation in length and branching or fenestrate colony form in which amount of bending of the tips, form a scal- both zooid skeletons and polypides seem loped margin to the lophophore (Fig. 3DV aligned to produce the most effective cur- The functional significance of this shape is rent flow through the colony (Fig. 13B). unknown, but as it occurred in a species The campylonemidan trend has also oc- which appeared to have reached a high level curred in the lunulitiform cheilostome, Dis- of behavioral integration, it probably plays coporella umbellala depressa, whose highly WINSTON: BRYOZOAN MORPHOLOGY AND FEEDING BEHAVIOR integrated colonies are free-living on sandy bottoms. This type of lophophore is found in cyclostomes, too, being particularly pro- nounced in species of Lichenopora (e.g. L. buskiana, L. inlricata) in which the zooids are arranged one above the other in a col- umn, with a channel between and a hollow center to the colony or sub-colony, so that truly colony-wide current flow obtains. Thus, similar lophophore conformations have evolved in all three orders of marine bryozoans, apparently in response to the need to more efficiently channel feeding currents.

Introvert Length The length and flexibility of the introvert region supporting the expanded lophophore crown is an important determinant of the Figure 4. Examples of bent-tentacled (campylo- type of behavior possible to the polypide. nemidan) lophophores: A. Aeverrillin ármala, The amount of introvert protruded from the B. Lichenopora buskiana, C. Reteporellina cve- zooid can range from none (Fig. 2A) in linae. cyclostomes, to short (Fig. 2B), to mod- erate (Fig. 3A), to long (Fig. 1, Fig. 3C, counts between 21 and 30, but the species D), in which case most of the gut can be with the highest tentacle number was a seen within the introvert. In general, a very ctenostome Sundanella sibogae (mean ten- short introvert means a greater restriction tacle number 31). in the degree of rotation afforded to the Variation in tentacle number may occur lophophore crown, but this depends also on within a species, and between polypides of the structure of the colony and other aspects the same colony of a species. lebram of behavior (see behavior section). (1973) pointed out that external factors (chiefly nutrition) can influence the ten- Lophophore Dimensions tacle number. My own work (Dudley, Tentacle Number 1973) on the growth of Conopeum tenuis- The number of tentacles in marine bryo- simum. has shown that this variability fol- lows an orderly pattern. Polypide size, ten- zoans varies from a minimum of eight to a tacle number and variability in tentacle maximum of 30 or more. Tentacle number is related both to taxonomic position (Win- number increase in the distal direction as ston, in press, for review) and to polypide zooid size increases, and variability is there- fore a function of genetic as well as environ- size. Of the species examined, 20 (eight ctenostomes, six cheilostomes and six cyclo- mental control of the colony. This pattern stomes) had mean tentacle numbers be- appears to apply to other species as well. tween eight and ten. In two ctenostomes In those species in which zooid size in- and 16 cheilostomes mean tentacle number creases toward the younger outer edges of was between 11 and 15, while in one cyclo- the colony, polypide size, tentacle number stome, one ctenostome and nine cheilo- and variation in tentacle number also in- stomes it was between 16 and 20. Six species crease. (all cheilostomes) had mean tentacle number It is possible that variation in tentacle 10 BULLETIN OF MARINE SCIENCE, VOL. 28, NO. 1, 1978

MEAN LOPHOPHORE DIAMETER

Figure 5. A. Distribution of moutii size in bryo- Figure 6. Sleganoporella inagnilahris and Clili- zoan species from Florida and Panama; B. Dis- donia pyriforrnis•contrasting lophophore size in tribution of lophophore size in bryozoan species thie largest and smallest species studied. from Florida and Panama.

Lophophore Size number within a colony is due to increase In the species I have studied from Florida in number of tentacles with aging of the and Panama mean lophophore diameter polypide. Tentacle length can increase as varied from 187 /xm to 1012 /.im (Fig. the polypide ages. For example, in many 5B), and as tropical species in most groups species, the young recently developed or re- of invertebrates are generally smaller than generated polypides (distinguishable by their greater transparency and the lack oE cold-water forms, the actual possible range food in their guts) has short, equi-tentacled in size is probably even greater. The largest lophophores, while older polypides in ad- species is 5 times the size of the smallest, jacent zooids have longer or obliquely trun- and this size differential produces even more cate lophophores. An increase in anal ten- important differences with respects to feed- tacle Length occurs especially wherever ing and behavior. Figure 6 shows the lopho- polypides are positioned so that these ten- phore of the smallest species studied (Chli- tacles must elongate to produce an effective donia pyriforrnis) in comparison with that incurrent cell. of the largest {Sleganoporella magnilabris). WINSTON: BRYOZOAN MORPHOLOGY AND FEEDING BEHAVIOR 11

Mouth Size and Shape There are four kinds of structures which may be presumed to have a sensory function The size and shape of the lophophore and (Fig. 7) : latero-frontal cilia, tentacle-tip the length of the tentacles may set physical cilia tufts, and abfrontal structures, either limits on the type and size of particles that bunches of stiff cilia, or single ciliary struc- can be collected (Strathmann, 1973), but tures which may be immovable or may show the most important physical limiting factor a whip-like activity. is, of course, imposed by the size of the mouth. In the 53 species measured mouth Latero-jrontal cilia size ranged from 15 to 91 ¡xm (Fig. 5A). Even without allowing for behavioral dif- The latero-frontal cilia were not noticed ferences between species, the food available by early workers and were first reported by to the first species is surely much more re- Lutaud (1955) for the cheilostome Electro stricted than that available to the second. pilosa. Since then they have been observed The mouth in marine bryozoans is usually on a ctenostome, Zoobotryon verlicillatum described as round (Hyman, 1959), but this (Bullivant, 1968), other cheilostomes, is not the case in all species observed. Spe- Membranipora villosa (Strathmann, 1973), cies with small mouths (less than 30 ,u,m in Cryptosula pallasiana (Gordon, 1974), and diam) all had round mouths. Species with the cyclostome, Crisia elongata (my obser- mouths of intermediate size (30 to 40 ^m in vations). These cilia, at least in Cryptosula diam) had mouths of shapes varying from and Electra, have two-branched rootlets, round to ovoid. In the larger species (mouths like the frontal cilia, but in living animals greater than 41 ^m in diameter) mouths they appear thicker than the frontal and lat- were always elongated either into key- eral cilia (fused cilia?) (Fig. 7A). Accord- hole or triangular shapes. In all but one ing to Gordon (1974) these structures in species (Gemelliporidra multilamellosa) this Cryptosula are not always immotile but may elongation was parallel to the elongation of make occasional flicking movements. How- the lophophore, in the dorsal-ventral direc- ever, in other species (pers. obs.) they do tion, in contrast to the lateral elongation of not show any motility (at least under the mouth and lophophore that occurs in phy- microscope). Strathmann (1973) has sug- lactolaemates, phoronids and brachiopods. gested that the latero-frontal cilia may serve as upstream particle sensors, governing the Comparative Tentacle Structure ciliary reversals which play a large role in the process of particle transport down the The fine structure of the gymnolaemate tentacles. If this is the case, these struc- tentacle has recently been examined in de- tures could (1) detect a wider range of food tail by several workers (Smith, 1973; Lut- particles than the lateral cilia alone, and (2) uad, 1973; Gordon, 1974), and it seems a if longer than the lateral cilia themselves, al- reasonably safe assumption that this general low the animals to detect particles that structure is common to all members of the would pass outside the range of the lateral group. The position and structure of the cilia, making it possible for them to utilize frontal and lateral cilia seem to be fairly a faster moving feeding current. Strathmann universal, as would be expected considering found species of Bugula to have some latero- that they are basic to the food collection frontal cilia 1.5 times as long as the lateral mechanism (Strathmann, 1973). It is with cilia. In the species of Bugula I have ob- the occurrence and distribution of possible served latero-frontal cilia are also longer sensory structures that most of the potential than lateral cilia in keeping with the ability for variation lies, and this, too, is expected of these species to create more rapid feeding on the basis of differences in ecology and currents than species with other colony types behavior. (see behavior section). 12 BULLETIN OF MARINE SCIENCE, VOL. 28, NO. 1, 1978 WINSTON: BRYOZOAN MORPHOLOGY AND FEEDING BEHAVIOR 13

End Tufts "armed at the back with about a dozen fine Clumps of cilia can be seen at the tips of hairline processes, which project at nearly the tentacles in the illustration of Electra pi- right angles from the tentacula, remaining losa given by Lister (1834, plate XII, Fig. motionless, while the cilia are in constant 2), although he did not mention them in the and active vibration." Farre did not discuss text. Gordon (1974) mentioned a clump of the possible sensory nature of these struc- cilia on the tips of Cryptosula tentacles, but tures, but by the time of Hincks (1880) it they are not mentioned by other recent au- seemed to be accepted that the tentacles thors, although they are present in all spe- "must also be regarded as tactile organs, and cies I have observed. In Crisia elongala in many species are furnished with special these end tufts are particularly long (Fig. appendages, by which their sensitiveness and 7B), three or four times the length of the power of detecting the presence of minute lateral cilia. If these are truly touch-sensory particles are greatly increased." cilia (or palpocils, to use Hincks' (1880) The most comprehensive description of term) it is not hard to imagine why they are these cilia is given by Gordon (1974) who concentrated on the tentacle tips. Most spe- found the abfrontal tentacle surface of cies of marine bryozoans utilize a testing Cryptosula to be adorned with short tufts of position before the lophophore is expanded about ten immotile cilia, alternating with (see Fig. lOE), during which time only the solitary motile cilia. BuUivant (1967) de- tips of the tentacles protrude from the ori- scribed pairs of bristles on the abfrontal ten- fice. Very likely it is these bunches of cilia tacle surface of Zoobolryon. In my obser- which enable the animal to monitor the vations I have found that all species possess movement and particle content of the sur- some abfrontal sensory structures. These rounding water. are most commonly tufts of various sizes with solitary bristles in between. Whip-like Abjronlal Structures structures similar to those of Cryptosula oc- Abfrontal structures are of several types: curred in Watersipora. clusters of immotile cilia (Fig. 7C) similar It seems probable that the latero-frontal to those found at the tentacle tips and soli- cilia could be functioning in controlling tary structures, either immotile cilia, or ciliary feeding, while the end bristles are also whip-like structures which display a flicking in a position to act as particle sensors, both activity (Fig. 7D). These abfrontal struc- in the testing and expanded positions, but tures were known to the 19th century biolo- the possible role of the abfrontal structures gists (Lister, 1834; Farre, 1837; Hincks, is not so clear. Possibly, if they sense water 1880), and their presence or absence in movement or particles, they could have various species was noted. Lister (1834) some role in controlling polypide orienta- was apparently the first to detect the pres- tion relative to other polypides and so func- ence of abfrontal "hairs" or cilia which he tioning in the formation of temporary clus- suggested might "give notice of anything ters, and permanent current cells. Thus, it coming within their touch." Farre (1837) would be interesting to examine their com- called these structures "spines," and de- parative structure and position in species scribed them in Bowerbankia densa that are known to have differing kinds of (= Bowerbankia imbricata) as follows: feeding behavior.

Figure 7. Tentacle sensory structures: A. Tentacle o£ Watersipora suhovoidea, B. Tentacle of Bug- iila neriiina, C. Tentacle of Bowerl^anlciii (Indian River sp.), D. Tentacles of Crisia elongala. B = ab- frontal "bristle"; ET = end tuft of cilia; Cl = abfrontal clumps of cilia; SC = non-motile abfrontal single cilia; W = abfrontal motile "whip" cilia; LC = band of lateral cilia; LFC = lateral-frontal cilia. 14 BULLET,N OF MARINE SCIENCE, VOL. 28, NO. 1, 1978

r

Figure 8. Basic components of polypide behavior. (1) Fully retracted, orifice closed; (2) opening of operculum or lip; (3) extrusion of tentacles through orifice in "testing" position; (4) extrusion of poly- pide; (5) expansion of lophophore; (6) bending of polypide in scanning activity; (7) retraction of loph- ophore.

INDIVIDUAL AND COLONIAL BEHAVIOR to the testing position and then expand (Fig. Parameters of Individual Behavior lOE, F). Extrusion of the polypide (4) consists of extending the compressed ten- Before species-specific differences in be- tacles of the lophophore and a varying havior can be explained it is necessary to amount of the introvert and upper region of understand the basic components of poly- the gut from the zooecium. This also usu- pide behavior. Figure 8 illustrates these ally occurs rather slowly, though the rate processes beginning with the protrusion of differs between species and also depends the polypide from the zooid. on how well the animals are adapted to the conditions in the experimental situation. At Extrusion of the Polypide and this stage the tentacles are still held tightly Expansion of the Lophophore together; if the tenacles are long they may be bunched or twisted together. Once the (1) When the polypide is fully retracted polypide extends as far as possible from the within the zooid, the orifice (governed by zooecium the lophophore unfurls (5). In operculum, lips, or terminal membrane) is those species with long tentacles it usually completely closed. The first hint of emer- takes a little more time for the ends of the gence of the polypide is (2) the opening of tentacles to untwist and straighten out. the operculum or lip or widening of the Animals may not expand the lophophore terminal membrane; this is generally ac- fully immediately, but instead may keep companied by a shifting of the polypide for- withdrawing or contracting the bell for a ward or upward from the fully retracted posi- period of time until they become accustomed tion. The extrusion of the tentacles through to laboratory (or natural?) conditions. the orifice (3) is usually a fairly slow move- When the lophophore is expanded (in chei- ment. The polypide may pause here in a "testing" position, as if it is sensing water lostomes and ctenostomes) the introvert movement or particles present•e.g. when may bend (6) in order to orient the open- food particles (Dunaliella) are added to the ing of the lophophore in various directions water zooids of many species rapidly emerge (Fig. lOA also). As noted earlier the de- WINSTON; BRYOZOAN MORPHOLOGY AND FEEDING BEHAVIOR 15 gree of flexibility depends partly on how associated with tentacle feeding or more much the polypide can be protruded from complex tentacle action than the first type). the zooecium (introvert length), but it also seems to be a species-specific component of Other Individual Actions behavior. (7) Retraction of the lophophore Other actions shown by polypides of one is a very rapid motion. The tentacles are or more of the species observed include in- compressed and snapped back in (mucous tegrated activity of several tentacles or of the surfaces and interior collagen rod giving whole lophophore. These include widening protection from the stress of this action). or contracting the lophophore bell (espe- Polypides may retract all the way back into cially in Beania spp.), bending the free ends the zooecium or may retract only partially of all the tentacles in, a movement which to the testing position, depending on the de- could occur either in rejecting heavy particle gree of disturbance. concentrations or in making a cage of the lophophore around particles. Several spe- Tentacle Flicking Activity cies, particularly Bugula neritina (Fig. 11), Flicking of the tentacles has been noted were observed actually using this cage to by many authors (Borg, 1926; Mangum and capture larger particles or protistans. "Wav- Schopf, 1967; Strathmann, 1973) to be one ing" or "fanning" tentacles•the movement of the most important components of poly- of several tentacles in and out in unison, pide behavior, and this character was anal- "avoidance" retractions, a motion in which yzed in several ways. Tentacle flicking the lophophore is rapidly pulled partway varies from species to species both in quan- back into the zooid leaving the bunched to- tity: the amount of fHcking, how often it gether ends of the tentacles extruded, and occurs, and in quality: the intensity and the "writhing," a continuation of this activity in length of tentacle involved. Speed may be which the partially withdrawn tentacles fast or slow, or a species may be capable of writhe about like snakes on the zooid sur- both fast and slow flicking. The intensity face also occurred in several species. A few varies from gentle to strong. Fast flicking is species could flatten all the tentacles out- usually a sharp, hard action almost too rapid ward so that the lophophore assumed a sau- to follow with the eye. Often the tentacle cer or disk shape (Celleporaria albirostris, appears to "bat" a particle in toward the Celleporina liassalli). mouth. In slow flicking it is possible to fol- Bending of the polypide occurs in many low the action of the tentacle as it curves species so that the lophophore can scan in around the particle and whips it down to- various directions (see Table Ï). Some spe- ward the mouth. Usually tentacles flick in cies (Fig. lOA, Bowerbankia) are capable toward the mouth, but they may also flick of multidirectional scanning, by bending and outward (shaking off unwanted particles). rotating the introvert region they can scan The tentacles of a few species seem to in a wide circle around the perimeter of the twitch the free ends from time to time (e.g. space occupied by the erect lophophore. Celleporina hassalli) or constantly (e.g. Many species which have a long introvert Zoobotryon verlicillalum). In most species (e.g. Beania intermedia, Sundanella sibogae, observed some kind of flicking activity oc- Conopeum spp.) practice multidirectional curred. In only a very few species (e.g. scanning. In a few species bending is not li- Rhynchozoon sp.) was tenacle flicking rare. mited to the introvert, for example Victorella The length of tentacle invloved in flicking pavida can bend both at the mouth and the action can vary from just the outer tip, to orifice level and the flexible zooid can bend, the outer third (the most common type, oc- too. Other species (usually those described curring in almost all species) to one half or below under Type II colony behavior) are even the full length of the tentacle (usually very limited in scanning activity, but the loph- 16 BULLETIN OF MARINE SCIENCE, VOL. 28, NO. I, 1978

ophore is capable of a slight degree of up and down motion (vertical scanning). In those type II species in which the introvert is very short (e.g. Synnotum aegyptiacum, Vilta- ticella contei) the lophophore can scan only very slightly. In other species, although the introvert may range from moderate to long the presence of colonial patterns involving incurrent and excurrent water pathways may mean that very little scanning is done by the polypides. When scanning activity does oc- cur in these species it is chiefly lateral scan- ning or bending of the lophophore.

Particle Rejection Mechanisms The force and persistence with which polypides of most species can reject un- wanted particles is astonishing. The mech- anisms used are varied, a particular species may reject particles by one or several meth- ods. Rejection mechanisms might be ex- pected to be most highly developed in those organisms found in the most particle-laden waters (e.g. Conopeum spp.), but to most species observed the abihty to reject un- wanted particles was clearly of great im- portance. Only two species, Alcyondium polyoum and Alcyondium polypylum be- have completely indiscriminantly in swal- lowing Dunaliella particles. Only when their guts were crammed completely full did they begin to reject particles between the bases of the tentacles. One species, Parasmittina nítida, was observed to swallow not only the Dunaliella particles, but also rough-edged detrital fragments also present in the water obtained from its natural habitat on the undei'sides of rocks in a sand-bottom inter- tidal area. Rejection of small particles com- monly occurred in streams (usually indicat- ing species with a pharyngeal rejection tract), or more or less individually between the bases of all of the tentacles (Fig. 9A). Large particles or masses of debris were FigLire 9. Rejection mechanisms. (A) Polypide commonly rejected by retraction of the of Meinbraiiipora tennis rejecting DiinalieUa par- lophophore (Fig. 9B, C). Some species ticles between the tentacle bases. (B) and (C) could reject both small and large particles Polypide of Membranipoia tennis rejecting a large in "puffs," an action which seemed to in- mass of debris by contraction and retraction of the lophophore. volve ciliary reversais, plus muscular action WINSTON: BRYOZOAN MORPHOLOGY AND FEEDING BEHAVIOR 17 of the pharynx and tentacles, forcing water cates. The zooecia are arranged alternately and the particle or particles to be ejected on the jointed branches and polypides are very forcibly from the vicinity of the lopho- well separated from each other even with the phore. In at least one species (see Sundanella lophophores expanded. The lophophore is sibogae) this form of rejection has appar- small (mean diam = 266 p.m) and equi-ten- ently evolved into a method for concentrat- tacled, and the eight tentacles are usually ing small particles for consumption. A few widely-spread. The mouth is round and tiny, denostóme species (Aeverrillia armata, averaging only 18 ^m in diam. Observations Amathia altérnala, Zoobolryon verliciUa- under the dissecting microscope and by mi- lum) utilized a "rejection action," widening crocinematography showed that Crisia feeds the bell, bending distal ends of tentacles out by almost passively filtering particles, only and then contracting the bell all in extremely occasionally directing them by a slight flick- rapid succession. Like puffing, this action ing of the outer ends of the tentacles. Ex- seemed to clear the lophophore of large, or amination of the polypides with the com- too many small particles. pound microscope showed that the species is admirably equipped to be a filterer, as it Movements of the Mouth and possesses long evenly spaced latero-frontal Pharynx Region cilia, which, when the lophophore is ex- panded, almost fill the space between adja- Though the mouth is always open it is cent tentacles. These latero-frontal cilia are capable of some constriction or dilation. graded in length from the tip to the mouth, Particles may be rejected after being en- and are spaced about 2 ^m apart, forming a gulfed by the mouth via a ventral ciliated re- lattice work which prevents most particles jection tract in the pharynx, or motile or- from escaping. The tiny lophophore creates ganisms, like Dunaliella, may manage to only a slow-moving current, but apparently swim out of the pharynx and away between suitably-sized particles which do come with- the tentacles during the interval between in reach of the filter can be efficiently swallowings to the caecum. In all species trapped. The abundance of Crisia among studied, a pharynxful of particles would be sponge and tunicate colonies indicates that collected before the muscular gulping action it takes advantage also of the currents pro- of the pharynx caused them to be swallowed duced by other organisms in obtaining par- into the caecum (or gizzard in certain ticles to filter. ctenostomes). In a tiny species like An- guinella pálmala, 3-4 Dunaliella particles Pasylhea tulipifera "tentacle feeder" (each 6-JO /j.m diameter) would fill the Figure lOD pharynx, while in large species like Bugula neritina a large bolus of particles would col- Other species, though of similar size and lect before swallowing took place. found in the same hydroid-stem microhab- itat as Crisia elongata, feed in a very differ- Examples of Individual Feeding ent manner. Such species, for example, Behaviors Pasythea tulipifera (Fig. lOD), Chlidonia pyriformis, Synnotum aegyptiacum and Vit- Crisia elongata "filterer" taticella contei, actively "grab" for particles, Figure lOF using the tentacles to roll or toss them into The bushy erect colonies of this cyclo- the mouth, and can hardly be classified as stome species occurred intertidally in Flor- filter feeders. ida on rock surfaces, among the roots of Pasythea tidipifera also possesses a colony large hydroids like Thyroscyphus ram.osus, with erect jointed branches, but in this spe- and seemed to be especially well-developed cies, the autozooids are arranged in triads, among masses of sponges and colonial tuni- so that the middle polypide of the triad faces BULLETIN OF MARINE SCIENCE, VOL. 28, NO. 1, 1978

Figure 10. Examples of individual behavior. A. Scanning activity by a polypide of Bowcrhankia (In- dian River sp.); B. Simdaiiella sihogne using tentacles to concentrate a ball of Dnnalielln particles at the base of the lophophore (arrow); C. Biigiila neritina making lophophore into a cage for the capture of active protists. Ends of tentacles are twisted together to prevent prey from escaping; D. Pasylhea inlipifera using tentacles to roll particles down into the mouth. E. Ciisia elongala polypide in testing position; F. Ciisia elongala polypide in expanded position. WINSTON: BRYOZOAN MORPHOLOGY AND FEEDING BEHAVIOR 19

in one direction and the outer two in the op- polypides scan the water in a circle for par- posite direction (Fig. lOD). The iopho- ticles, bending the flexible introvert. This phore is equi-tentacled, averaging 217 /xm in movement is quite slow (it may take a min- diam, and the ten tentacles, usually rather ute or more for the lophophore to complete widely spread, may be held straight, bent in the circle) but scanning goes on more or less at the tips, or most commonly the upper continuously. In these species adjacent (anal) tentacles curve in while the lower lophophores are quite well separated; they two curve out. Unlike Crisia, in which only may touch each other in scanning, but when the ends of the tentacles flick, tentacle ac- this happens they quickly retract. Occasion- tion in Pasythea ranges from a fast flicking ally two or three adjacent polypides bend of the outer tips to a slower bending in of toward each other to form a cluster which the whole outer half of the tentacle. Ob- produces a stronger current. In Bower- servations of feeding showed that while bankia scanning is a smooth action, but in ciliary currents alone resulted in some par- some other species characterized by scan- ticles reaching the mouth, the action of the ning (e.g. Victoreila, Zoobotryon) rotation tentacles is also important in rolling or push- of the lophophore is accompanied by a ing particles in. Often the anal tentacles dancing, jerky motion of the tentacles. were seen to be curved in just slightly while two to four abanal tentacles were being used Bugula neriiina "cage-captor" to stuEE DunaUella into the mouth. Because Figures IOC, 11 there is a short introvert region protruded At least one species, the cheilostome when the polypide is expanded, each poly- Bugula neriiina, appears to have developed pide is capable also of a certain amount of an adaptation for Zooplankton feeding, vertical scanning, extending the area in forming the tentacles of its lophophore into which it can capture particles. a cage with the tops of the tentacles twisted tightly together, and using this cage to trap Bowerbankia spp. "scanners" active cüiates and other protistans, which Figure lOA are then ingested. Bullivant (1967) saw Species in which the zooids are uniserial Bugula neriiina capture a tintinnid in this or very well separated from each other along manner, but did not believe it could be the stolons are often characterized by an in- normal mode of feeding. However, in my dividualized pattern chiefly composed of observations, this activity occurred con- scanning activity. In ctenostomes of the stantly. In any situation where large par- genus Bowerbankia both body wall and ticles and protistans were present, there were polypide show a great deal of flexibility. always some polypides of the colony making When the polypide is retracted the zooid is cages with their tentacles and feeding in this contracted and often flattened against the manner. While such carnivorous behavior substrate. When the polypide is expanded has not been reported for bryozoans, it has the zooid elongates and rises into a vertical been found for copepod crustaceans for ex- or diagonal position relative to the substrate. ample, that very small species are complete- In the species of Bowerbankia studied the ly herbivorous, but larger ones capture zoo- lophophore is equi-tentacled. The small plankton or other large particles as well as species Bowerbankia gracilis and Bower- filtering nannoplankton (Marshall, 1973). bankia imbrícala hold the tentacles straight, In fact, it has even been suggested (Parsons while on the larger (mean lophophore diam and LeBrasseur, 1970) that the larger spe- = 751 ^.m) polypide of Bowerbankia (In- cies of copepods require large food in ad- dian River species) the tentacles are bent dition to nannoplankton or must have some outward at the tips. All three species studied method of concentrating small food particles were quite similar in feeding behavior. The in order to maintain themselves. 20 BULLETIN OF MARINE SCIENCE, VOL. 28, NO. 1, 1978

Figure 11. Biigula iierilina behavior sequences; AI-A3 One lophophore (top center) forms a cage; B1-B3 One lophophore (top center) goes from expanded, to contracted, to retracted position.

Sundanella sibogae "particle juggler" ticles. In this species the zooids are ar- Figure lOB ranged uniserially and each polypide ap- One of the largest species studied, the pears to function individualistically. The ctenostome Sundanella sibogae, showed an equi-tentacled lophophore averages 820 ¡j.m unusual behavioral adaptation that might in diam, with 31 long slender tentacles, and reflect a need to concentrate small food par- a mouth averaging 82 ¡xm. in size. The poly- WINSTON: BRYOZOAN MORPHOLOGY AND FEEDING BEHAVIOR 21 pide constantly scans back and forth, bend- Tatle 2. Species grouped according to type of ing and moving the whole introvert and the colonial behavior pattern upper half of the flexible zooecium to com- plete its rotations, It is characterized by a Colonial Behavior Pattern/Colony very slow flicking of the tentacles which Type T Individual Dominant Currents bend in to about half their length, then Aeverrillia ármala 1 W slowly bend back out. When fed high con- Beania iiilei inedia 1 W centrations of Dunaliella, polypides of Sun- Bowerbankia gracilis 1 W danella used this tentacle flicking combined Bowerhankia (Indian River ;ipecies) 1 W with ciliary action to concentrate the par- Bowerbankia imbricóla 1 W Siindanella sibogae 1 W ticles into a large ball which tumbled about Zoobotryon verliciltaliiin 1,2 M in the region just above the mouth. This ball Amailiia altérnala 1,2 W of particles was rapidly ejected above the Nolella slipala 1,3 W lophophore, then sucked back down above Terebripora sp. 1,3 W the mouth, the juggling action being re- Viclorella pavida 1,3 W peated several times until the polypide fi- Type 11 Separated by Colony Structure nally withdrew back into the zooid, taking 2 the food bolus with it. Angiiinella pálmala W Biigida nerilina 2 S Biigitia sloloiiijeia 2 S Patterns of Colony Behavior Biigiila inrrila 2 S Even more striking than these Individual Candti simplex 2 s behaviors are the colony-wide adaptations Caidibugula dendrograpia 2 s which marine ectoprocts have evolved in Caidihngida pearsei 2 M Chlidonia pyriformis 2 W order to separate, increase and channel feed- Crisia elongnla 2 VV ing currents. Species can be grouped in Crisia (Pacific sp.) 2 W various levels of integration based on the Pasylhea lidipifera 2 M extent to which the individual has become Scriipocellaria regular is 2 S Syniioliim aegypliaciiin 2 W subordinated to the colony and the extent Telraplaria dicholoma 2 M to which individuals of the colony act to- Villalicella conlei 2 W gether to produce an effect beneficial to all. Margarella buski 2,3 M These categories are not mutually exclu- Releporellina evelinae 2,7 S sive and are meant to describe functional Type III Temporaiy Clustering Dominant differences in behavior, although there are taxonomic and phylogenetic implications as Alcyonidiimi polyoum 3 M well. Table 2 shows 55 species of marine AIcyonidiiiin polypyluni 3 S Beania liirlissima 3 M ectoprocts grouped according to type of Conopeiim seurali 3 S colony behavior pattern. Conopeimi lenuissimiim 3 S Eleclra bellida 3 M Colonies in Which the Individual is Membranipora arborescens 3 S Dominant Membranipora lemiis 3 S Thalamoporella falcifera 3 W Figure 12 Walersipora subovoidea 3 M Eleven of the species for which the col- Sleganoporella magnilabris 3,4 M ony behavior pattern was studied fell into Crassimarginalella lincla 3,5 S a group in which the individual zooid ap- Type ÏV Permanent Cluster, peared dominant and multidirectional Non-Skeletal scanning behavior predominated. Poly- Hippoporina verrilli 4 S pides in this type of colony were completely Membranipora luberculala 4 S or partially separated from each other and Parasmiltina nitida 4 S there was little or no interaction of feeding Schizoporella jloridana 4 s 22 BULLETIN OF MARINE SCIENCE, VOL. 28, NO. 1, 1978

Table 2. (Continued) duced only weak colony currents, it is prob-

Colonial Behavior ably better characterized by Bugula- or Type V. Permanent Cluster, Pattern/Colony Reteporellina-type colonies in which the Currents Irregular-Skeletal lophophore-covered branches draw a strong "Rhyncliozoon" sp. 4,5 • current of water through the meshwork of Trematooecia turrita 4,5 S Celleporina hassaUi 5 M the branches. Microcinematographic and Trematooecia aviciilifera 5,7 S dye studies of Bugula turrita colonies Type VI. Permanent Cluster, Regular-Skeletal showed clearly the current being drawn Gemelliporidra multilamellosa 6 S down through the spiral of the colony, im- Lichenopora buskiana 6 S pinging with great force on the sides of the Lichenopora inlricala 6 S branches where the lophophores protruded "Plagioecia" sp. 6 S and drifting gently around the non-polypide "Tuhulipora" sp. 6 M bearing outer surface of the branches. Aside Cellopornria albiroslris 3,0,7 S Discopoiella umbellala depressa 3,6,7 S from flicking of the tentacles the polypides in this species change their orientation very little. The lophophores are all obliquely currents among individuals in the same truncate, curving out slightly at the tips, and colony. Ten of the species in which this type are moderate in size (mean lophophore of behavior occurred were ctenostomes, but diam = 393 ¡xm, 14 tentacles). Their most this type of pattern can also occur in characteristic actions are forming a cage by cheilostomes {e.g. Beania intermedia). The putting the tips of the tentacles together currents produced by such colonies are gen- around a bolus of particles, and rejection of erally weak, although the currents produced particles by avoidance retractions (clearing by individual polypides may be quite strong, the meshwork of large particles) and the depending on the size of the polypide. In puffing away of large concentrations of small polypides of this type of colony the introvert particles which have accumulated in the is usually medium to long as figure lOA mouth area but cannot be swallowed be- shows. cause of a full pharynx. Colonies in Which Polypides are Separated Colonies Characterized by the Formation and Their Orientation is Controlled by the of Temporary Clusters of Polypides Skeletal Slructure of the Colony Figure 14A Figures 11, 13 In the third category (12 species) zooids In the second group (12 species) the are arranged in contiguous sheets and poly- colony appeared to dominate the individual. pides are capable of forming temporary clus- Individual polypides were separated to a ters of lophophores which may serve to in- greater or lesser degree. Polypide orienta- crease and direct feeding currents. This type tion was controlled by colony structure and of behavior is found in both encrusting only a limited range of polypide motion was ctenostomes and cheilostomes. The two possible. Within this framework polypide species of Conopeum studied (Conopeum size ranged from small to large, individual tenuissimum and Conopeum seurati) serve behavior from simple to complex, and colony as excellent examples of this type of be- current production from weak to strong. havior pattern. In Conopeum the polypides These colonies occurred in all three orders show a constant twitching activity, rapidly of marine ectoprocts and the type of be- flicking the outer third of the tentacles in havior appeared to be a functional conse- and out, while at the same time slightly ex- quence of erect branching colony form. panding and contracting the lophophore Though this group contained tentacle feed- bell. The polypides do a lot of popping in ers like Pasythea and Viitaticella which pro- and out of the zooids and when expanded WINSTON: BRYOZOAN MORPHOLOGY AND FEEDING BEHAVIOR 23

SOOpm t I Figure 12. Example of a Type I colony: Bowerhankia sp. (left) growing along branch of the erect colony of Margarella biiski (large polypide on right).

they form clusters in which three or four strong, apparently an advantage of this type lophophores are oriented toward each other of behavior. (Fig. 14A). In doing so they may lie with Colonies in Which Polypides Form Fixed the long introvert stretched out parallel to Clusters, but Without Reflection by the colony surface, and the outer tips of the Skeletal Morphology tentacles of some polypides may even be touching the surface of the colony. When Figure 14B-E the polypides first emerge from the zooid In other encrusting forms (four species) they scan in all directions and orient toward zooids were contiguous, and polypide orien- the nearest polypides which are also ex- tation was fixed, but was not reflected in zo- panded in order to make the clusters, but oecial (skeletal) morphology. In these spe- the associations change as one polypide re- cies, like the Membranipora membranácea tracts and is replaced by another in a slightly described by Banta, McKinney and Zimmer different position. The currents produced (1974), polypides are organized onto fixed by the clusters of polypides are stronger than clusters or incurrent cells, with excurrent those that could be produced by individual chimneys between them. In small circular polypides alone, and flow of water and tur- colonies of Hippoporina verrilli, for ex- bulent reworl

500 H "1 I I

Figure 13. Examples of Typs II colonies: A. Ciiilibiigiila dcndrograpw branch, showing obliquely truncate lophophores; B. Releporellina eveliiuie, showing orientation of campylonemidan lophophores into the space between two branches. WINSTON: BRYOZOAN MORPHOLOGY AND FEEDING BEHAVIOR 25

Figure 14. Examples of Type IIL IV and V colonies: A. Type III. Polypides of Memhraniporn len- uis forming temporary clusters; B. Type IV. Incurrent cell of Hippoporina verriUi colony; C. Type IV. Small colony of Scliizoporella floridana forming one incurrent cell; D. Incurrent cell (side view) of Membranipora niberculala; E. Incurrent cell and excurrent chimney in Sleganopoiella magnilabris (Type IV); F. Incurrent cell in Tremalooecia civiciiUfera (Type V). 26 BULLETIN OF MARINE SCIENCE, VOL. 28, NO. 1, Í978 current cell are equi-tentacled and the loph- current cells as in IV, but the formation of ophore is held straight up, in outer rows the these cells is enhanced by the skeletal form introverts bend toward the center of the cell, of the colony itself, with raised and sunken and the lophophores become more and more areas which also channel the current flow. obliquely truncate, so that in side view the One species found in Florida, Celleporina cup-shaped incurrent unit is clearly defined. hassalli, forms nodular colonies around the The most obliquely truncate lophophores bases of Thyroscyphus stems. In these col- are found on polypides bordering on and onies frontal budding of zooids and the bent away from the excurrent chimneys. curvature of the colony around the hydroid Small unilaminate colonies of Schizo- stem gives a bumpy irregular surface to the porella floridana (Fig. 14C) showed a simi- colonies. The skeleton and the orientation lar behavior pattern, though polypides of of the polypides (with both equi-tentacled this species are more active, showing a slight and obliquely truncate lophophores) to- amount of lateral scanning, and cage making gether form well-defined current cells. activity, using the cages to capture balls of Another species, Trematooecia aviculifera Dunaliella ejected by puffing rejections from (Fig. 14F) occurring on dead coral rubble adjacent lophophores. Polypides of Schi- in shallow water at Galeta Reef, Panama, zoporella also are capable of "writhing"• has colonies which form thick masses with pulling back part-way into the zooid, then raised and hollow areas, often concentrically lashing the tentacles around. arranged in circular colonies, but not always Membranipora luberculata, commonly regular in nature. Due to the frontal bud- found on floating sargasso weed, also shows ding of zooids, the zooecia may be oriented a type IV behavior pattern, though the poly- in various directions, but the polypides al- pides in this species seem much more active, ways form functional clusters. Figure 15A popping in and out of the zooids, and flick- shows the arrangement observed in one col- ing one or more tentacles to give the effect ony, in the very center of the colony no of constant flickering motion. In places lophophores are expanded and an exhalent where a group of polypides has degenerated current occurs; in the slope of the adjacent or the sargasso leaf is irregular, polypides furrow the lophophores tilt out, those at the may be influenced in their orientation; but bottom of the furrow are held straight up, except for these irregularities, unless some- those on the side of the next ridge tilt to- thing disturbs a large portion of the colony, ward the center again so that current chan- all polypides are usually expanded so the nels are created in the furrows. The lopho- colony surface shows a mass of lophophores phores in this species are large (mean all with tentacles just about touching. In lophophore diam = 757 p.m, 18 tentacles) side view (Fig. 14D) it can be seen that and obliquely truncate, and while they show those at the edges of incurrent cells (bor- some range of tentacle activity and were ca- dering excurrent chimneys) may have pable of cage-making behavior, they do not slightly longer anal tentacles and the scal- seem to scan at all. loped profiles of the incurrent cells are ob- vious. Colonies in Which Polypides Form Permanent Clusters Enhanced by a Colonies in Which the Polypides Form Regular Patterning of the Zoarium Fixed Clusters, and the Clusters are This type of pattern was observed in both Enhanced by an Irregular Patterning of cyclostomes and cheilostomes. Among the the Colony Skeleton cyclostomes Lichenopora buskiana offered Figure 14E&F one of the simplest examples of beautifully In the fifth kind of colony (five species functional behavior based on skeletal pat- studied) polypide orientation is fixed to form terning. The tiny colonies occurred on red WINSTON: BRYOZOA» MORPHOLOGY AND FEEDING BEHAVIOR 27

algae and the undersurfaces of corals at Ca- this position the concave polypide-covered leta Reef. Each colony consisted of several surface faced downward and SCUBA ob- radiating rows of zooids with channels servations showed the very large (mean formed by alveoli (non-zooidal areas of small lophophore diam = 850 /¿m, 24 tentacles) coelomic spaces surrounded by calcareous polypides were oriented into the channels walls) between them and a large hollow area between the regularly spaced mounds in the center of the colony formed by the formed by frontal budding of the skeleton. roof of the brood chamber. Dye studies showed that water currents entered at the Species in Which Polypides Have Some base of the colony and exited in little puffs Form of Group Actions from the excurrent chimney formed by the Several species with various types of be- hollow colony center. The presence of a havior patterns also appeared to show united lighter green color in the colony center also activities of several polypides at once that seemed to indicate that some reworking of indicated an even greater degree of behav- the water was taking place. ioral (neurophysiological) integration. Another striking example of regular skel- Some of these group actions appear to etal patterning was observed in Lichenopora be associated with rejection. In Reteporel- intricata. In this species, obtained both in- lina evelinae it was noted that group retrac- tertidally and subtidally in Panama Harbor, tions occurred when it was necessary to the complex colony consists of subcolonies clear the space between the branches of a with zooecia arranged in rows. In each row large particle. This is probably one of the the zooids were progressively higher toward simplest types of unified activity. the center of the colony. Between the radi- Celleporaria albirostris forms brilliant ating lobes filled with zooid tube rows are cerise incrustations on the undersides of channels floored by alveoli, and each sub- foliaceous corals at Caleta. The skeleton of colony has a hollow center. The polypides the colony is molded in a regular series of are densely packed, one above the other, knobs and channels. The polypides are ori- each has the two bottom tentacles spread ented along the channels so that these form apart and curved away, and the rest curving the incurrent or through-current pathways, around in a scoop shape. The lophophores and the knobs mark the excurrent channels. of the raised zooid rows orient toward each Long spines formed by the zooecial skele- other to form "food grooves" which move tons project upward among the introverts the particles along. No polypides orient on of the expanded polypides, but do not ex- the channels that lead into the colony or to- tend above the lophophores so their func- ward the excurrent siphon formed by the tion in creating turbulence or channelling depression in the subcolony center. A par- water is uncertain. The system for channel- ticle caught between the rows of zooecia in ling water currents is regularly structured, a channel gets carried out to the excurrent and the greater length of the introvert makes siphon. It might be expected that the chim- scanning and changing polypide orientation ney would be efficiently puffing water away, to the temporary cluster type of behavior but in the laboratory the effect of the chim- possible when this is advantageous (as when neys is to create a turbulence that causes not all polypides are expanded). There also water to be reworked by the colony. seems to be a trend toward unified action Massive corallophyllic cheilostomes also by polypides in which a whole group of showed type VI patterning. The large (up polypides can be seen to undergo an avoid- to several cm in diam) bracket shaped col- ance retraction with writhing of the ten- onies of Gemelliporidra multilamellosa were tacles. Since there was at no time a preda- found attached to the sides of ledges and tor or large particles observable that would under large coral heads at Galeta Reef. In have disturbed them, it seems possible that 28 BULLETIN OF MARINE SCIENCE, VOL. 28, NO. ), 1978 this activity might somehow aid in process- an anal-abanal (adneural-abneural) direc- ing water and moving it through the colony. tion, and therefore, bilaterally symmetrical. In addition, several times a group of about This shape develops in only some polypides six polypides was observed waving or bend- in colonies that are organized to form cur- ing the lophophores in the same direction rent-enhancing cells. In colonies where the at once. Such action could also be im- pattern of skeletal growth positions poly- portant in directing current flow. pides so as to cause the maximum undirec- Other species also showed unified poly- tional current flow, the lophophores of all pide activity. For example, in the lunuliti- the polypides are often slightly obliquely form cheilostome Discoporella umbeUala truncate. depressa which has a highly integrated type In some cheilostomes the large and many- VI colony pattern, the polypides can be ob- tentacled lophophore may be more elabo- served to execute a rapid action in which rate, having the ends of some of the tentacles the two abanal tentacles move sharply out- shortened and bent more sharply outwards ward, and then return to the normal expan- so that a scalloped edge is apparent. sion. This action could perhaps serve to The strongest trend toward bilateral loph- push the current toward the upper (or low- ophore symmetry is shown by species with er when the colony is convex surface down) bent-tentacled or campylonemidan lopho- rows of zooids. phores. This development for the directing of afferent and efferent currents has devel- DISCUSSION oped in all three marine bryozoan orders. Probably the most important implication In colonies with individualized behavior pat- of this work is that there is a greater range terns (e.g. Aeverrillia, Valkeria, Viclorella), in polypide morphology and a greater vari- it has apparently developed to aid the in- ety of behavioral activity in marine ecto- dividual in processing water in very low procts than was previously suspected, with current situations, e.g. among hydroid stems, similar behavioral and associated morpho- or in sheltered estuarine habitats. In other logical features occurring (apparently in- species, e.g. Reteporellina and Lichenopora, dependently) in more than one order. it is linked with the development of a strong- Although the ectoproct lophophore may ly unidirectional colonial current flow. In be simple in comparison with those of still other forms (Discoporella) a slightly phoronids and brachiopods, there has ap- campylonemidan lophophore seems linked parently been a strong trend in subsequent with a behavioral action (widening of the evolution toward development of more com- ventral tentacles) that serves to distribute plex lophophores. The presumption is that feeding currents regularly over the colony. the original lophophore structure was, as Thus, variation in lophophore shape ap- it is today in many small forms (including pears to have a functional significance. The stoloniferous ctenostomes and cyclostomes), presence of such variation within the struc- of the equi-tentacled type. This grade of tures imposed by the microscopic size of construction has developed in both small individual marine gymnolaemate and steno- and large species. In the large species the laemate bryozoans, indicates that the loph- ends of the tentacles may be held stiffly, ophore has undergone a radiation in size so that the lophophore is still a cup-shape, and shape. This radiation has been asso- or the free ends may be flexed in and then ciated with behavioral developments and out, giving the lophophore a campanulate with increase of colony coordination in chan- shape. nelling feeding currents. The development In the obliquely truncate grade of con- and radiation of tentacle sensory structures struction the lophophore has lost its radial may parallel the trends in lophophore de- symmetry, becoming obliquely truncated in velopment, i.e. one would expect more WINSTON: BRYOZOAN MORPHOLOGY AND FEEDING BEHAVIOR 29 sophisticated sensory structures, especially Trends in Individual Behavior abfrontal structures, in those forms which In general larger size seemed indicative demonstrate the highest degree of behavioral of more complex tentacle flicking activity• integration, but the comparative morphol- e.g. larger polypides exhibited a greater ogy of sensory structures is still too little range of reactions, moving tentacles fast to known for speculation. slow, or making movements involving vari- Measurements on lophophore dimensions (of species from several localities and all ous lengths of the tentacle in connection three marine orders) has also indicated the with maneuvers for the ingestion or rejec- amount of variation that occurs and what tion of particles. However, among small its significance might be. Ryland (1975) polypides, only those of the cheilostomes analyzed lophophore parameters in a single seemed to be capable of complex actions. community of marine bryozoans from the In the small cheilostome species the action New Zealand intertidal. His study showed of the tentacles in batting or rolling par- that while lophophore dimension varied ticles into the mouth played an important from species to species, dominance of a role in feeding whereas in small ctenostomes very few species ensured that most lopho- and cyclostomes ciliary activity was the most phores in the community were in a much important component of feeding behavior. more restricted range. Though Ryland did The function of some of the individual not measure mouth size directly he con- behavioral action is still not known but the cluded that the restricted size range of most significance of others is obvious. For exam- lophophores in the community indicated ple, many species attempt to put the distal that most species were utilizing a common ends of the tentacles together to form a food resource, consisting of particles with- cage. But only in the larger species, par- in a similarly restricted size range. My ticularly Bugula neritina, does this seem to study, on the other hand, has considered have evolved into a specialized method of the lophophore dimensions of as many bry- capturing prey. Small species could not get ozoans as possible from a variety of habitats the ends of the tentacles completely together in tropical and subtropical localities in an or twisted in place, and with only a small attempt to discover the range of variation number of tentacles there were large gaps possible and the relationships between poly- between the tentacles through which active pide size and feeding activity. The range of prey could escape. variation found, in fact, is similar to that Some widespread and elaborate individual found by Ryland in his New Zealand pop- activities seemed to be associated with the ulation (in my samples tentacle number rejection of particles. Some species showed ranged from 8-31, tentacle length from a special rejection action; in others there J 24-859 iJ.m and lophophore diam from was an avoidance retraction with the loph- 187-1,012 ixm, versus 8-25, 200-900 ¡xm, ophore being pulled partway back in the and 250-1,000 ^m for bryozoans from zooid to avoid big particles of debris. Dif- Echinoderm Reef Flat). But, I have em- ferent species could reject particles in phasized that this range in lophophore di- streams from the rejection tract, puffs, or mensions means that the largest species has between the bases of the tentacles, and some a much greater size range of particles avail- could utilize a combination of methods de- able to it (up to 90 ;am) than does the pending on the size and concentration of the smallest (less than 20 /xm), and that the particles. size differentia!, in conjunction with the be- havioral differences observed, indicates con- Colony Behavior siderable potential for partial separation of Of the species studied the ctenostome food resources among marine bryozoans. species appeared most limited in the ex- 30 liUl.LEI IN OF MARINE SCIENCE, VOL. 28, NO. 1, 1978

Ifnensjfication of individual of ihe colony the problem is to create some turbulence

ZOOID LINKS weak Slrcxig and movement of the water to begin with. ^ In species living in high-energy environ- ^ 1 ra Rower boinkia sDO Alcvoiirikini SPP LichfinODO'a S|>p ments the problem is to change the flow in •k Vidorella navida Conopeum spp the immediate vicinity of the colony from SÏ laminar to turbulent. The whirlpool pat- ^ Cellfiporina hiiSMlli BLfqiiki spp terns set up by colonies observed in the i !S Trfiniainoerid lutma laboratory that appear to be reworking of 0 i % the water, are apparently necessary in nature CelleporarJa altxrosins '=L Rfîtepoiiîllirw evfrlii\n« Trpniati-xier.la a\'tculifero to get food particles out of the macro-cur- ? DrscoiXJfclla Limhellaui Ü 2 depress,, rents into a microenvironment from which i the animals can extract them. Figure 15. Levels of morphological and behav- In the most highly integrated colonies ioral integration (with respect to feeding and cur- studied, joint actions of groups of polypides rent producing activities). aided in rejection of particles and process- ing of water currents. These colonies ap- pression of colony-wide behavior patterns peared to be approaching by various means (chiefly type I and type II), though in- the stage of integration described by Bek- dividual behavior could be quite complex lemishev (1970) in which "cormidia" or (e.g. Sundanella). Carnose ctenostomes colonial organs are composed of several and several species of encrusting cheilo- individuals. According to Beklemishev col- stomes showed temporary clustering of onies can become more integrated in three zooids, apparently a trend for increasing ways, by weakening of zooid individuality, current flow (and enhancing turbulence) in by intensification of colony individuality, colonies where zooids are contiguous. and by the development of cormidia for the With massive encrusting forms the trend fulfillment of colony functions. With re- is toward increasing the efficiency in chan- spect to feeding behavior and manipulation nelling the water through the colony, and of feeding currents various types of bryo- by creating more turbulence, perhaps bring- zoan colonies can perhaps be considered as ing a greater number of particles in. This a matrix, in which the degree of integration can be accomplished in several ways: by increases by the increasing linkage between orientation by the polypides alone, by ori- zooids (in the sense of closeness of spacing, entation of polypides and formation of contiguity, and regularity of patterning, as rough, skeletal clusters, and by more regu- all bryozoans are connected, anyway), and larly patterned arrangements of mounds by increasing of linkage between polypides, and channels, carried to the extreme in as shown by increasing colony control of something like Lichenopora intricata. polypide orientation (scanning = weak; The ability of these organisms to manip- temporary = moderate; fixed = strong), and ulate water currents is one of the most strik- by the increasing degree of behavioral in- ing observations of the study. The function tegration as indicated by unified activities of many colony patterns and behavioral ac- of the polypides. Such a matrix (Figure tivities appears to be to create turbulent 15), while no doubt too simplified to cover water flow. In order to separate food par- all cases, does give a picture of the position ticles from the water column all these ani- of different types of colonies studied. mals have to have some method of chang- ing water flow as it passes over the colony. ACKNOWLEDGMENTS In species living in very still water environ- In many respects I am indebted to 19th century ments (e.g. sea-grass beds, inside dead mol- zoophytologists, for this work is based on the Jusk shells, and in other cryptic habitats), classic technique observation of living animals. WINSTON: BRYOrOAN MORPHOLOGY AND FEEDING BEHAVIOR 31

It is the kind of research few 20th century people Polyzoa. Van Voorst. London. 1; 601 pp., have the necessary uninterrupted time for, and I and 2; 83 pi. am most grateful to the Smithsonian Institution Hyman, L. H. 1959. The invertebrates. V. for the support that allowed me this opportunity. 5. McGraw-Hill. New York. 783 pp. The research was carried out during the tenure of Jebram, D. 1973. Preliminary observations on a Smithsonian Postdoctoral Fellowship, at the the influences of food and other factors on the Fort Pierce Bureau, Fort Pierce, Florida, and a growth of Bryozoa. Kieler Meeresforsch. Travel Fellowship, at the Smithsonian Tropical 29; 50-57. Research Institute, Galeta Laboratory, Panama. Kraepelin, K. 1887. Die deutschen Susswasser- I would like to express my gratitude to all people bryozoen. I. Anatomisch-systematischer Teil, at those institutions who helped me with collect- Abhandl. Geb. Naturw., Hrsg. v. Naturw. ing, identifications and encouragement, especially Ver. Hamburg 10: 1-168. R. S. Boardman and A. H. Cheetham (USNM), Lister, J. J. 1834. Some observations on the K. J. Eckelbarger, M. E. Rice and M. A. Middle- structure and functions of tubular and cellu- ton (HBF), and G. Hendler (STRI). Thanks also lar polypi, and of ascidiae. Phil. Trans. go to T. I. M. Schopf and R. M. WooHacott for Roy. Soc. London 124: 365-388. reading and criticizing drafts of the manuscript. Loppens, K. 1908. Les bryozoaires d'eau douce. Contribution no. 64 from the Harbor Branch Ann. Biol. Lacustre 3; 141-188. Foundation, Inc., Fort Pierce, Florida. Lutaud, G. 1955. Sur la ciliature du tentacule chez les bryozaires chilostomes. Arch. Zool. Exp. Gen. 92; 13-19. LITERATURE CITED . 1973. L'innervation du lophophore chez le bryozoaire chilostome Eleclia pilosa (L. ). Banta, W. C, F. K. McKinney, and R. L. Zimmer. Z. Zellforsch. 140; 217-234. 1974. Bryozoan monticules: excurrent water Mangum, C. P., and T. J. M. Schopf. 1967. Is outlets? Science 185; 783-784. an ectoproct possible? Nature 213: 264- Beklemishev, W. N. 1970. Principles of Com- 266. parative Anatomy of Invertebrates (Transi. Marshall, S. M. 1973. Respiration and feeding 3d Russian ed., Moscow, 1964, by J. M. in copepods. Adv. Mai'. Biol. 11: 57-120. MacLennan; Z. Kabata, ed.) University of Parsons, T. R., and R. I. LeBrasseur. 1970. The Chicago Press, Chicago. 1: 490 pp., and 2: availability of food to different trophic levels 532 pp. in the marine food chain. Pages 325-343 in Borg, F. 1926. Studies on recent cyclostomatous J. H. Steele, ed. Marine Food Chains. Uni- Bryozoa. Zool. Bidrag. Uppsala 10: 181- versity of California Press. Berkeley. 551 pp. 507. Ryland, I. S. 1975. Parameters of the lopho- Bullivant, I. S. 1967. Aspects of feeding of the phore in relation to population structure in a bryozoan Zooholryoii veniciUaliim (Delle bryozoan community. Pioc. 9th Europ. Mar. Chiaje). Ph.D. Thesis. University of South- Biol. Symp.: 363-393. ern California. 165 pp. Silén, L. 1944. On the division and movements . 1968. The method of feeding of loph- of the alimentary canal of the Bryozoa. Ark. ophorates (Bryozoa, Phoronida, Brachiopoda). Zool. 35A; 1-41. N. Z. I. Mar. Freshwat. Res. 2: 135-146. Smith, L. W. 1973. Ultrastructure of the ten- Dudley, J. E. 1973. Observations on the re- tacles of Fliislrellidra liispida (Fabricius). production, early larval development, and Pages 335-343 in G. P. Larwood, ed. Living colony astogeny of Conopeiiin lenuissimiim and Fossil Bryozoa. Academic Press. New (Canu). Chesapeake Sei. 14; 270-278. York. 634 pp. Farre, A. 1837. Observations on the minute Strathmann, R. R. 1973. Function of lateral structure of some of the higher forms of cilia in suspension feeding of lophophorates polypi, with view of a more natural arrange- (Brachiopoda, Phoronida, Ectoprocta). Mar. ment of the class. Phil. Trans. Roy. Soc. Biol. 23; 129-136. London 127: 387-426. Winston, I. E. In press. "Feeding," in R. M. Gordon, D. P. 1974. Microarchitecture and WooHacott and R. L. Zimmer, eds. The Biol- function of the lophophore in the bryozoan ogy of Bryozoans. Academic Press. New York. Cryptosula patlasianci. Mar. Bio). 27: 147- 163. DATE ACCEPTED: January 28, 1977. Grant, R. E. 1827. Observations on the struc- ture and nature of Fkislrae. Edinburgh New ADDRESS: Department of EartI) and Pianelnry Sci- Philos. I. 3; 107-342. ences, Tlie Johns Hopkins University, Baltimore, Hincks, T. 1880. A history of the British marine Maryland 21218.