Marine Biology 80, 63-74 (1984)

Marine ............... = = BiOlOgy 9 Springer-Verlag 1984

The reproductive biology of echinothuriid and cidarid sea urchins from the deep sea (Rockall Trough, North-East Atlantic Ocean) P. A. Tyler 1 and J. D. Gage 2 Department of Oceanography, University College; Singleton Park, Swansea SA2 8PP, Glamorgan, South Wales, UK 2 Scottish Marine Biological Association; P.O. Box 3, Oban PA34 4AD, Argyll, Scotland, UK

Abstract

The reproductive biology of 5 species of echinothuriid (Phormosoma placenta, Calveriosoma hystrix, Araeosoma fenestrum, Sperosoma grimaldii and Hygrosoma petersii) and 2 species of cidarid (Cidaris cidaris and Poriocidaris purpurata) sea urchins from the deep sea (Rockall Trough) has been examined from samples collected during 1973-1983. In all species the gonads lie within the interambulacrum attached to aboral gonopores and when fully developed occupy most of the test not occupied by the gut or Aristotle's lantern. In all the species, initial oocyte development takes place along the germinal epithelium embedded in nutritive tissue. In all the echinothuriids and in Poriocidaris purpurata, the oocyte grows to ca. 200 to 450 ktm, at which stage vitellogenesis begins. Oocyte growth continues until a maximum egg size of 1 100 to 1 500 ~m is attained. In the echinothuriids, two types of nutritive tissue are found. In the early stages of gametogenesis the oocyte is surrounded by well-structured periodic acid Schiff (PAS)positive tissue. As the oocyte grows this tissue becomes vacuolated, suggesting that there is a transfer of nutriment to the developing oocyte. In Phormosoma placenta, unspawned oocytes are phagocytosed. There is no evidence of seasonality in any of the echinothuriid species or in Poriocidaris purpurata, Extrapolation with shallow-water echinothuriids suggests that larval development is lecithotrophic, omitting any planktotrophi c phase. Of the species examined, only Cidaris cidaris has a reproductive strategy which produces a known larva, although the limited samples did not permit any determination of seasonality in this deep-sea population.

Introduction

The soft and flexible body form of the family Echinothuriidae is atypical of sea urchins. Echinothuriids are

found mainly in the deep sea, although littoral forms are common in the Indo-Pacific Ocean (Mortensen, 1940). In the North-East Atlantic Ocean this family is represented by five species which have a depth range varying from ca. 600 to 3 000 m (Gage et al., in press). By contrast, sea urchins of the family Cidaridae have a solid sub-spherical test supporting a battery of large rigid spines. In the North-East Atlantic this family is represented by three species. Of these, Cidaris cidaris and Poriocidarispurpurata are slope species found down to 1 800 m, although the former may be found as shallow as 50 m (Mortensen, 1927). Compared to our knowledge of the reproductive biology of most sea-urchin families, our knowledge of reproduction in the families Echinothuriidae and Cidaridae is limited to occasional observations on deep-sea specimens (Mortensen, 1927; Ahlfeld, 1977), or autecological studies of shallow-water echinothuriids (Mori et al., 1980) and cidarids (Holland, 1967; Williams and Anderson, 1975; Dix, 1977). In all the echinothuriid species studied to date, an egg > 1 000ktm diam is produced (Mortensen, 1927; Mori etal., 1980). This led Mortensen (1927) to predict that these forms have "direct development without a pelagic stage". Amemiya and Tsuchiya (1979) followed the development of the shallow-water Asthenosoma ijimai and confirmed direct lecithotrophic development for this species and the formation of a "biscuit-shaped late blastula" not previously described for any echinoid. Analysis of the reproductive biology of shallow-water cidarids shows a variety of reproductive strategies. Some species produce small (
64

P.A. Tyler and J. D. Gage: Reproduction of echinothuriid and cidarid sea urchins

Observations of living echinothuriids recently have been made from the submersible "Alvin" (Grassle et al., 1975; Pawson, 1982). Grassle etal. showed Phormosoma placenta to be strongly aggregated. This pattern was attributed to their slow-moving deposit-feeding mode of life that was likened by Grassle et al. to the herds of buffalo once found on the grassy plains of North America. Mortensen (1938) has shown that certain deep-sea echinothuriids can feed on plant material of terrestrial origin, whilst others in the same area contain locally derived sedimentary material. Pawson (1982) found only mucusconsolidated mud balls in the guts of P. placenta, whilst fragments ofSargassum sp(p?), were found almost exclusively in the guts of Hygrosorna petersii. This paper presents the results of an analysis of the reproductive biology of 5 species of echinothuriid and 2 species of cidarid sea urchins collected by the Scottish Marine Biological Association in the Rockall Trough, North-East Atlantic Ocean.

Materials and methods

Samples of the echinothuriid sea urchins Phormosoma placenta, Calveriosoma hystrix, Araeosoma fenestrurn, Sperosoma grimaldii and Hygrosoma petersii, and the cidarid sea urchins Cidaris cidaris and Poriocidaris purpurata, were collected during 1973-1983 from the Rockall Trough, North-East Atlantic Ocean, in the course of a deep-sea sampling programme conducted by the Scottish Marine Biological Association (Gage et al., 1980, and in press; Gage and Tyler, 1982). Of these species, only Phormosoma placenta was collected regularly enough to permit an analysis of seasonality in oogenesis from any single station (Table 1). The other species were collected sporadically in small numbers from various stations (Tables 2 and 3). Specimens were collected from benthic hauls made with an epibenthic sledge, Agassiz trawl, single-warp otter trawl or Granton otter trawl. Specimens were sorted on deck and fixed in 8% seawater formalin and were later transferred to 70% alcohol for storage. Before dissection, the test diameter of the adults was recorded. We refer to this measurement in the echinothuriids as the "collapsed test diameter" because the imbricated plates of the test in all species examined allow the specimens to collapse under their own weight in air, thus giving a diameter slightly greater than that in life. For the echinothuriid species, the test was then cut along the ambitus and a randomly selected gonad was removed. In the cidarids, the Aristotle's lantern was removed by cutting round the peristomial membrane and the gonads were removed through this aperture. All gonads were processed through graded alcohols, cleared in xylene, embedded in paraffin wax and sectioned at 7 pm. Sections were stained with haematoxylin and eosin, toluidine blue, periodic acid Schiff (PAS), or methyl green pyronin.

Results

Family Echinothuriidae

Size at first reproduction In reproductive studies and analysis of population structure, it is important to determine at what size an animal reaches sexual maturity. In Phormosoma placenta (Table 4), the smallest specimen possessing developing gonads was 45 mm in diameter, although three specimens between 47 and 55 mm diam showed no evidence of reproductive development. All specimens > 59 mm diam were found to be at some stage of reproduction. In the other four species, individuals grow much larger before reaching sexual maturity: in Hygrosoma petersii, gonads do not occur in specimens of a test diameter <75 mm and i n the other three species gonads appear only after a test diameter of > 80 mm is reached (Table 4). In all five species, the largest adult collected was considerably smaller than the known maximum adult size (Mortensen, 1927).

Gross morphology The gonads of all species examined lie within the interambulacrum, attached to aboral gonopores. In newly developing specimens each gonad consists of thin strands near the apex of the test and, as the gonad develops, it grows orally along the inside of the test, attached to the coelomic epithelium by mesenterial strands. When fully developed the five gonads occupy a large proportion of the test volume. All five species examined were gonochoric and there was an equal proportion of each sex. Gonad morphology and gametogenic biology appeared to be similar in all five echinothuriid species examined.

Histological structure of the ovary The outer wall of the ovary consists of cuboidal coelomic epithelium (Fig. 1 A). The wall consists of an inner and outer coelomic sinus and connective tissue, although these structures are difficult to separate by light microscopy. The inner surface of the wall is lined with germinal epithelium from which the germ cells arise (Fig. 1 A). Immediately inside the ovary is a mass of "nutritive tissue" which appears to contain both nutritive and phagocytic cells. Embedded within this nutritive tissue are developing oocytes ranging from oogonia to fully grown primary oocytes (Fig. 1 D).

Oocyte development The oogenic cycle begins with the proliferation of small clusters of oogonia close to the germinal epithelium

Table 1 . Phormosa placenta. Positions o f collection stations a n d sex ratio o f this e c h i n o t h u r i i d sea u r c h i n ( i m - - i m m a t u r e ) . H a u l s were m a d e by AT: Agassiz trawl; ES: e p i b e n t h i c sledge; GT: G r a n t o n otter trawl; SWT: single-warp otter trawl Date

Station

Latitude (N)

L o n g i t u d e (W)

D e p t h (m)

1973 Sept. 23 Sept. 23

ES20 ES22

50046 ' 56041 ,

09~ 09~

1 271 1 028

1975 Mar. 19 Sept. 3

GT1 GT7

56040 , 56025 '

09o06 ' 09010 ,

1976 Apr. 7 J u n e 25 July 2 July 9 July 10 J u l y 11

GTI1 GT17 ES69 ES99 ES105 ATI07A

56035 ' 56~ , 59039 ' 60000 ' 58o27 ' 57007 '

09016 ' 09~ , 07~ ' 10o35 ' 12035 ' 12o06 ,

1977 Aug. 5 Aug. 6 Oct. 22

SWF10 SWTll SWT17

56o50 ' 56002 ' 56049 ,

10000 ' 11008 ' 09o51 '

1978 Apr. 19 Oct. 3t

AT144 SWT32

57 ~ 13' 56~ '

10020 ' 09~ ,

1979 Aug. 9

AT162

51050 '

13~

1980 Feb. 28 Mar. 3 May29 M a y 29

ES169 AT171 AT177 ES178

1981 Apr. 12 Aug. 18 1982 Aug. 5 Aug. 10 Aug. 11

' '

?

c~'

1 1

1 4

805 780

2 2

1 0

9 3 4 - 1 054 1 190-1 296 1 050 1 160 1 600 2 000

1 1 0 1 0 2

0 3 1 0 2 1

2 018 2 530 1 977

13 1 10

7 im 10

ca. 2 240 2 006

2 10

0 10

'

992

7

13

54~ ' 57~ ' 57~ ' 56033 `

12~ ' 10017 ' 10016 ' 09~ ,

2 910 2 225 2 220 997

1 1 1 4

im 0 0 0

AT186 AT193

57~ ' 57028 '

10019 ' 11008 '

2 170 616

1 13

0 7

AT223 AT228 AT230

59041 ' 57~ ' 56o44 '

07009 ' 09~ ' 09012 ,

1 075 2 026 1 210

8 8 0

12 12 i"

650640-

ca. ca. ca. ca.

Plus 3 i m m a t u r e

Table 2. Calveriosoma hystrix, Araeosomafenestrum, Sperosoma grimaldii a n d ratios o f these e c h i n o t h u r i i d sea urchins. A b b r e v i a t i o n s as in Table 1 Date

Hygrosoma petersii. Positions

o f collection stations a n d sex

Station

Latitude (iN)

L o n g i t u d e (W)

D e p t h (m)

?

dr

ES23 GT1 GT7 GT11 ES178 AT223 AT259

56o37 ' 56040 ' 56~ ' 56~ ' 56o33 ' 59o41 ' 57o27 '

09 ~ 10' 09o06 , 09o10 ' 09 ~ 16' 09 ~ 17' 07~ ' 12~ '

704 6 5 0 - 805 6 4 0 - 780 9 3 4 - 1 054 997 1 075 1 041

4 1 3 0 0 11 im

2 0 2 4 5 8 im

AT194

57~

II~

AT230

ES20 GT17 ES129 SWT27 OTSB5130 AT230 AT239

Calveriosoma hystrix Sept. 23, 1973 Mar. 19, 1975 Sept. 3, 1975 Apr. 7, 1976 M a y 29, 1980 Aug. 5, 1982 Aug. 1, 1983

A raeosomafenestrum Aug. 15, 198I

'

'

631

14

6

56044 '

09012 '

1 210

12

8

56~ ' 56o25 ' 54o39 ' 54o27 ' 54~ ' 56044 ' 57~ '

09 ~ 17' 09o25 ' 12~ ' 12~ ' 12 ~ 14' 09 ~ 12' 09~ '

Sperosoma grimaldii Aug. 11, 1982

Hygrosoma petersii Sept. 23, 1973 J u n e 25, 1976 Apr. 5, 1977 M a y 4, 1978 Feb. 15, 1981 Aug. 11, 1982 J u l y 24, 1983

a Plus 1 i m m a t u r e ; b h e r m a p h r o d i t e

1 271 1 190-1 296 ca. 2 900 2 965 2 925 1 210 9 9 0 - 1 145

1 1 0 0 0 0 1b

0 2 1 2 1 1"

66

P.A. Tyler and Ji D. Gage: Reproduction of echinothuriid and cidarid sea urchins

Table 3. Cidaris cidaris and Poriocidaris purpuratus. Positions of collection stations and sex ratio of these cidarid sea urchins. Abbreviations as in Table 1 Date

Station

Latitude

Longitude

Depth

(N)

(W)

(m)

9

8

GT7 GT11 GT12 GT13 AT162 AT223

56~ ' 56035' 56o37' 56036' 51 ~ 59o41'

09010 ' 09 ~16' 09~ ' 09~ ' 13 ~ 07~ '

640934-1 1 037-1 508992 1 075

756 054 500 543

1 0 1 0 1 1

2 1 0 1 0 0

AT223 AT239

59041 ' 57005'

07009' 09~ ,

1 075 990-1 145

0 1

5 0

Cidaris cidaris Sept. 3, 1975 Apr. 7, 1976 Apr. 8, 1976 Apr. 8, 1976 Aug. 9, 1979 Aug. 5, 1982

Poriocidaris purpurata Aug. 5, 1982 July 23, 1983

Table 4. Echinothuriids. Size at first reproduction and maximum size of each species examined Species

Size at first reproduction (mm)

Max. size collected a (ram)

Max. size known b (mm)

Total no. in sample

Phormosoma placenta Calveriosoma hystrix A raeosomafenestrum Sperosoma grimaldii Hygrosoma petersii

45 84 89 83 75

115 150 126 140 165

125 240 280 220 200

180 40 20 20 11

Collapsed test diameter b From Mortensen (1927) a

(Fig. 1 A). These oogonia have a large nucleus and relatively little cytoplasm (Fig. 1 B). As the oogonia develop into small primary oocytes of ca. 10 to 15 # m diam, they separate but remain close to the germinal epithelium (Fig. 1 B). In oocytes of this size range, the cytoplasm is basophilic and contains evenly distributed RNA. The cytoplasm increases in volume in relation to that of the nucleus, but remains basophilic and RNA-positive. The single nucleolus (Fig. 1 B) found in primary oocytes at this stage is also strongly RNA-positive. At this point the primary oocyte still lies close to the germinal epithelium and is embedded in nutritive tissue. The primary oocyte continues to grow until, at ca. 400 # m , there is a distinct change in the nature of the cytoplasm. The cytoplasm becomes less basophilic and gives a weaker reaction to methyl green pyronin, while PAS-positive material accumulates evenly throughout the cytoplasm. The cytoplasm remains strongly PAS-positive until m a x i m u m oocyte size is reached (Fig. 1 D). There is, however, a slight interspecific variation in the nature of the PAS-positive material. In Phormosoma placenta the PAS-positive material is very granular and stains very strongly, whilst in Calveriosoma hystrix and Araeosoma fenestrum the cytoplasmic PAS-positive material appears to be more vacuolar and hence does not stain as intensely. This accumulation o f PAS-positive granules in the cytoplasm of these developing oocytes suggests that vitellogenesis is in progress.

During this period of growth the developing oocytes migrate through the nutritive phagocytic tissue, towards the lumen of the acinus which they appear to almost fill (Fig. 1 D). Maximum egg size in each species is in excess of 1000 # m diam (Table 5). During the course of gonad development there is a variation in the type and properties of the nutritive tissue present. In all species examined, a "generative" tissue is recognizable plus a truly phagocytic, or breakdown tissue.

Table 5. Echinothuriid and cidarid sea urchins from the Rockall Trough. Oocyte size at onset of vitellogenesis (Vit.) and maximum oocyte size attained (Max.) Species

Vit. Cum)

Max. (~m)

400-450 400-450 500-550 350-400 ca. 400

1 100 1 250 1 250 1 100 1 150

40 200-250

110 1 500

Echinothuriids

Phormosomaplacenta Calveriosoma hystrix Araeosomafenestrum Sperosorna grimaldfi Hygrosomapetersii Cidarids

Cidaris cidaris Poriocidarispurpurata

P. A. Tyler and J. D. Gage: Reproduction of echinothuriid and cidarid sea urchins In Phormosoma placenta the early generative nutritive tissue has a distinct structure consisting of vacuolated acidophilic cells (Fig. 1 B, E) containing numerous PASpositive granules lining infoldings of the inner sac (Fig. 1E) into which the haemal sinus possibly extends. This generative tissue appears to become depleted and less structured in some sections of the ovary (Fig. 1 E). The PAS-positive staining is less intense where fewer granules are present, whilst the tissue appears to be much more vacuolated. We have interpreted this variation in genera-

67

tive tissue structure as a process involved in the nutriment of the developing oocyte. The second type of "nutritive tissue" is seen in Phormosoma placenta (Fig. 1 E). This consists of a mass of phagocytes breaking down relict ova for recycling within the ovary. The development of males appears to be similar to that of echinothuriids examined previously. Infoldings of the germinal epithelium contain large vacuolated cells (Fig. 2A) and the developing spermatogonia grow along

Fig. 1. Oocyte development in echinothuriid sea urchins. (A) Ovary wall of Calveriosoma hystrix (toluidine blue); (B) oogonial and early previtellogenic oocyte development in Phormosoma placenta (haematoxylin and eosin); (C) partially spent nutritive tissue in Araeosomafenestrum (toluidine blue); (D) general view of various stages of oocyte development in P. placenta (PAS); (E) nutritive and phagocytic tissue in P. placenta (toluidine blue), c: coelomic epithelium; ct: connective tissue; ge: germinal epithelium; is: fold in germinal epithelium; nt: nutritive tissue; o: oogonia; p: previtellogenie oocyte; ph: phagocytes; sp: spent nutritive tissue; t: well-structured nutritive tissue; v: vitellogenic oocyte; w: ovary wall. Scale bars = 100 ~m

68

P.A. Tyler and J. D. Gage: Reproduction of echinothuriid and cidarid sea urchins

Fig. 2. Histological structure of ovary and testis in echinothuriid and cidarid sea urchins. (A) Testis ofAraeosomafenestrum (haematoxylin and eosin); (B) ovary of Cidaris cidaris (PAS); (C) section of ovary of PoriocidarisI)urpurata (PAS and haemalum); (D) testis of P. purpurata (toluidine blue), d: developing spermatogonia; nt: nutritive tissue; p: previtellogenic oocyte; pg: pigment; v: vitellogenic oocyte; w: gonad wall; y: vacuolated cells; z: early spermatozoa. Scale bars = 100 ~m this vacuolated structure. The spermatozoon has a long conical head typical of the Echinoidea.

Oocyte size~frequency distributions Oocyte size/frequency histograms were constructed to determine the distribution of different oocyte sizes within each individual specimen and to test for any inter-sample variation in mean oocyte size that might indicate a reproductive seasonality. Most individual specimens of Phormosoma placenta show a very erratic distribution of oocyte sizes (Figs. 3, 4). Oocyte development begins in specimens having a diameter of between 40 and 50 ram. A large number of small

oocytes (<300~tm diam) are produced. Some of these oocytes undergo vitellogenesis and develop to a maximum size of 1 100 ~m. Evidence from the oocyte size-frequency distributions suggests that by the time a test diameter of 55 to 60 m m is reached, there is a relatively stable distribution of oocyte sizes - a large proportion having a diameter of < 3 0 0 p r o with relatively few progressing through vitellogenesis to maturity. As the ripe oocytes are spawned, the stock of remaining oocytes is supplemented by newly dividing germ cells. Gamete production is presumed to be a continuous process and we have found no evidence of senescence in any of the specimens examined. The oocyte size/frequency data (Fig. 5, 6, 7) and maximum oocyte size (Table 5) for the other species suggest a similar pattern of oocyte development. None of these

P. A. Tyler and J. D. Gage: Reproduction of echinothuriid and cidarid sea urchins

% % % -

23.9"73

,Ooot

,ooj

%~

% % %

7Oo~

Poo

%

%

%

%. %, %. %.

2~

69

,oo:--..>

% 0

i5 o~

e%~ , ,,%-

~Oco"

'%e%~-o%l~?w

1.`0

@o. % .h ZOo,"..

19-3"75

0

3"9"75

"-0% "~ f~2

25"6.76

11-7"76

%%oc3

h

5.8.77

% % -~ %

I

q

%

o

o,

%%%

% % % -~ % % %

%

% % g % % ep

0 Q

,% Fig. 3. Phormosomaylacenta.Oocyte size/frequency histograms for specimens trawled at various dates from 1973 to 1977. Each frequency distribution was derived from gonad taken from a single individual

species had ever been seen to spawn, but the evidence available suggests that only a few oocytes are spawned at any one time. Only one specimen of Hygrosoma petersii was found to be an intra-gonadal hermaphrodite, and in this specimen none of the oocytes developed beyond the previtellogenic stage.

Family Cidaridae

Gametogenesis in Cidaris cidaris From the limited samples available, gametogenesis in Cidaris cidaris appears to be typical of shallow-water regular echinoids. Primary oocytes proliferate around the

70

P.A. Tyler and J. D. Gage: Reproduction of echinothuriid and cidarid sea urchins

%]

~

L.

tD

::E o O

b')

SOo. L o, ~

'~

i

8 ~'% .,~% "t:

/~oj

-o/

t

700, /

=z:

% %, %, %,

b3

%,

co


o o

~oo

18"8"81 >

-%.

%

loo.

'~

O, (3""

~

%.t 200

~ ;.

I~ I/ IFI'L

--

'

% -

%

Q

,oo 4 %

o %

I ~, % ? ~oo..I h! ~ . . ' 4

,o.~y

->.-

10"8"82

5"8"82

Fig. 4. Phormosoma placenta. Oocyte size/frequency histograms for specimens trawled at various dates from 1978 to 1982. Each frequency distribution was derived from gonad taken from a single individual

periphery of the ovary and lie close to the germinal epithelium, surrounded by nutritive tissue (Fig. 2 B). The oocytes grow and undergo vitellogenesis when in excess of 40/~m diam. Most remain close to the germinal epithelium until they reach the maximum diameter of l l0ffm. All

oocytes observed had intact germinal vesicles, suggesting that reduction division had not occurred. Migration to the lumen of the ovary usually accompanies reduction division (Tyler and Gage, in press). Within the few females examined, oocyte development appears to be synchronous.

P. A. Tyler and J. D. Gage: Reproduction of echinothuriid and cidarid sea urchins

/" s.8.82

.~

.../.,

71

, oo/"1 ,,oo.~

~ ~

~

,,oo I

~ ,0oo~ ~ 13.9.7 s

% /

~ ,Ooo~

%

~,oo

~,oo.

'~oo

~

~

,ooo

Soo

,oo

~23.9.7a

Soo.

,oo,

c,)

o ~ . d "-1 --a~ ~. ~ ~ %%0^'~ ~ ' ~

7

"-.1

Fig. 5 Calveriosoma hystrix Oocyte size/frequencies of specimens trawled in Augusi and September of various years Each frequency distributed was derived from gonad taken from a single individual

>

% %, %.

%. iJ?

SOo . o o

%. Fig. 6. Araeosomafenestrum. Oocyte size/frequencies of single gonads from each of 14 individuals trawled on 11 August I982

%

The nutritive tissue is very PAS-positive, except in spent specimens where a pale brown pigment, possibly representing breakdown products, replaces the nutritive tissue. From the very limited data available, we would not predict any seasonality for this species, although the larva is found in the plankton from February to June (Mortensen, 1901).

Gametogenesis in Poriocidarispurpurata The gonads of Poriocidaris purpurata are purple brown in colour. In the developing ovary the oogonia are arranged peripherally along the germinal epithelium. These grow into primary previtellogenic oocytes and elongate along the germinal epithelium, to which they are still attached

72

P.A. Tyler and J. D. Gage: Reproduction of echinothuriid and cidarid sea urchins

% %

B

J co ilJ

8

%

8

,oot

,oo

-

%

%

15

SWT 17

10 5 0

>-

u

20

Z

uJ

15

D 0

10

uJ oc

5

u_

0

% % % 4oo % .

J ,

,

,

,

,

,

,

,

,

~

,

,

,

,

,

Fig. 7. Sperosoma grimaldii (A) and Hygrosoma petersii (B). Oocyte size-frequencies from gonads of specimens trawled on 11 August 1982 (A) and on 23 September 1973 (B)

completely occlude the lumen and contain yolk granules up to 5 # m diam. They are no longer embedded in nutritive tissue, but appear to be surrounded by a number of accessory cells. As the oocytes reach maximum size the nutritive tissue appears to become very vacuolated and the light brown pigmentation of the ovary is clearly seen (Fig. 2 C). From the scanty evidence available we find very little phagocytic activity, suggesting that most, if not all, oocytes are successfully spawned. The limited evidence suggests that there is some synchrony of gamete develo pment within one ovary, but not between specimens. In the early development of the testis the germinal epithelium is covered with a thick layer of nutritive tissue. The testis has distinct folds reminiscent of the colonettes of testes in asteroids (Fig. 2D) (Cognetti and Delavault, 1962). Spermatogenic development is typical of echinoids, and numerous conical-headed spermatozoa are produced.

n = 140

Population structure of Phormosoma placenta

TEST

DIAMETER

(mm)

Fig. 8. Phormosoma placenta. Population-size structure in three samples. AT: Agassiz trawl sample; SWT: single-warp otter trawl sample

(Fig. 2C). During this period they are embedded in the thick layer of well-structured nutritive tissue. These oocytes begin vitellogenesis at ca. 200 to 250 # m diam. By this time they lose their attachment to the germinal epithelium but remain embedded in nutritive tissue. The oocytes continue to grow during vitellogenesis and reach a maximum size of 1 500#m diam. At this size, they

We have attempted to determine population structure from the three largest samples of Phormosoma placenta (Fig. 8). The pattern observed in these samples is similar to that of Ophiornusium lymani (Gage and Tyler, 1982). How, ever, the smaller size range will be under-represented due to the mesh size of the Agassiz trawl. We believe that there may be relatively rapid growth up to 40 to 45 mm test diameter.. At this point gonad development begins and there is a slowing of somatic growth, resulting in an accumulation of adults in the 55 to 75 mm size range in test diameter. Above 75 mm test diameter the population numbers decline, and few individuals larger than 85 mm test diameter are found. The largest specimen had a test diameter of 115 mm.

P. A. Tyler and J. D. Gage: Reproduction of echinothuriid and cidarid sea urchins Discussion

The echinothuriids constitute an essentially deep-sea family of echinoids, although some Indo-Pacific species occur in shallow waters (Mortensen, 1927, 1940). Studies of reproduction in this group are few, but both the observations of Mori etal. (1980) and those of Mortensen (1927) indicate that echinothuriids produce large eggs. Most other regular echinoids studied to date, with the exception of some cidarids, produce small eggs (max. 120/~m diam), suggesting that planktotrophic larval development occurs. The maximum egg size of the echinothuriid species studied here ranged from 1 100 to 1 300#m in diameter, which is consistent with the maximum size observed for the shallow-water echinothuriids Asthenosoma ijimai and A raeosoma owstoni (Mori et al., 1980). The development of Asthenosoma ijimai has been described in detail by Amemiya and Tsuchiya (1979): lecithotrophic larval development gives rise to a "biscuit-shaped" late blastula. The large egg size and low fecundity of the deep-sea species examined in this paper indicate a similar form of direct development omitting a planktotrophic stage. The pattern of oocyte development seen in shallowwater and deep-sea species is similar. The oocyte develops within the nutritive/phagocytic tissue. At about 400 gin oocyte diameter, vitellogenesis begins and the oocyte ultimately reaches a maximum size of 1 100 to 1 300/~m diameter. The cytoplasm of well-developed oocytes are packed with PAS-positive granules. If the nutritive tissue is truly supportive, the well-structured acidophilic PAS-positire tissue is early nutritive tissue. As this stored energy is transferred to the developing oocyte, the nutritive tissue becomes highly vacuolated and itself appears to be "spent". In Phormosoma placenta, relict oocytes are resorbed by phagocytic activity for use in either restructuring phagocytic tissue or as the nutrient for further developing oocytes, From the oocyte size/frequency figures we observe that only a few oocytes mature at any one time. Observations on Phormosoma placenta from the submersible "Alvin" by Grassle et al. (1975) and Pawson (1982) differ with respect to the aggregatory behaviour shown by this species. In the dense aggregations observed by Grassle et aI., we assume fertilisation presents no problems as some members of the population will be reproductively mature at the same time. The production of a few eggs that undergo direct development, together with a behavioural pattern giving rise to dense aggregations, might preclude the necessity for reproductive seasonality where larval survival relies on sedimentation of surface-derived organic production. However, Ahlfeld (1977) suggested that there was some evidence that Hygrosoma petersii showed reproductive seasonality. Our samples of this species were too small to contradict this, but evidence for the other species suggests that no reproductive seasonality occurs. Our observations of reproduction ofcidarid sea urchins from the Rockall Trough support previous general observations of their reproductive biology. Cidaris cidaris has

73

known planktotrophic development (Mortensen, 1927), although we have never found this echinopluteus in oblique RMT 1 (rectangular midwater trawl) samples taken at Station "M" in the Rockall Trough. The large egg of Poriocidaris purpurata was reported by Mortensen (1927) and he comments that the female genital pores were also very large, extending even to the outside of the genital plate. This structure would facilitate the spawning of the exceptionally large oocyte produced by this species. Our observations of the gametogenic biology of deepsea echinothuriids and cidarids support observations of the gametogenic biology in other deep-sea echinoderms (Tyler etal., 1982). The dominant reproductive strategy is the production of a large egg and, hence, direct development with relatively little dispersal. The pattern seen in Cidaris cidaris, and other species that reproduce by a planktotrophic larva is an apparently highly successful anomaly amongst the prevailing lecithotrophic strategies of the deep-sea benthos.

Acknowledgements. We wish to thank Professor F. T. Banner and Dr. M. B. Collins for facilities in the Department of Oceanography; the Master and Crew of R.R.S. "Challenger" for help at sea; Mrs. M. Pearson for sorting and identifying samples and Mr. A. Muirhead for photographic and technical assistance. The study was supported by N.E.R.C. Grant GR3/4131 to PAT, which is gratefully acknowledged. The Scottish Marine Biological Association is grant-aided by the Natural Environment Research Council.

Literature cited

Ahlfeld, T. E.: A disparate seasonal study of reproduction of eight deep-sea macroinvertebrate species from the N. W. Atlantic, 105 pp. Unpublished Ph.D. thesis, Florida State University I977 Amemiya, S. and T. Tsuchiya: Development of the echinothurid sea urchin Asthenosoma ijimai. Mar. Biol. 52, 93-96 (1979) Cognetti, G. et R, Delavault: La sexualit6 des asterides. Cab. Biol. mar. 3, 157-182 (1962) Dix, T. D.: Reproduction in Tasmanian populations of Heliocidaris erythrogramma (Echinodermata: Echinometridae). Aust. J. mar. Freshwat. Res. 28, 509-520 (1977) Gage, J. D., M. Jensen, D. S. M. Billet and P. A. Tyler: Echinoderms of the Rockall Trough area. II. Echinoidea and Holothuroidea. Bull. Br. Mus. nat. Hist. (In press) Gage, J. D., R. H. Lightfoot, M. Pearson and P. A. Tyler: An introduction to a sample time-series of abyssal macrobenthos: methods and principal sources of variabil(ty. Oceanol. Acta 3, 109-176 (1980) Gage, J. D. and P. A. Tyler: Growth and reproduction of the deep-sea brittle star Ophiomusium lymani Wyville Thomson. Oceanol. Acta 5, 73-83 (1982) Grassle, F. J., H. L. Sanders, R. R. Hessler, G. T. Rowe and T. McLellan: Pattern and zonation: a study of the bathyal megafauna using the research submersible Alvin. Deep-Sea Res, 22, 457-481 (1975) Holland, N. D.: Gametogenesis during the annual reproductive cycle of a cidaroid sea urchin (Stytocidaris affinis). Biol. Bull. mar. biol. Lab., Woods Hole 133, 578-590 (1967)

74 Mori, T., T. Tsuchiya and S. Amemiya: Annual gonadal variation in sea urchins of the orders Echinothuridea and Echinoidea. Biol. Bull. mar. biol. Lab., Woods Hole 159, 728-736 (1980) Mortensen, T.: Echinodermlarven. Nord. Plankt. III(a), 1-30 (1901) Mortensen, T.: Handbook of the echinoderms of the British Isles, 471 pp. Oxford University Press 1927 Mortensen, T.: On the vegetarian diet of some deep-sea echinoids. Annotnes zoo1. jap. 17, 225-228 (1938) Mortensen, T.: A monograph of the Echinoidea. III. I Aulodonta, 370 pp. Copenhagen: Reitzel 1940 Pawson, D.: Deep-sea echinoderms in the tongue of the ocean, Bahamas Islands: a survey, using the research submersible Alvin. Mere. Aust. Mus. 16, 129-145 (1982) Prouho, H.: Recherches sur le Dorocidaris papillata et quelques autres 6chinides de la Mediterran~e. Archs Zool. exp. g6n. (Ser. 2) 5, 213-380 (1888)

Tyler, P. A. and J. D. Gage: Seasonal reproduction of Echinus affinis (Echinodermata: Echinoidea) in the Rockall Trough, N. E. Atlantic Ocean. Deep-Sea Res. (In press) Tyler, P. A., A. Grant, S. L. Pain and J. D. Gage: Is annual reproduction in deep-sea echinoderms a response to variability in their environment? Nature, Loud. 300, 747-750 (1982) Williams, D. H. C. and D. T. Anderson: The reproductive system, embryonic development, larval development and metamorphosis of the sea urchin Heliocidaris erythrogramma (Val.) (Echinoidea: Echinometridae). Aust. J. Zool. 23, 371-403 (1975)

Date of final manuscript acceptance: March 1, 1984. Communicated by J. Mauchline, Oban