J. CeUSci. 38, 117-132 (1977) I I 7

Printed in Great Britain © Company of Biologists Limited 1<)TJ

TISSUE ECONOMICS OF HYDRA:

REGULATION OF CELL CYCLE, ANIMAL SIZE

AND DEVELOPMENT BY CONTROLLED

FEEDING RATES

JOANN J. OTTO* AND RICHARD D. CAMPBELLDepartment of Developmental and Cell Biology, University of California at Irvine,Irvine, California 92717, U.S.A.

SUMMARY

Epithelial cell production and epithelial cell loss in 6 different size classes of Hydra attenuatawere examined to understand the relationships between growth and morphogenesis. The sizesof adult hydra, the sizes of their buds, and their budding rates are all nearly proportional to theamount of food the hydra eat. Hydra fed at high rates (4-25 Artemia nauplii per day) all havethe same epithelial cell cycle time (about 4 days). Budding accounts for most of their cell loss.Hydra fed 4-12 Artemia per day maintain a steady state condition in which tissue loss balancestissue growth. Animals fed 25 Artemia per day are not in a steady state growth condition andchange in size. At the lowest feeding rates (0-1 Artemia per day), the epithelial cell cycle time islengthened to about 16 days. Cell loss from the tentacles accounts for most of the cell loss, andthis loss is not completely balanced by growth. As a consequence these animals cease buddingand shrink in size.

INTRODUCTION

The morphology of hydra is dynamic. Cells proliferate throughout most of thepolyp (Campbell, 1967a; Clarkson & Wolpert, 1967; David & Campbell, 1972;Webster & Hamilton, 1972), and tissue is lost through the process of budding andthrough sloughing at the tentacle tips, hypostome, and basal disk (Brien & Reniers-Decoen, 1949). In the steady state, cell proliferation balances tissue loss, and thehydra's size remains constant.

Hydra are not, however, all the same size (Stiven, 1965; Schroeder, 1969; Bisbee,1973), and they may have varying numbers of developing buds. This suggests that thedynamic growth processes reach different balance points in different individuals. Inthis study, we found that hydra can reach a steady state size on a wide range of dif-ferent feeding rates, and that the steady state size is proportional to the amount of foodingested. We investigated and quantified the parameters of epithelial cell productionand epithelial cell loss in 6 different size classes of Hydra attenuata in order to under-stand relationships between growth and morphogenesis.

• Address for correspondence (and present address of J.J.O.): Joann J. Otto, Departmentof Biology, University of Pennsylvania, Philadelphia, Pennsylvania, 19.174, U.S.A.

118 J.J. Otto and R. D. Campbell

MATERIALS AND METHODS

Culture methods

Hydra attenuata with 1-2 buds were obtained from, a stock culture. They were maintainedin 'M ' solution using the methods of Lenhoff & Brown (1970) in closed 100-mm plastic Petridishes with 20 animals per dish in an incubator at 19 °C and a 12-h light/12-h dark cycle.

Groups of hydra were maintained on different feeding regimes. Four of the groups consistedof hydra which were individually fed 1, 4, 8, or 12 Artemia nauplii per day for at least 2 weeksprior to any experiments. One group of hydra was starved. The hydra in another group werefed as much as they would eat twice a day (estimated to be about 25 Artemia per day and con-sidered here to be 25 Artemia per day) for 2 weeks prior to the experiments.

Analysis of cell types and cell cycle

The number and types of cells in the hydra were determined using David's (1973) macerationand cell identification procedure. To determine the cell number of the adult, all buds olderthan stage 8 (the stage in which the bud has a constricted basal region, Otto & Campbell, 1977)were removed before maceration and counting.

To analyse daily epithelial cell production, the length of the epithelial cell cycle in each sizeclass of hydra was measured. For animals fed 1, 4, 8, or 12 Artemia per day, the epithelial cellcycle was measured after the animals had been on the feeding regime for 3-4 weeks. For animalsfed o or 25 Artemia per day, the cell cycle was measured after the animals had been on thefeeding regime for 2 weeks. For continuous and pulse-labelling experiments, [Me-'H]-thymidine (ICN) at 20/iCi/ml was injected into the gastric cavity (David & Campbell, 1972).Effective continuous labelling was achieved by repeating injections at intervals shorter than theknown length of the S phase in well fed animals (David & Campbell, 1972), usually at 6-hintervals. Animals were killed 1 h after their last injection with the final injection being 50 /tCi/ml. Three animals were used per sample. To prepare the labelled hydra for autoradiography,they were macerated, fixed in formalin, and then spread on gelatin-coated microscope slides.After drying for 1 day the slides were washed in running water for 0-5 h and then dipped inKodak NTB-3 emulsion. The slides were exposed for 2-6 months and developed.

The autoradiographs were analysed using bright-field and phase-contrast microscopy. Thepercentage of labelled cells was determined for each sample by counting 1000 epithelial cellnuclei.

Analysis of tissue loss

To analyse daily cell loss, we measured tissue loss through budding and through cell sloughingat the tentacle tips and basal disk. To measure cell loss through budding, we counted the cells instage 8 buds (Otto & Campbell, 1977) from animals on the different feeding regimes, anddetermined the budding rate by counting and removing detached buds daily.

We estimated the rate of cell loss at the tentacle tips from measurements of tissue turnovertime and of the number of cells in tentacles. Small marks of India ink (Campbell, 1973) wereplaced at the bases of the tentacles and observed daily over a period of a week as they weredisplaced outwards on the tentacles. A mark's position was measured with an ocular reticle asthe fractional distance along the tentacle. For each feeding regime, the displacement of anaverage of 6 marks on 3 hydra was observed. To determine the number of epithelial cells inthe marked tentacles, they were fixed in Bouin's fixative, embedded in paraffin, sectioned andstained with Mayer's haematoxylin. The number of epithelial cells was counted in 4 tentaclesfrom 3 hydra on each feeding regime.

Basal disk cell loss was estimated in 3 animals per feeding regime by placing several marksalong the peduncle and measuring the rate of downward displacement. The marks were observeddaily and their position recorded as a percentage of total peduncle or of total body length. Thenumber of cells in the peduncle was determined from macerated peduncles.

An independent assessment of the relative rates of distal and proximal tissue loss wasobtained by determining the position of the stationary zone in the gastric region (Campbell,

Tissue economics of Hydra "919676), again using the vital ink marking technique. For each feeding regime 3 animals, markedwith an average of 5 marks, were observed daily over a 2-week period to determine the positionof the stationary zone. The positions of the marks were measured (with an ocular reticle) aspercentages of the distance from the hypostome to the basal disk. The position above whichmarks moved distally and below which marks moved proximally was determined for each feed-ing regime.

Fig. 1. Hydra and buds from different feeding regimes. The feeding rate (Artemia/day)is shown between each bud (lower row) and parent (upper row). The bud from thestarved hydra was the last one produced by the animal before it ceased budding. Thescale represents 2 mm.

RESULTS

In these experiments, we cultured hydra of different sizes by feeding individualanimals on particular regimes for several weeks. The sizes and budding rates of thehydra changed during their first 2 weeks on the new feeding regimes and reached newconstant levels by the end of 2 weeks. At this time we judged that the hydra in mostgroups were in steady state by the constancy of cell number and budding rates asdescribed below. However, those hydra fed o, 1, and 25 Artemia per day never reachedsteady state sizes.

Fig. 1 shows hydra and their buds from the different cultures. The animals differin size and in the numbers of buds. Since the ectodermal and endodermal epithelial

120 J. J. Otto and R. D. Campbell

cells are responsible for hydra's structural framework, we have analysed the dailyepithelial cell production and epithelial cell loss in these different sized hydra in orderto determine the factors governing size. In our experiments no distinction was madebetween ectodermal and endodermal epithelial cells since they have the same cellcycle parameters in well fed hydra (David & Campbell, 1972) and exist in approxi-mately equal numbers in the animal (Bode et al. 1973)-

50 r

0 1 8 12Feeding rate. Anemia/day

Fig. 2. Number of epithelial cells per adult hydra (ordinate) for each feeding regime(abscissa). (#) , averaged data from. 3 experiments each with 3 animals per feedingregime. For hydra fed 8 or 12 Artemia per day, the number of epithelial cells is shown3 weeks (•) and 7 weeks (A) after beginning the new feeding regimes. For animalsfed 4 Artemia per day, counts are shown 4 weeks ( • ) and 5 weeks (A) after thebeginning of the new feeding regime. The similarities between cell counts made atdifferent times show that the hydra are in a steady state condition.

Hydra size

Hydra size (Fig. 2) was quantitated by counting the number of epithelial cells inthe animals after maintaining the animals for 2 weeks on these feeding regimes. Thesize of the hydra is roughly proportional to feeding rate. The animals fed 4, 8 or 12Artemia per day reached a steady state condition since cell counts at different timesafter the start of the feeding regimes were the same (Fig. 2). Animals fed o, 1, or25 Artemia per day never reached a steady state. The numbers of cells for thesefeeding regimes given in Fig. 2 were measured at the end of 2 weeks. The hydra fed

Tissue economics of Hydra 121

25 Artemia per day develop complex, changing morphologies including extra tentaclerings (Fig. 1). Starved animals dwindle in size.

Cell production

The daily epithelial cell production in these hydra was determined by analysis ofthe epithelial cell cycle length and from data on animal size. Continuous and pulselabelling experiments with [3H]thymidine allowed us to calculate the epithelial cellcycle length.

100

2 4 6

Time of continuous labelling, days

Fig. 3. Continuous labelling of epithelial cells with [3H]thymidine. Abscissa representsthe time after the beginning of labelling. The percentages of cells labelled (ordinate)are shown for hydra fed o, 1,4, 8, 12, and 25 shrimp/day. Lines are least-squareregression lines for each feeding regime using the data points for the first 3 days.

A continuous labelling experiment provides an estimate of the length of the cellcycle minus the length of the S phase. Fig. 3 shows the rate at which cells becamelabelled in a continuous label experiment. For animals fed 4 or more Artemia per dayit took 3-4 days for the labelling curves, as extrapolated from the data for the first3 days, to reach 100%. We used the data for the first 3 days since to determine theaverage cell cycle duration one must estimate the labelling rate before significantnumbers of cells begin their second cycle. Starved animals and those fed 1 Artemiaper day required, respectively, 13 and 14 days (extrapolated) for 100% of the epithelialcells to become labelled (or 13 and 11 days if we use the data points for days 6 and 5).

We estimated the length of S phase in 3 ways, and each result, as well as the averagevalues for S phase durations, is listed in Table I.

122 J. J. Otto and R. D. Campbell

Method i involves extrapolating the continuous labelling curves (Fig. 3) backwardsto their intersections with the time axis. The negative of this intersection representsthe duration of S phase in these cells (David & Campbell, 1972).

Method 2 involves using a similar analysis of continuous labelling data taken every4 h over a period of 24 h. These data are presented in Fig. 4 (solid points). The £phase durations estimated from least-square regression lines (not shown in Fig. 4)for these data are listed in Table I.

Table 1. Duration of S phase

Feeding

Artemialday

0

1

48

1 2

25

• The

Estimates of »S-phaseA

Method 1

i-81-2

o-S0 90 9

i-4

Method 2

55——0 40-3i-o

3 methods are described in the

duration, days

Method 3

———

0 8 3

o-3S0-7

accompanying

•

Average

3-71-2

o-50 7

o-Si-o

text.

Method 3 involves correlating continuous and pulse-labelling data obtained overa period of 24 h (Fig. 4). We used this method because the cell cycles in hydra arepartially synchronized (David & Campbell, 1972) and, therefore, Methods 1 and 2above are not strictly applicable. In Method 3 we compared the percentage of cellsmarked by a continuous labelling regime with the sum of cells marked independentlyby pulse-labelling at time o and in a different animal at some later time. After aninterval, tny equal to the S length, continuous labelling will yield the same number ofcells labelled as will two pulses given at to and ts. Continuous labelling for a shorterperiod, tx, will label fewer than the sum of cells labelled in separate animals by2 pulses at tn and tx. Similarly, a continuous labelling regime longer than S lengthwill label more cells than any 2 pulses. Therefore, we plotted the continuous labellingcurve with the curve derived from sums of 2 separate pulses of label data (one pulsewas at time o), and took the time when these curves intersect as the duration of theS phase. These values for 3 feeding regimes are listed in Table I.

These 3 methods give estimates which are roughly similar to one another, and leadto the conclusion that the duration of S phase is somewhat dependent on feedingregime and never constitutes a large part of the cell cycle (see Table 1, last column).

From the results of the continuous labelling experiments shown in Fig. 3 and ofthe S phase experiments summarized in Table 1, the total epithelial cell cycle can becalculated. For animals fed 4 or more Artemia per day the average time it took toreach 100% labelling was 3-4 days, and the S phase is estimated to Ia3t about 0-7 days(Table 1). Thus, the length of the cell cycle is about 4-5 days. For animals whichwere starved or fed 1 Artemia per day the cell cycle length is about 16 days. Theseresults indicate that over a wide range of size (animals fed 4-25 Artemia per day), the

Tissue economics of Hydra 123

epithelial cell cycle length does not vary. Thus, a difference in cell cycle time does notaccount for the maintenance of steady state size differences. These experiments wererepeated 3 times with similar results.

60

50

40

30

Ij 20

1 10

2 0

8 60

40

30

20

10

c - D

8 12 16 20 24 0 4Time, h

8 12 16 20 24

Fig. 4. Comparison of the percentages of cells labelled by continuous and pulseadministration of [sH]thymidine. 0 , continuous labelling beginning at time = o.O, sum of the percentages of those cells labelled by a pulse administered at theindicated time point and those labelled by the pulse administered at time = o. Theduration of the S phase is estimated as the time at which the 2 curves intersect (seetext). The pairs of percentages indicated by the asterisk are not significantly different(P = 0-05) by the difference of the means of 2 binomial populations, A, B, C, D wereon regimes of o, 4, 8 and 12 Artemia/day, respectively.

We calculated the daily cell production by combining our estimates of cell number(Fig. 1) and of cell cycle time. For this calculation we assumed that all of the gastriccolumn cells and 70% of the peduncle cells divide (David & Campbell, 1972). Theseresults are shown by the horizontal lines in Fig. 11 (p. 129). The animals fed more arelarger in size and produce correspondingly more cells per day.

Tissue loss

Tissue loss through budding. The cumulative number of buds produced by the hydraon the different feeding regimes in one experiment is shown in Fig. 5. The buddingrates of the hydra on the various feeding regimes do not diverge until about the 7th dayon the regime. Even starved animals continue to bud for 10 days. Within 2 weeks thebudding rates become stabilized at the new levels.

From the data shown in Fig. 5, the steady state budding rate for each feeding regimewas calculated (Fig. 6A). Budding rate is nearly proportional to feeding rate. The sizesof the buds were measured and found also to be related directly to feeding level

124 3- 3- 0**° end R- D- Campbell

(Fig. 6 A), and nearly proportional to the parent size. From the budding rates and budsize, the number of epithelial cells lost by a hydra per day through budding wascalculated. As shown in Fig. 6B, the larger the animal, the greater is the number ofcells lost per day by budding.

20

Fig. 5. Cumulative number of buds (ordinate) produced per hydra after different times(abscissa) after beginning the new feeding regimes of o, i, 4, 8, 12, and 25 Artemiajday. The number of buds was determined by removing and counting detached buds.

Tissue loss from the tentacle tips. Tentacle cells are displaced distally by the additionof epithelial cells to the base of the tentacles and the loss of cells from the tentacletips. To estimate tentacle turnover time, marks of India ink were placed at the baseof the tentacle, and the displacement of the marks was observed (Fig. 7). For each ofthe feeding regimes about 10% of the tentacle is lost per day (Fig. 8A). The number ofepithelial cells per tentacle for the different sized animals is also shown in Fig. 8A.From these data and from the average number of tentacles per animal (about 6 for allfeeding regimes), the number of epithelial cells lost by tentacle tissue turnover perday was calculated (Fig. 8 B). There is a general slight rise in the absolute rate oftentacle cell loss with increased feeding rates.

Tissue loss from the peduncle. The rate of tissue loss at the basal disk was estimated

Tissue economics of Hydra 125

by marking cells, in the peduncle and measuring their downward displacement(Fig. 9). Since there is some proliferation in the peduncle (David & Campbell, 1972),this method measures the rate of peduncle turnover. Fig. 10 A shows these measured

8 12Feeding rate, Artemia/day.

25

Fig. 6. Loss of epithelial cells into developing buds. A, no. of epithelial cells per bud(O) and the budding rate (A) as functions of feeding rate. B, no. of cells lost perday by an adult hydra through budding, calculated from data in A.

rates, as well as the number of epithelial cells comprising the peduncle for animals onthe different feeding regimes. From these data, we calculated the number of cells lostby the peduncle .per day for the various feeding regimes (Fig. IOB). Hydra fed4 Artemia per day apparently lose more peduncle cells per day than do otherhydra. We did not.estimate peduncle tissue loss in animals fed 25 Artemia per daybecause the markers near the top of the peduncle sometimes moved upwards ornot at all. This resembles a situation described by Shostak (1968) in H. viridis,where cell production within the peduncle exceeds the rate of sloughing at thebasal disk. As a result, a 'stationary region' (Campbell, 19676) is set up within thepeduncle.

0 CEL 28

126 J. J. Otto and R. D. Campbell

Tissue bss from the hypostome. We attempted to analyse cell loss at the hypostomeby the India ink marking technique. Six out of 8 marks observed for 5-10 daysshowed little or no displacement. Because of this slow turnover rate and the small sizeof the hypostome, we concluded it did not constitute a significant factor in the overallcell economy of these hydra.

1 0 0 r

1 2 3 4 5 0 1 2 3 4

Time, days

5 0 1 2 3 4

Fig. 7. Movement of representative marks in the tentacle. The displacement of intra-cellular carbon marks along the tentacle with time is shown for the feeding regimesof o, 1, 4, 8, 12 and 25 Artemia/day (A-F, respectively). Each curve represents onemark on a tentacle.

Balance of tissue production and loss

Fig. 11 compares our estimates of daily cell loss (vertical bars) and cell production(horizontal lines) in individual hydra. Both values rise sharply with feeding rate. Theestimates of cell loss and production generally agree, as would be expected since thehydra are in a steady state condition.

Some of the discrepancies between these estimates of cell loss and production(Fig. 11) can be understood. For the animals fed o and 1 Artemia per day, the cell lossis greater than cell production. This agrees well with our observations and those ofother investigators (Bode et al. 1973) that starved animals decrease in size. For theanimals fed 25 Artemia per day cell loss is also greater than cell production after2-3 weeks on the feeding regime. However, the relationship between loss and produc-tion changes constantly in these animals. During the first 2-3 weeks of heavy feedingthe animals grow quite large and bud heavily. Then the interstitial cell: epithelialcell ratio falls, nematocyte production slows and the animals eat less (Bode et al.,1977). During this time the animals decrease in size. These observations are consistentwith the relative cell production and loss in animals fed o, 1, and 25 shrimp per dayas presented in Fig. 11.

Some of the discrepancies in Fig. 11 probably represent errors in our estimates.

Tissue economics of Hydra 127

For animals fed 4-12 Artemia per day, the estimates of daily cell production aregreater than those of daily cell loss; however, these animals appear to be in steadystate as evidenced by constancy of sizes and budding rates (Figs. 2, 5).

8 12Feeding rate, Artemia/day

25

Fig. 8. Epithelial cell loss through tentacle turnover, A, % of tentacle lost per day (A)through turnover (from data in Fig. 7), and average no. of cells per tentacle (O), asfunctions of feeding rate. B, total number of cells lost per hydra per day throughtentacle turnover, calculated by combining the data in A with the average value of6 tentacles/hydra.

Stationary zone

An independent estimate of the pattern of tissue loss was made by determining theposition of the stationary zone. This is the axial position above which tissue movesdistally and below which tissue moves proximally. This position reflects the relativerates of tissue loss from the distal and proximal ends (Campbell, 19676).

Tissue regions were vitally marked along the body column of animals maintainedon different feeding regimes. From the movements of these marks (Fig. 12) we deter-mined the positions of the stationary regions (positions where marks diverged). Forstarved animals the stationary zone is located in a position about 70 % of the length

9-2

128 J. J. Otto and R. D. Campbell

y100

0 2 4 6 8 10 12 140 2 4 6 8 10 12 140 2 4Time, days

6 8 10 12 14

Fig. 9. Movement of representative marks in the peduncle. The displacement ofintracellular carbon marks along the peduncle with time is shown for the feedingregimes of o, 1, 4, 8, 12 and 25 Artemia/day (A-F, respectively). Each curve representsthe displacement of one mark.

100

CD

0 1 8 12

Feeding rate, Artemia/day

Fig. 10. Epithelial cell loss through peduncle turnover. A, peduncle turnover time (A)evidenced by displacement of tissue markers (from data in Fig. 9), and averagenumber of cells per peduncle (O). B, total no. of cells lost per hydra per day throughpeduncle turnover, calculated from the data in A.

Tissue economics of Hydra 129

the body column from the hypostome. The more an animal is fed the higher thestationary zone is located (Fig. 13). In fact, in the animals fed 25 Artemia per day thestationary zone sometimes moves above the original tentacle ring and tentacles fromthis ring are displaced proximally.

150r

100

>TJ•DauD

"IQ.

- i150

production

^•Cell loss:

I V Tentacle turnover

Budding

100

.5"3c

50

8 12

Feeding rate,Artemia/day25

Fig. 11. Comparison of epithelial cell production and loss as a function of feedingrate. The numbers of cells produced per hydra per day are shown by horizontal lines.The vertical bars represent cell loss through budding (stipple) and tentacle turnover(cross-hatch). Cell loss through peduncle turnover is negligible at this scale.

DISCUSSION

These investigations show that the amount of food eaten by hydra is a majordeterminant of the polyp's tissue dynamics. Hydra's steady state size, budding rateand bud size vary with the rate at which polyps ingest food.

We found this major influence of food to be surprising for 3 reasons. First, there isno mention in the literature or hydra's steady state size varying with feeding rate. Infact, hydra size is sometimes considered to be a species character (Forrest, 1963), but

130 J. J. Otto and R. D. Campbell

1002 4 6 8 10 12 140 2 4 6 8 10 12 140

Time, days2 4 6 8 10 12 14

Fig. 12. Movement of representative marks along the body column. The positions ofintracellular carbon marks which were placed at different levels of the body column areplotted against time for each of the feeding regimes of o, 1, 4, 8, 12, and 25 Arteinialday (A-F, respectively). Each curve represents the displacement of one mark.

0 1 8 12Feeding rate, Artemia/day

Fig. 13. Average measured positions of tissue movement stationary zones (O) asa function of feeding rate (from data in Fig. 12), and ( • ) expected positions of thestationary zones as calculated from the relative rates of proximal and distal cell loss(Campbell, 19676). The drawing at left illustrates body organization, but proportionsvaried with feeding regime; for example, in some animals fed 25 Artemia per day thetentacles were located more than 15 % of the way down the body column.

Tissue economics of Hydra 131

evidently size can be diagnostic only if current feeding levels are known. Second, wehad anticipated that budding would occur once a hydra reached a threshold size, andthat through budding a parent hydra would prevent its size from exceeding thiscritical point. Apparently the morphogenetic fields which determine bud positionsand adult size are not fixed for a species but become altered in response to feedinglevel. Finally, our findings indicate that the budding process does not consist ofautonomous recruitment of tissue by the incipient bud tip (as is implied by the studiesof Sanyal, 1966, and Yao, 1945) since bud size is proportional to parent size.

The most unexpected aspect of our data, however, was that cellular growth rate(measured as epithelial cell cycle duration) over a wide range of hydra size isindependent of feeding levels. This appears to contradict Schroeder's (1969) findingthat hydra populations exhibit a growth rate which is proportional to the amount offood ingested. Furthermore, it would appear that growth rate would necessarily beproportional to feeding rate since food ingested by hydra is converted into new hydratissue with high efficiency (Schroeder, 1969; Stiven, 1965). The solution to thisapparent contradiction lies in hydra's ability to change size in response to differingfeeding levels (Fig. 2) so that in the steady state the amount of food ingested perepithelial cell is independent of feeding. The cellular nutritional level is thereby moreor less uniform at different polyp feeding levels. Therefore, short-term populationstudies which relate the number of polyps to the bulk feeding rate may give complexdata, since both hydra size and polyp ingestion rates may be changing. The rate ofincrease of total hydra tissue is, however, proportional to the rate of food consumptionby the population (Schroeder, 1969).

How the hydra arrive at the various sizes is an intriguing problem. When ourexperimental feeding regimes were started, animals which were suddenly being fedmore than they were used to may have had a transient decrease in epithelial cell cycleduration, leading to an increase in cell number. Although the epithelial cell cycletime does not vary with feeding rate during steady state growth, there could betransient changes in the cell cycle. When starved animals are refed, there is a suddenincrease in mitotic index (Campbell, 1967a; Webster & Hamilton, 1972). In additioneven in steady state hydra the epithelial cell cycle time is variable if one considersindividual cell cycles (David & Campbell, 1972) even though the average cell cycletime is fixed. Therefore it is reasonable to expect transient cell cycle changes inresponse to altered feeding rate. There could also be transient changes in cell lossresponse to changes in ingestion rate. Any such change which was purely transientwould result in an altered hydra size which could then become stabilized if feedingconditions did not change further.

We thank Jack Dunne for help in experiments on the stationary region, and Drs Hans Bodeand Susan Bryant for their critical reading of the manuscript. This work was supportedby NIH grants MH-25252 and NS-12446 and NIH Training Grants HD-003-47-5 andGM 07-134-01.

132 J. J. Otto and R. D. Campbell

REFERENCES

BISBEE, J. (1973). Size determination in Hydra: The roles of growth and budding. J. Embryol.exp. Morph. 30, 1-19.

BODE, H., BERKING, S.( DAVID, C, GIERER, A., SCHALLER, H. & TRENKNER, E. (1973). Quanti-tative analysis of cell types during growth and morphogenesis in hydra. Wilhelm Roux Arch.EntwMech. Org. 171, 269-285.

BODE, H., FLICK, K. & BODE, P. (1977). Constraints on the relative sizes of the cell populationsin Hydra attemiata. J. Cell Sci. 24, 31-50.

BRIEN, P. & RENIERS-DECOEN, M. (1949). La croissance, la blastogdnese, l'ovog^nese chezHydra fusca (Pallas). Bull. biol. Fr. Belg. 83, 293-386.

CAMPBELL, R. (1967a). Tissue dynamics of steady state growth in Hydra Uttoralis. I. Patternsof cell division. Devi Biol. 15, 487-502.

CAMPBELL, R. (19676). Tissue dynamics of steady state growth in Hydra Uttoralis. II. Patternsof tissue movement. J. Morph. 121, 19—28.

CAMPBELL, R. (1973). Vital marking of single cells in developing tissues: India ink injection totrace tissue movement in hydra. J. Cell Sci. 13, 651-661.

CLARKSON, S. & WOLPERT, L. (1967). Bud morphogenesis in hydra. Nature, Land. 214, 780-783.DAVID, C. (1973). Quantitative method for maceration of hydra tissue. Wilhelm Roux Arch.

EntioMech. Org. 171, 259—268.DAVID, C. & CAMPBELL, R. (1972). Cell cycle kinetics and development of Hydra attemiata.

I. Epithelial cells. J. Cell Sci. 11, 557-568.FORREST, H. (1963). Taxonomic studies on hydras of North America. VIII. Description of two

new species, with new records and a key to the North American hydras. Trans. Am. microsc.Soc. 82, 6-17.

LENHOFF, H. & BROWN, R. (1970). Mass culture of hydra: An improved method and itsapplication to other aquatic invertebrates. Lab. Anim. 4, 139—154.

OTTO, J. & CAMPBELL, R. (1977). Budding in Hydra attenuata: Bud stages and fate map.J. exp. Zool. 200, 417-428.

SANYAL, S. (1966). Bud determination in hydra. Indian J. exp. Biol. 4, 88-92.SCHROEDER, L. (1969). Population growth efficiencies of laboratory Hydra pseudoligactis

(Hyman) populations. Ecology 50, 81-86.SHOSTAK, S. (1968). Growth in Hydra viridis. J. exp. Zool. 169, 431-446.STIVEN, A. (1965). The relationship between size, budding rate, and growth efficiency in three

species of hydra. Res. Popul. Ecol. 7, 1-15.WEBSTER, G. & HAMILTON, S. (1972). Budding in hydra: The role of cell multiplication and

cell movement in bud initiation. J. Embryol. exp. Morph. 27, 301-316.YAO, T. (1945). Studies on the organizer problem in Pelmatohydra oligactis. I. The induction

potency of implants and the nature of the induced hydranth. J. exp. Biol. 21, 147-150.

{Received 31 January ig'jy-Revised 2 May 1977)