-1- ABSTRACT hey Cirriformia spirabrancha, a marine polychaete, was proven to exist in aggregated densities in two field study areas lo- cated in Monterey Bay, California. The data was gathered by taking cores of the substrate and counting the C. spirabrancha in each sample; this method lso proved valuable in determining the density and geographical description of various aggregates. Many environmental factors were eliminated as effectors of aggregation since they were not able to vary significantly in the short distances involved between populations of very high densities (70/80 cm) and populations of very low densities. Other environmental variables that could vary with small changes in distances, organic content and particle size, were proven to have no correlation with variations of the worm's population density. 3 -2- INTRODUCTION The Cirratulid polychaete Cirriformia spirabrancha (Moore 1904) is a common inhabitant of the sulfide-rich sand and mud flat areas of Monterey Bay, occuring both intertidally and sub- tidally. Preliminary field observations indicated that popula- tions of this worm show a patched distribution over much of the intertidal zone. Although often assigned to the sub-class Sedent- aria, individual worms exhibit an ability to move considerable distances per unit time (Henderson, 1968); a distance which is often greater than the dimensions of the aggregates themselves. Little has been written concerning the population struct- ure of annelids and, in particular, marine polychaetes. Pre- vious work on C. spirabrancha by Maginitie (1935) in Elkhorn Slough, described populations in general terms only. This is also true fer the work of both Courtney,1958, and George,1964, on the English Cirratulid, Cirriformia tentaculata. The present investigation was initiated to answer two basic questions: de populations of Cirriformia, in fact agg regate, and if so what is the mechanism? The results of this study indicate aggregation does occur in this species and suggests a mechanism involving an attracting substance and the necessity of obtaining threshold population densities before aggregation 3 -3- can occur. MATERIALS AND METHODS The main field study area was established under Fisherman's Wharf Monterey, California; where populations of C. spirabran- cha were found at the O.O tidal level and lower. The presence of piers in the area facilitated establishing a grid pattern which could be utilized for statistical work, this also made it possible to record exactly where each sample was taken. Another area, about 400 yards from the wharf, where worms were found at the +1.0 tide level and lower, was also investigated exten- sively. Counting the worm's tentacles as a measure of population density was found to be infeasible due to the crowding often en- countered which made quantification difficult at best. Further- more, it was discovered that only about half of the total popula- tion in any given area displayed their tentacles above the sand at any time. (Henderson, 1968). At the same time, it was found, by otserving populations of Cirriformia established in special outdoor tanks at the Hopkins Marine Station, that O, content and light can affect the number of tentacles showing. (fig 1). For these reasons, sampling was accomplished by taking cores of a constant volumee (2.2 liters), and sifting the material in a 2.5 35 -4- mm mesh strainer leaving essentially all the Cirriformia to be hand counted. The worms were never found below 25 cm (Hender- son 1968, and Kee, 1968) thus, since each core was 28 cm deep, the worms counted were considered to represent the total number of animals below the cross-sectional area of the core, 80 cm. To eliminate any possible differences in population density due to tidal movement, all cores were taken at low tide when from 5 to 25 cm of water was standing over the substrate. To obtain information concerning the mean population den- sity of the sample area, cores were taken at regular intervals over the sample area starting from tidal mean (O.0') to -2.0 at both low and high tides (fig 2). The curve obtained at low tide was used as a standard with which to compare all additional data. However, it should be noted that the curve obtained at high tide is similar (fig 2). To insure randomness in testing for aggregation, the natural grid provided by the presence of the piers was used. The area was divided into 108 squares and these were numbered. 24 numbers were chosen from a table of random numbers and each core was taken midway between the piers in the square correspondingly numbered. Population densities were determined and the variance was found by the following method: 391 -5- f (x-m) 2. L n-1 x = number of worms per 80 cm f - frequency of cores with a value ofx m mean number of animals per 80 cm of the total number of samples taken n= total number of samples taken Two additional series of cores were taken; one every 2 1/2' perpendicular to the shoreline (-.3' to -1.5' tide level), and the other every 4' parallel to the shoreline (-.8' tide level). All samples were obtained in the area of high mean density (fig 2). Sand samples were taken with a smaller coring device (dia- meter 4 cm) to a depth of 20 cm. The population densities of each core were recorded and the sand cleared of all macro organ- isms and dried for 24 hours at 105°0. The samples were then sifted through a series of Tylor Screens and the particle size recorded in terms of ødry weight. Organic analysis was accomplished by the Walkley Black titration method (Morgans, 1956). 2 grams of dried sand were ground in a mortar and sifted to remove particles and organisms above .5 mm in diameter. 10 ml of IN KCr,O, and 20 ml of concen¬ trated sulfuric acid were added to each sample. This mixture was for 1/2 hour and then 200 ml of distilled water was added along 355 -6- 10 ml of 85% Phosphoric acid and 1 ml of the indicator diphenyl amine. FeSO, (approx IN) was titrated against each sample and standardized with a solution containing no sand. The total available organic carbon present was determined according th the following formula: V1 -V. X .003 X 100 = % org C V, - volume of FeSo Va - volume of KCr,O, .003 - conversion factor of ml to grams of C since 3 mg of C is taken up by 1 ml of IN KCr,O, The final value was multiplied by 1.3 to account for the 75% efficienoy of the technique (Morgans, 1956). A factor of 1.8 was used to transform the data from % organic carbon into % or- ganic matter available (Morgans, 1956). To test for threshold levels of aggregation and possible mechanisms attracting the worms, two small boxes (18" X 18" X 6") were constructed and placed in an open concrete tank at the Hop- kins Marine Station. Various substrates were placed in these experimental boxes and worms set in different patterns on top of the substrate and then left for 5 days. After 5 days, a masonite 38 -7- grid was sunk into the sand and the worms counted in each square. RESULTS Random sampling of the test area (24 samples) resulted in a variance to mean ratio of 23.56 or s'/m - 23.56 (fig 3). Ac- cording to the test, the variance to mean ratio may vary from a value of 1.0 for a randomly distributed (Poisson) population to values less than 1.0 for uniform populations and values greater than 1 for populations tending towards clumping. The variance to mean ratio recorded far exceeds the value denoting aggregation. Aggregation was likewise suggested by other statistical methods: 1) "k" of the negative binomial (Anscombe, 1949) k--1.4 where: k = 7-8 = poisson dist. s - X k = (7-8 denotes agg. X = mean se - variance 2) chie test X S (-1) s 24/4 where: Poisson dist. = 1 regular dist. =0 N= number of samples agg. dist.)1 Two additional samples, one run perpendicular to the shore (-.3 to -1.5 tide level) and one parallel to the beach (-.8' tide level) gave similar results; these had means of 32.09 and 357 -8- 40.13 and variance to mean ratios of 21.1 and 23.2 respectively. The means of all three sample series, then, were in close aggree- ment with those obtained earlier (fig 1). The results of series of samples taken across particular aggregates show that the number of worms per unit area varies little within an aggregation. Samples taken within aggregates had a mean of 63 animals/80 cm with a standard deviation of 6.3. Few samples were found with densities between 45 and 60 worms per unit area, thus 60 worms/80 cm was established as the lower limit of aggregation and 45 worms/80 cm as the highest mean density. The aggregates vary from 1 to 15 feet in length, W generally with a "fringe" area surrounding the aggregation. This fringe area fluctuatesbetween 4" and 2' in width often with another aggregate immediately adjacent or, conversely, it may be next to an area of very low density. Thus areas of aggregation may be within 4" of areas with few or no worms. No aggregation patterns as such were observed with the ex- ception to a relationship between the surface troughs and crests produced by wave action. Samples taken in a trough often would be within the aggregation limits and samples taken in adjacent crests would often centain appreximately 1/2 the number of animals per unit area as the adjacent troughs. Since the worms live in 3 the mud,the troughs would bring the worms closer to the surface for respiration and feeding purposes. Although this relation- ship was found extensively throughout the test areas, a signifi- cant number of aggregations appeared where the interface between white and black sand was relatively deep. likewise, some troughs contained few or no worms and, in some instances, an aggregation's border crossed a trough. Hense the observed relationship may not be as significant as originally thought. Since areas of aggregated Cirriformia and non-populated areas can be adjacent, many environmental variabbes such as tidal height, temperature, light intensity, salinity, and O, content must be eliminated as potential effectors of aggregation because significant differences would not be expected to exist over such small distances. However, two substrate properties were consid- ered since variations within small areas could occur. These two factors were particle size and organic content. Particles size: Within a limited area, cores were taken in aggregates, in fringe areas, and in areas with few or no animals. Figure 4 shows that there was no correlation between the sub- strate's particle size and tke pepulatien density of Cirriformia in the test area. Organic content: Figure 5 shows that variation in popula- 35 -10- tion densities of the substrate tested had no correlation with the variation in organic content. It was suggested during the course of the work that agg- regation might be due to chemical attraction in that even when animals were put in sea tanks under conditions where all obvious differences in the substrate and other environmental variables were eliminated, the worms would aggregate. To test this hypothesis, boxes were constructed in a courtyard pool at the Hopkins Marine Station. Two types of substrate were put in each box. Half of each box was filled with sand taken from aggregated areas than completely cleared of organisms and the other half was filled with sand collected from areas containing no populations of Cirriformia. However, in one of the boxes. bags containing Cirriformia extract were buried in the half containing sand that had come from areas with no worms. 250 worms were placed on the surface of the sand in the middle of The woons i each box. This was accomplished by placing,the center square of a masonite grid which was removed after 10 minutes when the animals had burrowed into the substrate. After 5 days the grid was sunk into the box and the number of worms in each square counted( fig 6). The results were: 1) few or no worms iould 36 -11- enter an area unless Cirriformia had previously occupied the sand and 2) the above relationship could be overcome by add- ing bags ofthe Cirriformia extract. One factor that could have attracted the Cirriformia te the sand from aggregated areas would be the organic content; however, Figure 5 shows that the sand from non-aggregated areas actually had sohigher level of or- ganic content. Hense some attracting substance may be given off by the adult organisms which stimulates aggregation. Although the worms seem to attract each other, a threshold density per unit urea may well have to be reached before agg regation can occur. In two additional experiments, substrates taken from areas populatid with Cirriformia were placed in the courtyard boxes and worms were subsequently put on the surface in twe different patterns: one representing a clumped distribu- tion andthe other representing a regular distribution (fig. 7-a,c). The worms were left for 5 days and then dug up (fig 7-b.d). In neither case did aggregation occur; however, when the total pop- ulation density was calculated for the worms in the two boxes. it was found that they were far below the levels associated with aggregation in the field (2.6 and 4.1 worms/80 cm). While in one box (B) the worms were initially aggregated in the center (62 worms/80 cm"), the worms nevertheless dispersed within the -12- 5 day period suggesting the absence of any chemical attractand even in high density populations. However, it is quite likely that any chemical attraction could have been hargely lost by dilution as it was necessary to wash the worms beføre counting them; hense no aggregation would be expected to occur during the initial period of the experiment. DISCUSSION Aggregated areas consistently range between 60 to 80 worms/ 80 cm with the mean of aggregation close to 70 worms/80 cm. Furthermore, 80 animals/ 80 cm appears to approximatesthe high- est density in the aggregates, which seems reasonable since this density represents 1 worm per gquare om of surface area of substrate. The mechanisms of formation of aggregates of Cirriformia ire not known and present data is insufficient to solve the mystery, However, 3 possible theories will be discussed according to how well they could apply to the worm's behavior as determined by field observations and the courtyard experiments. 1) Cirriformia may, like many annelids, have a preferent- ial larvae settlement pattern and, in this way, build up seper- ate aggregates. The center of the population could get so crowded tähttthis portion of the population has no possible way of moving. Some worms on the outside of the aggregation would be able to break 36e -13- off the main aggregate thus explaining the fringe area. How- ever this theory does not explain why, in the courtyard, the worms aggregated within 3 days (fig 1). 2) Cirriformia may, at some critical density actively seek each other by means of chemical attraction. In this case a threshold level or organisms would be needed to initiate aggregation which would explain why populations tend to agg- man regate less in density areas. A problem arises when con- sidering how, fringe areas would be formed when the other areas were clumping. 3) The third possibility is one of combining the pre- vious two theories. The worms may first preferentially settle as larvae, over the aggregations and later, as adult worms, gxcrete a chemical substance to keep the aggregae intact. Since many of the biological systems of Cirriformia are not Gdesa completely knewa, possible reasons for aggregation usually have to be overlooked. For instance, the worm may be aggregating for feeding purposes, but since it is not known what Cirriformia eat, inferences tying aggregation and feeding behavior would be useless. During sampling, it was noticed that aggregations were often free of other organisms. Judd's (1968) work of predation Sor -14- on Cirriformia suggested that few animals would eat the worms. The obvious small of Cirriformia was found to be produced by the presence of a substance tentatively described as a primary amine. Purified extract retained the ordor of Cirriformia and pieces of Notemastus soaked in the solution were rejected by very hungry fish. Controls were eaten readily. Therefore a possible res- son for aggregation is to concentrate the worms for a secretion used as a defense mechanism. It may be that, for purposes of biological economy, the same substance that is protecting Cirriformia from possible predators is also attracting Cirriformia to each other. -15- SUMMARY 1. Cirriformia have been shown to aggregate under natural conditions where aggregated densities range from 60 to 80 worms /80 cm. 2. There seems to be no specific pattern of the aggregates themselves, other than a weak correlation to the closeness of the dark layer of sand to the surface. 3. The number of worms per unit area stays roughly constant throughout an aggregate as does the 2:1 relationship between the density of the aggregate and the "fringe" area surronnding it. 4. Light, O,ccontent, tidal height, salinity, and temp- erature were eliminated as factors effecting aggregation. Part- icle size and Organic content were likewise shown not to in- fluence aggregation. 5. Adult Cirriformia appeared to be attrated to extract of the smme species. 6. A possible reason for aggregation of Cirriformia would be to concentrate a defense secretion; this secretion could also be an attracting substance in the species. 36. O ACKNOWLEDGEMENT This work was supported in part by the Undergraduate Research Participation program of the National Science Foundation Grand GY-4369. I would also like to thank the many others in the program whose help proved to be invaluable. A special thanks to Dr. Welton Lee, whose assistense was greatly appreciated. 366 1. (first three pictures) Cirriformia show more tentacles as the light intensity lessens. (second two pictures) an air hoseeintroduced over the substrate resulted in far fewer worms showing their tentacles after 2 days. 2. Pupulation survey of main field area (each point repre- sents 6 cores taken at the same approx. tidal region. 3. Results of Random Slampling. A = lower limit of agg- regation, B = Mean of area (fig 1), C = Mean of these samples. 4. Particle Size; A = aggregated area, B = aggregated area, C = fringe area of A and B, D = area with no worms. 9. Percent (dry weight) total available organic matter. A - sample taken from the top of the core B = sample taken from the bottom of the same core. C = sample taken from a mixed core. Beavy Dark areas = aggregated densities, Medium - Fringe populations and Clear areas = no worms present in the substrate. Middle group taken 4" from each other and last group is the two sands used in fig 6. 6ig 6 Worms set in middle of the boxes (a and b) with b containing 7 extract bags. Dark side = sand from areas with no Cirriformia, Light side = sand taken from aggregated areas. 7. 125 worms set in A in a regular pattern and densly in B. 36 3 JHOHS 1SIG ZUOH 3011 K0T - 01 H9IH — O — LHSISH TVOI1— 8- 1 i11fII vas -01 -07 -08 -Ot -05 -04 6 9 JHOHS 22 — O1 -08 -OE -Ot O9 -09 02 -08 D 0 12 6 AJaWON WVS 2 1116 v3S m N in 3 CUM % (DRY WEIGHTI 0» 0 60 372 OSCANC MATESAL ( E DRT UEICHT Prro PEREPEPPi 37 6.4 272.0 ... ... . ... *... .*. * - * * . ... . ..... ..... .... .. 3.7 5 DAYS 76 272.0 .. .*. ... - ... ... . .. - . .. . * - 60 2 . . . * * ... .. .. .. O — :: - .769 . . .. *. .. .* 5 DAYS . . . 124.0 . . .. - 2.2 378 O REFERENCE Anscombe, T. J. (1949) Diometrics 5, 165-73 Courtfey, Williamina (1958) "Certain Aspects of the Biology of Cirratulid Polychaetes" submitted doctorate thesis George "On Some Environmental Factors Affecting the (1964) Distribution of Cirriformia tentaculata at Hamble." Journal of the Marine Biology Association United Kingdom 44, 383-388 Henderson, P.E. Personal contact Kee, Bill Personal contact Magininitie "Ecological Aspects of a California Marine Estuary" (1935) The American Midland Naturalist 16, 5 pp 629-764 Morgens (1956) "Notes on the Analysis of Shallow-Water Soft Substrata Journal of Animal Ecology pp367 76