INTRODUCTION Tricolia pulloides is an abundant intertidal prosobranch gastropod inhabiting many algae of the mid-intertidal zone. In the area studied around Hopkins Marine Station of Stanford Univer- sity, Pacific Grove, California, most T. pulloides inhabit the abundant red alga, Gigartina papillata. Combined abundances of organism and habitat provide ample opportunity for the study of T. pulloides and its exploitation of various microhabitats within G. papillata. Goals of this study were: 1) to determine changes in distri¬ butions of populations of Tricolia pulloides on Gigartina papillata during changing conditions of light, surge, and exposure in the field; 2) to examine isolated responses to changes in each physical factor in the laboratory; 3) to explain observed field and labora¬ tory behavior in terms of avoidance of possible stresses brought on by these changing physical factors. MATERIALS AND METHODS Field Study Investigation of Tricolia pulloides' behavior began with an intensive field study of populations inhabiting pairs of algae at 1.6 m, 1. 8 m, and 2.0 m above mean lower low tide. Distributions of these populations were observed every 2-3 hours through one complete tidal cycle beginning at 1800 on 2 May. Algae were upright, 3-5 cm high, located on a sloping granite face with no adjacent algal cover or sediment. Waves rarely broke on the area, but there was strong surge at times. Each alga was inhabited by 20-50 snails. These populations were supplemented by approximately 20% to amplify distributional shifts. Population densities remained within reasonable limits, however; many other algae the same size were inhabited by nearly a hundred snails. Distributions of Tricolia pulloides within Gigartina papillata in the field and laboratory were quantified by subdividing the alga into three regions: 1) outer region, comprised of all broad, thin, leafy frond ends beyond at least one major branch point; 2) inner region, comprised of granite substratum immediately beneath the alga and thick, narrow frond stems up to the first major branch point; 3) middle region, comprised of all remaining frond area. Radial distributions of populations in all experiments were charac¬ terized by recording the number of snails in each of these areas. In the field study, exact counts were obtained of the number of snails in outer and inner regions. Numbers for middle regions, however, were estimated. The difficulty of gently searching through an alga with fine forceps while on SCUBA during strong surge made exact counts of larger middle populations virtually impossible. All other counts were exact. Several distinct changes in distributions of Tricolia pulloides populations were observed during the field study. However, environ¬ mental factors, particularly light, surge, and exposure, were fluctuating concurrently, so that distributional shifts could not be attributed conclusively to changes in any single environmental factor. Controlled laboratory experiments were necessary to deter¬ mine important environmental parameters and their effects on radial distributions of T. pulloides populations. Laboratory Experiments Response to Light: Initial laboratory experiments tested for a response by Tricolia pulloides on Gigartina papillata to light from above and from the side. In this and other laboratory experi¬ ments, snails were collected from the field, marked with silver nail polish, with no apparent detrimental effects, and submerged in 13°C sea water with G. papillata for 24 hours prior to the experiment. Four algae, similar to those chosen for the field study, were collected from the field complete with granite chip and holdfast. The algae were cleared of biota, anchored in model¬ ling clay, submerged, and stocked with 25 T. pulloides. Each system of four algae and 100 gastropods was equilibrated in darkness. After sixty minutes, numbers of T. pulloides on sides of fronds to be illuminated were counted under dim red light. Thirty watt microscope lamps, positioned 10 cm above one system and 10 cm to the side of the other, were turned on. Numbers of T. pulloides on lit sides of fronds were recorded at thirty minute intervals. The vertical experiment was terminated after 60 minutes. The lateral experiment was terminated after 150 minutes. Because no significant phototactic response was detected during the laterally-lit experiment, despite a significant downward popu¬ lation shift in the vertically-lit experiment (see RESULTS), a second laterally-lit phototaxis experiment was conducted, in an artificial environment. At approximately 1100, 1300, 1900, and 2300 on 1 June, roughly 100 snails were submerged in clear plastic jars filled with sea water. Jars were placed on a lab bench two meters from a window. The water was swirled until Tricolia pulloides were scattered about the jar bottoms. After 30 minutes approximately 80% of the gastropods had climbed or were climbing the jar walls. Paper divided into eight sectors of 45° each was taped around the jars. Numbers of gastropods on the jar walls in each sector was recorded. Response to Surge: Tricolia pulloides'response to surge also was tested in the laboratory. Surge was simulated by shuttling a plywood board attached to PVC pipe back and forth in an aquarium filled with 13°C sea water. On 21,22, and 23 May, five algae were collected from the field, cleared of biota, anchored in modelling clay, and tacked upright to the aquarium floor. One hundred Tricolia pulloides collected and marked the previous day were placed on the submerged algae.- After 120 minutes equilibration in calm water, radial distribution of T. pulloides on each alga was recorded. The plywood board, which extended about halfway down the water column, was moved back and forth across the entire length of the aquarium. Manual surge was accelerated gradually to ten cycles per minute. After 30 minutes the radial distribution of each population was recorded. A minute of manual surge separated each reading, to prevent unre¬ corded populations from shifting. After 30 minutes calm, distri¬ butions were recorded again. Response to Exposure: Finally, exposure was tested in the laboratory for its effect on radial distributions of Tricolia pulloides populations. In order to simulate the natural tidal cycle, a clock was positioned above an aquarium, with the aquarium drain hose attached by string to the hour hand of the clock. As the hour hand made a revolution every 12 hours, the drain hose, and hence aquarium water level, rose and fell. Thus algae gradually were exposed and submerged twice daily. On 1 June three pairs of algae were collected from the field, cleared of biota, and anchored in clay. One pair was placed on the aquarium floor, one pair on a brick, and one pair on two bricks in the aquarium. At 1700 each pair was stocked with 50 snails collected the previous day. Algae were submerged completely, in synchrony with the high tide at 1711. Populations were allowed to equilibrate until 1800, at which time radial distributions were recorded. Observations were made under constant light every two hours, through two high and low tides. Statistical Analyses Row-by-column contingency tests (Zar,1974) were performed on most data. Outer, middle, and inner regions or lit and shaded sides were rows. Levels of physical factors were columns. Thus RxC tests indicated levels at which distributions, rows, were con¬ tingent upon levels of physical factors,columns. A chi-square test (Zar,1974) was used to compare distribution around jar wall for the dark experiment to an even distribution. RESULTS Field Study Radial distributions recorded during field studies are shown in Figure 1. An RxC test was performed on data for submerged populations with subjective estimates of surge levels, calm, mod¬ erate, and rough, as columns. Radial distributions, rows, were contingent upon surge levels (p £ 0.001). An RxC analysis of inner and outer data, eliminating estimates of numbers of snails in middle regions, also indicated radial distributions were contingent upon surge levels (p + 0.001). An RxC test also was performed on all field data, with exposed and submerged as columns (see Figure 1). Radial distributions, rows, were contingent upon presence of water (p £ 0.001). Elimination of estimates of populations in middle regions produced similar results (p « 0.001). Laboratory Experiments Response to Light: Figure 2 shows distributions on lit and columns indicated radial distributions were independent of prior surge on 23 May (p » 0.25). On 22 May, though, distributions were contingent upon prior surge (p £ 0.025). However, it appears that atypical initial distribtuions on 22 May might have been the source of unexpected difference between initial and final distributions. Response to Exposure: Radial distributions for the exposure : experiments are presented in Figure 5. RxC tests with exposed and submerged as columns indicated radial distributions were contingent upon submersion (p £ 0.001). Interestingly, when data for the one alga which did not remain upright upon exposure is discarded, the G value nearly doubles (21.3 vs. 38). Qualitative observation showed gastropods on this alga moved inward to a lesser extent when exposed than did gastropods on those algae which remained upright. DISCUSSION In all experiments, Tricolia pulloides populations shifted to outer regions of Gigartina papillata during optimum conditions of calm submersion in dim light. Mooers (1979) has shown diatoms to be T. pulloides' principal food item, and Chun (1979) has shown that diatoms and T. pulloides' grazing tracks are found almost en¬ tirely on outer regions of a similar red alga, Rhodoglossum affine Thus, under optimum conditions, T. pulloides populations may move to outer areas of G. papillata to feed. Desiccation is a considerable threat to Tricolia pulloides: an 1D50 of 18 hours was reported for laboratory exposure (Firestone, 1979), and may be considerably less in bright sunlight. Thus, negative phototaxis seen in natural and artificial surround¬ ings would seem an adaptive response. The sensitivity of this response is indicated by jar experiments: areas of highest snail density on jar walls shifted east as areas of incident sunlight shifted west during the day (see Figure 3). Similarly, T. pulloides response to exposure seen both in the field and the laboratory would minimize the period in which the snail is entirely without moisture and subject to eyaporative water loss. Inward movement of Tricolia pulloides during strong surge minimizes the threat of being swept off the alga, since inner regions swing through a smaller arc and with less tangential velocity than frond ends. During trial surge experiments snails on outer areas of algae often were swept off if strong manual surge was begun suddenly. Thus, inward movement of T. pulloides populations during surge also appears to be an adaptive response. FIGURE LEGENDS Figure 1: Radial distributions of approximately 150 Tricolia pulloides on 3 pairs of Gigartina papillata at different tidal heights. Upper, middle, and lower sections of histograms represent, respectively, outer, middle, and inner algal régions. Widths of histogram sections are proportional to numbers of snails in appropriate regions of algae. Approximate total number of snails observed per reading shown below. Populations above dashed line were exposed when observed. Populations below dashed line were submerged when observed. Time line shown, black bar indicates darkness. Figure 2: Summed distributions of 100 Tricolia pulloides on lit side, clear bar, and shaded side, hatched bar, of fronds of four Gigartina papillata before and after exposure to light from above and from the side. Sizes of clear and hatched areas are proportional, respectively, to numbers of snails on lit and shaded sides of fronds. Exact numbers enclosed. Significance levels (p values) are from RxC contingency tests (Zar,1974) performed to determine if snail distributions are independent of lighting. Figure 3: Distribution of Tricolia pulloides populations around walls of clear jars after 30 minutes in sunlight. Total numbers of snails per experiment indicated in parentheses. Thicknesses of sectors are proportional to numbers of snails on jar walls in each of eight 45° sectors. Sig¬ nificance level (p value) for night experiment is from a chi-square test comparing this distribution to an even distribution. Significance levels (p values) for morning, afternoon, and evening experiments are from an RxC con¬ tingency test (Zar, 1974) performed to determine if snail distributions around jar walls were contingent upon directional sunlight. North indicated by arrow labeled N. Direction of incident sunlight shown by unmarked arrows. Figure 4: Summed radial distributions of 100 Tricolia pulloides on Gigartina papillata after 60 minutes calm, 30 minutes manual surge, and 30 minutes calm on three consecutive days. Upper, middle, and lower sections of histograms represent, respectively, outer, middle, and inner algal regions. Widths of histogram sections are proportional to numbers of snails in appropriate regions of the 5 algae. Actual numbers enclosed in appropriate sections. Significance levels (p values) are for RxC contingency tests (Zar, 1974) performed to determine if radial distributions were contingent upon manual surge. Figure 5: Radial distributions of 150 Tricolia pulloides on three pairs of Gigartinä papillata during a 24 hour simulated tide cycle. Upper, middle, and lower sections of histo¬ grams correspond, respectively, to outer, middle, and inner regions of pairs of algae placed at high, medium, and low levels in an aquarium. Nidths of histogram sections are proportional to summed numbers of snails in appropriate regions of the 3 pairs of algae. Actual numbers enclosed in appropriate sections. Populations above dashed line were exposed when observed. Populations below dashed line were submerged when observed. Time line and date shown below. LITERATURE CITED Chun, Stephen A., 1979. Qualitative and quantitative effects of herbivore grazing on the epiphytes on Rhodoglossum affine. Unpublished manuscript on file at Hopkins Marine Station, Stanford University, Pacific Grove, California. Firestone, Alan L., 1979. Distribution, abundance, and size-class distribution of the prosobranch gastropods Tricolia pulloides and Barleeia haliotiphila in the rocky intertidal zone at Mussel Point and Point Pinos. Unpublished manuscript on file at Hopkins Marine Station Library, Stanford University, Pacific Grove, California. Mooers, Mary G., 1979. Diet and reproductive biology of the rocky intertidal prosobranch gastropod Tricolia pulloides. Unpub¬ lished manuscript on file at Hopkins Marine Station Library, Stanford University, Pacific Grove, California. Zar, Jerrold H., 1974. Biostatistical analysis. Prentice-Hall, Inc. Inglewood Cliffs, New Jersey. 41-54 pp. 6 O 5 — L —.—.—— — — — L ———-—.—.—.—.——.—— — — — — — — " — 3- o- — — — PHOTOTAXIS ON THE PLANT LIGHT FROM ABOVE Distribution after 60 minutes Distribution prior to light (equilibrated 60 minutes) exposure to light from above 46 32 lit side — P.05 shaded side 54. 68 etel LICHT FROM THE SIDE Distribution prior to light (equilibrated 60 minutes) 134 66 P9.25 lit side shaded side Distribution after 150 minutes exposure to light from the side 139 6 A morning (n=105) afternoon (n=75) evening (n=91) night (n=85) PHOTOTARIS INAJAR top view after 30 minutes l . N p00 POO POO p).25 4 . 5 6 QO O C 0 O E 4 12 — 1 1 30 — 1 5 a 10 — C 0 1. o 8 09 o — 1 9 — 5 -— — c A 4 4 4 — — 1 — ——.—.—.— --—.—.— — —.—.—.—.—.— . — — — ——.—.—.—.—.—.——---. —---—.—. —.—. —. —. —. — — —.—.—.———— —; 1 ——.—.—.—.—-- . —.—.—. — — 7- 1 1 L — — — 1 — Ic 1 L — — 1 — — —