THE INTERACTIONS BETWEEN THE TIDEPOOL AS AN ORGANISM AND TWO COTTID SPECIES CLINOCOTTUS ANALIS AND OLIGOCOTTUS SNYDERI Kelly E. Jensen Biology 175H June 9, 1989 ABSTRACT This study describes the tidepool environment along the shore of Hopkins Marine Station of Stanford University and relates it to the population and distribution of the cottid species Clinocottus analis and Oligocottus snyderi . The relationship between the cottid mass per tidepool volume and the percent of oxygen consumed by cottids in a tidepool is approximated by the equation: 202-0.0023* (mass/vol) 60/ where oxygen is expressed in grams, but algae seems to be the primary consumer of oxygen. It is suggested that a potential problem for fish is having algae in the same tidepool as a means of protection and as a competing oxygen consumer. In addition, tidepool height, volume and range in temperature are discussed as they relate to the cottid population. INTRODUCTION Tidepool ecology is especially interesting due to the great fluctuations in tidepool environmental factors during the course of à tide. Every time a tidepool is separated from the ocean, it is subject to changes in temperature, oxygen levels, salinity, pH, and the affects of different organisms that entered it during the previous high tide. In addition, tidepools generally have a rich diversity of species related to their small area, a situation that could lead to potential competition for any of a number of resources: oxygen, nutrients, food sources, space, and light. This study raises the question; what role do these physical factors play in determining the species composition and abundance in tidepools? The family cottidae (sculpins) was chosen as the subject of this study because of the abundance of these fish in the tidepool and nearshore areas of the California coast (Bolin, 1944 and Eshmeyer, 1983). This study is designed to describe the cottid population and the tidepool environment along the coast of Hopkins Marine Station and to use this description to illustrate a number of factors that might be affecting where cottids live. The investigation focuses its attention on specific tidepool parameters that provide possible explanations for the cottid distribution, and these explanations are presented with the idea that tidepools and cottids are two interacting organisms. MATERIALS AND METHODS The study site at Hopkins Marine Station of Stanford University (36 38' N,121 56' W) consisted of fifteen tidepools which were chosen based on the following characteristics: complete isolation from the ocean at low tide, difference in gross morphology, and the ease with which an investigator could measure the physical characteristics of the pool. The heights of the pools in relation to mean low low water were measured using a transit. Tidepool surface areas were estimated by dividing the top of the pools into rough polygons and calculating the sum of the areas of these polygons. Volumes were measured by bailing part of the pools and siphoning the rest of the water. A marked bucket and a 500 milliliter graduated cylinder were used to record the amount of water actually removed from the tidepools, and the volume of any residual was visually estimated. The water removed was immediately replaced with fresh water. Carbon dioxide was bubbled through the pools for fifteen to twenty seconds to stun fish, which were then collected with dipnets. Fish were taken into the lab for length, mass and species determination and, in most cases, were immediately released into the tidepool in which they were captured. Fish from three tidepools were kept in the lab so that a recruitment study could be conducted. Recruitment was measured in two tidepools. Each was emptied of cottids when initially censused (May 12, 1989 and May 15, 1989), and subsequent habitation was followed daily and continued through June 1. 1989. A daily lowtide census was taken for each pool by capturing fishes with a dipnet, measuring their length, and determining their species. To reduce the amount of disturbance, fish were returned to the tidepools immediately following the census, and no carbon dioxide was used to anesthetize the fish. The lengths of the fish were converted to mass by using the empirically determined regression of the log of cottid length (cm) to the log of cottid mass (grams): LENGTH - 0.0098 * MASS3.10 (R2 - 0.962, p « 0.001), calculated from the length and mass of each cottid found in the fifteen tidepools originally censused. The total biomass of fish found at each census was calculated by adding the single cottid masses. Physical data were taken for six of the fifteen tidepools; readings were taken at low, mid, and high tides over a period of two days (May 1, 1989 to May 3, 1989). Dissolved oxygen and temperature were measured with a YSI meter (model 54), pH was measured with a Markson digital pH meter (model 90), and sälinity was determined by a Wescor 5500 vapor pressure osmometer. The mean level for each of these parameters was calculated from the readings taken when the pools were not inundated with fresh seawater. Possible relationships between the different parameters for the tidepools were determined using a linear regression technique. Single regressions were done to examine the correlation between two parameters independently from the remaining parameters; a correlation was considered significant when the R value was significant at the 0.05 level. The SPSS multiple regression program was used to determine the independent variables which most significantly affect one dependent variable. Variables included: volume, surface area, height, mean oxygen level, mean temperature, mean phl, mean salinity, number of cottid per pool. mass of cottid per pool, mean length of cottid per pool, and the mass of fish/pool volume. One multiple regression was calculated with number of fish per tidepool as the dependent variable, and the other with cottid mass per pool volume as the dependent variable. The relative importance of the independent variables was determined by examination of the significance of the T test in the regression output; only variables which produced a significance greater than or equal to 0.05 were considered in the equation. The rate of oxygen consumption by the representative organisms in a tidepool was determined experimentally. 59.54 grams of C analis were placed in a fourteen-liter tank of sand-filtered seawater. Oxygen readings were taken at intervals of thirty minutes for four and a half hours. The rate of oxygen depletion in grams per hour was determined by the following calculation: 1.umles/liter0.5733timeu 541 r- 0.984). The slope of this regression (-0.5733 umoles/ liter/hour) is the rate of total consumption. 2. umoles,/liters/hr 32 ug/umoles' 1 g/10ug 10° liters/m m of tank -g0 consumed/hour 3. For a given cottid biomass: go, consumed/hour/g cottid. The percent of oxygen consumed by cottids in an actual tidepool was calculated as follows: 1. g 0, consumed/hour/g cottid' g cottid in pool time interval between readings - g 0, consumed by cottids. 2. difference between initial and final oxygen readings umoles/liter * 32 ug/umole * 1 g/10° ug 10°liters/m2 * m2 of tidepool - total g0 consumed in tide pool. 3. g 0 consumed by cottid/g 0» consumed in tidepool' 100 - percent of 0, consumed by cottids. In order to eliminate the complicated effects of photosynthesis on oxygen consumption, oxygen readings taken at the beginning and end of one night (3:24pm May 3, 1989 and 4.30am May 4, 1989) were used in the calculations above to détermine the amount of oxygen consumed by cottids in six tidepools. A similar experimental design was used to determine the oxygen consumption of invertebrates and algae in one of the six tidepools used in the cottid experiment. The experimental biomasses were 642.6 grams of invertebrates and 333.1 grams of algae (Appendix 1). Two, twenty-liter tanks were used, and readings were taken every three hours for a twelve hour period. The regression equations for oxygen consumption were 02lumoles/liter - 7.22-0.48333'timehours (R2 - 0.896, p « 0.05) for invertebrates, and 102lumoles/liter - 8.42-0.495'timepours (Ré- 0.882, p - 0.05) for algae. Further oxygen readings were taken in two pools every hour for a seven hour period on a sunny day. Two readings per pool were used to calculate the rate of oxygen consumption in the pools and the percent of oxygen consumed by cottids during the day as opposed to during the night. RESULTS SECTION ONE: THE POPULATION Table 1 briefly describes tidepools in which fish were and were not found. Two species of cottid were found in this study; Clinocottus analis and Oligocottus snyderi (Figure 1). Of the sixty-nine fish captured, 904 were C analis (Figure 2). Six of the seven O. snyderi were found in the same tidepool; this tidepool contained no C. analis Figure 3 shows the mean lengths of the total cottid population and of each species separately. The mean lengths do not differ significantly (p»0.05), but C analis has a much greater range (1.8-11.7 cm) than O. snyderi (2.6-5.1 cm). The middle 70 of the total population of cottid. C. analis, and Q snyderi fall in about the same range (3.1-5.1 cm). The mean mass does not differ significantly (p»»0.05) between species (Figure 4). The range in mass for C analis is again much greater than O snyderi and the range in mass for the middle 70% of O. snyderi (0.5-1.3 g) falls in the lower part of the range for the middle 702 of C. analis (0.4-1.8 g). All of the single linear regressions for tidepool parameters were calculated to determine any independent correlations between parameters. Many were not significant, but two show significant correlations. Mean cottid length per tidepool tends to increase with tidepool volume (Figure 5). The total number of cottid per tidepool tends to decrease with the height of the tidepool (Figure 6). SECTION TWO: PHYSICAL FACTORS The results of the two multiple regressions calculated are summarized in Table 2. Mean oxygen level seems to be the most important variable affecting both the cottid mass/pool volume and the number of cottids per tidepool, but it appears to be more closely correlated to mass/pool volume. Figure 7 illustrates the independent relationship between the mean oxygen level and the mass of fish/volume of a tidepool. In Figure 8, the percent of oxygen consumed by cottids in six tidepools is shown in relation to the cottid mass / pool volume of those pools. This relationship is approximated by the equation: « 02 - 0.0023 * MASS/VOLI.47 (r2 - 0.987. p « 0.001) where oxygen is expressed in grams. The values for cottid oxygen consumption during the day are also shown in Figure 8. Although it appears that cottids consume less oxygen during the day, this difference is probably due to the effects of photosynthesis on oxygen levels. The relative amounts of oxygen consumed by sculpins, invertebrates, and algae during the night in one pool is shown in Table 3. Algae were apparently the primary consumers of oxygen; the oxygen consumed by algae was calculated to be 112.74, but this error may be due to the fact that the time points for the calculation were from two low tides, and the tidepool was replenished with fresh water and oxygen between those points. The difference in overall consumption when algae is respiring in the observed pool is probably still quite large. SECTION THREE: RECRUITMENT STUDY The change in total cottid biomass for one pool at daily intervals over a fifteen day period is shown in Figure 9. DISCUSSION This study has attempted to illustrate the enormous scope of physical and morphological characteristics of tidepools that could affect the distribution of tidepool sculpins. Although interaction among different organisms has been discussed in many contexts in the past, the ways in which tidepool fishes interact with tidepools as organisms is not well documented. There are at least two possible scenarios to explain the relationship that cottids have with the levels of oxygen in a tidepool. Because the correlation between the fish mass/volume ratio and the mean oxygen level of a tidepool is negative, a larger amount of fish in a tidepool corresponds to a lower mean level of oxygen. This suggests that cottids might be causing the changes in the amount of oxygen. Alternatively, other organisms in the pool might be more important in determining the fluctuations in oxygen level, and this level could be a limiting factor for the fish. This study shows that as the mass of fish in a tidepool (per pool volume) increases, the percent of oxygen consumed by fish also increases. But the fraction of total oxygen consumed by fish is small and algae appears to be the primary consumer of oxygen in a tidepool. Thus, I hypothesize that if there is causality in this correlation, the number of fish are responding to the level of oxygen rather than causing any change in it. Although the most stressful time for tidepool fishes in terms of oxygen availability is probably during the night between high tides, because algae are respiring and the pools are not receiving fresh water, it is difficult to state in the context of this study whether oxygen is the limiting factor for the amount of fish biomass that occurs in a certain tidepool. Regardless of the fact that algae and invertebrates use much more of the total oxygen overnight than cottids, oxygen levels in the tidepools where these fish reside might never drop to dangerous levels before they are replenished with fresh sea water or sunlight for photosynthesis. Further experimental studies are needed to settle this question. In the observed tidepool, algae is apparently the primary oxygen consumer by quite a large margin. If oxygen is in fact a significant limiting factor, then an interesting situation occurs, Algae provide an advantage for the fish as a means of protection from predators, but it also might provide a disadvantage by using large amounts of oxygen and possibly leaving the cottid with a less than ideal oxygen level. The above calculations for oxygen consumption in the observed tidepool were made using the fish biomass / pool volume of the unmanipulated pool. This biomass was equal to was the initial level of this pool while it was used in the recruitment study. Over the course of this study, biomass reached a maximum of almost exactly this level before dropping again and remaining quite low. (Figure 9). This information poses the question of whether there exists a maximum for cottid biomass in tidepools in relation to the amount of oxygen available to the fish. Determining whether different cottid species consume oxygen at significantly different rates and investigating the occurance of these species in tidepools with varying amounts of algae would be useful to answer this question. Further areas of investigation might include what, if any, mechanism the fish have to detect such a limit and whether fish are distributed accordingly, Other correlations between different tidepool parameters suggest possibilities for explaining the distribution of tidepool fishes. For example, the positive correlation between tidepool volume and mean cottid length might be another possible illustration of the distribution of cottids according to some limiting factor. Cottid length is related to biomass by the equation LENGTH - 0.0098 * MASS 3.10. This regression includes only those pools which contain cottid, which tended to be richer in algae and invertebrate diversity than those without fish (Table 1). In similar pools, the bigger the pool the more resources it contains. It follows that larger fish could reside in larger pools. Included in this explanation of where tidepool fishes occur is the question of where they do not occur in the intertidal zone, The negative correlation between tidepool height and number of 13 cottids illustrates that there are tidepools that are probably difficult for cottids to reach physically. Those pools which only get splashed occasionally due to their distance from the ocean are not only difficult to encounter, but may also have characteristics which are detrimental to tidepool fishes. For instance, one tidepool in the study area (which contained no fish) receives fresh seawater only at the highest tides when a large wave sends water through a small channel to the pool. Not only is the salinity of this pool higher than that of regular seawater, but algae growth is poor and the diversity of other organisms is relatively limited. Finally, the two species of cottid found in this census did not reside in the same tidepools. In the context of competition among similar species, this might suggest that, instead of sharing the resources in the same tidepool by using them at different times or by using separate resources (Yoshiyama 1977), these fishes are simply not residing in the same space. Another possibility, however, is that the environment affects these species in different ways. According to Nakamura (1976), O. snyderi have a lower tolerance of wide temperature extremes than other cottid species. The fact that the six of the seven Q. snyderi found in this study area were residing in a tidepool that was almost always receiving fresh ocean water and had the smallest temperature range over a tide cycle supports the possibility that 14 the fishes prefer to stay in such a pool. Because it is generally agreed that there is no single environmental factor that causes a certain distribution of intertidal fishes, this study has attempted to record the different characteristics of tidepools and examine those characteristics that appeared most important in determining where cottids were found. The scenarios presented to explain the distribution of cottids along the rocky intertidal zone of Hopkins Marine Station were developed with a specific idea in mind: that tidepools might be usefully observed as interactive organisms as well as environments. LITERATURE CITED Bolin, Rolf L. 1944. Cottid and Myctophid Fishes of California. Natural History Museum of Stanford University. 3: 67-70, 73-76. Eschmeyer, William N. Earl S. Herald, Howard Hammann.1983. A Field Guide to Pacific Coast Fishes of North America. Boston: Houghton Mifflin Company. pp. 161-185. Morris, Robert M. 1960. Temperature, salinity, and southern limits of three species of pacific cottid fishes. Limnology and Oceanography 5: 175-179. Morris, Robert M. 1961. Distribution and temperature sensitivity of some eastern pacific cottid fishes. Physiological Zoology, 34: 217-227. Nakamura, Royden. 1976. Temperature and vertical distribution of two tidepool fishes. Copeia. 1976: 143-152. Toshivama, Ronald Masaru. 1977. Competition and Rocky Intertidal Fishes. PhD thesis University of Michigan. Ann Arbor: Xerox University Microfilms. APPENDIX 1: INVERTEBRATES AND ALGAE USED IN OXYGEN CONSUMPTION EXPERIMENT INVERTEBRATES 642.6 grams 2 Pachygrapsus crassipes Stongylocentrotus purpuratus 2 Patiria miniata 38 Tegula funebralis 9 Pagunis samuelis 1 limpet (unknown species 1 Heptacarpus pictus 1 pelagic crustacean ALGAE 333.1 grams Pelvetia fastigiata Gigartina caniliculata Corallina pinnatifolia Corallina vancouveriensis Prionitis cornea Endocladia muricata Mastocarpus jardenii TABLE I. GENERAL TIDEPOOL DESCRIPTIONS DESCRIPTION roeoos TIDEPOOLS OF Po WITHNO COTTIO WITH COTTID DIVERSITYO INVERTEBRATES HIGH LOW GRDWTH OE POOR RICH ALGAE 2 1.807m HEIGUTS 21.682m MEAN SALINITY 1050.25 mo/ 2070.83 mol/ TABLE 2. MULTIPLE REGRESSION OUTPUTS DEPENDENT INDEPENDENT 8 SLOPE VARIABLES VARIABLE -2.85 0.758 NOMBER MEAN, OE COTTID LEVEL TEMPERATURE 11.24 0.227 -71.44 0.410 MEAN 2. MASS/VOL RATIO LEVEL SICNIFICANCE ooouc oouae o.o021 TABLE 3. OXYGEN CONSUMPTION IN ONE TIDEPOOL (NIGAT) ORGANISM 6RAMS O, CONSUMED %O, CONSUMEO COTTIOS 2.24 0.452 INVERTEBRATES 637.2 1.144 ALGAE 1589.0 100 FIGURE LEGENDS Figure 1. Clinocottus analis and Oligocottus snyderi. Bolin, 1944. pp.125 and 127. Figure 2. Species of sample population. Figure 3. Length description of sample cottid population. The range in lengths for the overall cottid population. C. analis and Q. snyderi are shown by a bar graph, and the mean lengths for each catagory are represented with a horizontal line. The shaded portion represents the middle 704 of each population. The value 702 was chosen because the Osnyderi population is composed of only seven fish; one fish was removed from the upper and lower limits. Figure 4. Mass description of sample cottid population. The upper and lower limits for the O snyderi population were removed to shoy the middle 702 of the population (shaded area) and this value was used in the other populations also. Figure 5. Tidepool volume versus mean cottid length. The regression equation and its significance is shown. Figure 6. Tidepool height versus number of cottid. The regression equation and its significance is shown. Figure 7. Mean oxygen level versus cottid mass/volume. Figure 8. Cottid biomass per pool volume versus 2 oxygen consumed by cottid. The circles represent oxygen consumed at night, and the squares represent consumption during the day . The equation of the curve is approximated by: LENGTH - 0.0023 * MASS/VOL 1.47 (R2 - 0.987, p «0.001) where 0» is expressed in grams. Figure 9. Cottid recruitment. The initial level represents the amount of cottid mass found in the initial census of the tidepool studied. FIGDRE 1. CLINOCOTTUS ANALIS AND OLIGOCOTTUS SNYDERI 2 . Clinocottus analis (Girard). Q Drau by Rolf L. Bolin. 4 11 kaaa ... Oligocoltus snyderi Creeley. Q Draun by Walter B. Schwarz. FIGURE 2. SPECIES OE SAMPLE COTTID POPULATION 00 60 2 20 COTTIO C.ANALIS O. SNYOERI EIGURE 3. LENGTH DESCRIPTION OF SAMPLE COTTID POPULATION 10% 01 SAMPLE POPULATION COTTIO C.ANALIS O. SNYDERI FIGURE 4. MASS DESCRIPTION OF SAMPLE COTTID POPULATION 107 0r SAMPLE POPULATION 9— COTTID C.ANALIS O.SNYDERI I U — O L 9 5 6 2 S5 o9 + () HISNTO NVa u 2 3 u 0 9 oooo- lo so iol — - 2 8 3 (W/S) TON/SSVN 9 0) — s2.. GI10 A9 OJMOSNOD NDDAXO e katakatakavavaaat- (SWV wo