THE EFFECTS OF DISPERSAL IN LABORATORY POPULATIONS OF THE POND SNAIL PHYSA GYRINA SAY Digertation for the Degree of Ph. D. MICHIGAN STATE UNIVERSIW JUDITH SAUVE WARNER 1975 u. .- .n."~.—¢¢~W . . / L ,.. We llllllllllllllllllilllllllllllll Leah-“ ; 31293011026220 hilt: 1‘3‘3333 ( 3.3 UAIAV‘ “"I' a:/ This is to certify that the thesis entitled The Effects of Dispersal in Laboratory Populations of the Pond Snail, Physa Gyrina Say presented by Judity Sauvé Warner has been accepted towards fulfillment of the requirements for Ph.D. Zoology Jegree in Mim Major professor Date November 14, 1975 0-7639 ‘J - “ ’ "- k ”05“? ' HOAGHSA U A BOOK BlNDERY "l8. LIBRARY SINGERS (Illflfllnnf All-Infin- ABSTRACT THE EFFECTS OF DISPERSAL IN LABORATORY POPULATIONS OF THE POND SNAIL, PHYSA GYRINA SAY By Judith Sauvé Warner Aquatic grazers generally feed on the substrate over which they move, so that their ability to detect differences in substrate .quality is not unexpected. However, their ability to distinguish quantitative differences in available food and the effects that this ability may have on their populational responses are largely unin- vestigated. This study examines the responses of such a grazer, the pond snail Physa gyrina Say, under controlled access to quantitative differences in a food substrate. Two replicated populations are compared: the controls, with 15 snails each, were subdivided into three groups restricted to discrete food levels; the experimental populations had unrestricted access to the three food levels. These food levels were discrete patches consisting of l, S, and 10 par- ticles of spinach, placed in separate feeding stations. On the short term, neither the shell growth, survivorship nor the number of eggs produced differed between the unrestricted l Judith Sauvé Warner and the control populations. However, the unrestricted snails dispersed non-randomly in the experimental populations, their densi- ties increasing with higher numbers of food particles per patch. In addition, the feeding frequencies of these unrestricted snails were more highly correlated with the number of food particles avail-4 able at the different patches. Although both populations had similar survivorships, the snails in the unrestricted populations ate and damaged less of their food each day. As a result, the experimental populations realized a greater carrying capacity than did the restricted population. THE EFFECTS OF DISPERSAL IN LABORATORY POPULATIONS OF THE POND SNAIL, PHYSA GYRINA SAY BY Judith Sauvé Warner A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Zoology 1975 ACKNOWLEDGMENTS I would like to thank my dissertation committee: Donald J. Hall, James Resh, and Robert J. Wetzel for their time and patience. My major professor, William E. Cooper, provided space and access to extensive computer time; he also supported my main—liner photocopy- ing habit. Financial support for this and other learning experiences was provided through Dr. Cooper by NSF Grant G120 (Design and Manage— ment of Environmental Systems) and NSF Grant GB15668 GB31018X (Coher- ent Areas Research Project in Freshwater Ecosystems). I've learned a lot from my association with these people and research groups, especially from the Ecology Group at Michigan State University. Most importantly, I've learned that each of us perceives our own reality. ii TABLE LIST OF TABLES. . . . . . . LIST OF FIGURES . . . . . . . INTRODUCTION. . . . . . . . . The Experimental System. Predictions. . . . . . . METHODS AND MATERIALS . . . . RESULTS 0 O O O O O O O O O Reproduction . . . . . . OF CONTENTS Growth and Survivorship. . . Dispersal in the Experimental Papulations. Food Consumption . . . . Feeding Activity . . . . DISCUSSION. . . . . . . . . . Populational Responses . Behavioral Responses . . CONCLUSIONS . . . . . . . . APPENDICES. O O O O O O O O O A. Statistical Model. . B. The Matching Law . . BIBLIOGRAPHY. . . . . . . . . iii Page iii iv L0 11 ll 15 17 21 25 32 32 35 39 40 4O 41 46 LIST OF TABLES Table Page 1. Growth and survivorship. Mean i l s.e. shell length (mm), and number of snails alive during the experiment. . . . 16 2. Linear correlation coefficients between the number of snails feeding and l) the number of snails present; and 2) the condition of the food particles after 24 hrs. of feeding activity. Bracketed coefficients are homogeneous (P = 0.05); ns = not significantly dif- ferent from zero... . . . . . . . . . . . 31 iv LIST OF FIGURES Figure Page 1. Mean number of eggs laid in the experimental (U) and control (R) populations at each food level (1,5,10) and in the trays (t) during the experiment. . . . . . . 13 2. Mean i S.E. number of snails present in the experimental (U) populations in the morning (am) and afternoon (pm) observations. Control (R) pepulations are averaged. over both observations. . . . . . . . . . . . . . . . . 20 3. The relative proportions of snails from each color category (Y = yellow; 0 = orange; R = red) at: A—-the feeding stations; B-—each food level (* signifies color category originally held at each food level); and C-—total distribution of snails among food levels . . . . . . . . . . . . . . . . . . . 23 4. The percent of spinach particles remaining after 24 hrs. of feeding activity in the experimental (U) and control (R) p0pulations. Stippled area is the percent damaged; clear area is the percent of intact particles . . . . . 27 5. Mean i l S.E. percent feeding at each food level in the experimental (U) and control (R) p0pulations in the 'morning (am) and afternoon (pm) observations. . . . . . 29 Appendix Bl. The matching between a) the proportion of snails feeding at two food levels (ordinate) with b) the preportion of food particles present at these same food levels. Opened circles = experimental populatipns; closed circles = control populations . . . . . . . . . . . . . 44 INTRODUCTION A major factor in the dynamics of natural populations and the regulation of animal numbers is the dispersal of individuals over the resource heterogeneities of their environment (Dayton 1971; Krebs et al. 1969; Meyers and Krebs 1971). This dispersal is an adaptive response especially in spatially heterogeneous environ—_ ments where there are real or apparent differenCes in the animals', limiting resources (Spright 1974). Yet the specific ecolbgical constraints effecting this dispersal activity can not always be in— ferred from the final distribution patterns alone (Connell 1963). For example, behavioral interactions such as competition can result in both random and non-random dispersion patterns (Pielou 1969); while both under- and overcrowding can result from dispersal in populations whose habitats vary in carrying capacities (Gadgil 1971). Still, dispersal or the lack of it has been shown important in labor- atory systems of predator—prey interactions (Huffaker 1958; Matsomoto and Huffaker 1973) and competitive interactions (Cotter 1974; Ghent 1966; Lomnicki and Slobodkin 1966; White and Huffaker 1969). ‘ Whether for predator-prey systems or for competitors, the predictive models for consumer strategies further underscore the l importance of dispersal in community ecology and population dynamics (MacArthur 1972; Pianka 1974). This is because dispersal involves not only the movement of animals and their distributions, but is also a factor in the resource allocation strategies of animals as consumers. Most of these models argue from the standpoint of food as the allocated resource, with the optimum strategy requiring effi- ciency in searching, pursuing and/or capturing food particles. But while many predators, including seed predators, have foraging be- haviors that are quantifiable in terms of the particle sizes they select (Holling 1966; Janzen 1971; MacArthur 1972), the foraging strategies of herbivores, especially grazers, appear more qualitative as they disperse over an apparent continuum of food resources (McNab 1963). Thus the foraging strategies proposed for herbivores have them Spending most of their time searching for food, selecting habi- tats and foraging as generalists on the quality or palatability of their food substrates (Hairston et al. 1960; Harper 1969; MacArthur' 1972). Not only have these models of resource allocation not been extensively tested for grazing species, but little is known about their behavioral and populational responses to quantitative food levels (Harper 1969). The Experimental System Here I examine the ability of an aquatic herbivore to distin- guish among food patches with different amounts of food in them, and compare the effects of such unrestricted foraging to control popula- tions restricted to fixed food levels. To test these effects, I chose a small aquatic gastropod, Physa gyrina Say, and a simple ex- perimental system. Phyga is a visual, gregarious pond snail with chemoreceptive abilities (Wells 1965); it disperses actively in field populationa‘and.can move more than ten cm in five'minutes‘ (Clampitt 1970). It can be sensitized to shocks but does not learn to avoid them (Wells and Buckley 1972; Wells and Wells 1971). Still, lghyga discriminates among species of emergent vegetation (Pip and Stewart 1974; E. Pip, pers. comm.) as it continually grazes the sub- strate over which it moves. In general, the foraging strategies of pond snails are rela- tively unknown, except for their responses to certain qualitative differences in available food (Eisenberg 1966; Townsend 1973). For this reason, I kept the complexity of the experimental systems to a minimum and restricted structural heterogeneity to the presence of glass finger bowls. These bowls served as feeding stations within the larger volume of the population tray. Replicated populations were set up to compare the population responses (reproduction, growth, and survivorship) and the feeding behaviors of Phyga as a grazer under controlled access to its food. The experimental or unre- stricted populations consisted of snails which were free to disperse among quantitative levels of a single food source (spinach), while the control or restricted populations were subdivided into grapps of snails confined to these different food levels. Two factors determined the length of the experimental run. First, the snails had to have enough time to respond, i.e. days rather than hours. Second, at 200 C, Phygafs eggs hatch in 14 days, but reproductive maturity is not reached for another 40 days (Clamr pitt 1970). However, survivorship for pre-reproductives is very sensitive to culturing conditions (R. M. Eisenberg pers. comm.) and mortality in either adults or pre-reproductives would supply an ex— traneous food source to the snails which readily feed on any organic material, thus confounding the populational responses compared here. Therefore I collected egg masses daily, did not allow reproductive recruitment into the populations, and terminated the experiment on the 14th day. Predictions Studies have shown that the densities of natural populations of pond snails can be regulated through the adjustments of their fecundity, growth, and survivorship around extant food supplies (Eisenberg 1966; 1970). Because these population responses are sen- sitive to food levels, at the outset two contrasting predictions were made for the populations designed in the present experiment. First, snails restricted to fixed food levels in the control popula- tions might consume more of their food simply because of their re— strictions to it. But if higher food consumption occurred, then these populations should have higher egg production, greater growth and/or survivorship. In contrast, all or some of these same popula- tion responses should be lower in unrestricted populations whose snails would be expending energy in dispersal and the seardh for food. Alternatively, snails dispersing among the different food levels in the experimental populations might more efficiently exr ploit the different food levels through their abilities to move in and out of the feeding stations. If the unrestricted snails disperse and thus displace intraspecific interference at the food patches by expanding their use of the spatial resources, then their opportuni- ties to feed at productive food levels with less interference might increase. If suCh feeding resulted in food consumption beyond the energetic demands of dispersal and maintenance, then their popula- tional responses in reproduction, growth and/or survivorship might increase. These same responses should be correspondingly lower in the groups of restricted snails because of the relatively small volumes in.which they live and the consequent interference over lfixed (and possibly limiting) food levels. METHODS AND MATERIALS Each population, replicated five times, consisted of fifteen adult snails in a plastic tray (41 x 46 x 11 cm) with three glass finger bowls serving as their feeding stations. These bowls were 10 cm in diameter, five cm deep and filled with 130 ml of filtered tap water. This water level was two cm below the bowl rims and was an effective barrier to emigration by the pond snails. The food particles, fresh spinach leaves out into 15 x 15 mm pieces, were about twice as large as the area covered by a single snail; there- fore, a snail on such a particle was feeding on a continuum rather than engulfing a single unit of food. In addition to its high food value to pond snails (Eisenberg 1966), fresh spinach does not disin- tegrate with the boiling required to prevent floatation of the par- ticles. Each of the three feeding stations in a tray held a different: number of spinach particles and designated a food level or patch size: low food - one particle; medium = five; and high = ten particles. In order that the three stations in each tray be about the same distances from its walls and from.the 25 watt incandescent light bulb over each tray, I placed them at the corners of an equilateral triangle in the 7 in the center of the tray, about three cm apart. The high food level was to the right, the low to the left and the medium, between and in front of the other two stations. The lights were kept on a 12—hr. light, lZ—hr. dark photoperiod, which also helped maintain the water temperatures within narrow limits (22 i 2'0 C). From a laboratory stock culture, I sorted 150 adult ghyga (mean length: 8.4 i 0.1 mm) randomly into three groups for marking. Individually identifying the fifteen snails in each population meant prohibitive bookkeeping, but both the growth and the distributions of subsets within the populations could be followed using three color categories. Therefore I marked snails with one of three different colors of Testors Pla Enamel (The Testor Corp., Rockford, Ill. 71101) and drew five snails from each category to make up a population. The snails of a given color code were arbitrarily assigned to a given food level: yellow to low food; orange to medium; and red to the high food level. At the beginning of the l4-day experimental run, the trays of the experimental or unrestricted populations were filled with 18 liters of tap water, filtered for visual particulate matter. This water level, about two cm above the bowl rims, permitted the snails to disperse freely among the feeding stations; but because the bowls were not touching, snails had to crawl out of one and onto the tray before entering another. The trays of the control populations were filled with 16 liters of water, about two cm below the bowl rims and so remained as restricted populations, subdivided into groups of five snails at each feeding station. Two observations were made each day. In the mornings, after about two hours of light and with spinach particles that had been ex— posed to about 24 hours of feeding activity, the following data were taken: densities at the food levels and in the trays; mortality; feeding densities; the number of foodparticles eaten, damaged (i.e. with rasped holes) and intact. Egg masses were then collected, fecal material and food particles removed, and fresh food and water supplied. About six hours later, the afternoon observations were made on densi- ties and feeding activity, but generally no feeding damage had occurred by this observation period. Phyga grazes constantly as it moves over its substrate, but a snail rarely moved immediately from a spinach particle once it was on it, even if another snail also moved onto it too. Therefore, a snail was recorded as feeding only when it was on a spinach particle. A split-plot experimental design (Appendix A) was used to ana- lyze the response variables (Gill and Hafs 1971; Kirk 1968). All data reported as percentages were transformed to arcsine units for the analyses of variance (Sokal and Rohlf 1969). Frequency data were analyzed using Chi square tests, which are reported as (X2 '(tab- aldf] < . . ular value) > X2 (Calculated value)). Where there were significant 10 differences in the response variables over food levels or with time, regression analyses were used to test the trends. Sample size for each mean was N = 65, unless otherwise noted. To characterize the distribution of snails in the unrestricted populations, I recorded the snails' positions by quadrants. These were visual divisions of the trays; first I used crossed diagonals from the corners, and as a second check, Cartesian coordinates through the center of the trays. Then the distributions of the snails among the different food levels, the tray quadrants, and their frequencies on the walls versus the tray floors were tested for randomness, using t-tests on the variance:mean ratios of densities (Kershaw 1964;'Stite- ler and Patil 1971). RESULTS Reproduction Under laboratory conditions, reproduction in P, gyrina can be affected by culturing conditions and snail densities (Clampitt 1970; van der Schalie and Berry 1973). But although in the present study the experimental populations produced a total of 8% more eggs and egg masses than did the controls, the average number of eggs laid each day (35 i 3), their clutch size (24 r 2) and infertility (38 i 3%) did not differ. Overall, each of these reproductive characteristics were poorly correlated with the food levels provided and with the amounts of food actually eaten or damaged in each population. Most of the eggs produced in the experimental populations were laid in the trays: but despite the poor correlations between egg production and food levels, the number of eggs laid at each food level differed significantly (P < 0.005; Figure 1). All reproduction in the control populations occurred in the feeding stations, and . though significantly higher than that at the comparable experimental food levels, the average number of eggs laid did not differ among food levels in the controls (P < 0.05; Figure l). 11 12 Figure l.--Mean number of eggs laid in the experimental (U) and control (R) populations; at each food level (1, 5, 10) and in the trays (t) during the experiment. NUMBER of £665 Y 13 100‘ 0"- o 0 5“ - ...,-.-.r-' "*“" P-°--|" 14 Over the fourteen day experiment, the egg production in each population increased linearly (P < 0.005) but their slopes (produc- tion rates) did not differ. Within the experimental population, how- ever, the number of eggs produced at the low food level did not in- crease over time, while the production rates at the medium and high food levels (though not different) were significantly lower than that in the trays (P < 0.001). Therefore, significantly more eggs were laid away from the feeding stations in the unrestricted populations . as the experiment progressed. In the control populations, eggs were laid at a signficantly lower rate at the low food level (P < 0.05), though this did not reduce the average number of eggs laid there. Overall, clutch size (eggs/egg mass) did not differ between the two papulations. However, larger clutches were laid at two sites: in the experimental p0pulations' trays where there was no food; and at the high food level in the control populations where these larger clutch sizes were highly correlated with food consumption. In con- trast, clutch sizes were significantly smaller at the low food level in the experimental populations and were not correlated with food consumption there. Neither the number nor the percentage of infertile eggs dif- fered between the two populations (P < 0.001); but among food levels in both populations, the percent of infertility was significantly different (P < 0.05). In both populations, it was highest at the 15 medium food level (44 i 3%), and lowest at the high food level (19 i 2% for the experimental populations; 26 i 2% for the controls). Growth and Survivorship Although shell length is considered a reliable index of age in populations of Physa and other snails (DeWitt'l955; Pollard 1973), growth in field populations of aquatic snails is affected by food limitations (Eisenberg 1966). Under laboratory conditions, growth is also affected by such culturing variables as the size distribu- tions and densities of the snails; food composition; volume; and conditioning periods (DeWitt 1954b, 1955; Eisenberg 1970; van der Schalie and Davis 1965). Fed ad—libitum amounts of lettuce, P, gyrina's fastest growth is in its first-five weeks after hatching and before it reaches sexual maturity at about 45 days and seven mm in length (Clampitt 1970; DeWitt 1954a, pers. comm.). There was no significant growth in either experimental or control populations (Table l), but these snails had matured under abundant food conditions in an uncrowded stock culture so that at- 8 mm in length, they were of_a size category that would be putting energy into reproduction rather than growth. Because they were equally free to disperse among all food levels, there was no reason to expect differential growth in the color categories of snails in TABLE 1. Growth and survivorship. and number of snails alive during the experiment. Mean i 1 S.E. shell length (mm); Populations Experimental Control Length overall on day l 8. i 0.06 8.5 i 0.08 day 14 8. i 0.04 8.6 i 0.05 color categories color categories on day 14: restricted to: yellow 8. i 0.05 low food (y) 8.3 i 0.07* orange 8. i 0.02 medium (0) 8.8 i 0.02 red 8. i 0.03 high (r) 8.8 i 0.02 Number alive (on day 6 of expt.) 14. i 1.8 13.6 i 1.7 Percent surviving ,to end of experiment 96. 86.6 *significantly different among food levels (P < 0.05). 17 the experimental populations, and there was none. However, in the control populations, the snails surviving at the low food level were smaller, while the snails at the medium and high food levels were larger than any of the color categories in either population (P < 0.05, Table 1). Therefore, snails restricted to the low food level not only did not grow, but apparently only the smaller individuals from the initial size distribution survived there to the end of the experiment. But snails specifically restricted to the higher food levels showed significant growth over the unrestricted populations. Four snails died in the experimental and ten snails in the control populations, so that the former had a significantly higher percent surviving to the end of the experiment (P = 0.05; Table 1). In the control populations, most of the mortality (8 snails) occurred at the low food level, but because most of these snails died after the eleventh day of the experiment, the mean number of snails alive during the experiment did not differ between the two pOpulations. Dispersal in the Experimental Populations As expected, the control feeding stations were effective barriers to emigration, and only four different snails moved from their stations into the population trays. Therefore mortality alone affected the number of snails in these restricted populations. 18 However, snails dispersed actively in the experimental populations, and this dispersal determined the snail densities among their food levels. These unrestricted snails did not favor the feeding stations over the tray quadrants but were equally frequent in the two general areas (x205[l] > 0.07). Also, they were randomly dispersed among the tray quadrants (x2 > 6.03), and did not frequent the walls over . 0.05[3] the floors of the trays (x2 [ :> 2.53). An obvious orientation to 0.051] . a specific area, say away from or near the overhead light bulb, there- fore did not occur. The unrestricted snails, however, were not randomly distrib- uted among food levels (x2 < 180.3). That is, their densities - 0.05[2] differed among food levels (P B 0.01; Figure 2) increasing, linearly, with the number of food particles present (P = 0.05). In addition, there was a diurnal shift in densities at the feeding stations, which reflects the fact that the snails did not remain in them for long periods and thus were not more frequent in the feeding stations than in the trays. Specifically, at each food level, there were always more snails present in the morning observations than in the after— noons (Figure 2). Significantly though;in each observation period, morning and afternoon, the densities at each food level did not vary significantly over the days of the experiment (P < 0.05). Therefore, each day the snails moved in and out of the unrestricted feeding 19 Figure 2.--Mean i l S.E. number of snails present in the experif mental (U) populations in the morning (am) and after- noon (pm) observations. Control (R) observations are averaged over both observations. 20 mom 9° mm" as: I r. 21 stations such that there were more snails present in them in the mornings than in the afternoons. But at each food level, there were no significant differences in densities among the morning observa- tions; similarly among the afternoon observations, densities at each food level did not vary significantly. Because the snail densities increased with the number of spinach particles available at the different food patches, it is unlikely that the original assignments of snails to the particular levels biased the counts. However, I checked each replicate for the distribution of snails from the three color categories; the variance: mean ratios of their densities (ca 0.3) at each food level were not significantly different from 1.0 (P < 0.05). Therefore the distribur tion of snails from each color category was random. If activity for a color category is taken as the percent of individuals tagged with that color that was present at the feeding stations, then clearly the colors did not affect their activity (Figure 3A), nor did the original assignment to a food level bias the activity of the snails (Figure 3B). Simply more snails occupied the higher food levels, without regard to their marking (Figure 30). Food Consumption Although their average densities did not differ, the experi- mental p0pulations ate (9 i 3%) and damaged (34 i 62) significantly 22 Figure 3.--The relative pr0portions of snails from each color category (Y - yellow; 0 3 orange; R - red) at: Ar-the feeding stations; B-—each food level (* . signifies the color category originally held at each food level); and C--tota1 distribution of snails among food levels. 23 ( < é?(/(///(((((//((§(({((/d {IIIXXXXXXW’ZXXRXXXRXXXXRI 12M‘::22iI);Ikex'z'z‘e‘e‘eziiméhealthily?»;::::r:::3‘&2:222‘22§>&2 .. Wmz122222222222.»nix/22222222222222)22222212221222222212)» O In ALIA IlDV FIVE TEN ONE TOTAL ACTIVITY 8 FOOD lEVELS COLOR CATEGORIES A 24 less of their food resources than did the closed populations (24 i 4; 51 i 82 respectively; P < 0.05). But unrestricted to food patches, the snails in the experimental populations ate approximately 10% of the spinach particles at each food level, while the percent eaten by the restricted groups in the control populations decreased signifi- cantly with the number of food particles provided (43 i 8; 30 i 6; 13 i 22 respectively; P < 0.005). Food consumption, in terms of the number of particles eaten and damaged, neither increased nor_decreased in the control popula- tions over the experiment. This rather constant food consumption was in spite of the growth at the higher food levels, and the significant mortality among the snails at the low food level. However, among the unrestricted food levels, there was a linear increase in the number of food particles eaten at both the medium and the high food levels, and this despite the constant estimates of snail densities at each of these levels (P - 0.005). Both damaged and intact spinach particles represented food to the snails, and after 24 hours of activity, the preportions of food still available was significantly higher in the experimental popula— tions (P < 0.005; Figure 4). Most of these particles were intact (ca 56%), while in the controls, most of the available food particles were damaged (51%). But while this damage was consistent across food 25 levels in the controls, it decreased significantly over food levels in the experimental populations (P < 0.005; Figure 4). Feedinngctivity In the afternoons, 6 hrs; after being supplied with fresh spinach, approximately 40% of the snails in each population were feeding (Figure 5). But in the mornings, some 24 hrs. after fresh food was supplied and after'Z hrs. of light, the proportion feeding (56%) was significantly higher in the experimental populations (P < 0.05). In the control populations, about 40% of the snails were feeding, except at the low food level (17%) where usually little remained of the single spinach particle (Figure 4). The unrestricted populations clearly differed in feeding pattern from the restricted pOpulations, and this may be reflected in the relative impact each pOpulation had on its food resources. A consistent percent feeding in the restricted populations may have resulted in the relatively uniform damage to its food particles across food levels. In contrast, the diurnal fluctuations in feed- ing patterns for the unrestricted populations may have resulted in more efficient feeding in terms of the amount of time spent on the food particles and therefore in less food damaged and eaten. Although significant in both populations, the correlations of feeding frequencies with snail densities, and with the amounts 26 Figure 4.--The percent of food partiCles remaining after 24 hrs. of feeding activity in the experimental (U) and control (R) populations. Stippled area is the percent damage; clear area is the per- cent intact. 27 :o o. m _ .23 .on oo— m..u>m.. GOO“. O— m — ico— ONINIVWBU OOOJ 28 Figure 5.—-Mean r l S.E. percent feeding at each food level in theeexperimental (U) and control (R) popula- tions in the morning (am) and afternoon (pm) observations. 29 WA>35>$>2>>$2>>>>t>>>>>>>>>>>§>52>>>>>>>>22 2 3Willi?I(Ki(«(444<<<<<<<<<>>>>>>tI>>>>>>>>h>>>>>>>>2>>>>>: 2 i::3:Ittezgrttzéit3<<<<6<<<<<)?i>>>>>>>>>2>>li>23>>>>>>>>2>>21??th :- 100' % I%”'T.>$) sum as; 30 of food available are not only higher in the experimental populations, but they also both show greater homogeneity over food levels (Table 2). In particular, the correlations between feeding and available food particles reflect the fact that Phy§a_conforms to the matching law (Baum 1974; Herrnstein 1970). This law predicts that the rela- tive rates of engaging in alternative activities, such as feeding at different food levels, will be proportional to the relative rates of I reinforcement, such as the number of spinach particles at these al- ternative food levels. Although this matching occurs in both p0pu- lations, the correlation is closer in the experimental populations, whose snails closely matched their feeding activity to the available spinach particles at each food level (Appendix 2). 31 TABLE 2. Linear correlation coefficients between the number of snails feeding and: l) the number of snails present; and 2) the condition of food particles after 24 hrs. of feeding activity. Bracketed coefficients are homogeneous (P = 0.05); ns = not significantly dif— ferent from zero. Number of Afternoon Morning Snails Feeding Observations Observations Snail Shail Damaged Intact Densities Densities Leaves Leaves Experimental populations at low 0.6301 0.6350 0.3579 0.2329nS medium 0.6486 0.6677 0.4674 0.4513 high 0.8663 0.7463 0.3413 0.3254 Control populations at low 0.3252 0.1311 0.6181 . 0.1943nS medium 0.2988 0.1990 0.2436 0.1373 ' ns ns high 0.3090 0.0224nS 0.4881 0.3607 DISCUSSION Populational Responses Knowing that dispersal is'a basic factor in the regulation of animal numbers, I expected clear differences in populational responses between the unrestricted snails and those confined to fixed spatial and food resources in the controls. In particular, pond snails under restricted Spatial and food resources should respond first in repro— duction, then growth, and finally in adult mortality (Cooley 1973; Eisenberg 1966). For example, Physa's reproduction is sensitive to culturing conditions: at fixed volumes and increasing snail densi- ties, egg production is reduced, while doubling the culturing volume at fixed densities does not affect production but infertility in- creases (DeWitt 1954c). Yet here the presence or absence of dis- persal in the large volume of the population trays altered neither the average egg production nor the degree of infertility. That is reproduction in the control populations was not affected by the relatively small volumes (130 ml) restricting groups of snails to fixed food levels, nor did egg infertility increase in the experi- mental populations with snails dispersing in the unrestricted tray volumes (18 liters). 32 33 That both populations produced about the same number of eggs might suggest that the snails were simply using energy reserves accumulated prior to the experiment; However, snails from the same culture, but in a separate, tandem experiment completely stOpped pro— ducing eggs after three days of fasting, suggesting that no signifi- cant amounts of accumulated energy were available for reproduction in these snails. In the size categories of snails used in the present experi— ment, Phygg can grow at a rate of’2 mm/week (DeWitt 1954a); yet there was no significant growth, overall, in either population here. This lack of growth is probably due to the fact that the snails, raised under abundant and high quality food conditions in a large and un- crowded culture had already matured with the early and fast growth characteristic of pond snails under such rearing conditions (Eisen- berg 1966, 1970). Still, within the closed populations, two growth responses are evident. First, snails surviving at the low food level were smaller than the average size that was originally drawn for the ex- periment; this shows a higher Survivorship but little growth for the smaller individuals restricted to-limited food. Second, the snails at the medium and high food levels were larger than the other color categories in either p0pulation so that Physa responded to more 34 abundant food by growing at the medium food level, and then with both growth and larger clutch sizes at the high food level. Mortality among adults is the final characteristic affected by stress on pond snail p0pulations and the stress of fixed food levels, particularly at the low food level, is apparent here. For although mortality did not differ on the average between the two pop- ulations, snails at the low food level in the_controls had, toward the end of the experiment, clearly the highest mortality. But though smaller and fewer, this same group of snails produced an average nume ber of eggs that was comparable to that of other snail groups in the 1‘ restricted populations. However, dead snails remained in the feeding stations between observations and these were readily fed on by the other snails. Therefore, mortality at the low food level may have provided a high energy food and made possible the continued high egg production in spite of the limited amount of food and reduced snail densities there. In sum, the effectiveness of dispersal in displacing the stress of food and spatial constraints on the population responses of an aquatic grazer are evident from these responses within rather than between the two populations. Without dispersal, and restricted to these limiting conditions, Physa first responds with lower surviv- orship in its larger individuals, but then at the medium food level, these restricted snails grow rather than increase their egg production. 35 Finally at the higher food level, this grazer produces larger clutches and grows rather than laying additional egg masses each day. Unrestricted, Phy§a_disperses non-randomly among food patches, evidently displacing the stress of fixed food and spatial resources with uniform growth and survivorship among the unrestricted snails. Behavioral Responses For aquatic gastropods, some of the most interesting obser- vations on dispersal come from work on their abilities to learn and the role of external stimuli, such as mucus trails, in their move— ments and orientations (Townsend 1973; Wells and Wells 1971). Snails produce mucus trails while moving and perhaps because they can fol- low these trails (Cook et al. 1969; Wells and Buckley 1972), some herbivorous snails orient to different kinds of food in the labora— tory (Townsend 1973) and select among natural food substrates (Clam- pitt 1970; Pip and Stewart 1974). However, such selection has not been documented beyond the distribution of snails over different qualities of food, and has shown little of the variation in individ- ual behaviors that regulates densities (Grime et al. 1970). Still, behavioral dichotomies, much as Wellington (1964) found in moth larvae, do appear in some snail populations where some individuals 36 wander constantly while others return consistently to the same spots (Breen 1971; Lomnicki 1969; Pollard 1973). Because the substrate over which a grazer moves is often its food, variations in both the quality and the quantity of foragable habitats is often indistinct, and the dietribution of resource habi- tats unassessed in terms of grazing strategies. In the present ex- periment, the quality of food and its distribution were controlled by the use of a single food, spinach, at the feeding stations. Thus, how frequently these unrestricted snails encountered and used the spatial and food resources available to them describes their forag- ing in terms of the environmental grain of these population trays (Holling 1966; Pianka 1974). That is, this grazer used the spatial resources of the feeding stations as a course-grained resource, fre- quenting them with disprOportionate densities. But the high corre- lation between feeding activity and the number of available food particles, clearly demonstrates that Phy§a_exploited the food par- ticles within these stations as a fine-grained resource. If, by occupying the different feeding stations as course- grained resources, Phyga shows some selectivity among these stations with different amounts of food, then its feeding behaviors could be interpreted broadly as a foraging strategy. Yet by regulating their densities at each food level on a diurnal cycle and over the experi- mental run, these unrestricted snails contradict the expectations 37 of an optimal foraging strategy. That is, animals, such as an aquatic grazer whose major foraging component is searching rather than pursu- ing or capturing food, should generalize by consuming a broad diet, but then specialize somewhat in the most productive food patches (Pianka 1974). Diurnal fluctuations at the feeding stations contra-' dict the expectation that having found a productive food patch, this Agrazer, like any consumer, would maximize its net resource gain by remaining at that patch. Contrast this feeding behavior with the relatively consistent feeding in the control populations where about 40% of the snails were , .feeding at each daily observation. These restricted populations had imposed upon them an Optimization of the profits and costs of search- ing; this Optomization is reflected in the greater growth and larger clutch sizes at the higherfood levels, and perhaps, by the continued high egg production even_at the low food level. Why then did the densities of snails fluctuate diurnally in the experimental populatiOns? Perhaps it is because these unrestric- ted systems permitted the snails to disperSe away from the feeding stations after they were disturbed during the morning cleanings. Or Iperhaps they showed a circadian rhythm involving feeding as other snails have (Malone and Nelson 1969); or variation in other activi- ties, for example, over the day, oviposition occurs most frequently in the early morning hours for Physa (DeWitt 1954a). Of course, the 38 snails may not have remained at the food patches simply because the food levels did not represent as sharp a gradient in size as I had intended. So that, though these snails foraged in a non-random manner, the accessibility of food and the levels used here may not have prompted an optimization in that foraging, as might be expected from models of particle size selection (Pianka 1974). Still, their non-random dispersal and the close matching of their feeding to available food shows that this grazer distinguishes among quanti- tative differences in an available food substrate. CONCLUSIONS Because the populational responses of pond snails are sensi- tive to extant food supplies (Eisenberg 1966), at the outset of this experiment two contrasting predictions were made. Neither prediction was met; rather, the two constraints: restriction to food levels, and unrestricted movements among food levels, resulted in similar, short-term responses for the two populations. At the basis of the two contrasting predictions was the ex- pectation that the experimental and control populations would exploit their food resources differently and according to their dispersal capabilities. This they did do, for although they both had the same potential carrying capacity in the food supplied to them, the experi- mental populations ate less of their food than did the controls. Effectively then, the experimental populations realized a higher carrying capacity as a result of their dispersal behaviors. There- fore, when able to disperse among different food levels, the grazer P, gyrina lost nothing in either reproduction or survivorship as a result of its movement, and in fact reduced its collective impact on its food resources. 39 APPENDICES APPENDIX A STATISTICAL MODEL = + A + B + c + AC + BC yijkl " 1 (1)3 k ( )ik ( ) + 01 + (AD)11 + (CD)kl + and”k + 800(1)jkl + Eijk£} (i)jk Source _ Expected Mean Squares A Treatments: A (experimental and control 02 + a; + (Ai) populations) . 2 + 2 error a. replicates B(i)j 0 GB - : 2 + 2 + 2 C Food levels Ck 0 CBC (Ck)/2 2 2 2 Food levels in papulations (AC)ik o + CBC + (Ac)ik/2 3 2 + 2 error c (BC)(i)jk 0 CBC D Time: D (13 days) 02 + 02 + (D)2/12 1 BD 1 2 2 2 Populations over days (AD)11 . o + OBD + (AD)11712 Food levels over days (CD) 02 + 02 + (CD)2 /24 k1 BD kl error d: BD , 02 + 02 (1)31 BD BCD 02 + 02 (i)jkl BCD Ai : Populations: experimental and control; 1 = 1,2 B(i)j: Replicates nested in populations: j‘= l,2,3,4,5 Ck : Food levels: low (1 particle);9medium (5 particles); high (10 par- ticles); k = 1,2,3 Dl : Time: thirteen days of the experiment 1 = l,...,l3 40 APPENDIX B THE MATCHING LAW The matching law predicts that the relative rates with which an animal engages in two alternative activities willbe proportional to the relative rates of reinforcement from such activities (Baum 1974; Herrnstein 1970). For example, giVen two patches of food with different amounts of food in them, an animal should feed at them in proportion to the amount of food it can receive in each. .In terms of the present experiment, if the number of snails feeding at the ith food level is (f1) of the total number of snails present in the p0pu- 1ation (n), then the relative feeding activity at the ith food level is F1 =bii. Similarly if the number of spinach particles present as intact and damaged at the ith food leveI’is“(ai) of the total number of particles fed to the p0pu1ation (r = 16), then the relative rein- a forcement ratio at the ith food level is Ri = £33 Under the matching law, the relative amount of feeding activity (F1) and the reinforce- =-ment rate (R1) at the ith food level will be proportional to those rates at the jth food level: F R 1 1 F, = R, (3' J J avhxn u chLJ” (1) or i The preportionality proposed by the matching law in equatiOn (2) is tested graphically with a log-log plot of the ratios: the 41 42 feeding activity at two alternative food levels is plotted on the ordinate, the reinforcement plotted on the abscissa. If there is a perfect matching or pr0portionality between the activity and rein- forcement ratios, then for every change in the relative reinforce- ments at two food levels, there will be a proportional change in the relative feeding activity at these two food levels. In other words, the plotted data will lie on a 457 line through the origin of the graph (Baum 1974)° For the matching function to be non-trivial, the feeding frequency (activity) must exceed the number of food particles (reinforcement) Hernnstein 1970). This caveat is satisfied here as the snails constantly scrape the substrate over which they move, so that feeding activity (scraping) far exceeds the reinforcements (spinach particles on which to feed) at each food patch. Therefore, in this experiment there are alternative feeding opportunities, not only between the tray areas where no spinach particles are present and the feeding stations with spinach, but also among the feeding stations themselves where different numbers of spinach particles are available. In Figure B1, the relative feeding activities (Fi/Fj) and food reinforcements (Ri/Rj) are plotted for the experimental (Opened circles) and control (dark circles) populations. These comparisons are plotted for the first eight days of the experiment, after which most of the points were redundant and therefore not plotted as they 43 Figure B1.--The matching between a) the proportion of snails feeding at two food levels (ordinate) with b) the proportion of food particles present at these same food levels. Opened circles = experimental popu- lations; closed circles = control p0pu1ations. 44 Immau.h¢(m no amniaz £02\30_ o; n. E P n b \\ \ \ \ \ \ \\ o \ \ \ \ O .\ \ O . \ 0.6\ O .50 O \ K o ;u_;\e:_v.e ”thEmquu—Z.m¢ III' In . UV. '0. re. OHIO!!! ALIAILDV 45 simply cluttered the plots (especially the comparisons between the medium and high food levels). ‘ Two points are obvious from Figure Bl. First, snails in the experimental populations closely matched their feeding activities to the spinach particles available to them. On the other hand, snails restricted to fixed food levels in the control populations showed much less correlation between their feeding and the number of food particles. These low correlations occur in both directions: a few. snails feeding in the presence of many spinach particles; and a large number of snails feeding on few spinach particles. Second, the close matching of feeding activity to reinforcement levels in the experif mental populations is consistent across all three food levels, but in the control populations, the closest matchings appear at the medium/high food comparisons. However, even in these comparisons, the matching is not close in the control populations. B IBL IOGRAPHY B IBLIOGRAPHY Baum, W. M. 1974. Choice in free-ranging wild pigeons. Science 185: 78‘79 o Breen, P. A. 1971. Homing behavior and population regulation in the limpet Acmaea (lelisella) digitalis., The Veliger 14:177-183. Clampitt, P. T. 1970. Comparative ecology of the snails (Physa gyrina and Physa integra (Basommatophora:Physidae).- Malacc— 1ogia 10:113-152. Connell, J. H. 1963. Territorial behavior and dispersion.in some marine invertebrates. Res. Pop. Ecol. 5:87—101. Cook, A., 0. s. Bamford, J.-D- B. Freeman and D. J1.Teidman. 1969. A study of the homing habit of the=1impet.' Anim. Behav., 17:330-359. Cooley, J. L. 1973. Effects of chronic environmental radiation on» a natural population of the aquatic snail Physa heterostropha. Rad. Res. 54:130-140. Cotter, W. B. 1974. Social facilitation and deve10pment in Ephestia kuhniella Z. Science 183:747-748. - . Dayton, P. K. 1971. Competition, disturbance and community organiza— tion: the provision and subsequent utilization of space in a rocky intertidal community. Ecol. Monog. 41:351-389. DeWitt, R. M. 1954a. Reproduction, embryonic development and growth. in the pond snail Physa gyrina Say. Trans. Amer. Micros. Soc. 73:124-137. DeWitt, R. M. 1954b. Reproductive capacity in a pulmonate snail (Physa gyrina Say). Amer. Nat. 88:159-164. DeWitt, R. M. 1954c. The intrinsic rate of natural increase in a pond snail (Physa gyrina Say). Amer. Nat. 88:353-359. 46 47 DeWitt, R. M. 1955. The ecology and life history of the pond snail Physa gyrina. Ecol. 36:40-44. Dinter, I. and P. J. Manos. 1972. Evidence for a pheromone in the marine periwinkle, Littorina littorea Linnaeus. The Veliger 15:45-47. Eisenberg, R. M. 1966. The regulation of density in a natural p0p— u1ation of the pond snail Lymnaea elodes. Ecol. 47:889-906. Eisenberg, R. M. 1970. The role of food in the regulation of the pond snail Lymnaea»elodes.' Ecol. 51:680-684. Gadgil, M. 1971. Dispersal:population consequences and evolution. Ecol. 52:253—261. Ghent, A. W. 1966. Studies of behavior of the Tribolium flour beetle. II,.Distributions in depth of T. castaneum and T, confusum in fractionable shell vials. Ecol. 47:355-367. Gill, J. L. and H. D. Hafs. 1971. Analysis of repeated measurements " of animals. J. Animal Sci. 33:331-336. Grime, J... P., G. M. Blythe and J. D. Thornton, 1970. Food selection by the snail Cepae nemoralis. Symp. Br. Ecol. Soc. No. 10: 73—990 Hairston, N. G., F. E. Smith and L. B. Slobodkin. 1960. Community structure, population control and competition. Amer. Nat. 94:421-425. Harper, J. L. 1969. The role of predation in vegetational diversity. Brookhaven Symp. Biol. 22:48-62. Herrnstein, R. J. 1970. On the law of effect. J. exp. Anal. Behav., 13:243-266. Holling, C. S. 1966. The functional response of invertebrate preda— tors to prey density. Mem. Entomol. Soc. Canada. 48:1—81. Huffaker, C. B. 1958. Experimental studies on predation:dispersion factors and predator-prey oscillations. Hilgardia 27:343-383. Janzen, D. H. 1971. Seed predation by animals. Ann. Rev. Ecol. Syst. 2:465-492. 48 Kershaw, K. A. 1964. Quantitative and Dynamic Ecology. Edward Arnold Publ. London 183 pp. Kirk, R. E. 1968. Experimental Design:procedures for the behavioral sciences. Brooks/Cole Publ. 00., Belmont Calif. 577 pp. Krebs, C. J., B. L. Keller, R. H. Tamarin. 1969. Microtus p0pu1a- tion biology:demographic changes in fluctuating populations -of M. ochrogaster«.and._M. pennsylvanicus-in-southsrn-Indiana. Ecol. 50:587-607. 'Lomnicki, A. 1969. Individual differences among adult members of a snail population. Nature 223-1073-1074. Lomnicki, A. and L. B. Slobodkin. 1966. Floating in Hydra littoralis. Ecol. 47:881-889. MacArthur, R. M. 1972. Geographical Ecology. Harper and Row Publ.,. N. Y.» 269 pp. Malone, C. R. and D. J. Nelson.‘ 1969. Feeding rates of freshwater snails (Goniobasis clavaeprmis) determined with cobalt60. Ecol. 728-730. Matsumoto, B. M. and C. B. Huffaker. 1973. Regulatory processes and p0pu1ation cyclicity in laboratory populations of Anagaster kuhniella (Zeller)(Lepidoptera:Phycitidae).- V. Host finding and parasitization in a "small" universe by an entomophagous parasite Venturia canescens (Gravenhorst)(Hymenopterazlchneu- monidae). Res. POpul. Ecol. 15:32-49. McNab, B. K. 1963. Bioenergetics and the determination of home range. size. Amer. Natur. 97:133-140. Myers, J. H. and C. J. Krebs. 1971. Genetic, behavioral and repro- ductive attributes of dispersing field voles Microtus pennsyl- vanicus and M. ochrogaster. Ecol. Monog. 41:53-78. Pianka, E. A. 1974. Evolutionary Ecology. Harper and Row Publ., N. Y. 356 pp. Pielou, E. C. 1969. An Introduction to Mathematical Ecology. Wiley- Interscience. N.Y. 286 pp. 49 Pip, E. and J. M. Stewart. 1974. The dynamics of two plant-snail interactions in certain aquatic systems in Manitoba. AIBS 25th Ann. Meeting June 1974. Abstract No. 152. Pollard, E. 1973. Growth classes in the adult Roman snail (Helix pomatia L.). Oecologia (Berl.) 12:209-212. Sokal, R. R. and F. J. Rohlf. 1969. Biometry. W. H. Freeman & Co., San Francisco. 776 pp. Spright, T. M. 1974. Sizes of populations of a marine snail.v Ecol. 55:712-729. Stiteler, W. M. and G. P. Patil. 1971. Variance—to-mean ratio and Morisita's index as measures of spatial patterns in ecological populations. Statistical Ecology Vol. 1:423-459. Penn. State Univ. Press. Univ. Park. Townsend, C. R. 1973. The food finding orientation mechanism of Biomphalaria glahrata (Say). Anim. Behav. 21:544-548. van der Schalie, H. and E. C. Berry. 1923. The effects of tempera- ture on growth and reproduction of aquatic snails. Sterkiana No. 50. 92 pp. van der Schalie, H. and G. M. Davis. 1965. Grdwth and stunting in Oncomelania (GastrOpoda:Hydrobiidae). Malacologia 3:81—102. Wellington, W. G. 1964. Qualitative changes in populations in un- stable environments. Cand. Entomol. 96:436-451. wells, M. J. 1965. Learning by marine invertebrates. Adv. marine Wells, J. J. and S. K. L. Buckley. 1972. Snails and trails. Anim. Behav. 20:345-355. Wells, M. J. and J. J. Wells. 1971. Conditioning and sensitization in snails. Anim. Behav. 19:305-312. White, B. G. and C. B. Huffaker. 1969. Regulatory processes and population cyclicity in laboratory p0pu1ations of Anagaster kuhniella (Aeller)(lepidoptera:Phycitidae). 1. Competition for food and predation. Res. Pop. Ecol. 11:57-83. "IIIIIILIIIIIIIIIIIIIIII