INFORMATION TO USERS This material was produced from a microfilm copy of the original document. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the original submitted. The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction. 1. The sign or "target" for pages apparently lacking from the document photographed is "Missing Page(s)". If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting thru an image and duplicating adjacent pages to insure you complete continuity. 2. When an image on the film is obliterated with a large round black mark, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image. You will find a good image of the page in the adjacent frame. 3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning" the material. It is customary to begin photoing at the upper left hand corner of a large sheet and to continue photoing from left to right in equal sections with a small overlap. If necessary, sectioning is continued again — beginning below the first row and continuing on until complete. 4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints of "photographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, title, author and specific pages you wish reproduced. 5. PLEASE NOTE: Some pages may have indistinct print. Filmed as received. University M icrofilm s International 300 North Zeeb Road Ann Arbor, Michigan 4B106 USA St. John's Road, Tyler's Green High Wycombe, Bucks, England HP10 8HR 78-3572 THRELKELD, Stephen Thomas, 1951THE MIDSUMMER DYNAMICS OF TWO DAPHNIA SPECIES IN WINTERGREEN LAKE, MICHIGAN. Michigan State University, Ph.D., 1977 Ecology University Microfilms International, Ann Arbor, Michigan 48106 THE MIDSUMMER DYNAMICS OF TWO DAPHNIA SPECIES IN WINTERGREEN LAKE, MICHIGAN By Stephen Thomas Threlkeld A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Zoology 1977 ABSTRACT THE MIDSUMMER DYNAMICS OF TWO DAPHNIA SPECIES IN WINTERGREEN LAKE, MICHIGAN By Stephen Thomas Threlkeld Midsummer reductions in densities of Daphnia popu­ lations are common and have been reported worldwide. Dur­ ing 1976 reductions were also found for Daphnia pulicaria and Daphnia galeata mendotae populations in Wintergreen Lake, Michigan. Previous attempts to determine the causes of such declines have indicated correlations between predator populations, the presence of unpalatable algae or high midsummer water temperatures, and the reduction or disappearance of the daphnid populations. In this study, in situ life table experiments were performed to assess the influence of natural food and temperature con­ ditions on population growth. The effect of predation by fish and Chaoborus was assessed by analysis of population size structure, predator gut contents, recovery of sedimented fecal material, and the age distribution of parthenogenic eggs carried by the females of the population. Stephen Thomas Threlkeld A weekly sampling program showed that D. pulicaria was dielly concentrated after early summer at depths between the anaerobic hypolimnion and the warmer epilimnion, consistent with previous information indicating it to be a cold-water species. However, in situ life table data showed that D. pulicaria was capable of sur­ viving in epilimnetic water up to 2 7°C, although repro­ duction was greatly reduced for a six day period in midJuly. This reduction in reproduction appeared to be the result of an interaction of high temperatures, declining standing crops of small algae and increasing amounts of Anabaena, Ceratium and Volvox. D. galeata mendotae did not show any adverse effect due to these mid-July algaetemperature conditions. Analysis of predator gut contents suggested that both daphnids were heavily preyed upon by bluegills (Lepomis macrochirus) . Shifts in the size structure of the daphnid populations were also consistent with intense size-selective predation by bluegills, as was the tendency for the age distribution of parthenogenic eggs to become skewed to younger eggs as the decline proceeded. Dif­ ferences in body size, epphipial production, and habitat preferences between the two species are also consistent with the hypothesis that planktivory is a strong selective force in their evolution. ACKNOWLEDGMENTS I would like to thank the members of my guidance committee, D. J. Hall, E. E. Werner, W. E. Cooper, R. G. Wetzel, and R. W. Merritt, during this study. for advice and encouragement M. Klug and J. Molongoski provided unpublished data from their concurrent studies on Winter­ green Lake. D. Howard provided the algal counts and assistance with Coulter Counter analyses of algal samples shown in Figure 15. W. C. Johnson assisted in the col­ lection of fish and arranged for the use of laboratory facilities at Wintergreen Lake. Several persons from the Ecology group at Michigan State University, in par­ ticular J. Dacey and D. Wagner, provided helpful comments at various stages of the research. To all of these people I am grateful. Financial support was provided by a Department of Zoology teaching assistantship and NSF grant BMS 7409013A01 to D. J. Hall and E. E. Werner. Last, but not least, I would like to acknowledge W. T. Edmondson who first introduced me to the topic of this dissertation, and my wife, Michele, for providing encouragement throughout this study. TABLE OF CONTENTS LIST OF TABLES . “............................... iv LIST OF FIGURES ..................................... v INTRODUCTION ........................................ 1 METHODS AND MATERIALS .............................. 5 GENERAL POPULATION DYNAMICS ....................... 14 IN SITU LIFE TABLE EXPERIMENTS .................... 25 Methods and Materials ........................... Results and Discussion........................... 26 29 POPULATION LOSSES ................................. 47 SPECIES SUCCESSION ................................. 55 EVOLUTION OF DAPHNIA LIFE HISTORY CHARACTERISTICS. 59 Daphnia pulicaria .............................. Daphnia galeata mendotae ....................... Discussion. . ................................. 60 71 71 SUMMARY AND CONCLUSIONS ........................... LIST OF REFERENCES ................................. iii 79 81 LIST OF TABLES Table 1. 2. 3. 4. Page Major hypotheses for the midsummer decline of Daphnia populations ........................... 3 Criteria for ageing parthenogenic eggs and per­ cent of total development time spent per s t a g e ........................................ 11 Mean number of Daphnia pulicaria and D. galeata mendotae per liter i n t h e epilimnion (E) and hypolimnion (H) of nine southwestern Michigan lakes during the day (D) and at night (N) in August, 1976 .................... 13 Summary of life history characteristics of Daphnia pulicaria and Daphnia galeata m e ndotae.................................... 73 iv . LIST OF FIGURES Figure 1. 2. 3. 4. 5. Page Bathymetric map of Wintergreen Lake showing sampling stations used in this study. The distribution of macrophytes is indicated by the shading. Contour intervals are shown in meters ................................ Top p a n e l : Mean number of Daphnia pulicaria (--- ) an(=i d . galeata mendotae (— ^ per liter as estimated by noon Co) and midnight (•) Van Dorn collections and by midafter­ noon (□ ) and midnight ( ■ ) vertical tow collections. For graphical clarity, Van Dorn and midnight vertical tow records are terminated on the last date each species was observed in those samples. The last date each species was observed in midafternoon vertical tows is indicated by a P (for D. pulicaria) or G (for D. galeata mendotae) on the abscissa. Bottom panels: Mean number of parthenogenic eggs per female D. pulicaria and D. galeata mendotae as estimated by four sampling regimes; symbols as used in top panel. ....................... 17 Depth-time distribution of isopleths of temperature (°C) (top panel) and dissolved oxygen (mg liter-1) (bottom panel) . 19 Depth-time distribution of isopleths of Daphnia pulicaria per liter at noon (top) and midnight (bottom). Water depths with­ out oxygen and greater than 22°C are indi­ cated by stipling and shading, respectively, in the lower panel ........................... 21 Depth-time distribution of Daphnia galeata mendotae per liter at noon (top) and mid­ night ( b o t t o m ) .............................. 24 v $ i 7 Figure 6. 7. 8. 9. 10. 11. 12. Page Isopleths of population growth rate (cohort r) for cohorts of Daphnia pulicaria as a function of the date the cohorts were started and the mean water temperature experienced by the cohort's individuals prior to maturation. Epilimnion incu­ bations are indicated by closed circles (•), metalimnion ( 2 . 5 m , 3 . 0 m , 3 . 5 m ) incubations by open circles (o), and hypolimnion incubations by closed squares ( ■ ) . 32 Age-specific reproduction ( m ) of Daphnia pulicaria cohorts incubated at 0.5 m . . 34 . Size structure of Daphnia pulicaria popu­ lation during July, 1976„ The solid vertical line indicates the minimum size of reproductive individuals in the popu­ lation; shaded areas indicate egg-bearing f e m a l e s ..................................... 36 Daily egg production (eggs*adult female alive”! • day” !) by Daphnia pulicaria in in situ life table experiments at all incubation depths. Shaded areas indicate eggs which were cast off in undeveloped con­ dition with molts; open areas indicate live births. No cohorts were started at 3.5 m after m i d - J u l y ....................... 39 Isopleths of population growth rate (cohort r) for cohorts of Daphnia galeata mendotae as a function of the date the cohorts were started and the mean water temperature experienced by the cohort's individuals prior to maturation. Symbols as in Figure 6. The three cohorts from 3 . 5 m having negative r values (see text) are o m i t t e d ..................................... 41 Size structure of Daphnia galeata mendotae population during July, 1976'. Format as in Figure 8 ..................................... 43 Brood size-body size relationships for Daphnia galeata mendotae population during July, 1976. Vertical bars include the mean ± one standard deviation .................... 46 vi Figure 13. 14. 15. 16. Page Distribution of dmax for Daphnia pulicaria (--- ) and Daphnia galeata mendotae (--from midafternoon (o) and midnight (•) vertical tow s a m p l e s ....................... 52 2 Photographs of replicate fields (0.107 mm ; 5.28 yl) of settled water samples of algae collected from 0.5 m during July, 1976 . . 65 Abundance of algal particles of three size fractions and three dominant genera col­ lected from 0.5 m during July, 1976. Note that Anabaena and Ceratium densities have been multiplied by 1000 for inclusion on these a x e s ................................. 67 The relationship between cohort r and tem­ perature of Daphnia pulicaria (•) and D. galeata mendotae (o) as determined in Tn situ life tableexperiments at all incu­ bation d e p t h s ..................................... 78 vii INTRODUCTION Seasonality is common to all natural ecological systems. To the extent that seasonal changes in abiotic and biotic aspects of natural environments are predictable, natural populations evolve life history characteristics (e.g., habitat preferences, reproductive strategies, body size) which permit them to persist under temporary con­ ditions unfavorable for continued population growth. Factors important in the seasonal fluctuations of natural populations must play an important role in the evolution of these life history characteristics. Although identifi­ cation of selective pressures and the evolution of life history tactics in terrestrial plants and animals has become a major focal point in contemporary ecology, the relationship between population regulation and life history evolution of planktonic organisms has not yet been adequately addressed. The major impediment to this synthesis in aquatic systems has been difficulty in the unequivocal identification of factors most important in seasonal population regulation. For example, although midsummer declines of Daphnia populations have been reported worldwide 1 (e.g., Birge 1896, Hall 1964, Wright 1965, Clark and Carter 1974, DeBernardi 1974, George and Edwards 1974, Fott 1975) no consensus exists on what is their cause. A number of hypotheses which have been proposed for such midsummer declines of Daphnia are summarized in Table 1. Basically these hypotheses may be classified as those related to (1) increased water temperature, ditions, (2) changing food con­ (3) increased activity or abundance of inverte­ brate predators (e.g., Chaoborus, Leptodora) , and (4) in­ creased activity or abundance of vertebrate predators (various fish species). Whether a general combination of these factors regulates the midsummer abundance of all Daphnia species or not is unknown. However, of the four classes of factors cited in Table 1, usually three or more are associated with each of the reported declines. The multiplicity of hypotheses for each species' dynamics has made thorough analyses difficult, and those without independent assessments of each hypothesis unconvincing. Thus, it is impossible to evaluate how important various behaviors, allocations of energy between reproduction and growth, or habitat choices are in permitting species populations to persist in aquatic systems. The present study attempts to combine an analysis of the midsummer decline of two Daphnia species caria Forbes and D. galeata mendotae Birge) (D. puli­ in Wintergreen Lake, Michigan, with an evaluation of factors influencing Table 1. Major hypotheses for the midsummer decline of Daphnia populations. Hypothesis Increased water temperature Changing food conditions Decline mediated by Restriction of vertical distribution to deeper, cooler waters Birge 1895, 1896, Hall 1964, Tappa 1965, Bell and Ward 1970, Haney and Hall 1975 Degeneration of parthenogenic eggs Hrbackova-Esslova 1962 Increased mortality Craddock 1975, Goss and Bunting 1976 Decreased assimilation efficiency Lampert 1977 Decreased filtering rate Burns 1969 Curtailment of individual growth and increased vulnerability to invertebrate predators Hall et al. 1976 Filtering rate inhibition Crowley 1973 Decreased assimilation efficiency Nowak 1975 Reproductive decline George and Edwards 1974, Clark and Carter 1974, Kwik and Carter 1975 Increased mortality due to: starvation toxicity of midsummer algae pH extremes resulting from increased photosynthesis Invertebrate predators Increased mortality associated with: increased abundance of predators increased activity (related to water temperature or predator growth) Vertebrate predators Reference Increased mortality associated with increased abundance and individual feeding activity (temperature or growth related) of predators Threlkeld 1976 Arnold 1971 O'Brien and DeNoyelles 1972 Wright 1965, Hall 1964, DeBernardi 1974 Fedorenko 1975 Hrbacek 1962, Galbraith 1967, Noble 1975, Clark 1975, Fott 1975, Fott et al. 1974, Baumann and Kitchell 1974, Wong and Ward 1972, Ward and Robinson 1974 4 the life history characteristics and midsummer distri­ bution of these two species in north temperate lakes. METHODS AND MATERIALS Wintergreen Lake is a small (15 ha, max. depth 6.3 m, average depth 3.5 m ) , hypereutrophic lake in southwestern Michigan located on the Kellogg Bird Sanctuary of the Kellogg Biological Station of Michigan State University (Manny 1972, Manny et al. 1977, Wetzel 1975, Wetzel et al. unpubl. data). Vast amounts of nutrients are introduced into the lake by migratory and resident bird populations which result in an annual mean primary productivity greater than 12 00 mg C* m (Manny 1972, Manny et al. 1975). -2 .day -1 Common midsummer algal genera include Aphanizomenon, Anabaena, Ceratium, and Microcystis transparency (Manny 1972, Crowley 1973); midsummer water (Secchi disk) is usually less than 1.0 m (Wetzel 1975, Wetzel et al. unpubl. data, Klug et al. unpubl. d a t a ) . Approximately 50% of the surface area of the lake is covered with floating and submergent vege­ tation (N u p h a r , N y mphaea, Potomageton, Hydrodictyon) . All limnological measures except where otherwise noted were obtained from a single station located in the deepest, limnetic region of the lake 5 (A in Figure 1). Figure 1 Bathymetric map of Wintergreen Lake showing sampling stations used in this study. The distribution of macrophytes is indicated by the shading. Contour intervals are shown in meters. 50 E L E V A T I O N 271m, A R E A 15 ha Figure 1 i 100 _J_ METERS 150 _I 8 Estimates of the amount and quality of food available to zooplankton in Wintergreen Lake were based on measurements of: (1) chlorophyll a and phaeopigments, and (2) algal size structure and taxonomic composition. Pigment analyses were performed on particulate matter collected biweekly at 1-meter depth intervals. Water samples were filtered through Millipore filters (HA, 0.45 y m ) , plant pigments extracted in 90% aqueous acetone and absorbances read using a Hitachi Perkin-Elmer Model 139 Spectrophotometer. Following acidification with 1 N HC1 the samples were read again. Chlorophyll a and phaeopigment concentrations were calculated from the absorbance equations of Wetzel and Westlake (1969). Algae, collected each week from 1-meter depth intervals, were preserved in Lugol's solution. surface meter Algal samples from the (0.5 m) taken from June 15 to August 3 were analyzed for size structure using a Model A Coulter Counter. In addition, July algal samples were settled and viewed with a Wild Inverted microscope; identifications were made according to Prescott (1970). Physical and chemical parameters measured each week included pH, dissolved oxygen water transparency thermistor). (Secchi d i s k ) , and temperature (YSI Temperature was measured each day during the summer months k (Winkler technique), (June-September). 9 Zooplankton were collected by two methods. Once a week, the vertical distribution and abundance of Daphnia was assessed by taking duplicate 4-liter Van Dorn samples at 1-meter depth intervals at both noon and midnight. Also, vertical tows with a 15-cm diameter, 145-ym mesh net were taken each week at midnight in triplicate at Station A and in duplicate on a biweekly schedule at three peripheral stations (B, C, and D in Figure 1). From July 10-31 duplicate vertical tows were also taken each day in midafternoon at Station A. All zooplankton were collected and concentrated on 145-iam mesh netting. Preservation in 4% sucrose (40 g/liter) formalin solution (Haney and Hall 19 73) prevented the loss of Daphnia eggs and distortion of the carapace by ballooning. Analysis of vertical tow samples from Station A involved taking sufficient 1-ml aliquots from the sample jars (50-100 ml) to count at least 200 individuals of each Daphnia species present. Daphnia were counted in a Petri dish modified with plastic baffles to serve as a Bogorov plankton counter (Gannon 1971), thus eliminating any remixing and accidental recounting of individual animals. In addition, the carapace length (exclusive of tail spine and head) of each of the 200+ individuals was measured to within 0.1 mm (0.2 mm size classes) and the number of eggs and their stage of embryonic development was noted. The stages of development of Daphnia F i 10 parthenogenic eggs were identified according to criteria in Table 2, modified from Green (1974) , and George and Edwards (1956), Lei and Clifford (1974) . Displacement volumes of midnight vertical tow collections of plankton from stations A, B, C, and D were measured in a graduated cylinder after removal of the preservative solution by filtration. Aliquots containing 100 or more Daphnia were also examined to see if the relative abundance of species was consistent between stations. Van Dorn samples were counted in their entirety for total individuals, parthenogenic eggs, gravid females, males and epphipial females for each Daphnia species present. Egg ratios (Edmondson 1974) were calculated for all Station A samples by dividing the number of partheno­ genic eggs present by the total number of females. ences for zooplankton identification were: Refer­ Daphnia puli­ caria (Brandlova et al. 1972, Hrbacek 1959), D. galeata mendotae (Brooks 1957), Chaoborus punctipennis 1972), all other zooplankton (Saether (Edmondson 1957). Fish were collected by gill net and hook and line, primarily in the limnetic zone of the lake. Stomachs were removed and preserved in 4% sucrose-formalin solution. From July 12-30, sedimenting seston was collected by duplicate glass jars (6.25 cm diameter, 12.5 cm depth) suspended at 4.5 m. The jars were collected daily and replaced with clean jars; the contents were concentrated Table 2. Criteria for ageing parthenogenic eggs and percent of total development time spent per stage. Relative Duration (Percent of total ± 2.0 std. errors) Stage Description 1 Egg stage, egg membrane intact; no differentia­ tion into body regions 30.2 ± 2.2 LC I, II; GE I; G I, II; OF 1-4; E 1-3 2 Embryo with or without head bulge; antennae becoming distinct; eye spots not yet visible 32.0 ± 1.4 LC III, IV; GE II; G III, IV; OF 5-7; E 4-5 3 Embryo with two small pink or red eyes 4 5 Embryo with two brown or black eyes Embryo with a single median black eye 8.7 ± 0.5 13.0 ± 0.5 16.0 ± 2.7 References3 LC V; GE III; G V, VI; OF 8; E 6 LC VI, VII; GE IV; G VII LC VIII; G VIII . OF 9-15; E 7-8 > LC: Lei and Clifford (1974); G E : George and Edwards (1974); G: Green (1956); OF: Obreshkove and Fraser (1940); E: Esslova (1959). Roman numerals refer to egg stages and arabic numerals to figures in the literature cited. 12 on 145-um netting and preserved in 4% sucrose-formalin solution. Analysis of these samples was limited to assess ing the relative numbers of Daphnia found in undigested and digested states (in fish fecal material); degradation of zooplankton chitin was not observed over the brief (<1 day) period the material may have remained in the collecting jars. In addition to the Wintergreen Lake studies, nine other lakes (Table 3) in southwestern Michigan of dif­ ferent depth and water transparency were sampled in August, 1976, to assess diel vertical distribution of any D. pulicaria and D. galeata mendotae present. cate vertical net tows Dupli­ (15-cm diameter, 145-ym mesh net) were taken during the day and at night in the epilimnion and the entire water column. Hypolimnion-residing zoo­ plankton were estimated by taking the difference between numbers collected in the epilimnion and in the total water column; these numbers and tow volumes were then used to calculate mean densities per liter of these species in the epilimnion and hypolimnion. Table 3. Mean number of Daphnia pulicaria and D. galeata mendotae per liter in the epilimnion (E) and hypolimnion (H) of nine southwestern Michigan lakes during the day (D) and at night (N) in August, 1976. Lake, Date Sampled Maximum Depth (m) Secchi Disk (m) Surface Alkalinity (mg CaCOj/l) Temperature Range (°C) Mean Number per Liter D. pulicaria D. galeata mendotae (D) (N) (D) (N) Bassett 16-VIII-76 9.5 2.9 143 (E) 20.8-21.5 (H) 9.0-20.8 0.6 4.4 0.0 6.6 36.2 10.9 171.2 40.7 Deep 16-VIII-76 9.0 5.0 149 (E) 21.4-22.0 (H) 10.2-21.4 0.0 11.3 0.4 3.9 16.4 8.7 126.2 0.0 Hamilton 18-VIII-76 11.0 3.0 198 (E) 21.5-22.9 (H) 6.3-21.5 0.0 2.4 0.9 5.3 2.5 2.6 33.3 2.2 Lawrence 20-VIII-76 12.0 3.5 194 (E) 21.4-23.2 (H) 8.3-21.4 0.0 6.3 0.9 5.3 10.4 6.5 53.1 0.0 Little Mill 20-VIII-76 8.0 4.4 154 (E) 22.6-23.8 (H) 9.5-22.6 0.0 11.3 0.8 7.1 17.5 35.9 84.0 0.0 MacDonald 16-VIII-76 3.2 3.2 135 (E) 21.7-22.3 not present Palmatier 18-VIII-76 10.0 4.2 176 (E) 20.5-22.2 (H) 8.3-20.5 8.0 4.5 165 (E) 21.8-23.7 (H) 12.0-21.8 15.0 5.0 115 (E) 20.9-22.5 (H) 5.0-20.9 Three Lakes 20-VIII-76 Warner 18-VIII-76 0.0 1.2 0.3 2.7 not present 0.0 2.5 0.0 3.3 not present 26.3 0.0 51.4 0.0 11.5 5.8 34.4 0.0 36.9 0.0 71.3 0.0 to GENERAL POPULATION DYNAMICS Four data sets were available for the study of the midsummer population dynamics of the two Daphnia species in Wintergreen Lake, resulting from noon and midnight Van Dorn collections and from vertical tow samples taken at midnight and from July 10-31 in midafternoon. The average coefficient of variation between subsamples from vertical tow collections was 21.7%, and 26.7% between the replicated tow net samples. The average coefficient of variation between replicate Van Dorn samples time specific) was 28.0%. (depth and In contrast to these relatively low coefficients of variation within sampling regimes, the variation between estimates derived from different regimes was substantial, amounting to 50.8% between day and night estimates derived from Van Dorn samples, 63.8% between day and night estimates from vertical tow samples, and 73.6% among all sampling regimes. Substantial variation is expected among collections made with dif­ ferent devices and at different times of the day land and Rognerud 1974, Smyly 1968, Szlauer 1964, Hodgkiss 1977). In spite of this variation, the 14 (Lange- 15 different sampling regimes concur in the timing and pre­ cipitous nature of the decline (Figure 2). That this decline occurred in a similar fashion on a lake-wide basis was confirmed by examining zooplankton collected at stations B, C, and D. Daphnia pulicaria appeared shortly after spring overturn (Figures 2, 3, and 4), hatching from epphipia. The population increased throughout April, reaching its maximum density in early May. resting eggs June. (epphipia) Production of males and occurred from early May until mid- Since only parthenogenic eggs were included in the calculation of the egg ratio, a sharp decline in the egg ratio was expected and observed during this period (Figure 2). Following the disappearance of males and epphipia-bearing females in June, a residual population resumed active parthenogenic reproduction, although the egg ratio showed a steady decline until late July when the population disappeared altogether. D. pulicaria was generally found at depths where water temperature did not exceed 22°C and where oxygen was present (Figure 4). There was little diel vertical migration of this population to warm epilimnion waters (Figure 4). Daphnia galeata mendotae did not appear until June, but increased dramatically thereafter. After reaching its peak density in mid-July this population 16 Figure 2. Top p a n e l : Mean number of Daphnia pulicaria (--- ) anci d . galeata mendotae (--- ) per liter as estimated by noon (o) and-midnight (•) Van Dorn collections and by midafternoon (□ ) and midnight ( ■ ) vertical tow collections. For graphical clarity, Van Dorn and midnight verti­ cal tow records are terminated on the last date each species was observed in those samples. The last date each species was observed in mid­ afternoon vertical tows is indicated by a P (for D. pulicaria) or G (for D. galeata mendotaeJ~ on the abscissa. Bottom panels: Mean number of parthenogenic eggs per female D. pul­ icaria and D. galeata mendotae as estimated by four sampling regimes; symbols as used in top panel. 17 10 EGGS PER FEMALE NUMBER PER LITER 100 .1 •*v 1 .1 APRIL MAY Figure 2 JUNE JULY Figure 3. Depth-time distribution of isopleths of temperature and dissolved oxygen (mg liter“l) (bottom panel). (°C) (top panel) 00 0 1 2 3 4 5 6 1 2 3 4 5 6 — ‘ APRIL 11 ru '— MAY — --------- j----------1---------- 1 ------- “— JUNE JULY Figure 3 AUG. SEPT. OCT. Figure 4. Depth-time distribution of isopleths of Daphnia pulicaria per liter at noon (top) and midnight (bottom). Water depths without oxygen and greater than 22°C are indicated by stipling and shading, respectively, in the lower panel. 0 1 2 3 4 5 6 1 2 3 4 5 anaerobic 6 APRIL MAY JUNE JULY Figure 4 AUG. SEPT. OCT. 22 disappeared in late July (Figures 2 and 5). weekly sampling regimes The three (midnight and noon Van Dorn samples, midnight vertical tows) did not provide suf­ ficient resolution of the rapidity of this species' decline; the vertical tow samples collected each day in midafternoon during July showed that this decline was extremely rapid and spread over a twelve day period (Figure 2). The egg ratio of D. galeata mendotae declined throughout July suggesting the importance of changing food conditions in their late July disappearance. This species was concentrated in the metalimnion during the day, but became more dispersed in the epilimnion at night (Figure 5). Figure 5. Depth-time distribution of Daphnia galeata mendotae per liter at noon (top) and midnight Tbottom). KJ U> 0 1 1200 2 3 4 TOO DEPTH (m ) 5 1 2400 100 2 50 30 3 4 5 6 APRIL M AY jUNE jULY Figure 5 AUG. SEPT. OCT. IN SITU LIFE TABLE EXPERIMENTS Basic to a discussion of a causal link between a declining egg ratio and a declining population is the assumption that some source of mortality is offsetting the remaining reproduction occurring in the population. In previous studies of the midsummer dynamics of various zooplankton species (Hall 1964, Wright 1965, DeBernardi 1974, Kerfoot 1975, Kwik and Carter 1975), the instan­ taneous rate of mortality d has been estimated by the difference equation, r = b-d, where r is the observed rate of population change and b is the instantaneous birth rate (based on egg ratio and development time information: son 1974). see Caswell 1972, Paloheimo 1974, Edmond­ Past interpretations that changes in d resulted from changes in predator-related mortality rest on the assumption that physiological mortality did not change during the period over which r and b were esti­ mated, and that shifts in the age structure of the popu­ lation did not result in biased estimates of birth rate. In this study a series of in situ life table experiments was performed to assess the predator-free, age-specific 25 26 response of both Daphnia species to natural food and temperature conditions from June 17 to September 12, 1976. This information provided a means of calculating popu­ lation growth rate and time- and age(size)-dependent physiological mortality. Methods and Materials Each life table experiment was started from indi­ vidual female Daphnia isolated from Wintergreen Lake water during this period and incubated (at Station A, see Figure 1) in a 75-ml clear glass jar with screw cap and placed at 0.5 m, 2.5 m, 3.0 m, or 3.5 m. Such females and all other cohorts derived from them were incubated in lake water collected at the same depths the jars were suspended and filtered through a 145-)jm mesh net to remove other zooplankton. When the brood females gave birth, they were removed and discarded, and the life history characteristics duction) (survivorship, growth, repro­ of their offspring were followed until all members of the cohort were dead or until September 12, when all remaining cohorts were terminated. During the life of the cohort members, newborn individuals were removed and used to start other cohorts or discarded. Animals were transferred daily by wide-bore pipette to clean jars and freshly collected water, and returned to their in situ incubation depths within two hours of their removal from the lake. This daily 27 transfer procedure permitted a constant renewal of water and algae as the animals might experience in the lake (minus predation). In addition, the transfer period per­ mitted daily assessment of the survivorship, reproduction and growth of the members of each cohort. Survivorship was measured by counting the number of live animals during transfer to the clean jars; collection of dead individuals at the bottom of the previous day's jars provided a check on these daily survivorship counts. Reproduction was measured during the daily removal of newborn individuals (easily distinguishable from adults by size). In addition, any dead newborn individuals were collected from the bottom of the jars and eggs cast off in undeveloped con­ dition with the molts were counted. Individual growth was assessed by measuring the molts (exclusive of tail spine) discarded during the previous day's incubations. The response of daphnids to different food conditions throughout the summer was assessed by continually start­ ing new cohorts (and thus having several cohorts of dif­ ferent age present on any date in the summer); and the entire set of cohorts provided a means by which the difference between daphnid response to natural conditions with and without predators might be assessed. In addition to the assessment of survivorship, growth and reproduction by animals in water from depths normally inhabited by them in the lake, animals were also 28 incubated in 0.5 m water at a depth of 4.5 m after the freshly collected surface water had been cooled to temper­ atures at 4.5 m. This particular manipulation provided a measure of the effect of temperature were identical) species. (where food conditions on life history characteristics of these As in the epilimnion (0.5 m) and metalimnion (2.5 m, 3.0 m, 3.5 m) experiments, incubations using 0.5 m water at 4.5 m were started on several dates during the summer to evaluate the effect of seasonal changes in natural food conditions. The life history characteristics of over 250 cohorts of the two species were determined during the period June 17-September 12. Information from these cohorts was combined according to cohort starting date and depth of incubation to give age-specific survivorship (1 ) and reproduction X (m ) schedules for each species. X Survivorship and repro- duction were then used to calculate population specific) growth rate r (hereafter, cohort r ) , where Z 1 m e“rx = 1.0 X X (cohort (Lotka 1945). Cohort densities were determined by the number of offspring produced per brood by the isolated brood females, and ranged from 1 to 10 animals/75 ml jar (13.3 to 133.3/liter). Initial cohort density did not significantly affect cohort r on several test dates during the summer, presumably because the rates of exploitation of food in the water at these densities 29 were sufficiently low compared to the 100% replacement of water per day. Population growth rate generally decreases (Frank, Boll and Kelly 19 57) or remains constant (as in present results) with increasing cohort density. Since the initial cohort densities in the in situ life table experiments were greater than or equal to Daphnia densities in the lake, they should, therefore, tend to provide con­ servative estimates of their growth capabilities under natural conditions. Results and Discussion Population growth rates (r) calculated from the cohort experiments were much higher throughout the summer than anticipated from the observed dynamics of the two Daphnia species in Wintergreen Lake. All cohorts of D . pulicaria started during the period June 17-September 12 were able to replace themselves (r ^ 0) and, in general, r values exceeded 0.1 Similarly, D. galeata (day ^ ) . mendotae showed positive r values throughout the summer with the exception of three cohorts incubated at 3.5 m in early July. The natural population of D. galeata mendotae survived longer than the members of these three cohorts, and all other cohorts of this species were capable of replacing themselves. (n = 58) Thus it appears that the natural milieu was suitable for the continued growth of both species throughout the summer. 30 Figure 6 shows that cohort r of D. pulicaria in epilimnion and metalimnion incubations was depressed in mid-July. Separate analyses of 1 X to Keyfitz 1968) and m X data (according showed that changes in age-specific sur­ vivorship from June to July only accounted for about 4% of the reduction in cohort r during this period. trast, about 75% action) In con­ (the remaining 21% being due to inter­ of the observed reduction in cohort r was associ­ ated with an age-independent reduction in reproduction for a six day period in mid-July (Figure 7). Cohorts begun shortly before or during this period thus took longer to begin reproduction, and cohort r was thus substantially reduced. The mid-July food and temperature conditions associated with this reduction in reproduction will be discussed later. In spite of this reproductive slowdown, individual growth was not slowed during this period. The size structure of the natural population during mid-July shows a gradual elimination of small-bodied individuals probably caused by reduced recruitment as seen in the life table experiments coupled with continued growth of existing juveniles into larger size classes (Figure 8). The successive elimination of large size classes shown in Figure 8 is unaccounted for by mortality estimates from the in situ life table experiments. The presence of egg-bearing individuals in the natural population (Figures 2 and 8) in mid-July appears Figure 6. Isopleths of population growth rate (cohort r) for cohorts of Daphnia pulicaria as a function of the date the cohorts were started and the mean water temperature experienced by the cohort's individuals prior to maturation. Epilimnion incubations are indicated by closed circles (•) , metalimnion (2.5 m, 3.0 m, 3.5 m) incubations by open circles (o) , and hypolimnion incubations by closed squares ( ■ ). MEAN TEMPERATURE (°C ) 25 •• • V* 20 15 20 JUNE 30 10 JULY DATE COHORT BEGUN Figure 6 20 30 Figure 7. Age-specific reproduction (m^) of Daphnia pulicaria cohorts incubated at 0.5 m. 1 35 O Daphnia pulicaria O mx 30 0 o ■ >5 3< M <5 1 < * <3 25 AGE (days) 0< ° £1 * J 20 W" u> 15 r r 10 r o o o o o oo o O O ■ II fill III • • 20 iu n e • 301 • 10 • 20 July DATE, 1976 Figure 7 30 10 August 35 Figure 8. Size structure of Daphnia pulicaria popu­ lation during July, 1976. The solid vertical line indicates the minimum size of reproductive individuals in the population; shaded areas indicate egg-bearing females. 36 D. pulicaria .4 .6 .8 1.0 1.2 1.4 1.6 1.8 7 -6 7 -1 9 .1 - .1 7 -1 0 FREQUENCY .1 - 7 -2 1 7 -1 3 .1 - l 'LI 7 -1 5 .1 7 -2 3 7 -1 7 .1 .1 - BODY SIZE Figure 8 37 to contradict the near-total elimination of reproduction observed at the same time in the life table experiments. During mid-July some eggs did appear in animals in the life table experiments but they were largely aborted with the cast molts (Figure 9); the eggs appeared quite normal and would have been included in an egg ratio calculation from a natural sample. In contrast to D. pulicaria, cohorts of D. galeata mendotae did not show any decline in cohort r in epilimnion incubations in mid-July (Figure 10). An analysis of the size structure of the D. galeata mendotae population dur­ ing its late July decline strongly suggests that sizespecific mortality of reproductive individuals was respon­ sible for its elimination, as the larger size classes were successively removed from the natural population as the decline proceeded (Figure 11) . This mortality cannot be explained by natural physiological mortality as measured in the life table experiments. The declining egg ratio of the natural population of D. galeata mendotae (Figure 2) during this period does not appear to be consistent with the continued ability of cohorts in the life table experiments to grow at r values of 0.15 to 0.3 (Figure 10). This apparent discrepancy is resolved by noting the removal of large-bodied, eggbearing females from the natural population during this period. Because the elimination of larger size classes 38 Figure 9. Daily egg production (eggs*adult female alive“l* day”l) by Daphnia pulicaria in in situ life table experiments at all incubation depths. Shaded areas indicate eggs which were cast off in undeveloped condition with molts; open areas indicate live births. No cohorts were started at 3.5 m after mid-July. 39 5.3 4-| 0 .5 5.5 4 2 0 6.3 3 .0 m 2 - DAILY EGG PRODUCTION 2.5 m 4 -1 3 .5 m no cohorts present 4 4.5 2- JULY Figure 9 AUGUST 5.9 Figure 10. Isopleths of population growth rate (cohort r) for cohorts of Daphnia galeata mendotae as a function of the date the cohorts were started and the mean water temperature experienced by the cohort's individuals prior to maturation. Symbols as in Figure 6. The three cohorts from 3.5 m having negative r values (see text) are omitted. TEMPERATURE (°C ) MEAN 25 0(0 20 15 1 10 JULY 20 DATE COHORT BEGUN Figure 10 30 42 Figure 11. Size structure of Daphnia galeata mendotae population during July/ 1976. Format as in Figure 8. 43 P. galeata mendotae FREQUENCY .4 .6 .8 1.0 1.2 1.4 1.6 7-10 7-21 7-13 7 -2 3 7-15 7 -2 5 7 -2 6 7-17 .1 - .1 - 7-27 7-19 1 .4 . - .3 BODY SIZE Figure 11 44 is so rapid in late July (the maximum size of D. galeata mendotae decreases from 1.8 mm to 0.8 mm in two weeks) the contribution of parthenogenic eggs by larger individuals is reduced more rapidly than the remainder of the popu­ lation can re-establish a stable age-size structure. The constancy of the relationship between brood size and body size (Figure 12) throughout this period supports the idea that the observed decline in the egg ratio of the natural population is not due to inadequacy of the natural food supply. It appears that the slope of the brood size- body size relationship is less sensitive to rapid tran­ sitions in the age or size structure of the population than is the traditional egg ratio calculation, and may be preferable as an indicator of population growth potential. The decline of both Daphnia populations in Wintergreen Lake is associated with adult mortality in excess of that observed in the in situ life table experiments. This mortality appears to be size-selective as larger size classes are eliminated prior to the removal of smaller ones. In addition, in the decline of D. puli- caria, juvenile size classes show a disproportionate loss due to continued individual growth and reduced repro­ duction by adults. 45 Figure 12. Brood size-body size relationships for Daphnia galeata mendotae population during July, 1976. Vertical bars include the mean ± one standard deviation. 46 .4 .6 .8 1.0 1.2 1.4 1.6 m m 7-13 7-21 =ti±TL 7-15 SIZE 2 BROOD 7-23 7-17 7-25 2- 2 7-19 7-26 - -p 2— - 2- BODY SIZE Figure 12 i —n POPULATION LOSSES Many of the hypothesized adverse effects (Table 1) of midsummer food and temperature conditions, had they been present, would have been reflected in the results obtained in the in situ life table experiments. In addition, physical limnological parameters of Wintergreen Lake (e.g., diel pH, weekly pH, dissolved oxygen) failed to show any marked deviations during the decline of the two Daphnia populations. However, potential sources of mortality excluded from the in situ life table experiments do include preda­ tion by the bluegill punctipennis (Lepomis macrochirus) , Chaoborus (Diptera) and Tropocyclops prasinus. Chao­ borus punctipennis did not show any increase during the Daphnia declines, as has been commonly observed for cladoceran predators of Daphnia (Hall 1964, Wright 1965, DeBernardi 1974). 0.84/liter) The stable numbers of Chaoborus (5.32 ± in June and July presumably arose from a balance of natural population growth and predation by yellow perch (Perea flavescens) ; collections of adult yellow perch (> 150 mm std. length) in July showed that their limnetic feeding activity was almost entirely 47 48 confined to Chaoborus. The increased water temperature in midsummer may nevertheless result in increased preda­ tion rates by these potential daphnid predators 1975). Their small head size (Fedorenko (< 1.2 mm) makes it unlikely that any adult daphnids will be eaten (Swift and Fedorenko 1975), and any population-wide predation impact must be exerted through the juvenile size classes of the two popu­ lations (DeBernardi 1974, Hall et al. 1976). size of Tropocyclops prasinus The small (< 0.9 mm) makes it even less likely that mortality of adult Daphnia will result from their presence. That invertebrate predation is not the predominant mechanism involved in the disappearance of the two Daphnia species is supported by the persistence of juvenile size classes of the smaller and presumably more vulnerable species (D. galeata mendotae) during the disappearance of the adults of both populations. Bluegill sunfish collected in mid-July were found to be eating all adult size classes of both Daphnia species. In addition, collections of sedimented seston during July showed that few undigested Daphnia reached the lake bottom compared to the thousands of Daphnia (> 90% of the total) found in fish fecal material. To the extent that these proportions reflect mortality of various kinds they further suggest the importance of planktivory relative to food, temperature, and physical factors in the decline of the two Daphnia populations. 49 The presumed importance of size-selective preda­ tion by bluegills in the observed declines of the two species of Daphnia can be tested further by examination of the age distribution of parthenogenic eggs carried by the females of each population. Edmondson (1968, 1974) discussed general considerations involved with departures from an even egg age distribution generated by a constant and continuous rate of egg laying without attendant popu­ lation growth. Conditions favorable to population growth will result in an egg age distribution dominated by young eggs; the production of new eggs will parallel the increase in reproductive females occurring during population growth and young eggs will always outnumber older eggs. For a declining population, however, two extreme outcomes of egg age distribution analysis are possible. Populations which decline from a reduction in the rate of egg laying per female (with a constant mortality schedule as in the zero population growth situation) will be dominated by eggs in later stages of development. Conversely, a popu­ lation which declines as a result of increased predation rate on egg-bearing individuals will show a preponderance of young eggs, since the probability that ovigerous females survive until their eggs hatch will decrease. Population declines resulting from interactions of increased predation and lowered egg laying rates will exhibit egg age distri­ butions intermediate between these two extremes. 50 Egg age distribution analyses were performed for both Daphnia populations during the months of June and July following the resumption of parthenogenesis by £• pulicaria. The relative age of the eggs was deter­ mined by criteria stated in Table 2. Deviations from an even egg age distribution were quantified as the maximum deviation (d ) of the observed cumulative egg age dismax ^^ ^ tribution from an even egg age distribution for each sampling date and time, as in the Kolmogorov-Smirnov test (Sokal and Rohlf 1969). Positive deviations were used to indicate dominance by young eggs, while negative values indicate a preponderance of older eggs. The range of values possible using the criteria in Table 2 is from +0.697 (only stage 1 eggs present) to -0.84 (only stage 5 eggs present). Deviation values for D. pulicaria showed a steady increase from late June until the population disappeared in late July, consistent with the predation model and opposite to that expected from a population declining simply from constant mortality rates coupled with reduced rates of reproduction (Figure 13). This increase in dmax suggests that although D. pulicaria reproduction was reduced during this period, predation-related mortality was of overwhelming importance in determining the egg age distribution. Figure 13. Distribution of dmax for Daphnia pulicaria (--- ) and Daphnia galeata mendotae (---) from midafternoon To) and midnight ("•) vertical tow samples. 7-i .6 .5 - X -4 -1 E ~0 .34 .2 .1 - ~ r 20 10 JUNE Figure 13 JULY 30 53 Deviation values for D. galeata mendotae cor­ responded to expectations for periods of population growth and decline due to predation (Figure 13). Popu­ lation size increased until about June 17, when the population size temporarily stabilized and then began to decline (Figure 2). Declining values prior to June 17 were expected from growing populations approach­ ing their maximum densities; this trend was reversed as the population of D. galeata mendotae began to decline. If this decline had been due to a slowdown in reproduction dmax _ , values should have decreased further (and become nega 3 tive) but instead they increased again, consistent with the effects of increasing predation. Just before their disappearance, mortality was sufficiently intense that only females bearing eggs in the first two stages of development (Table 2) were found in the population. Although the data are not extensive, diel changes in d max values are consistent with the interpretation given here (Figure 13). Limnetic activity by bluegills is generally restricted to daytime periods (Baumann and Kitchell 1974), so a lowering of the deviation of the egg age distribution from an even egg age distribution should be expected after nightfall. At this time, females with eggs in early stages of development will develop into later stages, and give rise to a more even egg age dis­ tribution. Synchrony in the timing of egg laying may 54 give rise to the diel changes in egg age distribution observed here, favoring newly laid eggs (positive dm a x ) after a short period of egg laying, and older stages (negative d ) at some later time. 3 max However, it is ' unlikely that simple diel periodicity in egg laying could give rise to the more complicated patterns which developed among samples collected at the same time of day during the decline of Daphnia in Wintergreen Lake. SPECIES SUCCESSION The succession of Daphnia pulicaria by D. galeata mendotae in Wintergreen Lake appears typical of species succession of zooplankton congeners throughout the world (Hutchinson 1967). Only in the near-coincidence of the decline of both species' populations does the sequence in Wintergreen Lake appear unique; usually periods of decline of zooplankton congeners are separated by several weeks or months. The closeness of the population declines to the reproductive slowdown of Daphnia pulicaria also bears further investigation. As was pointed out before, life table information showed that D. galeata mendotae was capable of continued reproduction throughout July, and that little variation in its capacity to increase (as measured by cohort r) occurred during the period of population decline (Figure 10). Another indication of continuing repro­ ductive potential is the consistency of the brood sizebody size relationship in field samples collected during July (Figure 12). In contrast, little is known of seasonal variation in predation pressure on this species, 55 56 although analyses of population size structure and egg age distribution suggest it to be very important. If predation was severe enough in late July to cause such a precipitous and total disappearance of this species, why was it capable of an equally dramatic increase in June and early July? One possible answer is that fish abundance and feeding activity in the limnetic zone increased in late July, resulting in increased mortality of D. galeata men­ dotae . Suggestive evidence that this may be true in Wintergreen Lake was obtained from trial runs of the sediment collecting jars 1976. (see p. 10) on May 4 and June 8, On neither of these dates did the sediment traps (n = 8) contain fish fecal material containing Daphnia as was observed during mid-July, although Daphnia molts and recently dead Daphnia were found in the traps. A seasonal which would result shift tolimnetic feeding by bluegills in these sediment trap results may be brought about by declining populations of littoral zone prey. Densities of Simocephalus serrulatus among the lily pads (N u p h a r , Nymphaea) (Figure 1) declined notice­ ably from June to July, and bluegills collected in midJune were found to be eating these littoral zone prey. Werner and Hall (pers. comm.) have observed a similar seasonal shift by bluegills to limnetic prey as littoral zone prey declined in abundance. Baumann and Kitchell 57 (1974) propose that diel migrations of bluegill between limnetic and littoral zones are related to the quantities of prey available in these areas. In addition, the rapid reduction of population densities of D. pulicaria in mid-July may cause bluegills already normally residing in the limnetic zone to include greater numbers of D. galeata mendotae in their diets. In the period prior to its reproductive depression, D. pulicaria was just able to maintain its population density, even though its egg ratio showed that it was capable of increasing in numbers. This lack of population increase suggests that a balance was present between predation on this species and its predator-free rate of increase. As this species declined ductive collapse) (following its repro­ planktivores present may have included (due to optimal foraging considerations: see Werner and Hall 1974) more of the smaller bodied but abundant D. galeata mendotae in their diets. That D. galeata mendotae suffered an increasing mortality during this period is born out by the reduction in large size classes present and by the increasing tendency for only females carrying younger, less advanced eggs to be represented in the popu­ lation. Several other observations of a decline of D. pulicaria before that of D. galeata mendotae (Birge 1896, Hall 1964, 1971, Noble 1975) are consistent with this view. Additional analyses of prey species selection by 58 planktivores on a seasonal basis would aid in evaluating the possibility suggested here that declines in single species populations may lead to predation-related declines in other species. EVOLUTION OF DAPHNIA LIFE HISTORY CHARACTERISTICS The preceding discussion of the midsummer dynamics of two Daphnia species in Wintergreen Lake suggests certain similarities and differences in the reasons for the decline of these populations. To the extent that these results identify factors potentially important in the regulation of these two species in north temperate zone lakes, they also provide a basis for discussion of the evolutionary response of these species to typical midsummer conditions. Certain features of the Wintergreen Lake system, such as size-selective predation by planktivorous fish and the tendency of D. pulicaria to avoid the epilimnion in midsummer, are also found in other lakes where these species occur. Similarly, the combination of bluegreen algae and high water temperatures associated with the abortion of parthenogenic eggs of D. pulicaria in Winter­ green Lake is common to many eutrophic lakes where this species occurs, and should be expected to play an important role in its evolution. 59 60 Daphnia pulicaria Assessment of the effects of midsummer conditions on D. pulicaria is hampered by potentially erroneous designations of several populations of similar appearing species. Brooks (19 57) considered D. pulicaria Forbes to be synonymous with D. pulex Leydig, and only recently have zooplankton workers in North America (Brandlova et al. 1972, Wong and Ward 1972, Ward and Robinson 1974) begun to routinely distinguish between these two species, as has been done much earlier in Europe Esslovci 1966). (Hrbdcek 1959, Hrbdekovci- As a result, little ecological data exist for clearly defined populations of D. pulicaria. (D. J. Hall, pers. comm.) Hrbdcek indicated that a Daphnia popu­ lation from Oneida Lake, New York, previously identified as D. pulex (Hall 1971, Clark 1975, Noble 1975) actually D. pulicaria. comm.) is Dodson's current studies (pers. of Daphnia systematics suggest that Birge's (1895, 1896) Lake Mendota population of D. pulicaria reassigned since that time to D. sch0dleri 1967) (Brooks 1957, Hutchinson is probably also referable to D. pulicaria. In addition, my own collections of Daphnia from lakes studied by Haney and Hall (1975) show that the D„ pulex popu­ lations reported by them are probably D. pulicaria instead (Table 3). Brandlova et al. (1972) indicated substantial physiological differences between D. pulex and D. pulicaria and concluded that the distribution of 61 these two species was pond and limnetic, respectively. In that numerous studies of D. pulex collections from limnetic regions Arnold 1971, Crowley 1973) (sic) are based on (e.g., Burns 1969, the amount of ecological information on D. pulicaria in North America may be greater than at first supposed. In the following dis­ cussion, information obtained from limnetic populations identified as either D. pulex or D. pulicaria will be used. Limnetic populations of this species complex (D. pulex or D. pulicaria) are generally confined to colder w a t e r s , either being found in spring and autumn pulses (Hall 1964, Fott 1975, Noble 1975), or confined to hypolimnetic regions in midsummer (Table 3, Birge 1895, 1896, Hall 1964, Haney and Hall 1975). Management of highly eutrophic lakes by total aeration (Shapiro and Pfannkuch 1973) or hypolimnion aeration (Fast 1971) consistently results in either a numerical increase in this species or expansion of its vertical distribution to include the oxygenated, cool-water habitats. Several components of the population growth (see review by Hall et al. 1976) of D. pulicaria have been shown to be negatively affected by typical midsummer resource, predator or temperature conditions, which may result in these distribution patterns and response to aeration. For example, the depressive effects of high temperatures (> 2 0 °C) and large-celled bluegreen algae 62 on filtering and assimilation rates have been documented (Burns 1969, Arnold 1971, Crowley 1973); in addition, Arnold has shown that these effects are manifest in reduced reproduction and survivorship. (1962) and Brandlova et al. Hrb^Skovd-EsslovS (1972) have reported that D. pulicaria is less efficient reproductively at 28°C than at either 20° or 24°C; Hrbackova-Esslovd also reported that eggs produced at 28°C often remained undeveloped. The present study showed that food-temperature conditions present in the epilimnion and metalimnion in mid-July in Wintergreen Lake resulted in reduced rates of egg production and increased abortion of these eggs. These events occurred in 0.5 m, 2.5 m, 3.0 m, and 3.5 m incubations, but not in 4.5 m incubations where 0.5 m water and food had been cooled to temperatures at 4.5 m. This pattern suggests that temperature was of major importance but temperatures earlier and later in the summer were equal to those of mid-July (23-27°C); only the food conditions differed between this period and those of later and earlier. The size structure of the algal community shifts during mid-July from numerous small algae (e.g., Dysmorphococcus, Cyclotella, Crypto- monas) to fewer large-celled algae Ceratium, Volvox) . (e.g., Anabaena, In late July increasing numbers of bacteria-detritus aggregates re-establish the high 63 densities of small (2.4-7.6ym diameter) particles. These shifts are apparent from photographs of settled water samples (Figure 14) and from size frequency distributions obtained by Coulter Counter analysis of samples from this period (Figure 15). Thus it appears that an interaction of food and temperature conditions brought about the reduction in egg laying and abortion of eggs observed in mid-July. Lampert (1977) recently demonstrated that algae of large cell size were required in greater abundance to offset reduced filtering efficiency by Daphnia pulex. However, at lower temperatures the balance between assimilation and respiration could be achieved at lower concentrations of a given algal species. In this respect, the results obtained in Wintergreen Lake provide a natural example of the interaction of food and temperature effects observed by Lampert under laboratory conditions. In spite of D. pulicaria1s ability to survive midsummer epilimnetic temperature and food conditions, they consistently avoided this habitat in Wintergreen Lake and in a number of nearby lakes (Table 3). The seasonal change in daytime vertical distribution of this population in Wintergreen Lake may be explained as a response to avoid visual predators which become more active in the limnetic zone during summer Kitchell 1974). (Baumann and However, the summer nighttime Figure 14. Photographs of replicate fields (0.107 m m 2 ; 5.28 yl) of settled water samples of algae collected from 0.5 m during July, 1976. 0.5 m JULY, 1976 6 13 20 27 Figure 14 66 Figure 15. Abundance of algal particles of three size fractions and three dominant genera collected from 0.5 m during July, 1976. Note that Anabaena and Ceratium densities have been multiplied by 1000 for inclusion on these axes. 67 10 Anabaena (x 103) 2.4 - 7.6 jjm CELLS PER ML. •D Ceratium (xio3) 4 _ 7.6 -15.2 Jim 10 > 1 5 .2 jim JULY Figure 15 68 distribution of the species (concentrated in the metal- imnion in Wintergreen Lake and in the hypolimnion when oxygen conditions permit) does not appear to be a strategy which results in maximization of population growth rate r as measured in the life table experiments. Only during a very short period in mid-July is cohort r higher in the hypolimnion incubations than in the epilimnion incu­ bations (Figure 6). Approximately 60% of the three-fold variation in cohort r of D. pulicaria in the epilimnion incubations can be accounted for by variation in food particle density starting date. (2.39-7.6 ym diameter) on the cohort In the hypolimnion incubations epilimnion water was also used) (where only 3.4% of cohort r variation was attributable to food particle density. In view of the great sensitivity of D. pulicaria to changes in food conditions at epilimnion temperatures, the best strategy for this species for sustaining high population growth may be to avoid water depths where high temper­ atures occur. This is in spite of the ability of D. pulicaria to achieve positive population growth in epilimnion incubations throughout the summer. Alternately, any temporary gain in population growth by D. pulicaria due to migration to warmer, epilimnion waters at night during good food periods may be offset by higher predation rates in these epil­ imnion waters. For example, bluegills may be sufficiently 69 effective in nocturnal feeding (cf. Keast 1970, Seaburg and Moyle 1964, Baumann and Kitchell 1974) that higher population growth rates are achieved by D. pulicaria by remaining in the hypolimnion or metalimnion. However, it is unclear if nocturnal feeding by bluegills occurred in this study, as collections made at dawn indicated a predominance of littoral zone feeding by these fish. The hypothesis that predation by vertically migrating Chaoborus may influence the vertical distribu­ tion of D. pulicaria is untested due to the avoidance ky Chaoborus of the Van Dorn samplers used in this study. However, the diel vertical migrations of D. galeata me n dotae suggest that diel variations in the depth distribu­ tion of Chaoborus are probably unimportant to the vertical migration of D. pulicaria. Another possible explanation for D. pulicaria*s nighttime distribution involves the presence of D. galeata m e ndotae. Although D. pulicaria survives in water from which D. galeata mendotae has been removed by filtration (see methods and materials section, In situ life table experiments), the possibility of either aggressive inter­ action or the secretion of allelochemic substances by D . galeata mendotae exists. Responses by zooplankton to physical disturbance by other zooplankton (Strickler and Bal 1973, Fryer 1957) or the secretion of water-borne organic compounds (Banta 1939, Banta and Brown 1929, 70 Dahl et al. 1970, Katona 1973, Griffiths and Frost 1976) are well documented. Strickler (pers. comm.) has observed that as part of the avoidance reaction of zooplankton prey to predators, feeding activity is reduced while swimming speed is increased. The effect of similar encounters among potential competitors on food intake and population growth has not yet been determined. However, the lack of a density effect cn cohort r in this study makes it unlikely that physical disturbance of D. pulicaria by D. galeata mendotae is of major importance in the diel vertical distribution of D. pulicaria. Any chemical compounds released by D. galeata m en­ dotae responsible for the restriction of D. pulicaria to the metalimnion must be volatile or rapidly broken down, or their presence in freshly collected water would have influenced the animals in the life table experiments. As the water was aired slightly (to remove zooplankton) any volatiles may have been lost. If the product is only produced at night when D. galeata mendotae is found in the epilimnion and rapidly broken down during the day (by sunlight, for example) it would not have been included in water collected in mid-afternoon and used in the in situ life table incubations of D. pulicaria. The role of allelochemic compounds in zooplankton suc­ cession obviously requires further study. 71 Daphnia galeata mendotae In contrast to D. pulicaria/ D. galeata mendotae is generally thought of as a warm-water species by virtue of its midsummer epilimnetic distribution (Hall 1964, Tappa 1965, Haney and Hall 1975, Table 3) and its capacity for active reproduction in midsummer 1964, 1971, Tappa 1965, Cummins et al. 1969). (Hall In the present study, D. galeata mendotae failed to show any mid-July reproductive decline in epilimnion incubations. Burns (1969) found no depressive effect of high temperature on filtering rate as she had for D. p u lex. D . galeata mendotae appears to be far more susceptible to predation than to food or temperature conditions, as indicated by Hall (1964, 1971), Suffern study. (1973) and in the present Whether or not its large-headed, summer morphs are less vulnerable to predation by invertebrates than noncyclomorphic forms is as yet untested. Discussion Differences between Daphnia pulicaria and D. galeata mendotae include preferred habitat, reproductive response to midsummer algae-temperature conditions and laboratory-measured filtering rates at various temper­ atures. Probably most significant, however, are dif­ ferences in the body size of these two species and their tendency to produce resting eggs or epphipia. Whether or not these differences are due to past evolutionary 72 pressures similar to those which resulted in the midsummer declines of these two species in Wintergreen Lake is, of course, unknown. However, a possible test of the gen­ erality of the results obtained in this study would be to examine current life history characteristics of these two species for consistency with the interpretations given earlier on the importance of various factors in bringing about the midsummer decline. As indicated in Table 4, D. pulicaria tends to produce a larger egg, mature at a larger size and ulti­ mately reach a larger body size than does D. galeata men­ dotae. This undoubtedly makes D. pulicaria more vulnerable to size-selective predators such as the bluegills found in Wintergreen Lake. The tendency for these two species to produce enlarged helmets in midsummer (cyclomorphosis) is also consistent with their vulnerability to sizeselective predators. D. galeata mendotae, by reducing its body length and producing a large helmet instead, may increase its reproductive rate (Brooks 1966) , while decreasing its vulnerability to both vertebrate and invertebrate predators (Brooks 1966, Dodson 1974). D. pulicaria, which is larger in size than D. galeata m e ndotae, does not produce enlarged helmets, presumably because the resulting reduction in body size or visibility due to use of this tactic is insufficient to substantially reduce mortality from size-selective predators. Instead, Table 4. Summary of life history characteristics of Daphnia pulicaria and Daphnia galeata mendotae. Characteristic D. pulicaria D. galeata mendotae 20°C 25°C Filtering rate temperature optimum Preferred midsummer habitat Metalimnion, hypolimnion Epilimnion, metalimnion Diel vertical migration Restricted to water cooler than 22°C Pronounced, to epilimnion Response to bluegreen algae, high temperature conditions Reduced reproduction, egg abortion No adverse effect Seasons of epphipia production Spring, autumn Autumn Egg size (Stage 1, max. dimension) 0.27 mm 0.19 mm Size at first reproduction 1.0-1.2 mm 0.8-1.0 mm Adult body size 1.0-2.4 mm 0.8-2.0 mm Slope of brood size-body size relationship Gentle Steep Cohort r versus temperature Identical relationship for both species 74 D. pulicaria appears to devote its energies to increase carapace length (exclusive of helmet) also increase brood size. and, indirectly, The tradeoffs involved with these strategies are discussed more fully in Hall et al. (1976). The other major difference between these two species is their propensity to produce resting eggs or epphipia. D. pulicaria produces resting eggs in spring and autumn (Stross 1973, this study), whereas D. galeata mendotae produces these resting eggs in autumn and Hall, 1975, Hall, pers. comm., this study). (Haney It appears that the tendency to produce resting eggs in spring is a characteristic associated with the species more likely to be eliminated in midsummer. Conditions which cause D. galeata mendotae to disappear in midsummer (as in Wintergreen Lake) may be sufficiently rare as to not have resulted in the evolution of an early-summer, epphipia-producing phenotype of this species. In lakes where D. galeata mendotae does not go extinct in mid­ summer, there is a reproductive advantage to producing parthenogenically until autumn. Only two resting eggs per every other instar are produced via epphipial repro­ duction, while the cumulative effect of parthenogenic population growth over an entire summer at a rate of 2-15 eggs per adult instar would probably overwhelm the net contribution of early-epphipia strategists. Only in 75 the case of a large-bodied species such as D. pulicaria likely to be eliminated during midsummer would it be profitable in spring and early summer to produce epphipia at the expense of more numerous parthenogenic offspring. Alternately, production of epphipia by D. pulicaria may be in response to temporary conditions not encountered during June and July when both species were present. Dur­ ing early May, chlorophyll a (corrected for phaeopigments) was undetectable and particulate carbon was at its springsummer low of s? 1 mg/liter (Klug et al. unpubl. data). Thus, production of epphipia may have been the only kind of reproduction resulting in viable offspring at that time. Since life tables were not run for either species during this period, it is difficult to assess whether continued parthenogenic reproduction would have resulted in population growth. Thus, care must be taken to avoid acceptance of all life history characteristics only as adaptations to intense predation without consideration of alternate hypotheses. A great deal has been written about size- related aspects of competitive abilities in zooplankton, and it is interesting to consider the life history char­ acteristics of these two species (Table 4) in light of competitive effects not observed in this study. For example, during the decline of these two species all size classes of D. galeata mendotae carried more eggs than 76 corresponding size classes of D. pulicaria. In response to this apparent competitive superiority of D. galeata mendotae it might be supposed that D. pulicaria evolved to be a larger animal, sacrificing high egg production for an increased ability to withstand periods of food shortage (Threlkeld 1976) or ingest a wider size spectrum of food particles (Burns 1968, Hall et al. 1976). Exami­ nation of the life table data collected in this study shows that this interpretation is only partially satis­ factory. Although it does appear that D. galeata mendotae is more efficient than D. pulicaria in egg production, this is not reflected in higher population growth rates; the relationships between cohort r and temperature for both species are indistinguishable (Figure 16). Thus, for resource and temperature conditions as experienced by these two species in Wintergreen Lake, their population growth rates are identical. Only in the face of predation do their strategies result in different population growth rates. Potential differences in competitive abilities of these two species may appear at different seasons or under different resource conditions, but were not observed in this study. In contrast, all aspects of the life history characteristics of these two species appear consistent with a dominant influence of planktivory in their evolution and seasonal association in Wintergreen Lake. Figure 16. The relationship between cohort r and temperature of Daphnia pulicaria (•) and D. galeata mendotae (o) as determined in in situ 1 i f«=> t^hT^ experiments at all incubation depths. •• •3-1 • COHORT • * •* ° o o • 0 • 2 - 1 •1 • *•&* h °* •* ^ o ^ • • * o o*o<5>0 Q O ° y ° o • 0