Jumping Rhythmicity in a Collembolid Population on a Central California Beach and the Effect of Temperature on Jumping Äctivity Michael D. DeLapa Spring, Hopkins Marine Station I. Introduction Emphasis on physiological adaptation as an evolutionary strategy has often overshadowed consideration of behavior as a means of adaptation. Adaptation, in either sense, is concerned with an organism's ability to exploit best a particular environmental niche. The marine collembolids, commonly called springtails, have adapted to survive in a biologically harsh environment and are one of the few macroscopic inhabitants of sandy beaches in central California. Marine collembolids are limited to this environment, and their delicate nature seems paradoxical in this context; able to survive the severe environmental conditions of sandy beaches, wind, temperature fluctuation, wave action, et cetera, collembolids desiccate within minutes of being removed from a humid environment. Even more difficult to explain is the behavior of jumping which is characteristic of many collembolids found on central California beaches. These collembolids have a jumping organ, the furculata, which allows them to spring into the air. The adaptive significance of jumping behavior is unclear. While Davenport (1903 ) suggests that jumping aids respiration, there is no published work on this to date. In the case of marine collembolids it is evident that the timing of such hehavior must be closely regulated— jumping at inopportune times, such as during high tide or midday, would be disasterous. No literature deals with rhythmic patterns of collembolid jumping activity, but field work conducted by Hopkins Marine Station students (spring, 1978) on a beach near Mussel Point in Pacific Grove, California suggested the presence of a temporal pattern of fluctuating jumping activity in the collembolid Archiostoma besselli. The following work was initiated to investigate the rhythmic nature of jumping behavior in the population Page 2 and to consider the effect of temperature on this behavior. The work was divided into a preliminary field study followed by experimental work in the laboratory. Page 3 II. Field Work A. Materials and Methods A pocket beach near Mussel Point, California was chosen as a study area for field investigation of collembolid jumping activity. A matrix system of hollow metal rods driven into the sand was laid out for reference in collecting animals (figure 1 ). Station numbers were assigned to the rods indicating meters away from a reference post located at the margin of the beach and the land vegetation. Traps were set out at regular time intervals during five field studies. The setting of traps proceeded as such: a three centimeter square of graph paper divided into one-tenth centimeter units was placed inside a 4.5 centimeter diameter jar lid (figure 2 ). One large drop of Mobil 10W-40 oil was spread over the entire surface of the lid, fully coating both the graph paper and the lid surface. The lids were placed approximately two meters away from one another at specific station sites for two minutes. To one side of the stations samples were designated "A", to the other side they were designated "B", When two lids were placed side by side for replicate samples they were labeled Al and A2 or Bl and B2. After the two minute interval, the lids were collected, brought into the laboratory, and surveyed under a dissecting scope for collembolids sticking in the oil. The counting methodology was such that all animals falling onto the graph paper and onto the white surface of the lid were counted unless there were more than about 300 specimens in which case four random square centimeters on the graph paper were counted and multiplied by pi (2.25)4 to estimate the total number on the lid surface. Page 4 II. Field Work B. Results The results from three of the five study periods appear in tablesl, 2 and 3. For the other two collection periods fewer than ten collembolids were caught in any two minute collection period. One of these study ; periods with negligible numbers took place on a windy morning (4:00 A.M. to 1:00 P.M. ) with a morning low tide; the other occurred on a calm foggy day with an afternoon low tide. For the three other collection periods jumping activity as a function of time for two representative stations has been graphed (Figures3, 4 and 5 ). For each station at a particular time samples from A and B were totalled, or if Al and A2 or Bl and B2 were collected, they were averaged, then summed. A tidal cycle is superimposed on the graphs (dashed line ). The air temperature at station O at the collection time was recorded and appear in the tables. Figures 6, 7 and 8 show the distribution of collembolids along the beach at the maximum activity period for each study period. During the five study periods, collembolids ranged from as high as station 9 to as low as station 26, with the greatest activity generally being found near station 19 in the middle of the tidal range. For the April 27-28 study period the largest number of specimens was collected at station 19 at 7:45 A.M. (April 28 ) slightly preceeding a low tide. The morning hours of this collection period were characterized by calm ocean condition, norwind and warm temperatures (approximately 16 degrees centigrade at the time of maximum activity ). For the May 1-2 study period, the largest sample was collected at station 12 at 11:00 P.M. (May 9). This study period was a night of rough ocean conditions, Page 5 no wind, and cool air temperatures (approximately 10 degrees centigrade throughout the night ). For the May 9 study period the most animals were collected at station 18 at 7:30 A.M., slightly following a low tide. This collection period was characterized by calm ocean conditions, no wind and cool air temperatures (approximately 10 degrees centigrade at the time of maximum activity). Several less formal collections were also undertaken with the following results. During a week of high winds, animals were found to occur in negligible (less than ten collected in a two minute interval) numbers at all times sampled during the day and night. An afternoon of light rain at a 4:00 P.M. low tide resulted in animals being active in numbers similar to those found in early morning periods. Collembolids were always found in negligible numbers on warm, sunny afternoons (10:00 A.M. to 5:00 P.M.). Page 6 II. Field Work C. Discussion Examination of tables 1, 2, and 3 and figures 3, 4 and5. reveals the complex nature of collembolid activity in the field as a function of time. In study period 1 (April 27-28 ) maximum activity for all stations lower than station 16 occurs at 7:45 A.M., slightly preceeding an 8:52 A.M. low tide. For stations15 and 16 maximum activity occurs at the 4:20 A.M. collection soon after these stations were uncovered by the ebbing tide. Five of the six stations sampled at 6:15 A.M. show a decrease in activity from 5:35 A.M., thus the appearance of a bimodal activity pattern for this particular study period. In study period 2 (May 1-2 ) a low tide occurs at 1:51 A.M. and maximum activity for stations 11, 13, 14, and 15 occurs at the 3:40 A.M. collection, while station 12 has maximum acivity at the onset of collecting ( 11:00 P.M.). Station 10, though not sampled continuously throughout the night, shows highest activity at 5:55 A.M., two hours after the other stations sampled. This study period was unique in several respects. It followed a week of storms, and the low tide fell up to station 16 because of the high waves. The wave action from the storm had removed nearly one foot of top sand from the beach, thus at station 15 there was a sudden dip in the contour of the beach. By the time study period 3 (May 9 ) took place, the previously exposed coarse sand was again covered by fine sand and the dip was gone. Figure 3 shows the most regular pattern of jumping activity observed at any time, a dramatic example of the usual pattern: all stations sampled (stations 12, 14, 15, 18, 20, and 22 ) peak suddenly at 7:30 A.M., slightly following a 7:00 A.M. low tide. Previous to the huge 7:30 A.M. peak, stations 12, 14 and 18 have a Page 7 slight peak (slight in comparison to the major peak, but still representing hundreds of animals ) of activity at 3:30 A.M. or 4:30 A.M., again suggesting some bimodality in activity. Examination of temperatures for the study periods reveals that for study periods 2 and 3 temperatures were in the same range, approximately 10 degrees centigrade during collections, however, the amount of activity was not similar. Study period 1 and study period 3 both had maximum activity near an early morning low tide, but temperatures were generally higher for study period 1 (approximately 17 degree centigrade at maximum activity ) and activity was much lower during this period as compared to study period 3. This suggests that temperature might be affecting the amount of activity. Low tides for periods 1 and 3 occur nearly two hours apart, yet activity is maximum for both periods about the same time, 7:30 A.M. This suggests that temperature, while possibly affecting the amount of activity, does not affect its timing. Again, examining data from study periods 1 and 3, the occurrence of some activity before the 7:30 A.M. peak is more prominent in study period 1, again being correlated with slightly higher temperatures in this period. In both cases, with an early moning low tide a small peak of activity; occurs on an outgoing tide and preceedsa peak of greater activity. Other environmental conditions also seemd to effect jumping behavior. Periods of winds resulted in very low activity (usually non-existent) of collembolids at any time of day or night. Humidity seemed to have an effect as well. On a rainy afternoon collembolids were observed to be active in numbers which had only been found in early morning periods. Foggy mornings seemed to bring collembolids out in the greatest numbers, but quantitive data on this is lacking. Page 8 The apparent positive effect that low temperatures have on jumping activity in the collembolid population studied and the lack of temperature effect on the timing of maximum jumping activity suggest the presence of an endogenous cuing mechanism in collembolid jumping activity. Work performed by Michelle McSpadden (1978 ) revealed the presence of such a mechanism, and laboratory work was initiated to investigate the effect of temperature on both the amount of activity and the timing of activity. Page 9 III. Laboratory Work A. Materials and Methods Collembolids were collected between 5:00 A.M. and 8:00 A.M. on May 24 near station 18. Plastic trays and coffee cans were sunk in the sand with an inch or so of seawater in the bottoms. A virtual monolayer of collembolids formed on the water surface, and the containers were brought into the laboratory. Using a mouth aspirator, collembolids were concentrated in a small jar from where they were spooned into six Tuperware containers painted flat blåck to simulate constant dark conditions. Infared diodes were positioned in the containers in a manner diagrammed in figure 9. The light detecting diodes were wired to amplifying circuitry (designed by Robin Burnett and built by me ) which then signalled a channel on an EstralineAngus event recorder, thus, activity, as defined by a partial interruption of the infared beam, was continuously recorded, In analyzing the data, the marks on the Estraline Angus chart were counted in hourly segments. Comparison of maximum activity for the different containers was obtained by combining the counts from three consecutive periods of highest activity. The containers with the collembolids ran for 24 hours in ambient laboratory conditions in order to establish relative numbers for latter comparisons of containers. A Rustrak temperature recording device monitored laboratory temperature. At 9:00 A.M. on May 25 the six containers were removed from ambient laboratory conditions and were placed in four different temperature regimes. Two containers were placed in a heated water bath, one container was placed in an insulated container subject to laboratory temperature (control ), two containers were placed in a 12.3 degree centigrade formatemp bath and one container was placed in a styrafoam Page 10 container with a frozen "blue ice" lid. The collembolids were allowed to run in the different temperature regimes for 24 hours until 9:00 A.M. on May 26 when they were returned to ambient laboratory conditions. Five of the six containers were allowed to run for 48 hour under these conditions. Collembolids in the blue ice container ran only 24 hours in the ambient laboratory conditions and werereturned to a very cold temperature regime (0 to 1 degree centigrade) after this period. They spent approximately 22 hours at these cold temperatures, then were allowed to slowly warm. Page 11 III. Laboratory Work B. Results and Discussion The results of the laboratory experiments appear in figures 10, 11, 12, and 13. The temperature in or near each container at the times of maximal activity appears at the top of the graphs. The darkened bar represents fime spent in the particular tmperature regime. The findings are summarized in table 4 and figure 14. A temperature response curve for collembolid activity is shown in figure 15. In all containers maximum activity during the first 24 hours occurs between 5:00 A.M. and 6:00 A.M. on May 25. (This peak of maximum activity will be referred to as the "A" peak.) Also in each case, a smaller activity peak (designated peak "B") occurs approximately 11 hours previous to peak A. These results are consistent with work done by McSpadden (1978 ) and support the conclusion that collembolid activity is, in part, controlled by an endogenous temporal program. McSpadden's work suggests that peak B might be a tidal component of a circadian clock, the A peak being the principal element of a circadian rhythm. In the containers placed in the heated water bath activity decreases to a very low level and does not reappear when the containers are returned to ambient laboratory conditions on the third day. This suggests that within this temperature range (28.8- 34.0 degrees centigrade) mortality is high. Maximum activity in the control container (placed in an insulated regime on the second day ) and in the container placed in the formatemp was confined to the early morning "gate" throughout the course of the experiment. This pattern of activity is consistent with MeSpadden's Page 12 work. The A/B ratio in the control container increasesfrom 2.5 the first day to 4.2 on the second day, possible the result of slightly warmer temperature in the insulated container. On the third and fourth days the ratio returns to approximately the value it was the first day. The overall amount of activity in this container decreases as a function of time, probably the result of mortality and escape. In the two containers placed in the formatemp bath activity increases on the second day by approximately 170% and 330% (peak A ). The timing of the A peak does not seem to be affected, that is, maximum activity still occurs in the early morning period. Also in this temperature regime, the A/B ratio increases significantly from its first day value, and a return of the containers to ambient condtions results in an eventual return of this ratio to approximately its first day value. The increase of the A/B ratio on the second day is due to peak A increasing nearly 5 times as much as peak B. The container placed in the blue ice box exhibited a 25% decrease in activity in this regime as compared to activity during the first day of ambient laboratory conditions. Once again, the timing of the A peak does not appear to be affected in this cold regime (May 25 ). The A/B ratio increases from 4.5 the first day to 12.8 the second day, with A increasing more than B. The unusual occurrence of exaggerated bimodality on the third day with a decrease in the A/B ratio to 1.3 could have several explanations. The cold conditions could have effected the metabolism of the collembolids such that upon reaching the first permissive point in their activity cycle, point B, they become more active than before. Another possible explanation is that the return to ambient conditions from a cold regime has affected the clock mechanism Page 13 independent of a simple physiological response; the collembolid clock might be causing the increased B peak activity rather than the animals' metabólism. Further investigation is needed to explain this phenomenon, Heckrotte's work ( 1960) with snake activity patterns suggests an analogy that might be useful in understanding changes in peak ratios, Heckrotte shows that a peak of activity occurs midday for low temperatures (21.1 degrees centigrade ) and an activity peak occurs in early morning for high temperature (35.5 degrees centigrade ), with shifting bimodal peaks for intermediate temperatures. The conclusion he reaches is that at each temperature activity is maximum at a time of day when temperature would be expected to be optimum for activity. Whether or not this sort of interpretation is applicable to collembolid activity remains to be shown. Another point of interest concerning the container kept in the blue ice box deals with its teturn to very cold conditions on the third day. The temperature in the box was maintained from 0 to 2 degrees centigrade for approximately 22 hours, then allowed to slowly warm. Activity peaked near 1:30 P.M. on the fourth day and did not show the the same sort of decline that had been seen in the other containers. This seems to suggest that the collembolid clock had been stopped by the period of very cold temperatures and started again as the temperature increased. There are some problems with this hypothesis in that the container was kept in cold conditions for approximately 22 hours, yet peak activity occurred approximately 7.5 hours later than would normally be expected (i.e. 7.5 hours after 6:00 A.M. ). This could be explained below a critical temperature if the clock only terminated in the 0 to 2 degree centigrade range. Because the cold regime fluctuated between these temperatures, it is possible that only 7.5 hours was Page 14 spent below the critical temperature. Further work is needed to investigate this aspect of temperature effects. The result that different temperature regimes effect the peak ratios suggests a hypothesis concerning the effect of temperature on the collembolid clock. If peak B is a tidal component of a circadian clock then it could be less obligatory and more responsive to environmental opportunities. The following algorithm might be at work: always be active in the early morning hours and also be active at low tide if condtitions (temperature, humidity, et cetera) are permissible. The observation of animals being active at a 4:00 P.M. low tide on a rainy day is supportive of this idea. When a low tide falls at early morning and conditions are right, one would expect the greatest activity- there is some suggestion of this from the field data that has been presented. The temperature reponse curve of collembolid activity (figure 15 ) displays a maximum at cool air temperatures (14.0 degrees centigrade in this particular experiment ). This result is in agreement with field data.thateshows greater activity at cooler air temperatures. No field work was done at temperatures below 10 degrees centigrade, but it would be interesting to see how cold temperatures could reach in the field before activity is suspended. Page 15 IV. Summary Data from the five field study periods, along with observations from less formal field studies, though complex, suggest formation of several generalizations: 1) Collembolid activity is generally greatest on calm, cool early mornings, with maximum activity occurring from 6:00 A.M. to 8:00 A.M. during a morning low tide. 2) Collembolids can also be actively jumping during the night, but to a much less degree and do not necessarily peak in activity at the low tide. Collembolids do not actively jump on calm, sunny days ( 10:00 A.M. 3) to 5:00 P.M.) everything else being constant. 4) The greatest number of active collembolids on the study beach was normally found near station 18, a position midway in the tidal zone. 5) Collembolid activity is significantly affected by environmental factors. Wind decreases the amount of activity and cool air temperatures appear to increase the amount of activity. 6) Temperature does not seem to affect the timing of the period of maximum activity, but warmer temperatures, while decreasing maximum activity, appear to increase earlier activity. The results of the laboratory experiments suggest the formation of several generalizations concerning the effect of temperature on collembolid jumping activity in the laboratory: 1) Collembolids show an endogenous pattern of maximum activity during the early morning hours peaking at approximately 6:00 A.M., and this activity is maximized in the laboratory at air temperatures near 14 degrees centigrade. Collembolids do not survive in the laboratory when sand 2) temperatures exceed 29 degrees centigrade. Very cold temperatures (O to 1 degree centigrade ) appear to 3) stop the collembolid clock and shift activity from the early morning gate. The A/B ratio is temperature sensitive, increasing significantly in response to a low temperature regime (11 to 14 degrees centigrade ). A return to ambient laboratory temperatures (21 to 26 degrees centigrade) from a cold regime (11 to 14 degrees centigrade ) results in a significant decrease in the A/B ratio within two days. Page 16 5) Different temperature regimes above 8 degrees centigrade affect the total amount of collembolid activity and the amount of activity at particular times (peaks A and B), but they do not affect the timing of maximal activity (peak A). e 17 V. Bibiliography Davenport, C. B. 1903. The Collembola of Cold Spring Beach with special reference to the movements of the Poduridae. Cold Spring Harbor Monographs II Heckrotte, C. 1960. The effect of environmental factors on the locomotory activity of the Plains garter smake. Ph.D. thesis University of Illinois. As referenced in: Sweeny, B. and J. W. Hastings. 1960. Effects of temperature upon diurnal rhythms. Symposia on Quantitative Biology. Vol. 25. VI. Acknowledgement: I would like to thank Michelle McSpadden for helping collect the field data and for sharing her findings with me. Judy Thompson and Susan Harris provided mental relief that enabled me to relax when time had run out. Consultations with Chuck Baxter were instructive even though they cost me several sixers. Dr. Donald Abbott was an inspiration and his incredible enthusiasm was a spiritual amphetamine to me during the pre-dawn hours. The most superlative thanks to Robin Burnett whose time and advice made this work possible. The lost sleep will never be missed. TABLE LEGEND Table 1- Data from 15 two minute collections of collembolids on April 27 to April 28. Stations collected from are designated ;; Air temperature at station O at the time of collecting appears under the collection time. The word "tide" appears at the station where the waves were splashing up to. Table 2- Data from 9 two minute collections of collembolids on May 1 to May 2. Stations collected from are designated. Aif temperature at station O at the time of collecting appears under the collection time. The word "tide" appears at the station where the waves were splashing up to. Table 3- Data from 10 two minute coleections of collembolids on May 9. Stations collected from are designated. Air temperature at station 0 at the time of collecting appears under the collection time. The word "tide" appears at the station where the waves were splashing up to. Table 4- A summary of the effects of temperature on collembolid activity. The temperature regime is indicated. Activity refers to the per cent change in recorded events on an Estraline Angus event recorder. In the A/B Ratio column the number on the left is the A/B ratio on one day, the arrow points to its value the following day. The days being reférred to are indicated above the two columns. sapsasas ooooooo oN 3 uooo oooooo o o aSPooo 2 s p ktaa- 8 tatatavakaava- SNN PNO 5 aa- a SbSo taa- o8o 8 0 0 60 Paus 8N 5 kaa- 5 3 888 8 ptkaaa- 8 0 8 8 8oo 5 - S i o s i 5 J 5 8 8 8. 5 2 o 9 FIGURE LEGEND Figure 1- A schematic diagram of the field matrix used to collect collembolids. Station 0 represents the highest station up the beach, at the margin of the sand and the land vegetation. The numbers of stations refer to meters away from station Q. High tide falls up to station 12, low tide receeds below station 24. The letters "A" and "B" represent collecting sites at each station and are approximately two meters apart. Figure 2- A schematic diagram of a collembolid trap used for field studies, In the top view, a 3 centimeter square of graph paper used for large count estimations is shown. The side view shows the 1.2 centimeter lip of the jar lid. Figure 3- A graph of the number of collembolids caught in a collembolid trap in a two minute interval as a function of time of day. The number of animals caught at two different stations, 15 and 18, is represented by either a filled-in circle or an outlined square. The tidal cycle is represented by the hashed line. Collections were taken on April 27 and April 28. Figure 4- A graph of the number of collembolids caught in a collembolid trap in a two minute interval as a function of time of day (May 1 to May 2 ). The number of animals caught at two different stations, 12 and 13, is represented by either a filled-in circle or an outlined square. The tidal cycle is represented by the hashed line. Figure 5- A graph of the number of collembolids caught in a collembolid trap in a two minute interval as a fnction of time of day (May 9 ). The number of animals caught at two different stations, 15 and 18, is represented by either a filled-in circle of an outlined square. The tidal cycle is represented by the hashed line. Figure 6- A bar histogram of the distribution of animals along the beach at the peak collection interval. The y axis represents numbers of collembolids caught in a two minute interval and The darkened bars thex axis is station along the beach. represent collections taken at the A position at the stations, the outlined bars represent the collections taken at the B pos-ition at the stations. Figure 7 A bar histogram of the distribution of animals along the beach at the peak collection interval. The y axis represents: numbers of collembolids caught in a two minute interval and thex axis is station position along the beach. The darkened bars represent the collections taken at the A position at the stations, the outlined bars represent the collections taken at the B position at the stations. Figure 8- A bar histogram of the distribution of animals along the beach at the peak collection interval. The y axis represents numbers of collembolids caught in a two minute interval and thex axis is station position along the beach. The darkened bars represent the collections taken at the A position at the stations, the outlined bars represent the collections taken at the B postion at the stations. Figure 9- A schematic diagram of a container with an infared diode used to contain collembolids during the course of the laboratory experiments. Figure 10-Two bar histograms of the amount of activity in 2 collembolid container as a function of time of day while subjected to a heated water bath regime. The hashed lines and the darkened bar represent the time interval that the container spent in the temperature regime. The temperaturee in or near the container at specific times appears across the top of the graph. The range of temperatures in the different conditions appears above as well. Figure 11-A bar histogram of the amount of activity in a collembolid container as a function of time of day in the control container. The hashed lines and the darkened bar represent the time interval that the container spent in the insulated: regime. The temperature in or near the container at specific times appears across the top of the graph. The range of temperatures in the different conditions appears on top of the graph as well. Figure 12-A bar histogram of the amount of activity in 2 collembolid container as a function of time of day while subject to a 12.3 degree centigrade Formatemp bath. The hashed lines and the darkened bar represent the time interval that the containers spent in the Formatemp regime. The temperature in or near the container at specific times appear across the top of the graph. The range of tem peratures in the different conditions appears on top of the graph as well. Figure 13-A bar histogram of the amount of activity in a collembolid container as a function fo time of day while subject to a regime of cold temperatures in a "blue ice" box. The hashed lines and the darkened bar represent the time intervals spen in the different cold regimes. The temperature in or near the container at specific times appears across the top of the graph. The range of temperatures in the different condtions appears on top of the graph as well. Figure 14-A summary of the response of collembolid activity to different temperature regimes. The arrow represents movement from one regime to another. Response of peak A changes and changes in the A/B ratio as a result of different temperature conditions is noted. The dashed line separates the contáiner kept in the blue ice box from the other containers since it went through several temperature changes. The dates referring to the different regimes for the blue ice box container are above the arrows. The temperature ranges in all the regimes are noted below the name of the regime. Figure 15- A graph of the response of collembolid activity to changes in temperatures. On the y axis the percent change in activity from activity shown under ambient laboratory temperatures is shown. On thex axis the temperature of the regime to which the collembolids were moved from the ambient laboratory conditions is shown. The hashed line is drawn to represent the temperature response curve of collembolids to the different regimes. c High Tide Low Tid FIGURE 1 FIELD MATRIX Station O Station 3 Station 6 Station 9 Station 12 Station 15 Station 18 Station 21 Station 24 ----— ————— —----— — — : . â 1.2 -3 cm Top View —— - — —- ——— — 1 - ----— --. — ..-. -----.— — Side View FICURE 2 COLLEMBOLID TRAP -75 Number of Collembolids — 8 D 0 3 6 Number of Collembolids — 9 9 5S De 1+ 3 Number of Collembolids D 0 200 8100 H 16 19 20 FIGURE 6 APRIL 27-28 (7:45 A.M. ) A B L H 22 26 FIGURE 7 MAY 1-2 ( 3:40 A.M.) A B □ A Station 1000 3 500 I 12 H 14 FIGURE 8 MAY 9 (7:30 A.M. ) H 4 A B L Epox 5.0 cm FIGURE 9 COLLENBOLID CONTAINER Plexiglass Rod 3.5 cm — Infared Diode N 1.5 cm + 5 - ata- Sand + aaa- . - Aetivrtyouns perne .. E F —L. — —. + 8 1. 1 : —— — .: —-— + — --- 3 — — ... —1— + — 8 E 1 I — — X 9 . — .— 1— . — — — X + —+ . — .... : X ... —— —— X : 2 + 1. 1 — — â — 2 . ... ——+ Activity (Counts per Hour E — —— — — — — — — — — — +—: — — — — — — — —— — —+ S . .. . . 2 .... . 1 09 —1 .... 1 — ... — — 1 —. — . ++ —f . X 1 . 1 —— — ....... —— . ... .. X 1—. — X X — — . 1 — 1 X ..... 1 — — — — — — — — — +1 — — — — - — — —:— — — — .. p — — — . 1 — . Activity (Counts per Hour) — —+ — .... —/— + 1 — .. — — — — — : L — L 8 1 — — — 0 : 1 :: + — / t E - — -. — — — — — — . — — — J + — + + X — 8 — 4 X — 1 + — — — ..: X ... 0 — L ... — 1 — — E e ... .. ... —1 X 8 ..... N F 1 ——. —:— — — | — — | — — — | — —— — —1— . : ... —— — —— —: — L ...:— 4 . Activity (Counts per Hour :.. . ... . ... ... — .. 7 + —: — â 1 —1 — — .... .. 3 1— 3 — :. —— — . 1 1 — X — X 15 + 1 ..: X ... —— — — — —1 ——— — ... — — 0 ... S — — | — / S 2 1 : — — 171 — — — 9 : — — . — — + — — — — — — —— — — — .. — — — —1— — 1 +0 — — —:— — ——— — — — — 1— 1 | 1 FIGURE 14 SUMMARY OF TEMPERATURE EFFECTS Ambient - — Heated Water Bath: cessation of activity, death 21-26 C 29-34 C — Ambient Insulated: consistent 21-26 C 23-26 C Ambient — Formatemp Bath: Peak A 7 A/B 21-26 C 14 C —--------- (May 24-25) Ambient — Blue Ice Box: Peak'AV A/B T 21-26 C 9-16 C (May 25-26) Blue Ice — Ambient: Peak A A/BJ 9-16 C 23-26 C (May 26-28) Ambient — Blue Ice Box: Peak Shift 23-26 C 0-2 C O. 2 8 /0 Percent Change in Activity (From May. 24 to May 25 )