xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xlink http:www.w3.org1999xlink xmlns:xsi http:www.w3.org2001XMLSchema-instance
leader 00000nas 2200000Ka 4500
controlfield tag 008 000000c19749999pautr p s 0 0eng d
datafield ind1 8 ind2 024
subfield code a M39-00405
Murphy, James, T.
Respuestas a los cambios en la ubicacin de la fuente de nctar por Xylophanes acrus (Sphyngidae) como una medida de la memoria espacial.
Response to changes in location of nectar source by Xylophanes acrus (Sphyngidae) as a measure of spatial memory
Moths in the family Sphyngidae are known to expend a considerable amount of energy in locating food resources. Considering the high investment of energy required to locate food resources within large areas, a valuable adaptation to aid in more efficiently locating resources might be the ability to remember resource locations in order to relocate them when necessary. The purpose of this study was to determine how Xylophanes acrus would respond to sensory cues and if these responses indicated an ability to develop a spatial memory for locations of nectar sources. Xylophanes acrus responses to artificial nectar sources were elicited using a natural flower attractant with an artificial nectar source attached. The presence of spatial memory was tested by manipulating the location of an artificial nectar source and recording visitation patterns of X. acrus within enclosures. The results indicated that X. acrus did show an ability to remember the location of nectar sources, as a significantly greater number of X. acrus individuals visited locations where nectar was located previously compared to new nectar sources in adjacent locations (P =
0.012). It would be highly adaptive if X. acrus could use this ability in a natural setting to more efficiently locate resources over large areas, thereby conserving energy that would otherwise be spent if relying solely on visual or olfactory sensory stimuli detection to relocate resources.
Las polillas en la familia Sphyngidae son conocidas por gastar una cantidad considerable de energa para localizar los recursos alimenticios. Si se considera la alta inversin de energa requerida para encontrar a los recursos alimenticios dentro de un rea grande, una adaptacin valiosa que ayuda en encontrar a los recursos ms eficientemente podra ser la habilidad para recordar las ubicaciones del recurso con el fin de reubicarlas cuando es necesario. El propsito de este estudio fue determinar como Xylophanes acrus responde a las pistas sensoriales y si estas respuestas indican una habilidad para desarrollar una memoria espacial para ubicaciones de fuentes de nctar. Las respuestas de Xylophanes acrus en las fuentes de nctar artificiales fueron provocadas usando un atrayente floral natural con una fuente de nctar artificial. La existencia de una memoria espacial fue examinada manipulando la ubicacin de una fuente de nctar artificial y anotando los patrones de visitas de X. acrus dentro del encierro. Los resultados indican que X. acrus muestra una habilidad para recordar la ubicacin de fuentes de nctar ya que un nmero significativamente mayor de individuos de X. acrus visitaron las ubicaciones donde el nctar fue ubicado anteriormente en comparacin con las fuentes de nctar en lugares adyacentes (p = 0.012). Sera altamente adaptativo si X. acrus pudiera usar esta habilidad en un ambiente natural para localizar a los recursos ms eficientemente en reas ms grandes, por lo tanto conservando la energa que podra de otra forma ser gastada si se basa solamente en la deteccin de estmulos sensoriales visuales u olfatorios para reubicar los recursos.
Text in English.
Monteverde Biological Station (Costa Rica)
Alimentos de animales
Estacin Biolgica de Monteverde (Costa Rica)
Tropical Ecology Summer 2004
Ecologa Tropical Verano 2004
t Monteverde Institute : Tropical Ecology
1 Response to changes in location of nectar source by Xylophanes acrus (Sphyngidae) as a measure James T. Murphy Department of Biochemistry, University of Wisconsin Madison ABSTRACT Moths in the family Sphyngidae are known to expend a considerable amount of energy in locating food resources. Considering the high investment of energy required to locate food resources within large areas, resource locations in order to relocate them when necessary. The purpose of this study was to determine how Xylophanes acrus would respond to sensory cues and if these responses indicated an ability to develop sources. Xylophanes acrus responses to artificial nectar sources were elicited using a natural flower attractant with an artificial nectar source attached. The presence of ource and recording visitation patterns of X. acrus within enclosures. The results indicated that X. acrus did show an ability to X. acrus individuals visited locations where n ectar was located previously compared to new nectar sources in adjacent locations (P = 0.012). It would be highly adaptive if X. acrus could use this ability in a natural setting to more efficiently locate resources over large areas, thereby conserving en ergy that would otherwise be spent if relying solely on visual or olfactory sensory stimuli detection to relocate resources. RESUMEN Las polillas en la familia Sphyngidae son conocidas por gastar una cantidad considerable de energa para localizar recurs os alimenticios. Si se considera la alta inversin de energa requerida para encontrar recursos alimenticios dentro de un rea grande, una adaptacin valiosa que ayuda en encontrar resursos ms icaciones del recurso con el fin de reubicarlas cuando es necessario. El propsito de este estudio fue determinar como Xylophanes acrus responde a ubicaciones de fuentes de nctar. Las respuestas de Xylophanes acrus a fuentes de nctar artificiales fueron provocadas usando un atrayente floral natural con una fuente de nctar artificial. La existencia de una ndo la ubicacin de una fuente de nctar artificial y anotando los patrones de visitas de X. acrus dentro del encierro. Los resultados indican que X. acrus muestra una ivamente mayor de individuos de X. acrus visitaron ubicaciones donde el nctar fue ubicado anteriormente en comparacin con fuentes de nctar en lugares adyacentes (p = 0.012). Sera altamente adaptativo si X. acrus pudiera usar esta habilidad en un ambie nte natural para localizar recursos ms eficientemente en reas ms grandes, por lo tanto conservando energa que podra de otra forma ser gastada si se basa solamente en la deteccin de estmulos sensoriales visuales u olfatorios para reubicar recursos. INTRODUCTION Many organisms are known to cover large ranges in search of resources (Janzen 1983). Moths in the family Sphyngidae are no exception, often traveling long distances in search of resources and possibly migrating in response to seasonal cycles of availability of these
2 resources (Murray et al. 2000). Sphyngid moths are sometimes referred to as energetically expensive. It may be advantageous for Sphyngid moths to g ather resources more efficiently by learning and remembering the locations of resources, rather than repeatedly relocating these resources by visual or olfactory cues. R.R. Baker, an English biologist who specializes in butterfly and moth navigation, hypot hesizes that the nocturnal moth uses extremely distant points of light as reference points when navigating from one location of interest to another (Janzen 1983). When a moth chooses an often brighter and nearer artificial light point, the moth flies towa rd the light until it arrives at the source. Then, finding itself immersed in artificial daylight, instinctively roosts (Janzen 1983). Janzen hypothesizes that if moths have a capacity for geographically oriented navigation using distant light points as guides, that this is interest to them (Janzen 1983). An alternative hypothesis is that moths simply respond to raw sensory cues when locating resources, essentially a repeated stimulus and response situation. There is some evidence that Sphyngids have the ability to learn the locations of nectar sources (Kelber 1996), however more research is needed. The purpose of this study was to determine how Xylophanes acrus res ponded to locations of nectar sources. It was hypothesized that X. acrus would be able to learn and xhibit a tendency to revisit these sources in subsequent trials using memory. This would provide evidence for an adaptation designed to minimize the amount of energy X. acrus uses in locating food resources. MATERIALS AND METHODS Study Site Research was conducted at the Monteverde Biological Research Station in Costa Rica from July 14 to August 1, 2004. Xylophanes acrus males and females were collected from the adjacent cloud forest using both ultraviolet and incandescent lighting at an elevation of 153 0 m. Enclosure Setup An indoor enclosure (dimensions: 2.0 x 1.5 x 0.8 m) was constructed using a suspended mosquito net secured to a solid tabletop with thumbtacks and duct tape. The enclosure volume allowed X. acrus to fly freely inside. Boundaries of regions, visual barriers, and sites were delineated with markers within the enclosure (Figure 1). Due to X. acrus response to light, especially in the ultraviolet to green spectrum, surroundings were cleared of ambient lights and observations were made u sing red light. Experiment 1: Response to Visual Cues Responses to visual cues were elicited by using a Guettarda poasana (Rubiaceae) flower or an artificial flower made of paper (Figure 2). In each trial, 3 6 X. acrus were placed within the starting ar ea and responses to Site A, B or C were recorded for 20 minutes. Trial 1 was a control to determine random movements or biases to any area of the
3 enclosure, with no visual cues at either site (Figure 2). In Trial 2, a G. poasana flower was placed inside a vial, which was then screwed tightly shut to eliminate olfactory cues, and placed at Site A (Figure 2). Xylophanes acrus were placed at Site C and responses were recorded (Figure 1). In Trial 3, the G. poasana flower was moved to Site B, and replaced b y the artificial flower in Site A (Figure 2). Experiment 2: Response to Olfactory Cues Responses to olfactory cues were elicited using crushed pieces of G. poasana flowers, which produce a strong sweet fragrance (Murray et al. 2000). The pieces were plac ed in a Petri dish and covered with a piece of cardboard to prevent X. acrus from responding to visual cues. Trials were identical to those in Experiment 1 with the exception of the use of olfactory cues instead of visual cues. In every trial, X. acrus w ere initially placed at Site C (Figure 1). Trial 1 was a control, with no cues at either site to test for specimen bias or random responses (Figure 4). In Trial 2, the olfactory cue was placed at Site A and responses were recorded (Figure 4). Five minut es were used to fan out the enclosure before proceeding with Trial 3. In Trial 3, the olfactory cue was relocated to Site B (Figures 4). Experiment 3: Response to Nectar Source Location Responses were elicited with an artificial nectar source created usi ng a G. poasana branch and flower. The artificial nectar sources were made using 2 3 cm straw segments (diameter = 0.8 cm) sealed at one end with a lighter and filled with a sucrose solution of about 28 33% created by mixing sugar and water and using a Re ichert hand refractometer to check concentration (Josens and Farina 2001). The sucrose solution was loaded into the straw segments using an eyedropper. The corolla of a G. poasana flower was removed and inserted into the straw segment containing the sucr ose solution. The straw was then taped onto the G. poasana inflorescence (Figure 6). Other blossoming flowers were removed to eliminate competition with the artificial nectar source, in order to ensure that an equal reward was available for any X. acrus that responded. Visual barriers were put in place to block X. acrus direct line of sight from Site C initially (Figure 1). Trials were identical to previous ones with the exception of two introduced visual barriers and the nectar source apparatus as the s ensory cue. After Trials 2 5, 5 minutes was used to set up the next trial and fan away residual olfactory stimuli. Trial 1 was a control to determine random responses and biases (Figure 7). Trials 2 and 3 were set up identically, each with the nectar so urce placed at Site A (Figure 7). In Trials 4 and 5, the nectar source was moved to Site B (Figure 7). In Trial 6, the nectar source was returned to Site A (Figure 7). Xylophanes acrus responses were recorded as discreet choices within the enclosure: Re gion A, B or C (Figure 1) and compared according to mean number of responses to the region containing the nectar source versus the regions lacking the nectar source (Figure 8), and graphed according to number of responses to each region in each trial regar dless of nectar source location (Figure 9). Analysis of Data ANOVA analyses were used to determine if there were significant effects of location on response (Ambrose and Ambrose 1995). A 2 x 2 contingency table was used to test for significance in the co ntrol trial (Trial 1) in Experiment 3 compared to responses in Trial
4 2. Flow charts outlining all chosen response sequences were created and Chi Square analysis was used to determine if the responses were significant. Flow charts outlining all chosen res ponse sequences were created and Chi Square analyses were used to determine if the results were significant (Zar 1984). Flow Chart 1 organized sequences based on response either to the region containing nectar source or the regions lacking nectar source ( Figure 10). Flow Chat 2 organized response sequences by three possible responses: taking nectar from the source, responding to the region containing the nectar source, or responding to the regions lacking it (Figure 11). RESULTS A total of 43 Xylophane s acrus individuals were collected and used in this study. In Experiment 1, mean value of number of responses to Site C (home) was far greater (mean = 35.7) than the mean for responses to both Site A (mean = 6.3) and Site B (mean = 2.7). A significant ef fect of response was found in all three trials using ANOVA (Trial 1, F = 181.208, P = <0.0001; Trial 2, F = 75.675, P = <0.0001; Trial 3, F = 47.624, P = <0.0001). Significant differences were found between responses to Sites A, B and C D (Figure 3). In all three trials, X. acrus individuals preferred Site C to both Sites A and B (Trials 1 3, C > A, P = <0.0001, C > B, P = <0.0001). In Trial 2, X. acrus showed preference for Site A over Site B, but a significant preference for Site B in Trial 3 was not found (Trial 2, A > B, P = 0.0083). The number of responses to Sites A, B and C were summed and graphed for each trial (Figure 3). In Experiment 2, the mean number of responses for all trials to Site C (mean = 27.3) was greater than the m ean responses to Site A (mean = 8.0) and Site B (mean = 7.7). A significant effect of response was found in all three trials using ANOVA (Trial 1, F = 105.882, P = <0.0001; Trial 2, F = 21.779, P = <0.0001; Trial 3, F = 5.846, P = 0.0037). In Experiment 2, significant differences were found between responses to Sites X. acrus individuals showed preference for Site C responses (C > A, P = <0.0001, C > B, P = <0.0001). In Trial 2, X. acrus individuals preferred Site C to both Site A and Site B (C > A, P = 0.0098, C > B, P = <0.0001). A significant preference for Site A, which contained the olfactory cue, over Site B, lacking the olfactory cue, was also observed in Trial 2 (A > B, P = 0.0001). In Tria l 3, X. acrus individuals preferred Site B to Site A, but no preference was found between Site B and Site C (C > A, P = 0.0026, B > A, P = 0.0053). Data on number of responses to each site was summed and graphed for each trial (Figure 5). In Experiment 3 a significant difference was observed between possible responses, showing that the X. acrus did exhibit a non random response to the nectar source ( 2 = 28.6, P = <0.0001). The mean for Site A was greatest (mean = 18.0), followed by Site C (mean = 15.3) followed by Site B (mean = 9.7). A significant response effect was found in Trials 1 5, however, not in Trial 6, using ANOVA (Trial 1, F = 410.524, P = <0.0001; Trial 2, F = 17.558, P = <0.0001; Trial 3, F = 29.450, P = <0.0001; Trial 4, F = 3.261, P = 0.0416; Trial 5, F = 8.906, P = 0.0002). Significant differences in number of responses between Sites A, B and C within trials were found (C > A and C > B, P = <0. 0001). In Trial 2, Site A was preferred over Site C and Site B, and Site C was preferred over Site B (A > C, P = 0.0230, C > B, P = 0.0005, A > B, P =
5 <0.0001). In Trial 3, Site A was again preferred to Site C and Site B, however, there was no appreciabl e preference between Site C and Site B (A > C and A > B, P = <0.0001). In Trial 4, Site A was preferred to Site C despite the change in nectar source location to Site B (A > C, P = 0.0120). In Trial 5, Site B was preferred to both Site A and Site C (B > C, P = <0.0001, B > A, P = 0.0044). In Trial 5, no significant preference was found between Site A and Site C. In Trial 6, Site A was preferred to Site C (A > C, P = 0.0226)(Figures 8, 9). All frequencies of responses in each trial was graphed with two interpretations: one comparing visits in each trial to the region containing the nectar source to visits to the region lacking the nectar source (non Region C (i.e. remaining within Region C)(Figure 8), and the other comparing numb er of visits to regions regardless of nectar source location (Figure 9). The summary of individual X. acrus responses in all six trials was organized into two flow charts (Figures 10, 11). In Flow Chart 1, two responses were possible in each frequencies were compared to expected numbers using a Chi Square analysis, which showed significant results overall ( 2 = 45.0, P = <0.0001). A second flow chart was created which organized the frequency of all possible sequences of responses considering frequencies were compared to expected, and a significant difference in response sequences was f ound ( 2 = 38.9, P = <0.0001). Sequences 1, 2 and 3 were most frequent in Figure 10, while Sequences 4, 5 and 6 were most frequent in Figure 11. DISCUSSION Experiments 1 and 2 were designed to test X. acrus response to two different components of sensor y input that are used in location of resources (Figures 2, 4). Responses to a genuine visual cue were determined to be statistically significant in Trial 2 of Experiment 1, which suggests that X. acrus did respond to visual cues and this should be conside red a factor in subsequent experiments. Experiment 2 tested olfactory cue responses using chemical attractants of G. poasana flowers (Figure 4). The results indicated that response to olfactory cues was more pronounced compared to visual cues. Significa nt responses to the olfactory cue occurred in Trials 2 and 3 of Experiment 2, which indicates the use of olfactory cues in locating resources (Figure 3). Experiment 3 measured X. acrus responses to artificial nectar sources that were manipulated to deter mine whether a learning trend existed in the responses to nectar learning nectar source location while it remained constant as indicated by the increase in number of respo nses to Site A between Trial 2 and Trial 3, and between Trials 5 and 6 (Figures 8, 9). The same trend existed in number of responses to Site B between Trials 4 and 5 (Figure 9). Further support for the hypothesis that X. acrus for nectar source locations was provided by the significant preference for Site A in Trial 4 even after the nectar source had been removed and put in Site B (Figures 7, 8). Even if olfactory cues remained in Site A during Trial 4, the stronger cue still s hould have been coming from Site B. Sequence 1 of Figure 10 was the sequence of responses that corresponded with the significant number of responses to Site A in Trial 4. This sequence suggested that there was a preference for Site A by those X. acrus tha t
6 responded to Site A in Trials 2 and 3. More evidence for this hypothesis came in the significance of Sequence 2 in Figure 11. These X. acrus took nectar from the source which would assure the association of a reward with this site. Another trend that suggested learning of nectar source location was provided by the decreased number of responses to Site A from Trials 4 to 5 (Figure 9). Therefore, the significant differences in sequences of responses and responses to sites within trials provided possible evidence for X. acrus because it would be an adaptive trait that could increase the efficiency with which X. acrus locate resources in the natural environment. Possessing a m emory of resource locations would lower the relative energetic cost of locating resources, especially considering the high energy mode of flight they exhibit and pattern of traveling long distances for food resources. Many factors must be considered in ev aluating the results of this study. The results seemed to be affected by lunar cycles to some degree; more X. acrus individuals were collected with lights when moonlight was at a minimum. Also, X. acrus activity and cooperation with experimental methods also seemed to be negatively correlated with ambient moonlight. There was a strong tendency for many X. acrus individuals to exhibit random responses. Also, there were significant numbers of individuals that did not respond to any sensory cues in any exp eriments. More recently gathered G. poasana flowers seemed to elicit more responses from X. acrus It was observed that X. acrus collected the same night that they were experimented upon seemed to exhibit no significant differences in responses from indi viduals that had been inside the enclosure for 24 hours or more. Finally, the location of the enclosure must be considered because of the effects starlight and moonlight seemed to have on X. acrus behavior. The enclosure was located less than 2 m from th e side of the lower lab that is almost entirely covered with windows. Light sources visible to X. acrus through these windows could have affected their behavior during trials. There are many opportunities for further research with X. acrus Beneficial fu ture research could explore how X. acrus respond to genuine olfactory cues compared to artificial olfactory cues. Another study could involve testing differences in the relative influence of visual and olfactory components on resource location between sma ller and larger volume enclosures. It is reasonable to question whether X. acrus rely more on one sensory component to detect more remote resource locations and more on another in closer proximity situations. Determining the rate of learning food source locations using visual compared to olfactory cues would also be significant in studying how X. acrus locate food resources. This study lent support to the hypothesis that X. acrus have an er studies might last depending upon the amount or quality of reward offered. Also noteworthy was the significant difference found in the effectiveness of attraction of X. acrus between the outdoor, incandescent lighting and florescent lighting of the lower lab compared to the ultraviolet light. A study designed to determine if different wavelengths of light elicit different responses and behaviors would also be interestin g. ACKNOWLEDGEMENTS
7 Much thanks to Carlos Guindon for his support and guidance throughout my study! Thanks to Karen Masters for her inspiration with statistical analysis methods you definitely made the final product more meaningful. Also, thanks to B ill Haber for consultation. Thanks to La Estacin Biologica de Monteverde for use of the equipment and facilities and for the continuing efforts to protect the environment. Thanks to the teaching assistants for tolerating me, and also for helping with pa per revisions and guidance. Thanks to all other students in my program for random support and help and comic relief. Finally, thanks to my parents, because without their support I would not even be here. LITERATURE CITED Ambrose, W. H., Ambrose, K. P 1995. A Handbook of Biological Investigation, 5 th Ed. Hunter Textbooks, Inc. Knoxville, Tennessee. Janzen, D. H. 1983. Why do moths come to light? D. H. Janzen (Ed). Costa Rican Natural History, p. 620. University of Chicago Press, Chicago, Il linois. Josens, R. B. and W. M. Farina. 2001. Nectar feeding by the hovering hawk moth Macroglossum stellatarum : intake rate as a function of viscosity and concentration of sucrose solutions. Journal of Comparative Physiology A Sensory, Neural and B ehavioral Physiology 187(8): 661 665. Kelber, A. 1996. Colour learning in the hawkmoth Macroglossum stellatarum. Journal of Experimental Biology 199 (5): 1127 1131. Murray, K. G., S. Kinsman, J. L. Bronstein. 2000. Hawkmoths (Sphingidae). N. M. Nad karni and N. T. Wheelwright (Ed). Monteverde, p. 248. Oxford University Press, New York, New York. Zar, J. H. 1984. Biostatistical Analysis, 2 nd Ed. Prentice Hall, Inc. Englewood Cliffs, New Jersey.
8 ( a ) ( b ) 2.0 m Site A 1.5 m Visual barrier 0.3 m Starting area (~0.36 m 2 ) = Site C Figure 1. Enclosure Setup: (a) Image of enclosure with Site C outlined in a dashed line. Xylophanes acrus individuals were placed here at the start of each trial, eliminating a direct line of sight to Sites A and B. (b) Footprint of the enclosure with all components and response regions considered in the data. Cues or nectar source were placed at either Site A or B. Regions A, B and C were the three possible responses in all experiments. Region A Region C Region B Site B Lines of sight 1.0 m
9 ( a ) ( b ) Trial 1 Trial 2 Trial 3 Artificial flower Real flower Figure 2. Experiment 1: (a) Images of the artificial and actual G. poasana flowers used as visual cues. Top image is the artificial flower (paper) inside the glass vials used to eliminate olfactory cues. Images underneath compare real and artificial flowers directly. (b) Setup of each trial showing visual cue locations.
10 Figure 3. Experiment 1: Responses of X. acrus to re al and artificial floral cues showing number of visits to each region for every trial. Note the majority of X. acrus did not respond to visual cues (i.e. stayed at Site C). There was only a small response to the real flower (Trial 2, Site A) and an insig nificant response to the artificial flower (Trial 3, Site A)(N = 43).
11 Trial 1 Trial 2 Trial 3 Olfactory cue (crushed G. poasana flower in Petri dish covered with cardboard) Figure 4. Experiment 2: setup and loca tions of olfactory cues in each trial. Trial 1 was the control: no olfactory cues. X. acrus were placed at Site C at the beginning of every trial (Figure 1). The vertical line between Trial 2 and 3 represents the 5 minute break used to fan away residual olfactory cues from previous trial. Figure 5. Experiment 2: number of responses to Regions A, B and C within the enclosure during each trial. Note the significant preference for Region A, containing the olfactory cue, over Regio n B in Trial 2, and preference for Region B in Trial 3 when then the cue was moved to Site B (N = 43).
12 Figure 6. Experiment 3: left, nectar source apparatus, right, nectar source positioned at Site B behind a visual barrier. Nectar source Trial 1 Trial 2 Trial 3 Trial 4 Trial 5 Trial 6 Figure 7. Experiment 3: setup for each trial. Trial 1 was the control to determine random responses and sample biases. There was a 5 minute break after Trials 2 5 to remove residual olfactory cues from previous trials. Xylophanes acrus were placed at Site C at the beginning of every trial, with the nectar source out of sight behind a visual barrier. Visual barriers remai ned in place in Regions A and B throughout Experiment 3 to control for any affect the barriers may have had on X. acrus preference.
13 Figure 8. Experiment 3: graph of each trial showing number of responses to Regions A, B or C. Black denotes responses to the region containing the nectar source, while grey denotes responses to the adjacent region lacking the nectar source (non Site C) White denotes responses to Region C. Note that the nectar source location was switched from Site A to B in Trial 4, and switched back to Site A in Trial 6. Note two trends: (1) the increase in Region A responses from Trial 2 to 3, also the increase i n Region B responses from Trial 4 to 5, (2) the high number of Region A responses directly after nectar source location switch in Trial 4 (N = 43 for all trials). Region containing nectar source Region lacking nectar source (non C) Region C
14 Figure 9. Experiment 3: number of responses in each trial based on region. Note the trend illustrated by the black columns of increasing frequency of responses to the region containing the nectar source while location remained constant from Trial 2 to 3 (Site A), and the same trend from Trial 4 to 5 (Site B)(as shown in fig. 7). The white columns represent responses to Region C, which were X. acrus individuals that remained relatively inactive.
15 Seq. 1 Seq. 2 Seq. 3 Figure 10. Experiment 3: Flow Chart 1, showing all X. acrus response sequences according to two possible responses in each trial: (1) response to region containing the nectar source regardless of whether or not necta r was taken, (2) response to region other than the one containing the nectar source (Figure 1, 7). To condense the chart, response sequences that were not taken were not represented in the chart, however these sequences were considered in statistical anal yses. Note: Especially significant sequences shaded in grey (black shaded box showed unusually small difference in number of responses between Trial 2 and 3).
16 Seq. 4 Seq. 5 Seq. 6 Seq. 4 Seq. 5 Seq. 6 Figure 11. Experiment 3: Flow Chart 2 tracks all X. acrus response sequences according to three possible responses in each trial: (1) response to region containing the nectar source and extension of prob oscis and taking nectar (FIR), (2) response to region containing the nectar source but not taking nectar, (3) response to region other than the one containing the nectar source (Figure 1, 7). To condense the chart, response sequences that were not taken w ere not represented in the chart, however these sequences were considered in statistical analyses. Note the sequences highlighted in grey (Sequences 4, 5 and 6). The black box highlights a particularly strong sequence of responses between Trial 3 NFI res ponses and Trial 4 NFI responses. Sequence 6 provided the best evidence that X. acrus use memory to locate resources.