- Email: [email protected]
The influence of wing area on the courtship behaviour of Drosophila melanogaster
THE INFLUENCE OF WING AREA ON THE COURTSHIP BEHAVIOUR OF Drosophila melanogaster BY ARTHUR W . EWING Department of Zoology, University of Edinburgh .
within similar limits in the courtship of flies from the same stocks . Also, what evidence there is from studies on Drosophila flight shows that increase in wing area results in decreased amplitude and frequency of wing beat (Reed, Williams & Chadwick, 1942) . If this were applicable to courtship vibration one would expect that the increased stimulation provided by males with large wings would to some extent be counteracted by decreased amplitude and frequency of the wing beat . Thus wing area and the amplitude and frequency of wing beat may be connected in such a way as to minimise rather than to exaggerate any effect of changed wing area on courtship stimulation . The fourth variable, the proportion of vibration, can be measured in individual courtships by the method described by Bastock & Manning (1955) . This factor has been shown to affect courtship success and Bastock (1956) demonstrated that males carrying the mutant gene "yellow" are sexually less successful than the wild type and that this is attributable to the "yellow" males' courtships containing shorter and more widely spaced bouts of vibration than that of the wild type . This paper describes the effects on courtship success of changing wing area by three methods .
Introduction During the courtship behaviour of Drosophila inelanogaster the male performs a movement named "vibration" in which he extends one wing, usually the one nearest to the female, and vibrates it in the vertical plane (Bastock & Manning, 1955) . It is probable that this movement plays a major role in sexually stimulating the female . Sturtevant (1915), for example, has shown that the mating success of wingless males is much less than that of normal ones . As several minutes of vibration are usually required before the female will copulate one assumes that vibration does not merely have a "trigger" function but acts additively in facilitating copulation . The nature of the stimulation provided by the wing display is not known with certainty but there is evidence that it is largely due to some mechanical effect of the wing vibration itself and not to visual stimuli (Bastock, 1956), or to scent wafted by the wings (Petit, 1958 ; Ewing & Manning, 1963) . One factor which might therefore affect the amount of stimulation produced is wing area and there is some indirect evidence that this is indeed so (Ewing, 1961) . Thus, by artificially changing the wing areas of the males by varying amounts and testing the effect of this procedure on total courtship time, it might be possible to assess the over-all importance of vibration and to see whether the stimuli from vibration do act additively in sexually arousing the female . Factors other than wing area which would affect the quantity of stimulation provided by vibration are as follows (a) The distance between the male and female during vibration . (b) The amplitude of wing beat . (c) The frequency of wing beat . (d) The proportion of courtship that consists of vibration . The first three factors are not easily measured . Males usually follow as closely behind the females as possible during courtship and therefore the distance between the sexes will vary
Rearing Temperature All the flies used in these experiments were derived from an outbred stock, Pacific, and were reared and experimented on under controlled conditions, the details of which have been given in a previous paper (Ewing, 1961) . Alpatov (1930) has demonstrated that the temperature during the larval and pupal life of nielanogaster affects the phenotypic expression of wing size in the adult . Thus flies reared at a low temperature have proportionately larger wings than those reared at higher temperatures . Culture bottles of stock flies were therefore set up at 18-+ l ±C and at 26+1 ±C and twenty days apart to ensure simultaneous hatching at the two temperatures . The flies were all tested at 26 ± C . 316
EWING : COURTSHIP BEHAVIOUR OF Drosophila melanogaster
In order to measure the increase in wing area due to rearing flies at 18 °C one wing was removed from ten males in each sample . These were mounted on microscope slides and with the aid of a camera lucida drawn on squared paper . The calculated areas of the wings were as follows : Males reared at 18 ° C 2300±25 units . Males reared at 26°C 1787+24 units . The unit of measurement was an arbitrary one and the mean increase in wing area of flies reared at 18°C over those reared at 26 ° C is 28 . 7 per cent . Males from the two samples were tested by means of the mass mating technique described by Manning (1961) . In this fifty pairs of virgin flies are introduced into a 250 ml . conical flask and the pairs removed with an aspirator when they copulate . These copulations are scored and a cumulative graph of total number of copulations against time can be drawn . This test was carried out twice using fifty males from each sample on each occasion . The females used were all reared at 26 ° C . The averaged results of these tests are illustrated in Fig . 1 . This shows
3l7
of vibration in each courtship calculated . The large winged males spend 28 . 5+9 . 7 per cent . of the courtship time vibrating and the males with the smaller wings, 31 . 4+6 . 6 per cent . Selection Two lines of flies which he had selected for high and low ratio of wing lengths to thorax length respectively, were kindly provided by Dr. F . W. Robertson of the Institute of Animal Genetics, Edinburgh . The results of this selection were that the high ratio line (HR) showed a mean increase of 12 per cent . in wing area and the low ratio line (LR) a decrease of 16 per cent . over controls while body size remained at approximately the same level as in the unselected controls (Robertson, 1962) . Three sets of mass matings were set up using fifty males from each line and from the controls each time . Control females were used for all the tests in case any change in female receptivity had occurred in the selected lines . Fig. 2 shows the averaged results of these experiments and it is obvious that the high ratio males have a higher mating speed than the controls and the low ratio males a lower one . Samples of single pair matings were also examined from each line and control . The mean 50
40 0 U 0
& 30 0 u
0 a. N E C
20
c c 10 time in minutes, from beginning of test .
Fig . 1 . Mass mating graphs of males reared at 18'C compared with controls reared at 26°C . The figures are derived from two separate tests, fifty males from each class being used on each occasion .
that the males reared at 18°C are more successful in obtaining mates than those reared at 26 ° C . Single pair courtships were also examined using ten males from each sample and the percentage
time in minutes,from start of test .
Fig . 2 . Mass mating graphs of males selected for high and low ratios of wing length to body size respectively compared with controls . The figures are derived from three separate tests .
318
ANIMAL BEHAVIOUR, XII, 2-3
percentage time spent in vibration in each sample was : HR males-29 , LR males-39 and controls-26 per cent. The low ratio males exhibited significantly more vibration than the high ratio males or control males (p<0 .001) and a possible explanation for this is given in the Discussion . One might expect that this increase in percentage vibration would compensate for the smaller wings of the LR males, but these two factors are not the only ones concerned with sexual success . During the course of selection other, undetected changes may have taken place and these might have had a deleterious effect on courtship. Amputation Amputation of portions of the wings is simply done using a scalpel while the flies are etherized . The point where radius (2+3) meets radius 1 and the medial cross vein were used as markers thus making it possible to remove, reasonably accurately, standard portions of the wings . Four classes of flies were examined ; controls, in which the wings were left intact, and three others with the wings cut at A, B, and C as illustrated in Fig . 3 . Flies with
q
B
C
Fig . 3 . Wing of Drosophila melanogaster showing the three positions at which amputations were carried out . their wings amputated at A were not entirely wingless and it is probable that some sexual stimulation was provided by the wing stumps . It is not, however, easy to remove the wings entirely without damaging the wing muscles in the thorax . Flies treated in this manner seem to court normally . Mayr (1950) and Manning (1959) have demonstrated that removal of the antennae and fore tarsi respectively do not aversely affect the courtship performance of the males and these opera ions are likely to be just as serious as amputation of portions of the wings . Camera lucida drawings of the wings of fifteen males from each of the classes were made and
their areas and standard deviation calculated as previously described . The area of stump left in the class A flies was estimated at 10 per cent . The percentage wing area remaining in the two classes with partially amputated wings was as follows, controls being taken to represent 100 per cent . Controls 100+0 . 63 per cent . "C" males 80 . 85+0 . 94 per cent . "B" males 41 . 57+4 . 79 per cent . Two separate sets of mass matings were set up using fifty males from each class on each occasion . Normal winged control females were used throughout . The averaged results of these mass matings are illustrated in Fig . 4 from which it can be seen that mating success increased with wing area . Fig . 5 shows more clearly the relationships between these two factors . It is a graph of the number of copulations that had occurred during the first sixteen minutes of each test plotted against wing area for the four classes of male . The line joining the points is almost straight and suggests very strongly that the stimulation provided by vibration is additive in its effect on the female . As calculated from the above graph the effect of entirely removing the wings would be to reduce mating speed by just over eighty per cent .,
4 6 8 10 12 14 time in minutes, from start of test
16
18
Fig . 4 . Mass mating graphs of males with their wings amputated at the three positions illustrated in Fig . 3 compared with controls . The figures are derived from two separate tests .
EWING : COURTSHIP BEHAVIOUR OF Drosophila melanogaster so
s
s'
0
20
40
60
80
100
percentage wing area remaining
Fig 5 . A graph, derived from Fig . 4, of wing area plotted against number of copulations at 16 minutes . The open circle denotes that the wing area at this point was estimated . suggesting that this percentage of the total stimulation provided by the males is via vibration . Discussion In a previous paper (Ewing, 1961), in which I described the courtship behaviour of lines of flies which had been selected for large and small body size, I showed that large males exhibited less vibration than controls but nevertheless were equally successful in obtaining mates . I suggested that this deficiency in the amount of vibration was being compensated for by the increased wing area of the large selected males . The small selected males exhibited more vibration than controls but in spite of this were sexually less successful . I provided some evidence to show that the increased vibration of small males had been selected to compensate for some unspecified deficiency in their courtship which was the consequence of selection for body size . This deficiency could, in part, be attributed to decrease in wing area . A similar argument to the one given above could account for the increased vibration found in the LR males where there may have been secondary selection pressure for this character resulting from the decreased wing area . The results described in this paper demonstrate that wing area is indeed an important factor in courtship success .
319
Summary 1 . In the male of Drosophila melanogaster the component of courtship behaviour which plays the major role in sexually stimulating the female is wing vibration . 2. Wing area was changed by three methods : (a) By lowering the temperature during preimaginal development . (b) By selective breeding. (c) By amputation . 3 . The sexual success of males with different wing areas was measured and it was found that males with large wings were more successful in obtaining mates than those with small or partially amputated wings . 4 . The relationship between wing area and sexual success appeared to be linear thus demonstrating the additive nature of the stimulation provided via vibration . 5 . An extrapolation from these results suggested that approximately 80 per cent . of the sexual stimulation is normally provided by wing vibration . 6 . The relevance of these results to some previously published work is discussed . Acknowledgements My thanks are due to Dr. A. Manning for his help throughout the course of this work and for critically reading the manuscript . I am also indebted to Dr . F. W . Robertson who allowed me to use some of his selected lines .
REFERENCES Alpatov, W . W . (1930) . Phenotypic variation in body and cell size of Drosophila melanogaster. Biol. Bull ., Wood's Hole, 58, 85-103 .
Bastock, M . (1956) . A gene mutation which changes a behaviour pattern . Evolution, 10, 421-439 . Bastock, M . & Manning, A . (1955) . The courtship of Drosophila melanogaster. Behaviour, 8, 85-111 .
Ewing, A . W . (1961) . Body size and courtship behaviour in Drosophila melanogaster. Anim . Behav ., 9, 93-99 .
Ewing, A . W . & Manning, A . (1963) . The effects of exogenous scent on the mating of Drosophila inelanogaster . Anim. Behav ., 11, 596-598 .
Manning, A . (1959) . The sexual isolation between Drosophila melanogaster and Drosophila simulans. Anim. Behav., 7, 60-65 .
Manning, A . (1961) . The effects of artificial selection for mating speed in Drosophila tnelanogaster. Anim. Behav ., 9, 82-92 .
Mayr, E . (1950) . The role of the antennae in the mating behaviour of female Drosophila . Evolution, 4, 149-154 .
Petit, C. (1958) . Le determinisme genetique et psychophysiologique de la competition sexuelle chez Drosophila melanogaster . Bull. biol., 92, 248-329 .
320
ANIMAL BEHAVIOUR, XII, 2-3
Reed, S. C., Williams, C . M . & Chadwick, L. E . (1942) . Frequency of wing beat as a character for separating species, races and geographical varieties of Drosophila . Genetics, 27, 349-361 .
Robertson, F . W . (1962) . Changing the relative size of body parts of Drosophila by selection . Genet . Res., Camb. . 3,169-180
Sturtevant, A . H . (1915) . Experiments on sex recogntion and the problem of sexual selection in Drosophila • J. anim. Behav ., 5, 351-366 . (Accepted for publication 25th February, 1964 ; Ms. number : 417) .
within similar limits in the courtship of flies from the same stocks . Also, what evidence there is from studies on Drosophila flight shows that increase in wing area results in decreased amplitude and frequency of wing beat (Reed, Williams & Chadwick, 1942) . If this were applicable to courtship vibration one would expect that the increased stimulation provided by males with large wings would to some extent be counteracted by decreased amplitude and frequency of the wing beat . Thus wing area and the amplitude and frequency of wing beat may be connected in such a way as to minimise rather than to exaggerate any effect of changed wing area on courtship stimulation . The fourth variable, the proportion of vibration, can be measured in individual courtships by the method described by Bastock & Manning (1955) . This factor has been shown to affect courtship success and Bastock (1956) demonstrated that males carrying the mutant gene "yellow" are sexually less successful than the wild type and that this is attributable to the "yellow" males' courtships containing shorter and more widely spaced bouts of vibration than that of the wild type . This paper describes the effects on courtship success of changing wing area by three methods .
Introduction During the courtship behaviour of Drosophila inelanogaster the male performs a movement named "vibration" in which he extends one wing, usually the one nearest to the female, and vibrates it in the vertical plane (Bastock & Manning, 1955) . It is probable that this movement plays a major role in sexually stimulating the female . Sturtevant (1915), for example, has shown that the mating success of wingless males is much less than that of normal ones . As several minutes of vibration are usually required before the female will copulate one assumes that vibration does not merely have a "trigger" function but acts additively in facilitating copulation . The nature of the stimulation provided by the wing display is not known with certainty but there is evidence that it is largely due to some mechanical effect of the wing vibration itself and not to visual stimuli (Bastock, 1956), or to scent wafted by the wings (Petit, 1958 ; Ewing & Manning, 1963) . One factor which might therefore affect the amount of stimulation produced is wing area and there is some indirect evidence that this is indeed so (Ewing, 1961) . Thus, by artificially changing the wing areas of the males by varying amounts and testing the effect of this procedure on total courtship time, it might be possible to assess the over-all importance of vibration and to see whether the stimuli from vibration do act additively in sexually arousing the female . Factors other than wing area which would affect the quantity of stimulation provided by vibration are as follows (a) The distance between the male and female during vibration . (b) The amplitude of wing beat . (c) The frequency of wing beat . (d) The proportion of courtship that consists of vibration . The first three factors are not easily measured . Males usually follow as closely behind the females as possible during courtship and therefore the distance between the sexes will vary
Rearing Temperature All the flies used in these experiments were derived from an outbred stock, Pacific, and were reared and experimented on under controlled conditions, the details of which have been given in a previous paper (Ewing, 1961) . Alpatov (1930) has demonstrated that the temperature during the larval and pupal life of nielanogaster affects the phenotypic expression of wing size in the adult . Thus flies reared at a low temperature have proportionately larger wings than those reared at higher temperatures . Culture bottles of stock flies were therefore set up at 18-+ l ±C and at 26+1 ±C and twenty days apart to ensure simultaneous hatching at the two temperatures . The flies were all tested at 26 ± C . 316
EWING : COURTSHIP BEHAVIOUR OF Drosophila melanogaster
In order to measure the increase in wing area due to rearing flies at 18 °C one wing was removed from ten males in each sample . These were mounted on microscope slides and with the aid of a camera lucida drawn on squared paper . The calculated areas of the wings were as follows : Males reared at 18 ° C 2300±25 units . Males reared at 26°C 1787+24 units . The unit of measurement was an arbitrary one and the mean increase in wing area of flies reared at 18°C over those reared at 26 ° C is 28 . 7 per cent . Males from the two samples were tested by means of the mass mating technique described by Manning (1961) . In this fifty pairs of virgin flies are introduced into a 250 ml . conical flask and the pairs removed with an aspirator when they copulate . These copulations are scored and a cumulative graph of total number of copulations against time can be drawn . This test was carried out twice using fifty males from each sample on each occasion . The females used were all reared at 26 ° C . The averaged results of these tests are illustrated in Fig . 1 . This shows
3l7
of vibration in each courtship calculated . The large winged males spend 28 . 5+9 . 7 per cent . of the courtship time vibrating and the males with the smaller wings, 31 . 4+6 . 6 per cent . Selection Two lines of flies which he had selected for high and low ratio of wing lengths to thorax length respectively, were kindly provided by Dr. F . W. Robertson of the Institute of Animal Genetics, Edinburgh . The results of this selection were that the high ratio line (HR) showed a mean increase of 12 per cent . in wing area and the low ratio line (LR) a decrease of 16 per cent . over controls while body size remained at approximately the same level as in the unselected controls (Robertson, 1962) . Three sets of mass matings were set up using fifty males from each line and from the controls each time . Control females were used for all the tests in case any change in female receptivity had occurred in the selected lines . Fig. 2 shows the averaged results of these experiments and it is obvious that the high ratio males have a higher mating speed than the controls and the low ratio males a lower one . Samples of single pair matings were also examined from each line and control . The mean 50
40 0 U 0
& 30 0 u
0 a. N E C
20
c c 10 time in minutes, from beginning of test .
Fig . 1 . Mass mating graphs of males reared at 18'C compared with controls reared at 26°C . The figures are derived from two separate tests, fifty males from each class being used on each occasion .
that the males reared at 18°C are more successful in obtaining mates than those reared at 26 ° C . Single pair courtships were also examined using ten males from each sample and the percentage
time in minutes,from start of test .
Fig . 2 . Mass mating graphs of males selected for high and low ratios of wing length to body size respectively compared with controls . The figures are derived from three separate tests .
318
ANIMAL BEHAVIOUR, XII, 2-3
percentage time spent in vibration in each sample was : HR males-29 , LR males-39 and controls-26 per cent. The low ratio males exhibited significantly more vibration than the high ratio males or control males (p<0 .001) and a possible explanation for this is given in the Discussion . One might expect that this increase in percentage vibration would compensate for the smaller wings of the LR males, but these two factors are not the only ones concerned with sexual success . During the course of selection other, undetected changes may have taken place and these might have had a deleterious effect on courtship. Amputation Amputation of portions of the wings is simply done using a scalpel while the flies are etherized . The point where radius (2+3) meets radius 1 and the medial cross vein were used as markers thus making it possible to remove, reasonably accurately, standard portions of the wings . Four classes of flies were examined ; controls, in which the wings were left intact, and three others with the wings cut at A, B, and C as illustrated in Fig . 3 . Flies with
q
B
C
Fig . 3 . Wing of Drosophila melanogaster showing the three positions at which amputations were carried out . their wings amputated at A were not entirely wingless and it is probable that some sexual stimulation was provided by the wing stumps . It is not, however, easy to remove the wings entirely without damaging the wing muscles in the thorax . Flies treated in this manner seem to court normally . Mayr (1950) and Manning (1959) have demonstrated that removal of the antennae and fore tarsi respectively do not aversely affect the courtship performance of the males and these opera ions are likely to be just as serious as amputation of portions of the wings . Camera lucida drawings of the wings of fifteen males from each of the classes were made and
their areas and standard deviation calculated as previously described . The area of stump left in the class A flies was estimated at 10 per cent . The percentage wing area remaining in the two classes with partially amputated wings was as follows, controls being taken to represent 100 per cent . Controls 100+0 . 63 per cent . "C" males 80 . 85+0 . 94 per cent . "B" males 41 . 57+4 . 79 per cent . Two separate sets of mass matings were set up using fifty males from each class on each occasion . Normal winged control females were used throughout . The averaged results of these mass matings are illustrated in Fig . 4 from which it can be seen that mating success increased with wing area . Fig . 5 shows more clearly the relationships between these two factors . It is a graph of the number of copulations that had occurred during the first sixteen minutes of each test plotted against wing area for the four classes of male . The line joining the points is almost straight and suggests very strongly that the stimulation provided by vibration is additive in its effect on the female . As calculated from the above graph the effect of entirely removing the wings would be to reduce mating speed by just over eighty per cent .,
4 6 8 10 12 14 time in minutes, from start of test
16
18
Fig . 4 . Mass mating graphs of males with their wings amputated at the three positions illustrated in Fig . 3 compared with controls . The figures are derived from two separate tests .
EWING : COURTSHIP BEHAVIOUR OF Drosophila melanogaster so
s
s'
0
20
40
60
80
100
percentage wing area remaining
Fig 5 . A graph, derived from Fig . 4, of wing area plotted against number of copulations at 16 minutes . The open circle denotes that the wing area at this point was estimated . suggesting that this percentage of the total stimulation provided by the males is via vibration . Discussion In a previous paper (Ewing, 1961), in which I described the courtship behaviour of lines of flies which had been selected for large and small body size, I showed that large males exhibited less vibration than controls but nevertheless were equally successful in obtaining mates . I suggested that this deficiency in the amount of vibration was being compensated for by the increased wing area of the large selected males . The small selected males exhibited more vibration than controls but in spite of this were sexually less successful . I provided some evidence to show that the increased vibration of small males had been selected to compensate for some unspecified deficiency in their courtship which was the consequence of selection for body size . This deficiency could, in part, be attributed to decrease in wing area . A similar argument to the one given above could account for the increased vibration found in the LR males where there may have been secondary selection pressure for this character resulting from the decreased wing area . The results described in this paper demonstrate that wing area is indeed an important factor in courtship success .
319
Summary 1 . In the male of Drosophila melanogaster the component of courtship behaviour which plays the major role in sexually stimulating the female is wing vibration . 2. Wing area was changed by three methods : (a) By lowering the temperature during preimaginal development . (b) By selective breeding. (c) By amputation . 3 . The sexual success of males with different wing areas was measured and it was found that males with large wings were more successful in obtaining mates than those with small or partially amputated wings . 4 . The relationship between wing area and sexual success appeared to be linear thus demonstrating the additive nature of the stimulation provided via vibration . 5 . An extrapolation from these results suggested that approximately 80 per cent . of the sexual stimulation is normally provided by wing vibration . 6 . The relevance of these results to some previously published work is discussed . Acknowledgements My thanks are due to Dr. A. Manning for his help throughout the course of this work and for critically reading the manuscript . I am also indebted to Dr . F. W . Robertson who allowed me to use some of his selected lines .
REFERENCES Alpatov, W . W . (1930) . Phenotypic variation in body and cell size of Drosophila melanogaster. Biol. Bull ., Wood's Hole, 58, 85-103 .
Bastock, M . (1956) . A gene mutation which changes a behaviour pattern . Evolution, 10, 421-439 . Bastock, M . & Manning, A . (1955) . The courtship of Drosophila melanogaster. Behaviour, 8, 85-111 .
Ewing, A . W . (1961) . Body size and courtship behaviour in Drosophila melanogaster. Anim . Behav ., 9, 93-99 .
Ewing, A . W . & Manning, A . (1963) . The effects of exogenous scent on the mating of Drosophila inelanogaster . Anim. Behav ., 11, 596-598 .
Manning, A . (1959) . The sexual isolation between Drosophila melanogaster and Drosophila simulans. Anim. Behav., 7, 60-65 .
Manning, A . (1961) . The effects of artificial selection for mating speed in Drosophila tnelanogaster. Anim. Behav ., 9, 82-92 .
Mayr, E . (1950) . The role of the antennae in the mating behaviour of female Drosophila . Evolution, 4, 149-154 .
Petit, C. (1958) . Le determinisme genetique et psychophysiologique de la competition sexuelle chez Drosophila melanogaster . Bull. biol., 92, 248-329 .
320
ANIMAL BEHAVIOUR, XII, 2-3
Reed, S. C., Williams, C . M . & Chadwick, L. E . (1942) . Frequency of wing beat as a character for separating species, races and geographical varieties of Drosophila . Genetics, 27, 349-361 .
Robertson, F . W . (1962) . Changing the relative size of body parts of Drosophila by selection . Genet . Res., Camb. . 3,169-180
Sturtevant, A . H . (1915) . Experiments on sex recogntion and the problem of sexual selection in Drosophila • J. anim. Behav ., 5, 351-366 . (Accepted for publication 25th February, 1964 ; Ms. number : 417) .