Divergent Selection for Geotactic Response and Evolution of Reproductive Isolation in
Sympatric and Allopatric Populations of Houseflies
Author(s): L. E. Hurd and Robert M. Eisenberg
Source: The American Naturalist, Vol. 109, No. 967 (May - Jun., 1975), pp. 353-358
Published by: The University of Chicago Press for The American Society of Naturalists
Stable URL: http://www.jstor.org/stable/2459699
Accessed: 20-04-2017 15:28 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact [email protected].
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
The American Society of Naturalists, The University of Chicago Press are collaborating with
JSTOR to digitize, preserve and extend access to The American Naturalist
This content downloaded from 132.174.254.72 on Thu, 20 Apr 2017 15:28:51 UTC
All use subject to http://about.jstor.org/terms
Vol. 109, No. 967 The American Naturalist May-June 1975
DIVERGENT SELECTION FOR GEOTACTIC RESPONSE
AND EVOLUTION OF REPRODUCTIVE ISOLATION IN
SYMPATRIC AND ALLOPATRIC POPULATIONS OF HOUSEFLIES
L. E. HURD* AND ROBERT M. EISENBERG
Department of Entomology, Cornell University,
Ithaca, New York 14850, and
Department of Biological Sciences, University of Delaware,
Newark, Delaware 19711
Population biologists have long argued the relative merits of allopatry versus
sympatry with regard to the evolution of reproductive isolation and consequent
speciation. That reproductive isolation can occur between populations of a
species which are geographically isolated is hardly subject to dispute. However,
a controversy exists as to whether gene flow under sympatric conditions will
permit reproductive isolation to evolve "within the dispersal area of the offspring of a single deme" (Mayr 1963, p. 451). Although Mayr (1963) has developed extensive arguments against sympatric origins of species, it has been
shown that environmental discontinuity can select for physiological barriers
to gene flow between adjacent populations, which decreases the reproductive
inefficiency of producing unfit hybrids (Antonovics 1968; McNeilly and Antonovics 1968). Other workers have demonstrated that gene flow may not be an
effective deterrent to reproductive isolation (Endler 1973; Millicent and Thoday
1961; Streams and Pimentel 1961; Thoday and Boam 1959; Thoday and
Gibson 1970).
Pimentel et al. (1967) proposed a model of sympatric speciation in which
reproductive isolation of subpopulations follows the establishment of races in
adjacent habitats, in response to disruptive selective pressures imposed by these
habitats. In this model, hybrid individuals resulting from gene flow between
habitats are selected against continuously, with the consequent strengthening
of divergence between races. Thus the frequency of hybrids decreases with time,
leading to reproductive isolation and eventual speciation.
Recently, Soans et al. (1974) experimented with the evolution of reproductive
isolation in allopatric and sympatric (3000 gene flow) subpopulations of houseflies from a single deme which were subjected to geotactic selection. Their
results indicated that reproductive isolation occurred under both allopatric
and sympatric conditions with a selection pressure of 9500.
Our study was similar to that of Soans et al. (1974) with the important
exception that the sympatric populations of houseflies were subjected to 5000
* Present address: Department of Biological Sciences, University of Delaware, Newark,
Delaware 19711.
Amer. Natur. 1975. Vol. 109, pp. 353-358.
? 1975 by The University of Chicago. All rights reserved.
353
This content downloaded from 132.174.254.72 on Thu, 20 Apr 2017 15:28:51 UTC
All use subject to http://about.jstor.org/terms
354 THE AMERICAN NATURALIST
gene flow in an attempt to test the effect of maximum possible potential
genetic dispersal on disruptive selection. The term "gene flow" as used by Soans
et al., as well as in this paper, refers to the degree of genetic exchange theoretically possible under the experimental conditions and not to the amount of
exchange actually achieved.
MATERIALS AND METHODS
Fly Rearing
The flies (Musca domestic) used in this experiment were from a wild
population collected in Ithaca, New York. About 100 pairs were placed in
ventilated plastic cages (14 x 20 x 10 cm) illuminated by an overhead
fluorescent light and maintained at approximately 25? C. Powdered milk and
sugar cubes provided food, and inverted, water-filled flasks stoppered with
cellucotton provided moisture. Shell vials of larval rearing medium (Ralston
Purina Co., CSMA Fly Larvae Medium mixed with water and dry baker's yeast)
served as oviposition sites for females.
After 24 h, the vials were transferred to larger plastic containers (30 x 22 x
10 cm) filled with larval rearing medium. About 5,000 pupae were collected at
the end of 8 days and held until adult flies emerged.
Flies which emerged were sexed within 24 h, and 1,000 flies of each sex were
held in the larger plastic cages and provided with food and moisture under a
12-12 photoperiod. All handling of adults was carried out while they were
lightly anesthetized with moist carbon dioxide.
Selection
The selection apparatus used consisted of a rectangular container 60 cm
high x 14 cm on each side, to which a tube 20 cm long x 4 cm in diameter
was attached at each end. A ventilated 16-cm-diameter plastic box was attached
to both top and bottom. One thousand flies of either sex were introduced into
the chamber from a mason jar inserted into a hole in the side of the chamber,
midway between top and bottom (see Soans et al. 1974).
Geoselection was carried out under red light at approximately 25? C. The first
50 flies to travel either to the top or bottom collecting boxes were removed.
These flies provided the starting population for each of three selection regimes:
1. Race A (allopatric).-The first 50 flies of each sex to exhibit positive (+)
geotaxis were removed, and their response time was recorded. The sexes then
were united in a single cage, and their offspring provided the next generation to
undergo selection for (+) geotaxis.
2. Race B (allopatric).-This group was treated exactly as race A, except that
only the first 50 flies which exhibited a negative (-) geotaxis were allowed to
mate.
3. Race C (sympatric).-This group was treated similarly, except that the
first 50 flies of each sex to exhibit (+) geotaxis and (-) geotaxis both were
collected. Then 25 flies were chosen randomly out of each set of 50, and these
This content downloaded from 132.174.254.72 on Thu, 20 Apr 2017 15:28:51 UTC
All use subject to http://about.jstor.org/terms
SELECTION AND REPRODUCTIVE ISOLATION IN HOUSEFLIES 355
100 flies were placed in a cage together. This was done in order to satisfy the
conditions of 5000 potential gene flow in sympatry.
The selection pressure on all races was 95%.
Reproductive Isolation
At the end of 16 generations, tests were performed on each race to determine
the degree of reproductive isolation, following the final selection for geotaxis.
Although response time was recorded only for the first 50 flies responding in
each race, the number of flies actually collected was increased to 75 in order to
obtain a sufficient number of flies for replicate tests. Races were marked with
different colors of enamel paint on the thorax. The following tests then were
carried out under a fluorescent light at 25? C in an otherwise darkened room:
Test 1.-One male of each race was paired with a female of its own and
another race in a plastic vial (5 cm high x 2.5 cm diameter). Ten replicates
were run, and the first mating in each vial was recorded as either homogamic or
heterogamic.
Test 2.-Twenty-five male and 25 female flies of races A and B, or of C( +)
and C(-), were placed in a plastic cage (14 x 20 x 10 cm). A count was then
made of the number of homogamic and heterogamic matings taking place in the
first 30 min of observation.
Test 1 was also performed on races A and B prior to final geoselection during
the sixteenth generation for the purpose of comparison with the results obtained
by Soans et al. (1974). This was impossible in the case of race C, since, unlike
the experiment of Soans et al., the present design incorporated both (A+) and
(-) in the same population under conditions of 50% potential gene flow. The
rationale for this procedure is that the collecting boxes on the top and bottom
of the chamber represent different habitats, in the sense that they impose
different selective pressures. Reproductive isolation tests were carried out after
this selection had been exerted, on samples taken from these "habitats" since
these were the only flies which were allowed to mate and reproduce. Sampling
from the population before selection necessarily imposes a high probability of
including in the tests flies which would never survive to reproduce following
selection. In nature the question of evolution of reproductive isolation is a
moot one for individuals which are selected out of a population before mating
can occur. Yates correction was applied to all x2 calculations. The index of
isolation (I) employed was adapted from Stalker (1942):
I No. of homogamic matings - no. of heterogamic matings
Total no. of matings
The range of I values is from -1 (all heterogamic matings) to + 1 (all homogamic matings) and will be 0 in the case of panmixia.
RESULTS
At the start of the experiment, approximately 3 h was required to collect
(+) geotactic flies from the original population, and about 41 h was needed for
the (-) geotactic flies. The (+) geotactic flies of race C began to exhibit a
This content downloaded from 132.174.254.72 on Thu, 20 Apr 2017 15:28:51 UTC
All use subject to http://about.jstor.org/terms
356 THE AMERICAN NATURALIST
TABLE 1
TESTS FOR HOMOGAMIO AND HETEROGAMIC MATINGS
AS A MEASURE OF REPRODUCTIVE ISOLATION
No. OF FLIES MATED No. OF MATINGS
ISOLATION
Male
Female
Homo
Hetero
INDEX
(I)
x2
P
Test 1:
LA LA
1
B
10
0
1.0
10.6 <.005
1B
1B
1
A
8
2
0.6
1 C(+) 1 C(+)
1
C(-)
10
0
1.0
13.0 <.005
1 C(-) 1 C(-)
LC(+)
9
1
0.8
7
0.56
Test 2:
25 A 25 A
25
B
25
B
25
9.0
<.005
25 C(+) 25 C(+)
25
C(-)
25
C(-)
29
9
0.52
9.5
<
.005
response time similar to the allopatric race A after only four generations,
although the (-) flies in race C were somewhat erratic in comparison to race B
and required more time to converge. By the end of 16 generations the initial
effect of 50% potential gene flow had apparently subsided, and all flies were
down to approximately 10 min response time. In addition, by the twelfth
generation, the requisite 50 flies were collected from races A or B before any
flies were observed to exhibit a geotactic response opposite to that selected for.
The general trend was for male flies to respond faster than females. Another
trend was for ( + ) flies in races A and C to respond faster than (-) flies in races
B and C, although this was not as consistent as the difference in response time
between sexes within races.
The results of the tests for reproductive isolation are summarized in table 1.
The tendency toward reproductive isolation was highly significant (P < .005)
for both allopatric and sympatric races of flies. Within the test vials there was
no obvious preference exhibited for vertical position for any of the races.
The I values for both allopatric and sympatric races were about equivalent
in both tests performed. Test 1, which was also run prior to geotactic selection
in generation 16 for races A and B, showed significant (X2 7.4; P < .025)
reproductive isolation with I values of 0.8 (race A) and 0.6 (race B).
DISCUSSION
Our results suggest that reproductive isolation has evolved in sympatric
populations of houseflies under conditions of 5000 potential gene flow, subjected
to 95% selective pressure. After 16 generations of selection both sympatric and
This content downloaded from 132.174.254.72 on Thu, 20 Apr 2017 15:28:51 UTC
All use subject to http://about.jstor.org/terms
SELECTION AND REPRODUCTIVE ISOLATION IN HOUSEFLIES 357
allopatric races achieved about the same response time in geotactic behavior,
indicating that the effect of gene flow was relatively short-lived. In addition,
both sympatric and allopatric races exhibited approximately the same extent
of reproductive isolation when subjected to tests for mate selection at the end
of the experiment.
These results tend to support the findings of Soans et al. (1974) that sym-
patric populations of flies exposed to 30% potential gene flow evolved reproductive isolation comparable to allopatric populations, after 38 generations.
However, there are a number of important differences between these two
studies. Their flies showed a longer initial geotactic response time, a slower rate
of evolution toward a decreased response time, and a tendency for (-) geotactic
flies to respond faster than (+) flies. In addition, their allopatric populations
still exhibited mixed geotactic responses at the end of 38 generations, whereas
our allopatric populations responded in a homogeneous fashion from the
twelfth generation.
Some of these differences probably are a function of the variability between
the natural populations from which the experimental animals were initially
chosen. It is unfortunate that this makes it impossible to single out the differences between the two studies on the basis of the differences in the magnitude
of potential gene flow alone. However, both studies strongly support the experimental evidence of Koopman (1950), Wallace (1954), Antonovics (1968),
and others that natural selection acts to strengthen barriers to gene flow
between diverging populations, which increases reproductive efficiency by
selecting out unfit hybrids (Dobzhansky 1970). This study does not, however,
supply support for the model of sympatric speciation proposed by Pimentel
et al. (1967), as both allopatric and sympatric populations appear to have
evolved at similar rates.
Considering the rapid rate at which reproductive isolation occurs in houseflies
under geotactic selection, the question is raised as to how long it would take
for complete speciation to occur (D. Pimentel, personal communication), as
well as to the nature of the mechanisms involved. It is unlikely that differences
in geotactic responses alone could account for the observed results, as flies
were forced to frequent common regions of their cages in order to feed, water,
and oviposit. This contention is supported by the results of the mating tests
run within race C. It is more likely that by selecting for geotactic response,
some other (e.g., behavioral) response which served to differentiate mating
types was responsible for the degree of reproductive isolation observed here.
SUMMARY
Experimental populations of houseflies subjected to 95% selective pressure
for geotactic preference under conditions of 50% potential gene flow and of
allopatry evolved reproductive isolation after only 16 generations. No significant
difference was found between sympatric and allopatric populations. Our study
does not support the hypothesis that incipient reproductive isolation occurs
more rapidly under conditions of sympatry than allopatry.
This content downloaded from 132.174.254.72 on Thu, 20 Apr 2017 15:28:51 UTC
All use subject to http://about.jstor.org/terms
358 THE AMERICAN NATURALIST
ACKNOWLEDGMENTS
We are grateful to D. Pimentel, L. L. Wolf, V. A. Lotrich, and H. V. Cornell
for their criticisms of an earlier draft of this manuscript. Support for this study
and subsequent writing came from a grant from the Ford Foundation to D.
Pimentel at Cornell University, and from a UNIDEL grant to the Department
of Biological Sciences of the University of Delaware.
LITERATURE CITED
Antonovics, J. 1968. Evolution in closely adjacent plant populations. V. Heredity 23:
219-238.
Dobzhansky, Th. 1970. Genetics in the evolutionary process. Columbia University Press,
New York.
Endler, J. A. 1973. Gene flow and population differentiation. Science 179:243-250.
Koopman, K. 1950. Natural selection for reproductive isolation between Drosophila
pseudobscura and Drosophila persimilis. Evolution 4:135-148.
McNeilly, T., and J. Antonovics. 1968. Evolution in closely adjacent plant populations. IV.
Heredity 23: 205-218.
Mayr, E. 1963. Animal species and evolution. Belknap, Cambridge.
Millicent, E., and J. M. Thoday. 1961. Effects of disruptive selection. Heredity 16: 199-217.
Pimentel, D., G. J. C. Smith, and J. S. Soans. 1967. A population model of sympatric
speciation. Amer. Natur. 101:493-504.
Soans, A. B., D. Pimentel, and J. S. Soans. 1974. Evolution of reproductive isolation in
allopatric and sympatric populations. Amer. Natur. 108:117-124.
Stalker, J. D. 1942. Sexual isolation studies in the species complex Drosophila virilis.
Genetics 27:238-267.
Streams, F. A., and D. Pimentel. 1961. Effects of immigration on the evolution of populations.
Amer. Natur. 95:201-210.
Thoday, J. M., and T. B. Boam. 1959. Effects of disruptive selection. II. Polymorphism
and divergence without isolation. Heredity 13:205-218.
Thoday, J. M., and J. B. Gibson. 1970. The probability of isolation by disruptive selection.
Amer. Natur. 104:219-230.
Wallace, B. 1954. Genetic divergence of isolated populations of Drosophila melanogaster.
Caryologia 6 (Suppl.):761-764.
This content downloaded from 132.174.254.72 on Thu, 20 Apr 2017 15:28:51 UTC
All use subject to http://about.jstor.org/terms