Behav Genet (2014) 44:144–154 DOI 10.1007/s10519-013-9639-1
ORIGINAL RESEARCH
The Deleterious Effects of High Inbreeding on Male Drosophila melanogaster Attractiveness are Observed Under Competitive but not Under Non-competitive Conditions Terhi M. Valtonen • Derek A. Roff Markus J. Rantala
•
Received: 19 April 2013 / Accepted: 24 December 2013 / Published online: 12 January 2014 Ó Springer Science+Business Media New York 2014
Abstract In order for the male courtship traits to honestly signal quality they need to be condition-dependent. Moreover, if these traits capture genetic variation in condition they should resemble life-history traits in being subject to strong directional selection and, consequently, suffer strong inbreeding depression. In this study we investigated the effect of high inbreeding on male attractiveness by assessing mating success, mating speed and copulation duration of inbred, outbred and crossbred (constructed by crossing separate, randomly chosen inbred lines) males of Drosophila melanogaster. When set to compete against a standardized competitor and compared to the success rate of the crossbred lines, inbreeding significantly reduced male mating success. Under competition, outbred males initiated copulation significantly sooner than crossbred and inbred males. Under non-competitive conditions, no effect of inbreeding was found on either mating speed or copulation duration. Both mating success and mating speed showed much higher inbreeding depression than male size. Keywords Attractiveness Copulation duration Courtship Honest signals of quality Inbreeding depression Mating success Mating speed
Edited by Yong-Kyu Kim. T. M. Valtonen (&) M. J. Rantala Department of Biology, Section of Ecology, University of Turku, 20014 Turku, Finland e-mail: [email protected] D. A. Roff Department of Biology, University of California, Riverside, CA 92521, USA
123
Introduction In species where males contribute only genes to the next generation, mate choice has been hypothesized to benefit females genetically (Fisher 1930; Zahavi 1975; Tomkins et al. 2004; Taylor et al. 2007). If there is heritable variation in those male traits that females find attractive, females mated with attractive males will have sexy sons that leave more descendants. Females will get this indirect benefit even if they are only passively attracted to these traits (Kotiaho and Puurtinen 2007). For the genetic benefits to arise such traits need to be honest signals of quality. They need to be condition-dependent and have associated costs, so that only individuals in good condition or those that are good at accumulating resources can afford to spend resources on them. Consequently, also condition/resource acquisition ability has to be heritable for the genetic benefits to arise (Zahavi 1975; Rowe and Houle 1996; Cotton et al. 2004; Lawniczak et al. 2007). Female preference for elaborate male sexual traits has been documented for a number of species (Andersson 1994). Due to female preference for elaborate male sexual traits, such characteristics are likely to be under directional selection (Andersson 1994). Furthermore, if sexual ornaments capture genetic variation in condition (Rowe and Houle 1996), they should resemble life-history traits in being subject to strong directional selection (Prokop et al. 2010). Strong directional selection purges all but the most recessive deleterious mutations and consequently, fitnessrelated traits typically show considerable directional dominance (Lynch and Walsh 1998; van Oosterhout et al. 2003; Crnokrak and Roff 1995; Roff and Emerson 2006). The phenotypic effects of these recessive mutations will be expressed during inbreeding, and traits with a high degree of directional dominance will suffer strong inbreeding
Behav Genet (2014) 44:144–154
depression (Roff 1997; Roff and Emerson 2006). Inbreeding leads to increased levels of homozygosity which in turn decreases fitness, and trait values in general, by either or both the unmasking of recessive deleterious alleles and reducing the frequency of heterozygotes (Charlesworth and Charlesworth 1987, 1999; Roff 2002; Charlesworth and Willis 2009; Kristensen et al. 2010). Consequently, inbreeding depression in sexually selected traits can result from changes in heterozygosity at loci that directly code for sexual traits. On the other hand, if sexually selected traits are condition-dependent, and inbreeding reduces condition, then sexually selected traits may also signal the negative effects of reduced heterozygosity at loci underlying condition (Rowe and Houle 1996; Prokop et al. 2010). Several studies indeed report substantial declines in male sexually selected traits and male attractiveness with inbreeding. For example, in guppies (Poecilia reticulata) inbreeding (F = 0.25–0.59) has been found to affect male ornamental color patterns (van Oosterhout et al. 2003; Zajitschek and Brooks 2010; but see Mariette et al. 2006) and male courtship behavior (van Oosterhout et al. 2003), and result in a strong decline in male sexual motivation (courtship intensity) and mating success (number of successful copulations and insemination success) (Mariette et al. 2006). Moreover, female guppies have been found to prefer outbred to inbred (F = 0.375) males in mate choice trials (Zajitschek and Brooks 2010). Similarly, in the least killifish (Heterandria formosa) inbreeding (F = 0.25) has been found to decrease the frequency of male mating attempts (Ala-Honkola et al. 2009). In the black field cricket (Teleogryllus commodus) inbreeding was found to have no effect on the attractiveness of the advertisement calls. However, because inbred (F = 0.25) male crickets had a far lower call rate (but see Drayton et al. 2007) than outbred males, inbreeding was considered to lead to a reduction in the number of females mating with inbred T. commodus males (Drayton et al. 2010). Joron and Brakefield (2003) showed that outbred male Bicyclus anynana butterflies have a higher mating success than inbred males, and that the deleterious effects of inbreeding on mating success are significantly stronger in semi-natural conditions than in the laboratory—highly inbred males (F = 0.375) achieved a mating success equivalent to 50 % of that of outbred males; males of the intermediate level of inbreeding (F = 0.25) had a mating success equivalent to 70 % of that of outbred males. In zebra finches (Taeniopygia guttata) inbreeding has been found to affect both male attractiveness to females and female choice behavior (F = 0.25) (Bolund et al. 2010). Also female house mice (Mus musculus musculus) and female mealworm beetles (Tenebrio molitor) have been found to be more attracted to the odors of outbred compared to inbred (F = 25) males (Ilmonen et al. 2009; Po¨lkki et al. 2012).
145
In the common fruit fly (Drosophila melanogaster), males have to perform a complex courtship (a series of behaviors) and in order to be successful they have to appeal to the female by visual, chemical, tactile and acoustical signals (e.g. Bennet-Clark and Ewing 1969; Averhoff and Richardson 1974; Markow 1981; Hall 1994; Rybak et al. 2002a, b). Using lines of D. melanogaster made isogenic for chromosome 2, Miller et al. (1993) investigated the effects of inbreeding on male mating ability and courtship behavior, and found significant reductions in overall mating ability and impaired mating behavior in the inbred lines. Similarly, Pendlebury and Kidwell (1974) and Sharp (1984) documented a reduction in male mating success (male competitive mating ability) with inbreeding in D. melanogaster. However, whereas Pendlebury and Kidwell (1974) reported a drastic reduction in competitive ability and mating rate at a rather low level of inbreeding (F = 0.25) and no further decline in fitness at higher levels of inbreeding (F = 0.50–0.951), Sharp (1984) reported severe inbreeding depression in male mating ability with greater negative effects at higher inbreeding levels (F = 0.25–0.951). Parsons (1964) made a complete set of diallel crosses among six inbred lines of D. melanogaster (full-sib mated for more than 120 generations) and found that the crossbred pairs had a much higher mating frequency than the inbred pairs. In a study by Averhoff and Richardson (1974) both courtship activity and time to copulation were found to be directly correlated with the degree of inbreeding (up to 25 generations of full-sib mating). Deleterious effects of inbreeding on male sexual characteristics have also been documented in other Drosophila species. For instance, in Drosophila subodscura the mating success of inbred males was lower than that of outbred males due to a poor courtship performance of the inbred males (Maynard Smith 1956). In Drosophila simulans, significant inbreeding depression (F = 0.25) has been found in both male attractiveness (copulation latency) and male fertility (the number of offspring a female produced in her lifetime after a single copulation) (Okada et al. 2011) and in Drosophila montana, inbreeding (20 generations of brother-sister mating) was found to decrease male courtship song frequency (Aspi 2000). An important and increasingly recognized phenomenon is that the effects of inbreeding may only be detected under naturally relevant context of competition (Meagher et al. 2000; Michalczyk et al. 2010; Simmons 2011) and/or environmental stress (e.g. Armbuster and Reed 2005; Joron and Brakefield 2003). Meagher et al. (2000) showed that under seminatural conditions inbred (F = 0.25) male house mice (Mus domesticus) sire only one-fifth as many surviving offspring as outbred males because of their poor competitive ability and survivorship—i.e. inbred males suffered high mortality and were poor at obtaining
123
146
territories and surviving during the defense of territories. Under laboratory conditions, inbreeding had only relatively minor effects on mice reproductive success in this species and no effect on survivorship (Meagher et al. 2000). Severe fitness consequences of inbreeding, when measured under competitive conditions, have also been reported for D. melanogaster (Latter and Sved 1994). In species where females mate with multiple partners, sexual selection continues after copulation in the form of sperm competition and cryptic female choice (Parker 1970; Simmons 2011). Michalczyk et al. (2010) investigated the effect of inbreeding (eight generations of sibling mating) on male reproductive fitness in the flour beetle (Tribolium castaneum) and found that male ability to achieve normal fertilization success was depressed only when conditions of sperm competition were generated. Whereas Mack et al. (2002) report that female D. melanogaster may adaptively bias sperm use against related males, in a study that replicated the experiment of Mack et al. (2002) no such effect was found (Ala-Honkola et al. 2011). According to AlaHonkola et al. (2011) the number of studies that have found a negative association between male–female relatedness and competitive fertilization success is roughly equal to the number of studies that have not found such a relationship. As demonstrated in the examples above, inbreeding depression on male secondary sexual traits is likely to impair an individual’s attractiveness to females and hence, compromise male mating success. To date, most studies have investigated the effect of low to moderate levels of inbreeding on male attractiveness. Here, using highly inbred lines of D. melanogaster (30 generations of brothersister mating, F *1) we examine the effect of high inbreeding on male attractiveness by examining mating success, mating speed and copulation duration of inbred, outbred and crossbred (constructed by crossing separate, randomly chosen inbred lines) flies. The effect of inbreeding on mating success is assessed in a competitive mating situation against a standardized competitor. Mating speed is assessed under both competitive and non-competitive conditions. Copulation duration is measured under non-competitive conditions. In Drosophila, mating speed (the time from female introduction to commencement of copulation) serves as a standard measure of female preference and male attractiveness, and females are expected to mate faster with more attractive males (Taylor et al. 2007; Okada et al. 2011). Consequently, if there is inbreeding depression, we expect mating speed to slow down and copulation duration, mating success and male size to decrease. The crossbred flies are used to examine whether crossing highly inbred individuals from different inbred lines is sufficient to reverse the deleterious effects of inbreeding on male attractiveness.
123
Behav Genet (2014) 44:144–154
Materials and Methods Flies and Husbandry The laboratory base population and inbred/outbred lines of D. melanogaster are the same as those used by Valtonen et al. (2011a, b). To establish a stock population and inbred/outbred lines approximately 500 females were collected by baits from an apple grove at Lappi in Southern Finland in 2006. Since their establishment in the laboratory, the base population has been expanded and maintained in large glass jars with a standing adult population of several thousand individuals. The inbred and outbred lines were generated using the crossing design of Roff (1998, 2002; see also Wright et al. 2008; Okada et al. 2011; Po¨lkki et al. 2012). First, 20 females were allowed to lay eggs in baker’s yeast supplemented vials (one female per vial). Upon adult eclosion the next generation flies were collected as virgin and male–female pairs were generated to construct twenty full-sibling families. Each family was housed in separate culture vials. Second, the families were randomly grouped into pairs (altogether ten pairs) and from each group two inbred families were formed by full-sib mating and two outbred families were formed by reciprocal matings of a male and a female from each family within the group (see Fig. 1). From then onwards, full-sibling mating was used to continue the inbred lines; the outbred lines were continued by mating a female from an outbred line with a randomly chosen male from the stock (for more details see Valtonen et al. 2011a). The maintenance of the lines was continued for approximately thirty generations before the study commenced (for the inbred lines inbreeding coefficient F *1). The crossbred lines were constructed by crossing separate, randomly chosen inbred lines. Only the first generation progeny of the crossbred lines was used in the study. Some of the inbred/outbred lines were lost and some were in asynchrony with most of the other lines. The lines used in this experiment are listed in Tables 1 and 2. In the competitive mating success experiment svspa–pol mutant flies were used. The mutant fly stock (stock number: 1566) was provided by the Bloomington Drosophila Stock Center. The svspa–pol mutant flies have smaller eyes than normal flies and their corneal lenses and pseudocones are blurred and irregular and are hence easy to identify under a microscope. The stock is described in Flybase (http://flybase.bio.indiana.edu). According to Grossfield (1975) flies homozygous for svspa–pol lack a measurable ERG, which suggests that the spa–pol mutation may render flies blind. Whether or not these mutant flies are weaker competitors due to their genotypic or phenotypic characteristics than normal flies does not, however, play an important role in this study. It is also important to note that
Behav Genet (2014) 44:144–154
147
Fig. 1 Schematic illustration of the breeding design for a single ‘group’ (X, x female; Y, y male; upper case indicates crosses between the two families, A and B, to produce outbred progeny, and lower case indicates full-sib matings, producing inbred progeny)
in this study we are not investigating the competitive abilities of this mutant strain. Because we wanted the treatment flies to behave as normal as possible, the second male used in the mating success trials needed to be one that was easy to distinguish from the treatment flies. We did not want to mark the flies because marking may affect their behavior. In the competitive mating success trials mutant flies engaged in a normal courtship and in some cases really fierce male–male interactions were observed between the two competing males. When these flies competed against inbred flies, they won 42 % of the trials (see Fig. 3). To our opinion this mutant strain provided a good reference point for comparison. All flies were maintained at room temperature (22–24 °C) under continuous light. Larvae were reared on 10 g agar, 80 g cornmeal, 20 g brewer’s yeast, 1.5 dl syrup, 10 ml nipagin, 1 l water diet and adults were fed baker’s yeast. These conditions were maintained throughout the study. Ice and CO2 were used in handling the flies. Experimental Flies For the present study, newly eclosed adults from inbred and outbred lines and from the stock were collected. When the collected flies were 4 days old inbred, outbred and crossbred breeding pairs were set up as described above by mating the outbred females to the stock males (outbred), and the inbred females to their brothers (inbred) or to a male from another randomly chosen inbred line (crossbred). Each pair was permitted to reproduce for 24 h in a 30 ml vial supplemented with baker’s yeast. The following day 30 eggs from each pair were collected and placed into fresh 30 ml vials supplemented with 10 ml food for the larvae with baker’s yeast for the adult flies on top. Upon eclosion to the adult stage, the next generation of male flies (experimental flies) were collected as virgins. Simultaneously, newly enclosed mutant males and wild type
females were collected from the stock populations. The collected flies were housed individually in vials containing baker’s yeast until used in the experiment at the age of 4–7 days (post eclosion). Competitive Mating Success and Mating Speed To assess the effect of inbreeding on male mating success and mating speed, inbred, outbred and crossbred flies were set to compete with mutant males for wild type females. Consequently, the attractiveness of the males was assessed with the help of unrelated, outbred females derived from the stock population. In each trial one inbred, outbred or crossbred male and one mutant male was confronted with a wild type female in a 30 ml vial. The transfer of the flies into the new vial was accomplished without anesthesia. The vials were capped with cotton plugs so that the flies had enough space to move freely. Mating speed, the time taken for one of the males to start copulating with the female, was recorded and the winner was identified under a light microscope. The triad was observed for 90 min. Males that did not mate were censored in the analyses (see Table 1). Thorax length, as an estimate of body size, was measured after the experiment under a light microscope using an ocular micrometer (all males, including those that did not mate, were measured and included in the analysis). A total of 142 inbred, 138 outbred and 131 crossbred males were included in the experiment (Table 1). Mating Speed and Copulation Duration Under Noncompetitive Conditions Mating speed and copulation duration under non-competitive conditions were assayed following the protocol of the competitive mating success and mating speed experiment. However, this time only one male and one female were placed in the 30 ml vial. The amount of time from the
123
148
Behav Genet (2014) 44:144–154
Table 1 Competitive mating success and mating speed Line
Mating success (n)
Mating speed (n)
Table 1 continued Censored (n)
Inbred
Line
Mating success (n)
Mating speed (n)
Censored (n)
OUTB 1M
8
4
4
INB 1G
7
3
0
OUTB 1P
15
10
1
INB 1H
10
9
1
OUTB 1S
20
14
0
INB 1I
6
5
0
OUTB 2G
19
14
1
INB 1J
6
4
0
OUTB 2I
8
3
0
INB 1K
13
12
1
Total (n)
138
94
9
INB 1L INB 1N
13 5
11 3
2 1
Sample sizes and lines used in the experiment
INB 1O
4
1
0
INB 1Q
18
9
2
INB 1R
10
10
0
INB 2A
6
4
0
INB 2B
8
4
0
INB 2C
10
4
1
INB 2D
16
1
1
INB 2G
10
2
2
Total (n)
142
82
11
a
Male line 9 female line
beginning time of a mating until the end time of a mating represented the duration of copulation. Again, males that did not mate within 90 min were censored in the analyses (see Table 2). A total of 38 inbred, 71 outbred and 67 crossbred males were included in the experiment (Table 2). Statistics
Crossbreda INB 1A 9 INB 2C
6
6
0
INB 1G 9 INB 1N
8
6
0
INB 1H 9 INB 2D
7
6
1
INB 1H 9 INB 2G INB 1I 9 INB 1L
6 7
4 5
0 0
INB 1J 9 INB 1H
6
6
0
INB 1K 9 INB 1Q
5
3
1
INB 1L 9 INB 1G
1
1
0
INB 1L 9 INB 1H
7
5
0 0
INB 1N 9 INB 1K
3
3
INB 1N 9 INB IR
10
10
0
INB 1O 9 INB 2D
8
5
1
INB 1Q 9 INB 1H
4
2
0
INB 1Q 9 INB 1I
8
5
1
INB 1Q 9 INB 2D
4
4
0
INB 1R 9 INB 2D
3
2
2
INB 2B 9 INB 1Q
8
7
0
INB 2B 9 INB2D
8
8
0
INB 2D 9 INB 1H
3
3
1
INB 2E 9 INB 1A INB 2E 9 INB 1C
7 4
2 2
0 0
INB 2G 9 INB 2D
8
4
2
131
99
9
OUTB 1A
18
11
0
OUTB 1F
18
13
1
OUTB 1H
9
6
0
Total (n) Outbred
OUTB 1I
8
5
1
OUTN 1J
15
14
1
123
The analysis of genetic effects on competitive mating speed was carried out using the univariate analysis of variance with treatment (= inbred, outbred, crossbred) as a fixed factor and difference in body size between the competing males as a covariate (Treatment male size - Mutant male size = Difference in body size). Differences among inbred lines are expected to arise during the inbreeding process and hence, it is to be expected that there will be significant differences among the different lines within each treatment. Consequently, line was nested within treatment and included in the model as a random effect. Mating speed was ln transformed to meet the assumptions of parametric tests. Mating speed and copulation duration under non-competitive conditions were analyzed using the univariate analysis of variance with treatment as a fixed factor and line nested within treatment as a random effect. Both mating speed and copulation duration were ln transformed to meet the assumptions of parametric tests. The equality of variances was tested with Levene’s test and Tukey’s HSD test was used for the pairwise comparisons. Mating success is a binomial variable and was analyzed using a mixed model anova (lmer package in R) with treatment as a fixed effect, difference in body size as a covariate, line as a random effect and a binomial error structure. The mixed model was compared to the fixed effects model (i.e. no line effect) using the results from fitting a general linear model without random effects and using the log-likelihood ratio test (note that the mixed model was fitted using maximum likelihood not restricted maximum likelihood):
Behav Genet (2014) 44:144–154
149
model1
ORIGINAL RESEARCH
The Deleterious Effects of High Inbreeding on Male Drosophila melanogaster Attractiveness are Observed Under Competitive but not Under Non-competitive Conditions Terhi M. Valtonen • Derek A. Roff Markus J. Rantala
•
Received: 19 April 2013 / Accepted: 24 December 2013 / Published online: 12 January 2014 Ó Springer Science+Business Media New York 2014
Abstract In order for the male courtship traits to honestly signal quality they need to be condition-dependent. Moreover, if these traits capture genetic variation in condition they should resemble life-history traits in being subject to strong directional selection and, consequently, suffer strong inbreeding depression. In this study we investigated the effect of high inbreeding on male attractiveness by assessing mating success, mating speed and copulation duration of inbred, outbred and crossbred (constructed by crossing separate, randomly chosen inbred lines) males of Drosophila melanogaster. When set to compete against a standardized competitor and compared to the success rate of the crossbred lines, inbreeding significantly reduced male mating success. Under competition, outbred males initiated copulation significantly sooner than crossbred and inbred males. Under non-competitive conditions, no effect of inbreeding was found on either mating speed or copulation duration. Both mating success and mating speed showed much higher inbreeding depression than male size. Keywords Attractiveness Copulation duration Courtship Honest signals of quality Inbreeding depression Mating success Mating speed
Edited by Yong-Kyu Kim. T. M. Valtonen (&) M. J. Rantala Department of Biology, Section of Ecology, University of Turku, 20014 Turku, Finland e-mail: [email protected] D. A. Roff Department of Biology, University of California, Riverside, CA 92521, USA
123
Introduction In species where males contribute only genes to the next generation, mate choice has been hypothesized to benefit females genetically (Fisher 1930; Zahavi 1975; Tomkins et al. 2004; Taylor et al. 2007). If there is heritable variation in those male traits that females find attractive, females mated with attractive males will have sexy sons that leave more descendants. Females will get this indirect benefit even if they are only passively attracted to these traits (Kotiaho and Puurtinen 2007). For the genetic benefits to arise such traits need to be honest signals of quality. They need to be condition-dependent and have associated costs, so that only individuals in good condition or those that are good at accumulating resources can afford to spend resources on them. Consequently, also condition/resource acquisition ability has to be heritable for the genetic benefits to arise (Zahavi 1975; Rowe and Houle 1996; Cotton et al. 2004; Lawniczak et al. 2007). Female preference for elaborate male sexual traits has been documented for a number of species (Andersson 1994). Due to female preference for elaborate male sexual traits, such characteristics are likely to be under directional selection (Andersson 1994). Furthermore, if sexual ornaments capture genetic variation in condition (Rowe and Houle 1996), they should resemble life-history traits in being subject to strong directional selection (Prokop et al. 2010). Strong directional selection purges all but the most recessive deleterious mutations and consequently, fitnessrelated traits typically show considerable directional dominance (Lynch and Walsh 1998; van Oosterhout et al. 2003; Crnokrak and Roff 1995; Roff and Emerson 2006). The phenotypic effects of these recessive mutations will be expressed during inbreeding, and traits with a high degree of directional dominance will suffer strong inbreeding
Behav Genet (2014) 44:144–154
depression (Roff 1997; Roff and Emerson 2006). Inbreeding leads to increased levels of homozygosity which in turn decreases fitness, and trait values in general, by either or both the unmasking of recessive deleterious alleles and reducing the frequency of heterozygotes (Charlesworth and Charlesworth 1987, 1999; Roff 2002; Charlesworth and Willis 2009; Kristensen et al. 2010). Consequently, inbreeding depression in sexually selected traits can result from changes in heterozygosity at loci that directly code for sexual traits. On the other hand, if sexually selected traits are condition-dependent, and inbreeding reduces condition, then sexually selected traits may also signal the negative effects of reduced heterozygosity at loci underlying condition (Rowe and Houle 1996; Prokop et al. 2010). Several studies indeed report substantial declines in male sexually selected traits and male attractiveness with inbreeding. For example, in guppies (Poecilia reticulata) inbreeding (F = 0.25–0.59) has been found to affect male ornamental color patterns (van Oosterhout et al. 2003; Zajitschek and Brooks 2010; but see Mariette et al. 2006) and male courtship behavior (van Oosterhout et al. 2003), and result in a strong decline in male sexual motivation (courtship intensity) and mating success (number of successful copulations and insemination success) (Mariette et al. 2006). Moreover, female guppies have been found to prefer outbred to inbred (F = 0.375) males in mate choice trials (Zajitschek and Brooks 2010). Similarly, in the least killifish (Heterandria formosa) inbreeding (F = 0.25) has been found to decrease the frequency of male mating attempts (Ala-Honkola et al. 2009). In the black field cricket (Teleogryllus commodus) inbreeding was found to have no effect on the attractiveness of the advertisement calls. However, because inbred (F = 0.25) male crickets had a far lower call rate (but see Drayton et al. 2007) than outbred males, inbreeding was considered to lead to a reduction in the number of females mating with inbred T. commodus males (Drayton et al. 2010). Joron and Brakefield (2003) showed that outbred male Bicyclus anynana butterflies have a higher mating success than inbred males, and that the deleterious effects of inbreeding on mating success are significantly stronger in semi-natural conditions than in the laboratory—highly inbred males (F = 0.375) achieved a mating success equivalent to 50 % of that of outbred males; males of the intermediate level of inbreeding (F = 0.25) had a mating success equivalent to 70 % of that of outbred males. In zebra finches (Taeniopygia guttata) inbreeding has been found to affect both male attractiveness to females and female choice behavior (F = 0.25) (Bolund et al. 2010). Also female house mice (Mus musculus musculus) and female mealworm beetles (Tenebrio molitor) have been found to be more attracted to the odors of outbred compared to inbred (F = 25) males (Ilmonen et al. 2009; Po¨lkki et al. 2012).
145
In the common fruit fly (Drosophila melanogaster), males have to perform a complex courtship (a series of behaviors) and in order to be successful they have to appeal to the female by visual, chemical, tactile and acoustical signals (e.g. Bennet-Clark and Ewing 1969; Averhoff and Richardson 1974; Markow 1981; Hall 1994; Rybak et al. 2002a, b). Using lines of D. melanogaster made isogenic for chromosome 2, Miller et al. (1993) investigated the effects of inbreeding on male mating ability and courtship behavior, and found significant reductions in overall mating ability and impaired mating behavior in the inbred lines. Similarly, Pendlebury and Kidwell (1974) and Sharp (1984) documented a reduction in male mating success (male competitive mating ability) with inbreeding in D. melanogaster. However, whereas Pendlebury and Kidwell (1974) reported a drastic reduction in competitive ability and mating rate at a rather low level of inbreeding (F = 0.25) and no further decline in fitness at higher levels of inbreeding (F = 0.50–0.951), Sharp (1984) reported severe inbreeding depression in male mating ability with greater negative effects at higher inbreeding levels (F = 0.25–0.951). Parsons (1964) made a complete set of diallel crosses among six inbred lines of D. melanogaster (full-sib mated for more than 120 generations) and found that the crossbred pairs had a much higher mating frequency than the inbred pairs. In a study by Averhoff and Richardson (1974) both courtship activity and time to copulation were found to be directly correlated with the degree of inbreeding (up to 25 generations of full-sib mating). Deleterious effects of inbreeding on male sexual characteristics have also been documented in other Drosophila species. For instance, in Drosophila subodscura the mating success of inbred males was lower than that of outbred males due to a poor courtship performance of the inbred males (Maynard Smith 1956). In Drosophila simulans, significant inbreeding depression (F = 0.25) has been found in both male attractiveness (copulation latency) and male fertility (the number of offspring a female produced in her lifetime after a single copulation) (Okada et al. 2011) and in Drosophila montana, inbreeding (20 generations of brother-sister mating) was found to decrease male courtship song frequency (Aspi 2000). An important and increasingly recognized phenomenon is that the effects of inbreeding may only be detected under naturally relevant context of competition (Meagher et al. 2000; Michalczyk et al. 2010; Simmons 2011) and/or environmental stress (e.g. Armbuster and Reed 2005; Joron and Brakefield 2003). Meagher et al. (2000) showed that under seminatural conditions inbred (F = 0.25) male house mice (Mus domesticus) sire only one-fifth as many surviving offspring as outbred males because of their poor competitive ability and survivorship—i.e. inbred males suffered high mortality and were poor at obtaining
123
146
territories and surviving during the defense of territories. Under laboratory conditions, inbreeding had only relatively minor effects on mice reproductive success in this species and no effect on survivorship (Meagher et al. 2000). Severe fitness consequences of inbreeding, when measured under competitive conditions, have also been reported for D. melanogaster (Latter and Sved 1994). In species where females mate with multiple partners, sexual selection continues after copulation in the form of sperm competition and cryptic female choice (Parker 1970; Simmons 2011). Michalczyk et al. (2010) investigated the effect of inbreeding (eight generations of sibling mating) on male reproductive fitness in the flour beetle (Tribolium castaneum) and found that male ability to achieve normal fertilization success was depressed only when conditions of sperm competition were generated. Whereas Mack et al. (2002) report that female D. melanogaster may adaptively bias sperm use against related males, in a study that replicated the experiment of Mack et al. (2002) no such effect was found (Ala-Honkola et al. 2011). According to AlaHonkola et al. (2011) the number of studies that have found a negative association between male–female relatedness and competitive fertilization success is roughly equal to the number of studies that have not found such a relationship. As demonstrated in the examples above, inbreeding depression on male secondary sexual traits is likely to impair an individual’s attractiveness to females and hence, compromise male mating success. To date, most studies have investigated the effect of low to moderate levels of inbreeding on male attractiveness. Here, using highly inbred lines of D. melanogaster (30 generations of brothersister mating, F *1) we examine the effect of high inbreeding on male attractiveness by examining mating success, mating speed and copulation duration of inbred, outbred and crossbred (constructed by crossing separate, randomly chosen inbred lines) flies. The effect of inbreeding on mating success is assessed in a competitive mating situation against a standardized competitor. Mating speed is assessed under both competitive and non-competitive conditions. Copulation duration is measured under non-competitive conditions. In Drosophila, mating speed (the time from female introduction to commencement of copulation) serves as a standard measure of female preference and male attractiveness, and females are expected to mate faster with more attractive males (Taylor et al. 2007; Okada et al. 2011). Consequently, if there is inbreeding depression, we expect mating speed to slow down and copulation duration, mating success and male size to decrease. The crossbred flies are used to examine whether crossing highly inbred individuals from different inbred lines is sufficient to reverse the deleterious effects of inbreeding on male attractiveness.
123
Behav Genet (2014) 44:144–154
Materials and Methods Flies and Husbandry The laboratory base population and inbred/outbred lines of D. melanogaster are the same as those used by Valtonen et al. (2011a, b). To establish a stock population and inbred/outbred lines approximately 500 females were collected by baits from an apple grove at Lappi in Southern Finland in 2006. Since their establishment in the laboratory, the base population has been expanded and maintained in large glass jars with a standing adult population of several thousand individuals. The inbred and outbred lines were generated using the crossing design of Roff (1998, 2002; see also Wright et al. 2008; Okada et al. 2011; Po¨lkki et al. 2012). First, 20 females were allowed to lay eggs in baker’s yeast supplemented vials (one female per vial). Upon adult eclosion the next generation flies were collected as virgin and male–female pairs were generated to construct twenty full-sibling families. Each family was housed in separate culture vials. Second, the families were randomly grouped into pairs (altogether ten pairs) and from each group two inbred families were formed by full-sib mating and two outbred families were formed by reciprocal matings of a male and a female from each family within the group (see Fig. 1). From then onwards, full-sibling mating was used to continue the inbred lines; the outbred lines were continued by mating a female from an outbred line with a randomly chosen male from the stock (for more details see Valtonen et al. 2011a). The maintenance of the lines was continued for approximately thirty generations before the study commenced (for the inbred lines inbreeding coefficient F *1). The crossbred lines were constructed by crossing separate, randomly chosen inbred lines. Only the first generation progeny of the crossbred lines was used in the study. Some of the inbred/outbred lines were lost and some were in asynchrony with most of the other lines. The lines used in this experiment are listed in Tables 1 and 2. In the competitive mating success experiment svspa–pol mutant flies were used. The mutant fly stock (stock number: 1566) was provided by the Bloomington Drosophila Stock Center. The svspa–pol mutant flies have smaller eyes than normal flies and their corneal lenses and pseudocones are blurred and irregular and are hence easy to identify under a microscope. The stock is described in Flybase (http://flybase.bio.indiana.edu). According to Grossfield (1975) flies homozygous for svspa–pol lack a measurable ERG, which suggests that the spa–pol mutation may render flies blind. Whether or not these mutant flies are weaker competitors due to their genotypic or phenotypic characteristics than normal flies does not, however, play an important role in this study. It is also important to note that
Behav Genet (2014) 44:144–154
147
Fig. 1 Schematic illustration of the breeding design for a single ‘group’ (X, x female; Y, y male; upper case indicates crosses between the two families, A and B, to produce outbred progeny, and lower case indicates full-sib matings, producing inbred progeny)
in this study we are not investigating the competitive abilities of this mutant strain. Because we wanted the treatment flies to behave as normal as possible, the second male used in the mating success trials needed to be one that was easy to distinguish from the treatment flies. We did not want to mark the flies because marking may affect their behavior. In the competitive mating success trials mutant flies engaged in a normal courtship and in some cases really fierce male–male interactions were observed between the two competing males. When these flies competed against inbred flies, they won 42 % of the trials (see Fig. 3). To our opinion this mutant strain provided a good reference point for comparison. All flies were maintained at room temperature (22–24 °C) under continuous light. Larvae were reared on 10 g agar, 80 g cornmeal, 20 g brewer’s yeast, 1.5 dl syrup, 10 ml nipagin, 1 l water diet and adults were fed baker’s yeast. These conditions were maintained throughout the study. Ice and CO2 were used in handling the flies. Experimental Flies For the present study, newly eclosed adults from inbred and outbred lines and from the stock were collected. When the collected flies were 4 days old inbred, outbred and crossbred breeding pairs were set up as described above by mating the outbred females to the stock males (outbred), and the inbred females to their brothers (inbred) or to a male from another randomly chosen inbred line (crossbred). Each pair was permitted to reproduce for 24 h in a 30 ml vial supplemented with baker’s yeast. The following day 30 eggs from each pair were collected and placed into fresh 30 ml vials supplemented with 10 ml food for the larvae with baker’s yeast for the adult flies on top. Upon eclosion to the adult stage, the next generation of male flies (experimental flies) were collected as virgins. Simultaneously, newly enclosed mutant males and wild type
females were collected from the stock populations. The collected flies were housed individually in vials containing baker’s yeast until used in the experiment at the age of 4–7 days (post eclosion). Competitive Mating Success and Mating Speed To assess the effect of inbreeding on male mating success and mating speed, inbred, outbred and crossbred flies were set to compete with mutant males for wild type females. Consequently, the attractiveness of the males was assessed with the help of unrelated, outbred females derived from the stock population. In each trial one inbred, outbred or crossbred male and one mutant male was confronted with a wild type female in a 30 ml vial. The transfer of the flies into the new vial was accomplished without anesthesia. The vials were capped with cotton plugs so that the flies had enough space to move freely. Mating speed, the time taken for one of the males to start copulating with the female, was recorded and the winner was identified under a light microscope. The triad was observed for 90 min. Males that did not mate were censored in the analyses (see Table 1). Thorax length, as an estimate of body size, was measured after the experiment under a light microscope using an ocular micrometer (all males, including those that did not mate, were measured and included in the analysis). A total of 142 inbred, 138 outbred and 131 crossbred males were included in the experiment (Table 1). Mating Speed and Copulation Duration Under Noncompetitive Conditions Mating speed and copulation duration under non-competitive conditions were assayed following the protocol of the competitive mating success and mating speed experiment. However, this time only one male and one female were placed in the 30 ml vial. The amount of time from the
123
148
Behav Genet (2014) 44:144–154
Table 1 Competitive mating success and mating speed Line
Mating success (n)
Mating speed (n)
Table 1 continued Censored (n)
Inbred
Line
Mating success (n)
Mating speed (n)
Censored (n)
OUTB 1M
8
4
4
INB 1G
7
3
0
OUTB 1P
15
10
1
INB 1H
10
9
1
OUTB 1S
20
14
0
INB 1I
6
5
0
OUTB 2G
19
14
1
INB 1J
6
4
0
OUTB 2I
8
3
0
INB 1K
13
12
1
Total (n)
138
94
9
INB 1L INB 1N
13 5
11 3
2 1
Sample sizes and lines used in the experiment
INB 1O
4
1
0
INB 1Q
18
9
2
INB 1R
10
10
0
INB 2A
6
4
0
INB 2B
8
4
0
INB 2C
10
4
1
INB 2D
16
1
1
INB 2G
10
2
2
Total (n)
142
82
11
a
Male line 9 female line
beginning time of a mating until the end time of a mating represented the duration of copulation. Again, males that did not mate within 90 min were censored in the analyses (see Table 2). A total of 38 inbred, 71 outbred and 67 crossbred males were included in the experiment (Table 2). Statistics
Crossbreda INB 1A 9 INB 2C
6
6
0
INB 1G 9 INB 1N
8
6
0
INB 1H 9 INB 2D
7
6
1
INB 1H 9 INB 2G INB 1I 9 INB 1L
6 7
4 5
0 0
INB 1J 9 INB 1H
6
6
0
INB 1K 9 INB 1Q
5
3
1
INB 1L 9 INB 1G
1
1
0
INB 1L 9 INB 1H
7
5
0 0
INB 1N 9 INB 1K
3
3
INB 1N 9 INB IR
10
10
0
INB 1O 9 INB 2D
8
5
1
INB 1Q 9 INB 1H
4
2
0
INB 1Q 9 INB 1I
8
5
1
INB 1Q 9 INB 2D
4
4
0
INB 1R 9 INB 2D
3
2
2
INB 2B 9 INB 1Q
8
7
0
INB 2B 9 INB2D
8
8
0
INB 2D 9 INB 1H
3
3
1
INB 2E 9 INB 1A INB 2E 9 INB 1C
7 4
2 2
0 0
INB 2G 9 INB 2D
8
4
2
131
99
9
OUTB 1A
18
11
0
OUTB 1F
18
13
1
OUTB 1H
9
6
0
Total (n) Outbred
OUTB 1I
8
5
1
OUTN 1J
15
14
1
123
The analysis of genetic effects on competitive mating speed was carried out using the univariate analysis of variance with treatment (= inbred, outbred, crossbred) as a fixed factor and difference in body size between the competing males as a covariate (Treatment male size - Mutant male size = Difference in body size). Differences among inbred lines are expected to arise during the inbreeding process and hence, it is to be expected that there will be significant differences among the different lines within each treatment. Consequently, line was nested within treatment and included in the model as a random effect. Mating speed was ln transformed to meet the assumptions of parametric tests. Mating speed and copulation duration under non-competitive conditions were analyzed using the univariate analysis of variance with treatment as a fixed factor and line nested within treatment as a random effect. Both mating speed and copulation duration were ln transformed to meet the assumptions of parametric tests. The equality of variances was tested with Levene’s test and Tukey’s HSD test was used for the pairwise comparisons. Mating success is a binomial variable and was analyzed using a mixed model anova (lmer package in R) with treatment as a fixed effect, difference in body size as a covariate, line as a random effect and a binomial error structure. The mixed model was compared to the fixed effects model (i.e. no line effect) using the results from fitting a general linear model without random effects and using the log-likelihood ratio test (note that the mixed model was fitted using maximum likelihood not restricted maximum likelihood):
Behav Genet (2014) 44:144–154
149
model1
