Mate-choice copying in Drosophila melanogaster: Impact of demonstration conditions and male–male competition

Highlights

• Two aspects of mate-choice copying methodology were investigated in fruit flies.

• We provided social information sequentially or simultaneously.

• We allowed or not male–male competition during the mate-choice.

• Mate-choice copying was found under sequential demonstrations only.

• Male–male competition did not impact mate-choice copying.

Abstract

Individuals of many species, including invertebrates, have been shown to use social information in mate choice, notably by extracting information from the mating performance of opposite sex conspecifics, a process called “mate-choice copying” (MCC). Here, we performed four experiments with Drosophila melanogaster to investigate two aspects of MCC methodology: whether (i) providing positive and negative social information simultaneously or sequentially during the demonstration phase of the protocol, and (ii) male–male competition during the mate-choice test, affect MCC. We found that the simultaneous provision of positive and negative information during demonstrations hampered female MCC performance, compared to the sequential provision of information. This can be interpreted in two alternative, yet not exclusive, ways: (i) attentional mechanisms may restrict the focus of the brain to one source of information at a time, and/or (ii) the shorter duration of demonstrations in the simultaneous protocol may have not permit full social learning use and may explain the non-detection of MCC in that protocol. Moreover, we did not detect any significant effect of male–male competition on female choice. This study thus provides further evidence for MCC in D. melanogaster and expands on the necessary methodology for detailed studies.

Introduction

Organisms need to continually assess environmental cues to increase the accuracy in their appraisal of the environment (Wagner and Danchin, 2010). This allows them to adaptively adjust their behaviour to current conditions (Dall et al., 2005, Danchin et al., 2004). However, sources and types of information available in the environment are diverse, both qualitatively and quantitatively, and it is still not clear how these numerous sources of information are perceived and processed, especially when they are contradictory. One intuitive view is that the more information available, the better the decision should be. For example, in the context of mate choice, multiple sources of information (i.e. signals or cues) have been shown to provide a more accurate estimation of the overall quality of a potential mate (Candolin, 2003, Moller and Pomiankowski, 1993, Scheuber et al., 2004). However, an increasing number of studies suggest that a too high information flow may hamper learning (Dukas and Real, 1993, van Swinderen, 2007, Weiss and Papaj, 2003), which may be rooted in two different constraints that are not mutually exclusive: a perception/sensory constraint and/or a processing constraint.

Processing constraints may result from a limited rate of information being processed by the brain (the “limited attention” hypothesis: Dukas, 2002, Dukas and Kamil, 2000). Alternatively, an organism may only attend to a subset of the available sources of information. For instance, organisms may filter out sources of information by focusing only on portions of the visual field at any given moment (spatially selective attention: Sareen et al., 2011), or by attending to only one or a few stimuli and ignoring the others (stimulus selective attention: Dukas, 2002). Sensory constraint, limited attention, and/or selective attention may strongly affect the simultaneous execution of two or more cognitive tasks; organisms are therefore expected to sort the different sources of information, focusing on the most relevant ones (Dukas, 2002).

According to Wagner and Danchin (2010), detectable information can be divided into two broad categories. First, organisms may interact directly with the physical environment thereby obtaining non-social information (Blanchet et al., 2010, Wagner and Danchin, 2010). Second, organisms may acquire social information by monitoring others interacting with their abiotic or biotic environment, including conspecifics (Bonnie and Earley, 2007, Dall et al., 2005, Wagner and Danchin, 2010). Social information may offer unique benefits by providing information about the quality and temporal predictability of the environment (Valone and Templeton, 2002), allowing organisms to make informed decisions (Reed et al., 1999).

Mate choice is a major fitness-affecting decision in sexually reproducing organisms. Classical sexual selection theories assume that females have genetically heritable preferences (Agrawal, 2001, Kirkpatrick and Ryan, 1991) and make independent choices, meaning that the choice of a given female is not context dependent and does not depend on the choices of other females (Alonzo, 2008, Wade and Pruett-Jones, 1990, Westneat et al., 2000). However, it has been demonstrated in many vertebrates and few invertebrates that females can extract information about male quality by observing the male’s mating performance and use this information to develop a preference for a given male, or male phenotype (Candolin, 2003, Danchin et al., 2004, Galef and White, 2000, Mery et al., 2009, Qvarnström, 2001, Witte and Noltemeier, 2002). This process is called mate-choice copying (hereafter MCC), and over the last two decades increasing attention has been devoted to exploring possible social influences on the development of mate preferences leading to non-independent mate choice (Alonzo, 2008, Losey et al., 1986, Pruett-Jones, 1992, Stohr, 1998, Westneat et al., 2000). Many studies have presented convincing evidence for MCC in a wide array of taxa, including humans (e.g. birds: White and Galef, 1999; fish: Witte and Ryan, 2002; mammals: Galef et al., 2008; humans: Waynforth, 2007, and even one non-social insect species (Drosophila melanogaster: Dukas, 2005, Mery et al., 2009).

In species where mating preference can be tested only once (e.g. because of the reluctance of females to copulate twice), MCC can be investigated using a two-phase experimental protocol. In the first phase (the demonstration), a naive female (called observer female) can gather social information about potential mates by observing them interacting with another female. A male copulating while being observed provides positive social information for its ability to attract mates and secure copulation, with male rejection providing negative social information. Demonstration methodology can vary greatly, for instance, by being performed sequentially (i.e. one male is observed at a time, as in Mery et al., 2009) or simultaneously (e.g. several demonstrations running jointly: Auld et al., 2009, Loyau et al., 2012, Witte and Ryan, 2002 and see Vakirtzis, 2011 for a review).

In the second phase (the mate-choice test), the preference of the observer female is assessed by offering her the “choice” between the same males she observed during the demonstration (MCC). Alternatively, the female can be presented with two other males of the same contrasting phenotypes as those used during demonstration. This latter protocol tests whether the observer female learnt the general rule of preferring males of a given phenotype, which corresponds to a generalised version of MCC. During this mate-choice test, males sometimes cannot interact, thus avoiding male–male competition. The female’s interest and willingness to mate with a particular male is then measured as the time she spends close to that male (i.e. latency: White and Galef, 1999). Even though this measure has been widely used in MCC experiments (Dugatkin and Godin, 1992, Frommen et al., 2009, Galef and White, 1998, Godin et al., 2005, Witte and Ryan, 2002), it has also been widely criticised as it remains an indirect measure of female mate preference (see Walling et al., 2010 for a review). Authors thus recommend using actual copulation tests whenever possible.

Here we report on four experiments that aimed to test the impact of two aspects of the demonstration protocol on the capacity of females to perform MCC in D. melanogaster. We created two male phenotypes by dusting them with green or pink powders. We first tested whether a simultaneous demonstration limits female ability to perform and generalise MCC, compared with a sequential demonstration protocol as described above. The sensory constraint/limited attention framework (Clark and Dukas, 2003, Dukas, 2002, Schmieder et al., 2012) predicts more efficient MCC when negative and positive social information about male attractiveness is provided sequentially.

Second, we intended to analyse the potential confounding effects of the male–male competition occurring during the mate-choice test on MCC. Male–male competition may limit observer female access to a given male, or information on male competitiveness may confuse her. Therefore, we manipulated the level of male–male competition during the mate-choice test by randomly assigning half of the observer females to one of two treatments: (i) females were offered the choice between a green and a pink male (competition situation) and (ii) females were randomly put with a green or a pink male (absence of competition). We expected MCC to be stronger in the absence of male–male competition, which we assessed by measuring the latency between the beginning of the test and the onset of the copulation.