Compensation for poor signal quality

Updated: Nov 1, 2019

Do animals compensate for poor quality (or missing) signals by investing more in another?

Oscillogram of a synthetic cricket courtship song (created by yours truly on Matlab, Raven, and Audacity)

Oftentimes in sexually signaling animals, signals exist across multiple sensory modalities (acoustic, olfactory, visual, gustatory). There are multiple hypothesis about how animals use different signaling modalities. For example, they may use each modality to send multiple messages to the receiver, or to reinforce one strong message (Hebets & Papaj 2005). I hypothesize that crickets compensate for poor quality, or absent, sexual signals by investing more in another modality.

The crickets have two primary sexual signals: acoustic and olfactory. Males produce songs to attract potential mates (acoustic signals), but also use cuticular hydrocarbons for the same purpose (olfactory signals). The Pacific field cricket is found throughout Australasia, but is also found on the Hawai'ian archipelago. In Hawai'i, the crickets coincide with an acoustically orienting parasitoid fly (Ormia ochracea) that tracks male calling song to find hosts for its parasitic larvae. A novel mutation arose in the Hawai'ian crickets, eliminating males' ability to sing and protecting the crickets from the fly. I am testing whether silent crickets compensate for their lack of song by investing more in attractive CHC profiles. Simultaneously, I am assessing this tradeoff in crickets in the non-parasitized populations of the Cook Islands, predicting to find that crickets with poor quality song also invest more in attractive CHCs.

Several studies have assessed the role of cuticular hydrocarbons as sexual signals; yet most of these studies have excluded the role of other sensory modalities in CHC sexual displays. The current literature on song and CHCs is also mixed; some studies demonstrate no relationship between the two (Simmons et al. 2013, Gray et al. 2014), while others find evidence that CHCs are redundant signals to song (Rybak et al. 2002). Furthermore, the two studies on courtship song and CHCs in T. oceanicus only look at Cook Island or Hawaiian males, but not both. In T. oceanicus, song is only a sexual signal, while individuals have a primary physiological need for CHCs. Since consequently there is high potential for trait loss in song and low potential in CHCs, I expect CHCs to serve as compensatory signals for song, acting as a safeguard against imperfect signal coding. I will test the hypotheses that 1) obligately silent males compensate for the lack of song by investing more in costly but attractive CHCs and 2) similar compensation occurs in males that do not carry the flatwing mutation, i.e. normal-wings and males from the non-parasitized Cook Islands, where males with lesser quality courtship song may invest more in CHCs as a sexual trait.

Hypothesis 1: Obligately silent males compensate for the lack of song by investing more in CHCs.

Methods: We collected normal- and flatwing males (n = 40 NW, 21 FW) as described in Aim 1 from Hilo (big island) and Oahu. We did not destructively sample in the relatively small Kauai population, and there are almost no flatwings in Hilo (Zuk et al. 2018), so our FW samples come from Oahu.

Results: We see no difference between CHC profiles of normal- and flatwing males (t-test, p=0.26, t=1.15, df=55.58) (Figure 3a,b). It appears that flatwing males do not compensate for lack of song by investing more in CHCs; these signaling modalities may send multiple messages and communicate different information to the female, i.e. courtship song indicates male quality (Rebar et al. 2011), while CHCs indicate relatedness (Thomas and Simmons 2008).

Hypothesis 2: Singing males compensate for poor quality song by investing in attractive CHCs.

Methods: The Zuk lab has frequently-replenished colonies of Hawaiian and Cook Island crickets in Caron Insect Growth Chambers with 12:12 light:dark cycles at 75% humidity and 250C (Bailey and Zuk 2008). I will assess the relationship between the short-range courtship song and CHCs in 20 singing males each from the Hilo, Oahu, Aitutaki, Mangaia, and Rarotonga lab colonies. I will induce courtship song by allowing a male to sense a female through mesh and record the song for one consecutive minute with an AKG D9000 microphone and Marantz PMD670 solid-state recorder.

I am building on recent advances in CHC sampling methods to develop a nondestructive extraction protocol. I will rub a glass filament brush over the cricket exoskeleton to remove CHCs and then swirl the brush in n-hexane for 5 minutes. After I record the courtship song, I will nondestructively sample the male’s CHCs. Then, I will conduct no-choice female preference trials, pairing males with different adult females from the same population, controlled for age and from a different rearing container. I will measure the time it takes for males to produce a courtship song, female latency to touch antennae and sample CHCs, female latency to mount (Shackleton et al. 2005, Rebar et al. 2011), mating success (transfer of a spermatophore), and female spermatophore retention time (Rebar et al. 2011), all of which are established pre- and post-copulatory preference metrics. After the mating trial ends, I will first weigh and then place the male into a 25% humidity incubator (high desiccation risk). Preliminary data show that 50% of crickets die within 24 hours of this treatment. Every half an hour for 24 hours, I will check for mortality. Once a male dies, I will immediately weigh the cricket again. By subtracting the end mass from the beginning mass (known as wet mass), I calculate “body water available”, or BWA (Arcaz et al. 2016). I will then dry the crickets with Drierite dessicant for one week, after which I will re-weigh the samples to obtain the dry mass and calculate the “body water content at death”, or BWCD. By subtracting BWA from BWCD, I will calculate water loss, my metric for desiccation tolerance (Arcaz et al. 2016).

At the end of this experiment, I will have a) song recordings, b) CHC profiles, c) measures for female pre- and post-copulatory preference, and d) measures of male desiccation tolerance. I will analyze CHC profiles as described in Aim 1. I will analyze song characteristics in Raven Pro 1.4 (Cornell Laboratory of Ornithology, Ithaca, NY, U.S.A.). These characteristics include the duty cycle (sound per unit time), carrier frequency, length of the song, intervals between different songs, and length and content of the different components of a song. Females prefer longer courtship songs with higher duty cycles (Rebar et al. 2009), but I will assess female preference for all song characteristics. I will then analyze the relationship between song quality and CHC profile, as well as between female preference, CHC profile, and desiccation tolerance. Lastly, I will compare regional variation in courtship song elements to the observed variation in CHCs from Aim 1 to assess the selective forces acting on these signals.

Expected Results: Song and CHCs evolve under different selective pressures: singing is conspicuous, while attractive CHCs can jeopardize waterproofing function. Males may optimize signal expression by investing more in attractive CHCs when unable to produce high quality songs. Consequently, I expect a tradeoff between song quality and CHC investment, as well as CHC investment, attractiveness, and desiccation resistance. I predict that the proportion of short-chain CHCs increases with a decrease in courtship song length and duty cycle, and any other components of courtship song that females are attracted to. I subsequently predict that males with more attractive CHCs will attract females more and have lower desiccation tolerance, i.e. higher water loss and lower survivorship in low humidity environments and higher attractiveness. However, since preliminary results show no difference between normal- and flatwing CHCs, there may be no compensatory relationship here either. If this is the case, I expect to see directional female preference for attractive courtship song elements and short-chain CHCs, but no correlation between attractiveness of the two signals.