Morphometrics
Workers of this species exhibit distinct variation in body size with readily identifiable major and minor workers. We measured facet number, facet diameter and identified their relationship with head width in major and minor workers. Head width is often used as a body-size proxy in ants (Kaspari and Weiser 1999) (Narendra et al. 2011). We took photographs of the dorsal surface of the heads of 30 minor and 44 major workers using a digital camera (D5600 DSLR, Nikon) and measured the widest part of their head using ImageJ (National Institutes of Health, Bethesda, MD, US).
To measure the facet number and diameter of each ant, we prepared eye replicas using established techniques. Detailed methods described here (Ramirez-Esquivel et al. 2017). Each eye replica was photographed with a digital camera (Olympus DP21, Olympus Australia, Victoria) attached to a light microscope (Olympus BX40, Olympus Australia, Victoria). The total number of facets was counted in all individuals. Diameter of eighty facets (representing approximately 10% of total facets) from the entire eye was measured in five minor workers and six major workers.
Electroretinography (ERG)
The temporal characteristics of O. smaragdina were determined by measuring the impulse response and the flicker fusion frequency (FFF) using electroretinography (ERG). Electrophysiological methods were as described previously (Ogawa et al. 2022). Briefly, animals were immobilised on ice for 5–10 minutes before removing their antennae and legs. Each individual ant was fixed, dorsal side up, to a plastic stage with bees’ wax before being mounted in a Faraday cage. A platinum wire of 0.127 mm diameter was inserted into the mesosoma and served as the indifferent electrode. As an active electrode, another platinum wire was placed against the lateral surface of the compound eye with conductive gel (Livingstone International Pty Ltd., New South Wales, Australia). ERGs were recorded through a differential amplifier (×1000; DAM50, World Precision Instruments Inc., FL, USA), with high- and low-pass hardware filter cut-off frequencies of 0.1 and 100 Hz respectively, connected to a computer via a data acquisition unit (Micro1401-3, Cambridge Electronic Design Ltd., Cambridge, England).
Light stimulus was provided to the frontal area of the compound eye with a 5mm diameter cool white light emitting diode (LED; C503C-WAS-CBADA151, Cree Inc, Durham, NC, USA). The LED was two degrees from the animal’s perspective that was set at 14cm from the animal at 10° elevation. All experiments were performed at room temperature (21–25°C) in the dark. Animals were dark-adapted for 20 min before each experiment. We carried out these experiments during the day (0900–1700 hrs), to ensure it aligns with diurnal activity rhythm that these ants exhibit.
The amplitude of the impulse response was measured as the voltage responses to a 1 ms square wave flash of light followed by 2 seconds of darkness. The response was averaged over 100 repetitions. The light source produced the maximum irradiance of (5.81×10− 5 W/cm2) at the surface of the eye (ILT1700, International Light Technologies). To identify the temporal characteristics of the impulse response we measured the following parameters at the highest light intensity, peak amplitude (mV), response latency (ms), time to peak amplitude (ms) and response duration (ms). The peak response amplitude was measured as the minimum amplitude of hyperpolarizing response of the eye. Response latency was defined as the time taken for the response to exceed 3 standard deviations of noise after stimulus onset. The standard deviation of the noise was calculated from all voltage changes in the last 500 ms before stimulus onset. Time to peak amplitude taken as the time of stimulus onset to the response peak. The duration of the impulse response was measured as the full-width of the response at half the maximum amplitude.
The FFF was estimated as the lowest temporal frequency at which the ERG reached a criterion threshold. The experimental design has been described in detail in a previous study (Warrington et al. 2017). Briefly, the visual stimulus followed a square-wave flicker over a range of stimulation frequencies from 2 to 200 Hz. Each frequency was presented for 20 s and the average response amplitude calculated using a Fast Fourier Transform. FFF were measured at 11 different light levels over a 5-log unit intensity range (1.33×10− 9 to 5.81×10− 5 W/cm2), increasing in 0.5 log unit steps apart from the lowest stimulus intensity (relative intensity at 0.00002). To evaluate any degradation of the response over time, the FFF at the highest intensity was tested before starting the series of FFF measures with 20 minutes dark adaptation in between. At high light intensities, the LED generated a measurable electrical artefact that looked like the response of the eye. The largest possible artefact was measured as the maximum signal recorded at the highest light intensity by covering the LED with a black cloth and then used as the response threshold (give threshold). The FFF was defined as the frequency at which the response power (log10 of the response amplitude power) crossed the threshold for each animal (see Fig. 1 in Warrington et al., 2017).
Pattern electroretinography (pERG)
Pattern electroretinograms were used to assess the spatial resolving power and contrast sensitivity of O. smaragdina. Detailed methods are described previously (Palavalli-Nettimi et al. 2019; Ogawa et al. 2019; Ryan et al. 2020). Briefly, animals were fixed on a plastic stage as described above and we used the same set of electrodes. Responses were amplified by a differential amplifier and sent to a 16-bit analog-to-digital converter device (USB-6353, National Instruments, Austin, TX, USA) connected to a computer. Individual animals were placed 30 cm from a white screen (51 cm width × 81 cm height), where the black and white sinusoidal grating stimuli were projected by a digital light processing projector (W1210ST, BenQ corporation, Taipei, Taiwan).
The stimuli were generated using Psychtoolbox 3 (Michelson 1927). The mean irradiance of the grating stimuli was 1.75×10− 4 W/cm2 measured using a calibrated radiometer (ILT1700, International Light Technologies, Peabody, MA, US). A temporal frequency of 2 Hz was used for all stimuli.
Prior to initiating recordings, a uniform grey stimulus with the same mean irradiance as the grating stimuli was presented to the ants for 20 minutes. They were then presented with 11 spatial frequencies (0.58, 0.53, 0.47, 0.42, 0.37, 0.32, 0.26, 0.21, 0.16, 0.11, 0.05 cpd), and up to five contrasts (95%, 75, 50, 25, 12.5) with the same mean irradiance for each spatial frequency. To evaluate any degradation of the response over time, the spatial frequencies of the gratings were presented in the order of decreasing frequencies of every second spatial frequency. The interleaved spatial frequencies were then presented in ascending order. At each spatial frequency, all five different contrasts were tested in decreasing order. For each combination of the stimuli, the response for five seconds each was recorded fifteen times to average them in the time domain. The averaged responses were then analysed using a Fast Fourier Transform, FFT. The non-visual electric signal (i.e., background noise) was measured as a control at two spatial frequencies (0.1 and 0.05 cpd) at 95% contrast with a black board used to shield the ant from the visual stimuli before and after the experimental series. The maximum signal out of the four control runs was used as the noise threshold.
For each eye that we carried out pERG recordings, we prepared eye replicas to determine the total facet number and diameter of 30 facets arbitrarily selected from the medio-frontal region. We also measured the head width of these ants.
Estimation of spatial resolving power and contrast threshold
An F-test was used to assess whether the response signal at the second harmonic (4 Hz) of the FFT response spectrum differed significantly from ten neighbouring frequencies, five on either side, for each spatial frequency and contrast combination. Spatial resolving power and contrast threshold were obtained by interpolating from the last point above the noise threshold whose amplitude at 4 Hz was also significantly greater than the ten surrounding frequencies, and the first point below the noise threshold. If the first point below the noise threshold was not significantly greater than the ten surrounding frequencies, the last point above the threshold was considered as the spatial resolving power, without interpolating between two data points. Contrast sensitivity is defined as the inverse of contrast threshold.
Statistical analysis
We determined the relationship between facet numbers, head width and worker caste. We assessed this with a linear mixed model using the maximum likelihood (ML) estimation method implemented in the nlme package of RStudio (Version 1.1.419, RStudio, Inc. Boston, MA, US). Head width and worker caste and their interaction were used as fixed effects in the model. Animal identity was used as a random effect.
To assess the relationship between facet diameter and worker castes, we used a linear mixed model. Worker caste was used as the fixed effects and animal identity as a random effect in the model.
A linear mixed effects model was used for testing whether the FFF differed according to stimulation light intensities. Stimulation light intensity was used as fixed effects and animal identity was used as a random effect. We used a linear model to evaluate any degradation of the response over time by comparing the first FFF at the brightest light, which was recorded prior to a series of various light intensities, to the last recording of FFF at the same light intensity.