Lochlan W
@lochlanw
@insect-vision.bsky.social Uni Konstanz PhD student with the International Max Planck Research School (QBEE) 🦋🧠🛩️ Intrigued by invertebrate sensorimotor control, vision, quantitative ethology 🇨🇦
IOW, hawkmoths appear to use a unified eye-proboscis control axis for flat surface interactions, and they strongly conserve this axis when perturbed instead of remapping sensorimotor control to the other body side. Indeed, we find the eye-proboscis coupling becomes STRONGER after perturbation
What DID exhibit a drastic change were the positions relative to the stripe, as well as the orientation of the head relative to the disc centre. ↪️ Meaning: instead of adjusting the proboscis, moths swapped sides and rotated to view the proboscis with the unpainted portion of the eye!
Finally, we sought to address a fundamental chicken-or-egg question 🥚🐔: does the proboscis bias drive the eye bias, or vice versa? Using monocular occlusion on the frontoventral region of the dominant eye, we were surprised to find that the proboscis... didn't move into view of the free eye. ❓
With this method, we found that the viewing tendency nicely matched the side of the proboscis bias. Even further, we found that on a contact-by-contact level, the moths tend to align where they look with where they touch - a striking similarity to the vertebrate strategy I mentioned before.
We then used a ray-based approach to create an estimate of the visual field and the hawkmoths' gaze, quantifying at which angle the stripe pattern is predominantly being viewed, and in which eye.
We found a strong clue suggesting that hawkmoths do the same. Left- and right-biased individuals positioned themselves distinctly on opposite sides of the stripe, corresponding to their bias. Thus, the proboscis and pattern were on the same body side, potentially more in view of a preferred eye. 🤔
A majority of the literature suggests that motor lateralization has consequences for related behaviours. So what does proboscis lateralization DO? Surprisingly, we found that proboscis lateralization propagates into which visual feature (e.g. an edge) hawkmoths target on the pattern surface!
We found that even naive animals exhibit lateralization that persisted over days. But even more interestingly, we found differences in the variability of proboscis positioning between experienced and naive, suggesting potential experience-dependent refinement. 🐣🐥
This led us to ask - how stable is this lateralization? Do individuals retain their unique way of controlling the proboscis, or is it an experience-dependent feature? Well, we tested the same individuals over consecutive days, and found that what they did on the first day PERSISTS over time! 🕰️
Not only that, but a continuum of lateralization directions and strengths emerge, creating subgroups of left-lateralized, right-lateralized, and unbiased proboscis control. This resembles the variation of handedness commonly found in vertebrates, like us humans! 🙋
We found that hawkmoths indeed show lateralization in how they position their proboscis while probing! Through thousands of individual proboscis contacts per animal, comparing the position relative to the head axis shows each individual has a unique way of placing their proboscis 👀
Hummingbird hawkmoths rely strongly on visual cues, such as flower patterns, to continuously guide their proboscis while searching for nectary. 🪻 We gave hawkmoths flat artificial 'flowers' with a simple directional pattern and quantified how they controlled the proboscis during interactions.
Ranging from molluscs to humans, lateralized appendage control shows up again and again. Even an 'unpaired' appendage, like an elephant trunk 🐘, can be lateralized. But do we find similar properties in an unpaired appendage of invertebrates, for example a hawkmoth proboscis? 👅
A very quick (and dirty) test of OCTRON on hawkmoth data 🦋 From installation to getting quite good CoM tracking, maybe 45 minutes total (with GPU and minimal training data)
Attending #ASABWinter2025 this coming week? Feel free to stop by my poster and check out how we used novel visual field estimations for freely-flying hawkmoths to investigate lateralized sensorimotor control! Excited to share the results from the past two years of work 🦋👁️👅🦾 and happy to chat!
Was an action packed and exciting 5 weeks spent at @kitp-ucsb.bsky.social for the #kitp_qbio program! Looking forward to seeing where active sensing navigates to in the coming years
5-minute chalk talk with @kitp-ucsb.bsky.social QBIO team today discussing my work with hawkmoths, visuomotor control, and pattern vision 🦋👁️🌺
@katrinvogt.bsky.social representing @uni-konstanz.de at the #KITP Neurosensing/QBio program, presenting work on the social life of fly larvae! 🪰🪱
Captured a couple shots of this cool lady today. No idea the species though 🕷️
Very cool brief visit with the @multipleye-lab.bsky.social 🕷️👁️ forgot to take a spider photo but the fish collection was quite cool 🐟
Sneak peek of some exciting work (click to view full panel) I'll be sharing on my poster at the Foraging and Information Seeking conference in Lyon next week! Let's discuss individuality, sensorimotor control, computer vision, and insect foraging 🤩👨🔬
Discussed plant-pollinator interactions over the past two days in Munich, with the Bavarian Academy of Science! I presented my PhD work and discussed how modern computational tools can better investigate the interaction between pollinating insects 🐝🪰🦋 and flowers 🌷🌹🌸🌼🌻🌺
Been hard to find the same community on here as on Tw!tter, but celebrating starting my 2nd year of the PhD with lots of neat results from my first project 😁 if you’re interested in thinking about creating some open-source analysis tools using existing methods (e.g. DeepLabCut), let’s talk ⚒️
