iiinova — where your digital self keeps evolving

Tanya Strydom @tanyadoesscience.bsky.social · 02/10/2026

Me (after having spent ~2 days on a slide deck, including colour palette specific re-rendering of images and figures): I have the desire to rework my colour palette 🙃🙃🙃

000

Tanya Strydom @tanyadoesscience.bsky.social · 26/08/2026

Working on layouts for a slide deck and honestly this one might be at its peak already. 😂

010

Tanya Strydom @tanyadoesscience.bsky.social · 23/06/2026

Me: yeah I’m just not feeling the #wc2026 vibes this year Also Me: is obsessed with the typeface used for the Norway kit. 🇳🇴 Imma need journals to get real chill real quick about non standard typefaces…

010

Tanya Strydom @tanyadoesscience.bsky.social · 22/06/2026

To me, the challenge is figuring out how these different representations relate to one another (and whether we can move between them in a principled way), rather than treating any single network as the definitive description of reality.

000

Tanya Strydom @tanyadoesscience.bsky.social · 22/06/2026

My current view is that different network approaches may be capturing different aspects of the same system. A network assembled from observations accumulated over years is different from short-term observations. None are necessarily wrong, but they may be used to answer subtly different questions.

100

Tanya Strydom @tanyadoesscience.bsky.social · 22/06/2026

It also gets a bit circular because we often use empirical networks to validate predicted ones, despite empirical networks being products of particular sampling windows and methodologies rather than direct observations of some underlying 'true network'.

110

Tanya Strydom @tanyadoesscience.bsky.social · 22/06/2026

Part of what makes this so difficult is the multiple layers of uncertainty. As you said, empirical networks are shaped by many factors. But even if we hold those things constant, different approaches to constructing networks can still lead us to different conclusions.

100

Tanya Strydom @tanyadoesscience.bsky.social · 22/06/2026

Oh yeah it's for sure the thing that keeps me up at night and I've yet to come to a satisfying answer. The more I think about it, the less convinced I am that the question is 'which network is correct?' but rather 'what aspect of interaction reality does this network actually represent?'

110

Tanya Strydom @tanyadoesscience.bsky.social · 22/06/2026

And finally 🙌 Huge thanks to the absolute dream team behind this one 💚  🐚 @barankarapunar.bsky.social  🕸️ Jennifer Dunne 🦴 Pincelli Hull  🌊 Crispin Little  🦈 @pimientoc.bsky.social  🌍 Andy Ridgewell  🌱 @beckerhopper.bsky.social  🦕 @profalexdunhill.bsky.social

030

Tanya Strydom @tanyadoesscience.bsky.social · 22/06/2026

So the takeaway isn't: ❌ "Which model is correct?" It's: ✅ "Which model is correct for your question?" Network reconstruction isn't just a technical step. It's part of the hypothesis. Choose accordingly 🎯 📄 doi.org/10.32942/X2X...

doi.org

Are We Mapping Ecosystems or Models? Framework Choices Dominate Food Web Topology and Extinction Inferences

110

Tanya Strydom @tanyadoesscience.bsky.social · 22/06/2026

Before anyone panics 😅 This doesn't mean food web models are bad. It means they're hypotheses. Every reconstruction method encodes a different view of how interactions happen 🧠 Trait constraints 🦷   Energetics ⚡  Topology 📈 They're answering slightly different questions.

120

Tanya Strydom @tanyadoesscience.bsky.social · 22/06/2026

And here's the bit that made us do a double take 😳 The reconstruction method explained MORE variation in network structure than the extinction event itself. Read that again. 🫠 The modelling framework had a bigger effect on the inferred food web than the ecological turnover we were studying.

121

Tanya Strydom @tanyadoesscience.bsky.social · 22/06/2026

💀 Question 3: Do they identify the same drivers of extinction? Again: 🚫 Not really. Different network reconstructions pointed to different mechanisms of collapse and recovery. Same fossils.  Same species.  Different stories.

110

Tanya Strydom @tanyadoesscience.bsky.social · 22/06/2026

🍽️ Question 2: Do they predict the same species interactions? Also nope. Even when two methods produced networks that looked structurally similar... ...they often disagreed about who eats who 😬 So similar-looking food webs can hide very different ecological stories underneath.

110

Tanya Strydom @tanyadoesscience.bsky.social · 22/06/2026

📊 Question 1: Do the networks actually look different? Yep. Very much so. Different reconstruction methods produced completely different food-web structures 😵‍💫 BUT Methods built on *similar* underlying ideas tended to cluster together 👯 (ADBM 🤝 ATN) So the assumptions matter.

110

Tanya Strydom @tanyadoesscience.bsky.social · 22/06/2026

Most studies ask: 💀 How did food webs change in an extinction event? We flipped the question 🔄 We took a marine community spanning the Early Toarcian extinction (~183 million years ago) And asked: If we reconstruct the food web 6 different ways... Do we reach the same conclusions? (Spoiler: lol no)

121

Tanya Strydom @tanyadoesscience.bsky.social · 22/06/2026

Ecologists love networks 🕸️ But here's the thing: There are LOTS of ways to build them. Traits 🦷  Energetics ⚡  Or just plain topology 📈 And each one comes with different assumptions🔧 We wanted to know: 👉 If we build the same ecosystem differently... do we get the same story?

120

Tanya Strydom @tanyadoesscience.bsky.social · 22/06/2026

🚨📄 NEW PREPRINT ALERT 🚨📄 🌐 Are we mapping ecosystems… or just mapping models? 🤔 We took ONE Jurassic fossil community 🦑🐚🐟 and built food webs SIX different ways. The result? 😳 The way you build the network can matter MORE than the change you're trying to study. 🔗 doi.org/10.32942/X2X... 👇

3116

Tanya Strydom @tanyadoesscience.bsky.social · 16/06/2026

When you receive a rejection for one fellowship application while in the throes of putting together a different fellowship application

010

Reposted by Tanya Strydom

Prof Alex Dunhill @profalexdunhill.bsky.social · 22/05/2026

New R package 📦: pfwim reconstructs ancient food webs from traits + body size rules 🌿🐳🕸️ cran.r-project.org/web/packages... by @tanyadoesscience.bsky.social @jackolivershaw.bsky.social @barankarapunar.bsky.social @beckerhopper.bsky.social with lots of help from @cemac.bsky.social

178

Tanya Strydom @tanyadoesscience.bsky.social · 12/05/2026

It’s been so cool to follow the work coming from @gabdans.bsky.social. Awesome to see someone pushing the field into the applied space and being intentional about thinking about the end use of the cool science that we do!

020

Reposted by Tanya Strydom

Matt Kerr @mattkerr.bsky.social · 11/05/2026

This work was masterfully lead by @jonastrepel.bsky.social, and features amazing art by @tanyadoesscience.bsky.social (attached here). It was a pleasure to be part of the discussions - and we hope to spark a wider conversation on developing robust frameworks for evaluating impact.

Artistic renderings of a community at three different stages, shown as three panels called "Evolutionary Baseline", "Current Baseline", and "Large Herbivore Reintroductions". The left panel, evolutionary baseline, includes large and megaherbivores - specifically a mammoth and wild horses. The middle panel, current baseline, shows the state after the extirpation or extinction of most large herbivore species, with only a few deer left and much more forest vegetation. The right panel shows the state after the introduction or reintroduction of large herbivores, with wild cattle and horses creating a more open landscape.

181

Tanya Strydom @tanyadoesscience.bsky.social · 25/03/2026

Once again busy with totally serious science...

Screenshot from the title of a GitHun pull request. The title of the pull request is 💅✨ massive yassification✨💅

020

Tanya Strydom @tanyadoesscience.bsky.social · 20/03/2026

More Friday 17:00 thoughts starting to unspool 🙃

020

Tanya Strydom @tanyadoesscience.bsky.social · 26/02/2026

Sending love to 6-year-old me who was obsessed with dinosaurs 🦖🦕 Quick science PSA: not actually dinosaurs LOL — Permian marine communities 😂🙈 (Science accuracy > vibes) Still wild to be studying deep time like this... 🥹✨

1531

Tanya Strydom @tanyadoesscience.bsky.social · 26/02/2026

🚨 NEW MUD DINOSAUR SCIENCE 🦕☄️ Obsessed with seeing networks being built, smashed and stress-tested with real data - not just debated from the sidelines 👀 (self-aware self-own 🙋). 🚀 Paleoecology leveled ⬆️. Big questions + awesome squad → measurable, mind-blowning results 🤯 Here’s what went down 👇✨

031

Tanya Strydom @tanyadoesscience.bsky.social · 22/01/2026

Thanks @gracielle.science ☺️☺️☺️ I'm in my Gen-Z science comm era 🧬💻🦄

120

Tanya Strydom @tanyadoesscience.bsky.social · 22/01/2026

Shoutout time 📣: Big love to my co-authors @dralexdunhill.bsky.social , Jennifer Dunne, @ctrlalttim.com and @beckerhopper.bsky.social. Couldn’t have done it without you. 🙌💖

120

Tanya Strydom @tanyadoesscience.bsky.social · 22/01/2026

Bottom line: Networks are picky, assumptions are queen, and our framework is a cheat code. 🫡

A conceptual diagram illustrating the hierarchical transition of ecological networks. It flows from a broad "Global Metaweb" at the top (all possible interactions), down to "Regional Metawebs" (filtered by species co-occurrence in yellow and orange regions), and finally to specific "Realised Webs" at the bottom (interactions filtered by local community context like abundance and energy gain).

110

Tanya Strydom @tanyadoesscience.bsky.social · 22/01/2026

4️⃣ A unified framework improves application and progress: We built something of a 'network GPS' 🗺️🕸️ to help pick the right network for the right question. This keeps everyone from getting lost in the jungle of misinterpretation 🌳😵‍💫 and makes progress actually progress 📈🚀.

110

Tanya Strydom @tanyadoesscience.bsky.social · 22/01/2026

3️⃣ Explicit assumptions are essential for evaluation: If you don’t know what assumptions went into making your network, you’re basically guessing 🎲❌🌱. Spell them out ✍️📝 so your analysis isn’t just chaos in disguise ⚡🕸️.

110

Tanya Strydom @tanyadoesscience.bsky.social · 22/01/2026

2️⃣ Network representations are not interchangeable: There’s no one-size-fits-all 🤷‍♀️🧢. Your favorite network might be great for unicorns 🦄 but terrible for horses 🐴. Pick your network like you pick your Netflix shows 🍿📺—what you’re actually trying to watch.

110

Tanya Strydom @tanyadoesscience.bsky.social · 22/01/2026

1️⃣ Network representations encode different processes: Different networks = different vibes. Some networks are about who eats who 🍴🦌🦅, some are about who hangs out with whom 🫂🌿. Knowing the network’s 'personality' is key to asking the right questions 🧠✨.

110

Tanya Strydom @tanyadoesscience.bsky.social · 22/01/2026

Preprint alert 🚨 (yes, I still go online, shocker) Scaling from Metawebs to Realised Webs: A Hierarchical Approach to Network Ecology TL;DR: Networks are like… weirdly finicky and if you want your network analysis to not be a sad chaos, check this preprint: [doi.org/10.32942/X2JW8K] 🏠💻

193

Tanya Strydom @tanyadoesscience.bsky.social · 12/12/2025

It should be illegal to have cool science brainwave moments after 16:00 on a Friday - if only to try and keep the work-life balance somewhat respectable

020

Tanya Strydom @tanyadoesscience.bsky.social · 10/11/2025

@hadley.nz , you’re so real for this 😭🙌 Lowkey spent my whole day glowin’ up my code with the new `genzplyr` pack — it’s giving ✨clean syntax✨ fr. hadley.github.io/genzplyr/ind...

hadley.github.io

dplyr but make it bussin fr fr no cap

`genzplyr` is an alternative syntax for `dplyr` that replaces boring old function names with GenZ slang. Your data wrangling is about to hit different.

051

Reposted by Tanya Strydom

Saoirse Kelleher @saoirsekelleher.eco · 17/10/2025

Happy to announce the next edition of the @sortee.bsky.social Workshop & Webinar series on October 27 at 16:00UTC, ‘The Risks and Rewards of Large Language Models in (open) Scientific Research’ with @tanyadoesscience.bsky.social !! Sign up here: events.humanitix.com/sortee-webin...

events.humanitix.com

SORTEE Webinar: The Risks and Rewards of Large Language Models in (Open) Scientific Research

SORTEE Workshop & Webinar Series

044

Tanya Strydom @tanyadoesscience.bsky.social · 28/07/2025

Whoever decided to host the poster session in the Natural history museum you are my hero! 🤩

A photo with the claws of a dinosaur in the foreground. In the background there are people milling about and chatting

040

Tanya Strydom @tanyadoesscience.bsky.social · 28/07/2025

Stickers placed = conference mode loading 🤓🤓🤓 Looking forward to learning about DINOSAURS!!!

Photo of a name tag for a conference about palaeontology. The name tag also features a collection of stickers related to palaeontology

010

Tanya Strydom @tanyadoesscience.bsky.social · 06/06/2025

Best of both worlds? Start with the Smörgåstårta, followed by Fish n Chips (with mushy peas obvi), and round it out with the Prinsesstårta. Simple!

100

Tanya Strydom @tanyadoesscience.bsky.social · 06/06/2025

They've done it again... This time with adding (unnecessary) visual detail to the zero commit days. Inconsolable 😭😭😭

010

Tanya Strydom @tanyadoesscience.bsky.social · 06/05/2025

“Back in time for tea and medals” Celebrating the end of a fun fieldwork campaign the British way™

A photo of a cup of tea next to a plate with a selection of different biscuits as well as a tin of Fortnum & Mason breakfast blend tea

000

Tanya Strydom @tanyadoesscience.bsky.social · 25/03/2025

I feel like GitHub has (very subtly) modified the gradient they use for ones contributions calendar... I am going to need 2-4 business days to come to terms with this…

000

Reposted by Tanya Strydom

Ashkaan K. Fahimipour @akfbio.bsky.social · 19/03/2025

New paper! (arxiv.org/abs/2503.14093) I haven’t been this excited about one in a while, and I think the theory @mhab.bsky.social develops here will be a big step forward for the field. Why? Let me explain 🧵...

13216

Tanya Strydom @tanyadoesscience.bsky.social · 27/02/2025

I had a class of 4th graders and 10th graders! And yeah for sure! 🥰 Or they're asking 'big questions' and you're questioning why we as a field haven't thought of it yet!

110

Tanya Strydom @tanyadoesscience.bsky.social · 26/02/2025

Thanks @gracielle.science for passively convincing me to register! xx

110

Tanya Strydom @tanyadoesscience.bsky.social · 26/02/2025

I've just had two @skypeascientist.bsky.social sessions today, and honestly it might just be the most rewarding part of this job. 🥲 Highly recommended for anyone who has the bandwidth/time for some science outreach

1244