Mark Bitter

ecoevorxiv.orgCommon, consequential, and actively maintained: a neo-balance view of genetic variation

New Perspective w/ @petrovadmitri.bsky.social and Paul Schmidt! We ask what evolution-in-action studies tell us about an old question: what genetic variation matters for fitness, and what maintains it? ecoevorxiv.org/repository/v...

Mark Bitter reposted

Matt Sasaki

@mattsasaki.bsky.social

Our physiological time series for the calanoid copepods of Lake Champlain is out today in Ecology! We spent a year tracking upper thermal limits, showing distinct differences between species as well as pretty substantial variation within species. doi.org/10.1002/ecy....

Mark Bitter reposted

Shaili Mathur

@shaili-mathur.bsky.social

I’m excited to share our new preprint, where we ask: to what extent can evolution in fluctuating environments be predicted? Spoiler: far more than the strong history-dependence of fitness would suggest! www.biorxiv.org/content/10.6...

biorxiv.orgOrder invariance of fitness makes evolution predictable despite strong history-dependent effectsEnvironmental change is ubiquitous in nature, yet understanding its effect on the predictability of evolution remains a challenge. Previous work in microbes has shown that growth and fitness in a part...

Mark Bitter reposted

Max Reuter

@maxreuter.bsky.social

Our paper on detecting (or not) balancing selection is now out, in all its typeset beauty, in the October issue of @molbioevol.bsky.social: academic.oup.com/mbe/article/...

Max Reuter@maxreuter.bsky.social

A new preprint from the lab, with postdoc @deboraycb.bsky.social and collaborators @aidaandres.bsky.social and Tim Connallon: “Characterising the detectable and invisible fractions of genomic loci under balancing selection” www.biorxiv.org/content/10.6...

biorxiv.orgbiorxiv.org

Mark Bitter reposted

Tom Murphy 🏳️‍🌈🇵🇸

@thommurphy.bsky.social

the Perspective is a super interesting read. I haven't waded into the arcana of evolutionary theory since my days as an EEB major in college but this new data is a tremendously interesting intervention in the field and really shows how technology and statistics can overdetermine questions of theory

Mark Bitter @markcbitter.bsky.social

ecoevorxiv.orgCommon, consequential, and actively maintained: a neo-balance view of genetic variation

ecoevorxiv.orgCommon, consequential, and actively maintained: a neo-balance view of genetic variation

We are grateful to the many colleagues who gave us feedback on this piece, to John Kelly for suggesting the term “neo-balance,” and to all members, past and present, of the Petrov and Schmidt labs!

Our hope with this Perspective is to open a conversation. We would love to hear what you think and what remains to be done!

ecoevorxiv.orgCommon, consequential, and actively maintained: a neo-balance view of genetic variation

We also want to know whether these alleles are old or recent, whether they are broadly pleiotropic, whether they respond to specific conditions or to general ecological shifts, and how natural selection coarse-grains the environment to generate recurrent fluctuating selection.

And we need theory for a regime where many large-effect loci experience fluctuating selection at the same time, and for how populations can maintain them without an unsustainable genetic load.

ecoevorxiv.orgCommon, consequential, and actively maintained: a neo-balance view of genetic variation

Where should the field go? We argue that time series data have a lot to teach us. We should use them to find and fine-map causal variants and then understand their functional and fitness effects.

We need to understand what stabilizes common large-effect alleles, as fluctuating selection by itself is likely not enough. Many mechanisms are possible and we need data to tell which operate: spatially varying selection, dominance reversal, storage effects, frequency dependence?

ecoevorxiv.orgCommon, consequential, and actively maintained: a neo-balance view of genetic variation

And how important is fluctuating selection on large-effect loci for demographic inference, human disease risk, how species will respond to environmental change, and patterns like Lewontin’s paradox?

Is balancing selection of this kind common only in some species, say those with very large N, or is it pervasive, as the human ancient DNA data suggest?

ecoevorxiv.orgCommon, consequential, and actively maintained: a neo-balance view of genetic variation

We propose a “neo-balance” framework. It keeps most of the neo-classical model (purifying selection, positive selection, drift, linked selection) but gives balancing selection a substantial role in maintaining common, large-effect variation.

Many questions remain. For instance, how do we square GWAS findings of mostly modest effects on any one trait with the large allele frequency fluctuations, and thus strong selection, that we see in time series?

ecoevorxiv.orgCommon, consequential, and actively maintained: a neo-balance view of genetic variation

We also discuss similar findings from midges (doi.org/10.1093/icb/icac098), Arabidopsis in outdoor experiments (doi.org/10.1126/science.adz0777), and the Florida Scrub-Jay (doi.org/10.1101/2025.09.09.673786).

In many of these systems, selection on the same variants changes in strength and often in direction over time. This is what you expect if fitness effects depend on context, which is the core premise of balance theory. And you can only see it by measuring allele frequencies as evolution happens.

ecoevorxiv.orgCommon, consequential, and actively maintained: a neo-balance view of genetic variation

Across very different organisms and timescales, these studies find the same thing: hundreds of common polymorphisms with large selection coefficients (often 10% or more). Under constant directional selection such variants should fix within hundreds of generations, but they have not.

Examples in the paper: Daphnia in a pond over 10 years (a; doi.org/10.1073/pnas.2307107121), Drosophila across seasons (b; doi.org/10.1371/journal.pgen.1004775), monkeyflowers over 23 years (c; doi.org/10.1002/evl3.308), and humans via ancient DNA (d; doi.org/10.1038/s41586-026-10358-1).

ecoevorxiv.orgCommon, consequential, and actively maintained: a neo-balance view of genetic variation

And in species w/ high recombination, like Drosophila, even old balanced polymorphisms leave a signal in a region too small to detect. Brandt et al. (doi.org/10.64898/2026.01.13.698512) showed w/ simulations that over two-thirds of long-term balanced loci are missed even under favorable conditions.

doi.orgdoi.org

Recently a different approach has become feasible: tracking allele frequencies as populations evolve in their natural settings. This lets us measure selection on individual variants directly, without relying on signatures that may have decayed or not yet accumulated.

ecoevorxiv.orgCommon, consequential, and actively maintained: a neo-balance view of genetic variation

There are classic cases of balancing selection (sickle-cell/malaria, G6PD, MHC), but evidence that it is a major force genome-wide and across taxa has been lacking. We think this may be a problem of power, not biology.

The main static signature of balancing selection is an allele that stays at intermediate frequency for longer than expected under neutrality. In humans this means at least ~1 million years. Balancing selection that started more recently leaves no such signature at all.

ecoevorxiv.orgCommon, consequential, and actively maintained: a neo-balance view of genetic variation

If balance theory is right, many common polymorphisms should have large fitness effects, be maintained by selection, and switch between being advantageous and deleterious depending on context. This can happen through selection varying in time or space, frequency dependence, dominance reversals, etc.

This debate has mostly been tested by inference from static, contemporary DNA sequences. This work established the central roles of purifying selection and drift and showed that positive selection is common. Scans for balancing selection found much less.

ecoevorxiv.orgCommon, consequential, and actively maintained: a neo-balance view of genetic variation

The "neo-classical" view, today’s standard model, holds that each functional locus has a single best allele. Selection mostly removes variation: purifying selection purges deleterious mutations, and positive selection occasionally fixes beneficial ones. Most common variants are effectively neutral.

The main alternative is “balance theory,” championed by Dobzhansky. It rejects the idea of a single best allele. At many loci alleles are beneficial in some contexts and deleterious in others, and selection actively maintains them.

ecoevorxiv.orgCommon, consequential, and actively maintained: a neo-balance view of genetic variation

With this paper, it was really fun to step back and think about the bigger picture: what do the strong selection and rapid adaptation we've quantified in D. melanogaster mean for the nature of genetic variation, and for fields like demographic inference, GWAS, and conservation genomics?

The Perspective centers on two competing views that fueled fierce debate throughout the mid-20th century (Muller vs. Dobzhansky, among many others). That debate was never fully settled. In brief:

ecoevorxiv.orgCommon, consequential, and actively maintained: a neo-balance view of genetic variation

doi.orgContinuously fluctuating selection reveals fine granularity of adaptation - NatureNatural environmental and ecological shifts impose sufficiently strong selection to drive exceptionally rapid, parallel and fluctuating adaptive tracking in a Drosophila melangaster mesocosm.

I've spent ~6 years quantifying selection in genetically diverse D. melanogaster populations (e.g., doi.org/10.1038/s41586-024-07834-x), building on work by A. Bergland, S. Rajpurohit, @heather-machado.bsky.social, @greensi.bsky.social, S. Rudman, @emilylbehrman.bsky.social, and M. Wittmann.