Strength in Numbers: Red Galaxies Bolster the Cosmic Star Formation Rate Density at z > 3

A. Carnall; A. M. Koekemoer; C. Conselice; C. T. Donnan; D.J. McLeod; D. Magee; D. Scholte; F. Cullen; F. Liu; G. D. Illingworth

arxiv: 2508.05740 · v1 · submitted 2025-08-07 · 🌌 astro-ph.GA

Strength in Numbers: Red Galaxies Bolster the Cosmic Star Formation Rate Density at z > 3

L. Barrufet , J.S. Dunlop , R. Begley , S. Flury , D.J. McLeod , K. Arellano-Cordova , A. Carnall , F. Cullen

show 15 more authors

C. T. Donnan F. Liu R. McLure D. Scholte T. M. Stanton R. Cochrane C. Conselice R. Ellis P. G. P\'erez-Gonz\'alez R. Gottumukkala N. A. Grogin G. D. Illingworth A. M. Koekemoer D. Magee M. Michalowski

This is my paper

Pith reviewed 2026-05-18 23:52 UTC · model grok-4.3

classification 🌌 astro-ph.GA

keywords red galaxiesdust-reddened galaxiescosmic star formation rate densityhigh-redshift galaxiesJWST observationsquiescent galaxiesgalaxy evolutiondust-obscured star formation

0 comments

The pith

Dust-reddened red galaxies supply nearly 40 percent of the cosmic star formation rate density at z~5 and can explain the later population of quiescent galaxies.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper examines a sample of 777 red galaxies selected from JWST NIRCam and HST data using color criteria that pick out dust-reddened objects. These galaxies prove to be massive and actively star-forming, with star formation rates and masses far above the average for the full JWST population. They dominate the high-mass end of galaxies at redshifts from 3 to 7 and deliver a star formation rate density of 3.9 times 10 to the minus 2 solar masses per year per cubic megaparsec at z~4. At z~5 they provide almost 40 percent of the total cosmic star formation rate density, more than ten times the contribution from bright submillimeter galaxies. Their abundance at z~6 also matches the numbers required to produce the observed quiescent galaxies at z>3 after typical quenching over roughly one billion years.

Core claim

A sample of 777 red galaxies spanning 1 < z < 8, selected from PRIMER JWST NIRCam and HST COSMOS data with robust colour criteria for dust-reddened rest-frame UV-optical emission, shows through SED fitting that these are star-forming systems with median SFR ~40 M⊙ yr^{-1} and log(M*/M⊙) = 10.3. They dominate the high-mass end, comprising 72% of galaxies with log(M/M⊙) > 10 at z = 3.3 and rising to 91% by z ~ 7. The number density of massive red star-forming galaxies at z ~ 6 is sufficient to explain the abundance of quiescent galaxies at z > 3 given typical quenching timescales of ~1 Gyr. This abundance produces a substantial contribution to the cosmic star-formation rate density of 3.9^{+0.

What carries the argument

The sample of 777 dust-reddened red galaxies identified by robust colour criteria from JWST and HST data, with star formation rates and stellar masses derived from spectral energy distribution fitting.

If this is right

At z ~ 4 the red galaxies contribute rho_SFR = 3.9^{+0.6}_{-0.5} × 10^{-2} M_⊙ yr^{-1} Mpc^{-3}.
At z ~ 5 the red galaxies supply nearly 40 % of the total cosmic star formation rate density.
Their contribution exceeds that of bright sub(mm)-selected dusty star-forming galaxies by more than an order of magnitude.
The number density of massive red star-forming galaxies at z ~ 6 is sufficient to explain the abundance of quiescent galaxies at z > 3 after ~1 Gyr of quenching.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Deeper and wider ALMA surveys could directly detect dust emission from these red galaxies to independently verify their star formation rates.
Galaxy formation models will need to produce a larger population of massive dust-obscured star-forming systems at z > 3 to match these observations.
The red galaxies may represent a key transitional population that quenches to form the quiescent systems observed at lower redshifts.

Load-bearing premise

The color criteria used to select the 777 red galaxies from PRIMER JWST NIRCam and HST COSMOS data accurately identify dust-reddened star-forming galaxies without significant contamination from other populations or incompleteness, and the SED fitting reliably determines their star formation rates and stellar masses.

What would settle it

If spectroscopic follow-up or deeper multi-wavelength data shows that a substantial fraction of the color-selected red galaxies have low or zero ongoing star formation rates, the claimed large contribution to the cosmic star formation rate density at z > 3 would not hold.

read the original abstract

A comprehensive account of the cosmic star-formation history demands an accurate census of dust-enshrouded star formation over cosmic time. We provide strong new constraints from a large sample of 777 red galaxies, selected based on their dust-reddened, rest-frame UV-optical emission. This sample of 777 galaxies spans $1 < z < 8$ and is selected from PRIMER JWST NIRCam and HST COSMOS optical data, ensuring robust colour criteria. The SEDs indicate that these dust-reddened galaxies are star-forming, with median $\mathrm{SFR \sim 40M_{\odot}yr^{-1}}$ and stellar mass $\log(M_{*}/M_{\odot}) = 10.3^{+0.6}_{-0.8}$; each exceeds the corresponding medians of the full JWST-detected population by over two dex. Our sample thus clearly shows that red galaxies dominate the high-mass end: they comprise 72 \% of galaxies with $\log(M/M_{\odot}) > 10$ at $z = 3.3$, rising to 91\% by $z \sim 7$ (albeit with large uncertainties at the highest redshifts). Crucially, we find that the number density of massive red star-forming galaxies at $z \sim 6$ is sufficient to explain the abundance of quiescent galaxies at $z > 3$, consistent with typical quenching timescales allowed in the $\mathrm{\sim 1Gyr}$ interval from $z \sim 6$ to $z \sim 3$. This large abundance yields a substantial contribution to the cosmic star-formation rate density: at $z \sim 4$, red galaxies provide $\mathrm {\rho_{SFR} = 3.9^{+0.6}_{-0.5} \times 10^{-2} M_{\odot} yr^{-1}Mpc^{-3}}$, and at $z \sim 5$ they supply nearly 40 \% of the total $\rho_{SFR}$. This exceeds the contribution of bright sub(mm)-selected dusty star-forming galaxies by more than an order of magnitude. Future deeper and wider ALMA surveys will provide further opportunities to strengthen and extend our results in our quest to fully quantify the contribution of dust-obscured activity to $\rho_{\mathrm{SFR}}$ at high redshifts.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

This JWST paper shows red galaxies from a 777-object sample dominate the high-mass end and supply a large share of rho_SFR at z>3, but the color cuts and SED assumptions need tighter checks before the numbers can be taken as firm.

read the letter

The key point is that a sample of 777 dust-reddened galaxies selected from PRIMER JWST NIRCam plus HST COSMOS data is used to argue that these systems account for rho_SFR = 3.9^{+0.6}_{-0.5} x 10^{-2} M_sun yr^{-1} Mpc^{-3} at z~4 and nearly 40% of the total at z~5, while their number density at z~6 is enough to supply the quiescent population seen at z>3 after typical quenching times. That is the central new claim and it is quantified directly from the observed counts and summed SFRs rather than from any prior model fit. The paper does a solid job of demonstrating that the red subset sits well above the median mass and SFR of the broader JWST-detected population and that it makes up 72% of log M >10 galaxies already at z=3.3, rising further at higher redshift. Those abundance statements rest on straightforward number-density calculations from the new photometry. The soft spots sit in the mapping from observed color to intrinsic properties. The selection relies on rest-frame UV-optical cuts whose completeness and contamination rate are not validated against spectroscopy or mocks in the abstract, and the SED-derived SFRs (median ~40 M_sun yr^{-1}) depend on the chosen attenuation curve and SFH priors. If either step admits a non-negligible fraction of lower-SFR or quiescent interlopers, both the fractional rho_SFR contribution and the progenitor link weaken. Uncertainties are already flagged as large at z~7, which is appropriate. This work is aimed at people modeling the cosmic star-formation history and the build-up of massive galaxies at early times. Readers who need updated observational anchors for dust-obscured activity will find the reported fractions useful, provided they examine the methods section for the exact color thresholds and SED assumptions. It deserves a serious referee because the sample size is substantial and the implications for galaxy evolution are direct, even though the referee will need to press on the robustness of the selection and the error budget on the SFRs.

Referee Report

4 major / 3 minor

Summary. The manuscript analyzes a sample of 777 red galaxies at 1 < z < 8 selected via rest-frame UV-optical color cuts from PRIMER JWST NIRCam and HST COSMOS photometry. The authors conclude that these dust-reddened systems are star-forming (median SFR ~40 M_⊙ yr^{-1}, log M_* ~10.3), dominate the high-mass end (72% of log M > 10 at z=3.3, rising to 91% at z~7), supply ρ_SFR = 3.9^{+0.6}_{-0.5} × 10^{-2} M_⊙ yr^{-1} Mpc^{-3} at z~4 and nearly 40% of total ρ_SFR at z~5 (exceeding bright sub(mm) DSFGs by >10×), and that their number density at z~6 suffices to explain the abundance of quiescent galaxies at z>3 given ~1 Gyr quenching timescales.

Significance. If robust, the results would substantially increase the inferred contribution of dust-obscured star formation to the cosmic SFR density at z>3, implying that color-selected red galaxies are a dominant channel missed by submillimeter surveys and providing a plausible progenitor population for early quiescent systems. This would tighten constraints on the high-redshift SFR history and galaxy quenching models.

major comments (4)

[cosmic star-formation rate density section] The central ρ_SFR values and fractional contributions (abstract and § on cosmic SFR density) rest on summing SED-derived SFRs for the color-selected sample; without explicit completeness corrections, effective volume calculations, or direct comparison to a total ρ_SFR reference (e.g., from the full JWST population or literature compilations), it is unclear whether the quoted 3.9^{+0.6}_{-0.5}×10^{-2} and ~40% figures are load-bearing or sensitive to selection incompleteness.
[sample selection] The color criteria used to isolate the 777 galaxies (sample selection section) are asserted to identify dust-reddened star-forming systems, yet the manuscript provides no quantitative contamination estimates from quiescent, AGN, or low-SFR interlopers, nor validation against spectroscopic samples or mock catalogs. This directly undermines the purity assumption underlying both the SFR density claims and the number-density comparison to z>3 quiescent galaxies.
[SED fitting and physical properties] The median SFR ~40 M_⊙ yr^{-1} and stellar-mass values (SED analysis section) are obtained via SED fitting; the paper should test robustness to attenuation-curve choice and SFH priors, as systematic offsets would propagate into the high-mass dominance percentages (72% at z=3.3, 91% at z~7) and the progenitor-link argument.
[discussion of progenitor link] The claim that the number density of massive red star-forming galaxies at z~6 is sufficient to explain quiescent galaxies at z>3 (discussion section) relies on an assumed ~1 Gyr quenching timescale; a quantitative comparison with observed quiescent number densities, including Poisson errors and redshift-bin matching, is required to make this link load-bearing rather than suggestive.

minor comments (3)

[sample selection] A color-color diagram or selection boundary figure would clarify the robustness of the criteria used for the 777-galaxy sample.
[abstract and results] The abstract states that SEDs 'confirm star-forming status,' but a brief table or text quantifying the fraction of galaxies with specific SFR above a threshold would strengthen this.
[introduction and discussion] Expand citations to recent JWST studies of dusty or red high-z populations to better contextualize the novelty relative to sub(mm) DSFG work.

Simulated Author's Rebuttal

4 responses · 0 unresolved

We thank the referee for their constructive and detailed comments. We address each major point below and outline the revisions we will make to the manuscript.

read point-by-point responses

Referee: The central ρ_SFR values and fractional contributions rest on summing SED-derived SFRs for the color-selected sample without explicit completeness corrections, effective volume calculations, or direct comparison to a total ρ_SFR reference.

Authors: We agree this requires clarification. The ρ_SFR was computed by summing SFRs and dividing by the survey volume. In the revision, we will add explicit completeness corrections from mock catalogs, detail the effective volume, and compare to the total ρ_SFR from the full JWST sample and literature to confirm the robustness of our 3.9e-2 value and 40% contribution. revision: yes
Referee: The color criteria are asserted to identify dust-reddened star-forming systems, yet no quantitative contamination estimates from quiescent, AGN, or low-SFR interlopers, nor validation against spectroscopic samples or mock catalogs are provided.

Authors: We will include quantitative estimates of contamination in the revised manuscript by analyzing spectroscopic validation samples and running mock catalog tests to quantify the purity of the selection and support our assumptions. revision: yes
Referee: The median SFR and stellar-mass values from SED fitting should be tested for robustness to attenuation-curve choice and SFH priors, as offsets would affect the high-mass dominance and progenitor arguments.

Authors: We will add robustness tests using alternative attenuation curves and SFH priors. The revised paper will demonstrate that the median SFR of ~40 M_sun/yr, masses, and dominance fractions are robust to these choices. revision: yes
Referee: The progenitor link claim relies on an assumed ~1 Gyr quenching timescale without a quantitative comparison to observed quiescent number densities including errors and bin matching.

Authors: We will provide a quantitative comparison in the discussion, including Poisson errors on number densities and redshift bin matching to literature quiescent samples, to make the link more robust. revision: yes

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper derives its ρ_SFR values, fractional contributions, and number-density comparisons directly from color-based selection of 777 galaxies in new PRIMER JWST NIRCam + HST COSMOS photometry, followed by SED fitting for SFR and M* and straightforward summation/counting. No equations or steps reduce by construction to fitted inputs renamed as predictions, self-definitions, or load-bearing self-citations; the central claims rest on empirical measurements of the observed sample rather than internal loops equivalent to the inputs.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 0 invented entities

The central claims rest on the accuracy of photometric selection criteria, SED modeling assumptions for deriving SFR and masses, and standard conversions from observed fluxes to physical quantities using a cosmological model.

free parameters (1)

Color selection thresholds for red galaxies
Specific color cuts are applied to identify dust-reddened objects; these thresholds are chosen based on expected reddening but may involve some optimization to the data.

axioms (2)

standard math Standard flat Lambda-CDM cosmology for distance and volume calculations
Invoked when converting redshifts to comoving volumes and number densities.
domain assumption Assumptions in SED fitting including dust attenuation laws and star formation history templates
Used to derive SFR and stellar mass from photometry; these are standard but model-dependent.

pith-pipeline@v0.9.0 · 6111 in / 1743 out tokens · 60505 ms · 2026-05-18T23:52:54.995947+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Foundation/RealityFromDistinction.lean reality_from_one_distinction unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

red galaxies provide ρ_SFR = 3.9^{+0.6}_{-0.5} × 10^{-2} M_⊙ yr^{-1} Mpc^{-3} at z∼4

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.