Environmental dependence of the Mass-Metallicity Star Formation Relations at z=4-10 with JWST

Caio Moreira Goolsby; Chandana Hegde; Christopher J. Conselice; Duncan Austin; James Arcidiacono; Lewi Westcott; Nathan Adams; Qiong Li; Shuqi Fu; Tom Harvey

arxiv: 2604.25632 · v1 · submitted 2026-04-28 · 🌌 astro-ph.GA

Environmental dependence of the Mass-Metallicity Star Formation Relations at z=4-10 with JWST

Qiong Li , Christopher J. Conselice , Lewi Westcott , Duncan Austin , Tom Harvey , Nathan Adams , Vadim Rusakov , James Arcidiacono

show 3 more authors

Caio Moreira Goolsby Chandana Hegde Shuqi Fu

This is my paper

Pith reviewed 2026-05-07 15:45 UTC · model grok-4.3

classification 🌌 astro-ph.GA

keywords galaxy environmentmass-metallicity relationhigh-redshift galaxiesJWSTstar formation ratechemical enrichmentreionizationgalaxy evolution

0 comments

The pith

Environment accelerates both star formation and chemical enrichment in galaxies at z=4-10.

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

The paper investigates how galaxy surroundings shape the mass-metallicity relation and star-formation activity in the early universe using JWST data from the CEERS and JADES surveys. It compiles measurements for 225 galaxies at redshifts 4 to 10 and defines environment through the projected surface density of the fifth-nearest neighbor within a redshift window of plus or minus 0.25. Galaxies in denser regions appear more metal-rich by 0.1 to 0.2 dex at fixed stellar mass, with the offset growing when star-formation rate is included, and the overall star-formation rate density rising by a factor of two to three at z around 6 to 9. The mass-metallicity slopes stay comparable across environments, while the sample as a whole sits below the local fundamental metallicity relation with a reduced deficit in dense areas. These patterns indicate that crowded environments promote faster metal buildup and star production during reionization.

Core claim

At 4.5<z<7 galaxies in dense regions are more metal-rich at fixed stellar mass by 0.1-0.2 dex while MZR slopes remain similar across environments; including SFR widens the separation and the full sample lies 0.2-0.3 dex below the local Te-based FMR with a smaller deficit in overdense regions. Metallicity rises weakly with effective radius up to 1 kpc then flattens with modest residual trends at fixed mass. A positive age-metallicity relation appears in both environments and is steeper in the field. Star-formation rate density is higher by a factor of 2-3 in overdense regions at z=6-9. The results indicate that environment accelerates both star formation and chemical enrichment during the re-

What carries the argument

Projected fifth-nearest-neighbour surface density Σ5 within Δz=±0.25, used to separate dense and field environments and to correlate with gas-phase metallicity, SFR, and stellar mass.

If this is right

Chemical enrichment proceeds more efficiently in overdense regions at fixed mass and SFR.
Star-formation rate density rises by a factor of 2-3 in dense environments at z=6-9.
The fundamental metallicity relation offset from the local value is smaller in overdense regions.
Metallicity shows only modest residual dependence on galaxy size once mass and environment are fixed.
A positive age-metallicity trend exists at 5<z<10 and is steeper for field galaxies.

Where Pith is reading between the lines

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

The environmental boost to enrichment could shorten the time needed for galaxies to reach the metallicities required for efficient cooling and further star formation.
Patchy reionization might be influenced if dense regions complete their metal enrichment and ionizing-photon production earlier than field regions.
Simulations that seed early galaxies with realistic large-scale density fields should reproduce an offset in the high-redshift MZR.
ALMA follow-up on molecular-gas content in the same dense versus field samples could test whether the accelerated enrichment is accompanied by higher gas fractions.

Load-bearing premise

The projected fifth-nearest-neighbour surface density within Δz=±0.25 accurately captures true three-dimensional environmental density without major projection effects or redshift uncertainties biasing the classification.

What would settle it

No metallicity offset between dense and field galaxies when the same sample is re-analyzed with spectroscopic redshifts of higher precision or a true three-dimensional density estimator.

Figures

Figures reproduced from arXiv: 2604.25632 by Caio Moreira Goolsby, Chandana Hegde, Christopher J. Conselice, Duncan Austin, James Arcidiacono, Lewi Westcott, Nathan Adams, Qiong Li, Shuqi Fu, Tom Harvey, Vadim Rusakov.

**Figure 1.** Figure 1: Spectroscopic redshift distributions of galaxies in the CEERS and JADES samples used in this work. The stacked histograms show the number of galaxies as a function of redshift for the two JWST fields, with CEERS shown in blue and JADES in red. 25th percentile, and overdense (protocluster-like) if Σ5 is above the 75th percentile. Galaxies in the intermediate percentiles are not used in the primary environme… view at source ↗

**Figure 4.** Figure 4: Bootstrap–MCMC posterior distribution of the intercept difference Δ𝑏 = 𝑏cluster −𝑏field after 3𝜎 clipping. The median (Δ𝑏 = 0.14 dex) and 95% credible interval are indicated by the red line and shaded region, respectively. 3.2 Environmental dependence of the mass–metallicity–SFR relation While the mass–metallicity relation provides a first-order description of chemical enrichment, it does not fully captur… view at source ↗

**Figure 6.** Figure 6: presents the distribution of galaxies in the log(SFR)– log(𝑀★/𝑀⊙) plane, color-coded by environment. Field (low-density) galaxies are shown in blue, while protocluster (high-density) members are shown in red. Individual error bars indicate measurement uncertainties in both stellar mass and SFR. The grayscale color bar encodes the gas-phase metallicity, 12 + log(O/H). At fixed 𝑀★ and SFR, galaxies in overd… view at source ↗

**Figure 7.** Figure 7: Left: Gas-phase metallicity (12 + log(O/H)) as a function of stellar mass, color-coded by star formation rate (SFR) measured within a sliding window of width Δ log(SFR) = 0.8 dex. Right: The same metallicity as a function of log(SFR), color-coded by stellar mass within a sliding window of width Δ log(𝑀★/𝑀⊙ ) = 1.2 dex. Curves represent the median metallicities computed in overlapping moving bins (±0.2 dex … view at source ↗

**Figure 8.** Figure 8: shows the resulting offsets Δ𝑍 as a function of redshift, compared with literature data and predictions from hydrodynamical TNG simulations (Garcia et al. 2025). We find that our full sample (purple) lies on average Δ𝑍 ≃ −0.25 dex below the local Andrews & Martini (2013) relation, in broad agreement with the metal dilution reported at 𝑧 ≳ 4–6 in recent JWST studies (e.g. Heintz et al. 2024; Curti et al. 20… view at source ↗

**Figure 9.** Figure 9: (a) Metallicity–size relation and (b) mass-controlled residual metallicity–size relation for CEERS and JADES galaxies. is largely driven by the mutual correlation of both metallicity and size with stellar mass, with galaxy size introducing at most a weak secondary modulation. This environmental contrast indicates that chemical enrichment in the field remains more susceptible to processes related to gas acc… view at source ↗

**Figure 10.** Figure 10: Gas-phase metallicity as a function of the stellar mass surfacedensity log10 (𝑀∗/𝑅 2 𝑒 ) for galaxies. Blue circles show field (low-density) galaxies and red squares show protocluster (high-density) galaxies. Error bars include uncertainties in 12+log(O/H) and the propagated uncertainties in log10 (𝑀∗/𝑅 2 𝑒 ). Large pentagons mark binned means in log10 (𝑀∗/𝑅 2 𝑒 ) with the shaded bands indicating the sta… view at source ↗

**Figure 11.** Figure 11: Age–metallicity relation for galaxies in low- and high-density environments in the CEERS and JADES fields. Blue circles represent field galaxies (low-density environments), while red squares denote protocluster galaxies (high-density environments). The solid lines show the best-fit relations obtained from orthogonal distance regression in the log10 (𝑡mw/Gyr)– metallicity plane, accounting for uncertainti… view at source ↗

**Figure 12.** Figure 12: Environmental dependence of the mass–metallicity relation in the SPHINX simulations. Panels show the stellar mass versus gas-phase metallicity, expressed as 12+log(O/H), derived using different nebular-line-based metallicity indicators. Galaxies in overdense and underdense environments are shown by red squares and blue circles, respectively. Small symbols represent individual galaxies, while large symbols… view at source ↗

**Figure 13.** Figure 13: Mass–metallicity relation (MZR) comparison between SPHINX simulations and JWST observations (CEERS and JADES) at 𝑧 ∼ 4–7 in different environments. Blue circles and red squares show the observational results for underdense (field) and overdense (protocluster) galaxies, respectively. Grey circles and squares show the corresponding SPHINX simulation results. Solid lines indicate the best-fitting linear rel… view at source ↗

read the original abstract

We study how environment affects the mass-metallicity relation (MZR) at $z=4$-$10$ using deep imaging and spectroscopy from the James Webb Space Telescope (JWST). Combining CEERS and JADES, we compile a sample of 225 galaxies with stellar masses, star-formation rates, and gas-phase metallicities. We characterize environment using the projected fifth-nearest-neighbour surface density, $\Sigma_{5}$, within $\Delta z=\pm0.25$. At $4.5<z<7$, we find that galaxies in dense regions are more metal-rich at fixed $M_\star$ by $\sim0.1$-0.2 dex, while the slopes of the MZR remain similar across environments. Including SFR increases the separation, suggesting more efficient chemical enrichment in overdense regions. Compared to the local $T_e$-based FMR, our full sample lies $\simeq0.2$-0.3 dex below the $z\sim0$ relation, with a smaller deficit in overdense environments. We also examine how metallicity relates to galaxy size using NIRCam-based effective radii. Metallicity increases weakly with size up to $R_e\sim1$ kpc and then flattens, with only a modest residual trend at fixed $M_\star$ and little environmental dependence. Using mass-weighted stellar ages at $5<z<10$, we find a positive age-metallicity relation in both environments, steeper in the field. Finally, we find that the star-formation rate density is higher in overdense regions at $z\simeq6$-9 by a factor of $\sim2$-3. Overall, our results suggest that environment accelerates both star formation and chemical enrichment during the epoch of reionization. Future wide-area JWST spectroscopy, combined with ALMA and Euclid, will better constrain the role of environment in early galaxy evolution.

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

The paper reports modest environmental offsets in metallicity and SFRD at z=4-10 but the projected density classifier is too loose to support the main claim.

read the letter

The main thing to know is that this work finds galaxies in higher projected density regions at z=4-10 are metal-richer by 0.1-0.2 dex and show 2-3 times higher star-formation rate density, with the offset growing when SFR is included. That is the headline result from the JWST CEERS+JADES sample of 225 galaxies. What the paper actually does is extend the mass-metallicity and fundamental metallicity relations to these redshifts while splitting the sample by local environment, then add checks on metallicity versus galaxy size and on the age-metallicity relation. The data compilation from new JWST spectroscopy and the multi-angle look at the same galaxies are the concrete advances; no prior work cited in the abstract had done the environmental split at z>4 with these tracers. The trends are presented as direct observational comparisons rather than model fits, which keeps the circularity low. The age-metallicity slope difference between environments and the SFRD enhancement are the parts that feel most worth following up. The soft spot is the environment metric. They use projected fifth-nearest-neighbor surface density inside a Δz=±0.25 window, which at z~6 spans 150-200 Mpc along the line of sight. That scale is far larger than protocluster sizes, so a non-negligible fraction of the high-Σ5 galaxies are likely interlopers. The abstract gives no mock tests for photo-z scatter or spectroscopic validation of the neighbors, and the reported offsets are small enough that projection contamination could create or erase them. Without those checks the central interpretation that environment accelerates enrichment and star formation rests on an untested assumption. Sample completeness and metallicity calibration systematics are also not quantified in the summary, though that is secondary to the density issue. This is for people who track high-redshift galaxy assembly and reionization models. A reader who needs the latest JWST trends on MZR and SFRD at z>4 will get usable numbers to compare against simulations, even if the environmental part requires caution. The work is coherent on its own terms and uses real new data on a question that matters inside the subfield, so it deserves a serious referee rather than a desk reject. The density proxy and its validation should be the main point of scrutiny in review.

Referee Report

2 major / 2 minor

Summary. The paper analyzes JWST CEERS and JADES data for a sample of 225 galaxies at z=4-10 to study environmental effects on the mass-metallicity relation (MZR), fundamental metallicity relation (FMR), galaxy sizes, stellar ages, and star-formation rate density (SFRD). Environment is defined using the projected fifth-nearest-neighbor surface density Σ5 within Δz=±0.25. Key results include 0.1-0.2 dex higher metallicities in dense regions at fixed M*, increased separation when SFR is included, a smaller FMR deficit relative to z~0 in overdense areas, weak positive metallicity-size trends with little environmental dependence, steeper age-metallicity relations in the field, and 2-3× higher SFRD in overdense regions at z~6-9. The authors conclude that environment accelerates both star formation and chemical enrichment during reionization.

Significance. If the environmental classifications prove robust, the work would offer timely empirical constraints on how overdense regions influence early galaxy chemical enrichment and star formation at z>4, extending local scaling relations into the epoch of reionization with a multi-faceted JWST dataset. The direct observational trends, inclusion of SFR and age diagnostics, and comparison to the local Te-based FMR are strengths that could inform models of protocluster evolution. The sample compilation from two major JWST programs adds value for the field.

major comments (2)

[§3 (Environment Characterization)] §3 (Environment Characterization): The projected Σ5 metric is computed within Δz=±0.25. At z≈6 this slice spans ~150-200 Mpc comoving along the line of sight, far exceeding typical protocluster scales of 10-30 Mpc. Without explicit validation against mocks that include realistic photometric redshift scatter for the neighbor sample, or cross-checks with spectroscopic redshifts, a substantial fraction of high-Σ5 galaxies may be interlopers. This directly undermines the load-bearing assumption that 'dense' classifications trace genuine 3D overdensities, potentially diluting or creating the reported 0.1-0.2 dex metallicity offset and factor 2-3 SFRD enhancement.
[Results (MZR, FMR, and SFRD sections)] Results (MZR, FMR, and SFRD sections): The abstract and trends are presented without reported error bars on the 0.1-0.2 dex offsets, without quantified sample completeness or selection functions, and without tests for systematics in the gas-phase metallicity calibrations. The statement that 'including SFR increases the separation' is post-hoc and unaccompanied by any assessment of covariance or selection bias between SFR and environment. These omissions are load-bearing for the central claim that environment accelerates enrichment and star formation.

minor comments (2)

[Abstract] The abstract would be clearer if it specified the exact subsample sizes for dense versus field galaxies and the precise redshift intervals used for each quantitative result (e.g., the SFRD factor at z=6-9).
[Figures] Figures showing MZR and SFRD comparisons should include error bars, bin occupancies, and explicit labels for the environmental subsamples.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed and constructive report. The comments highlight important limitations in our environment characterization and the presentation of quantitative results. We address each point below and will incorporate revisions to improve clarity and robustness.

read point-by-point responses

Referee: The projected Σ5 metric is computed within Δz=±0.25. At z≈6 this slice spans ~150-200 Mpc comoving along the line of sight, far exceeding typical protocluster scales of 10-30 Mpc. Without explicit validation against mocks that include realistic photometric redshift scatter for the neighbor sample, or cross-checks with spectroscopic redshifts, a substantial fraction of high-Σ5 galaxies may be interlopers. This directly undermines the load-bearing assumption that 'dense' classifications trace genuine 3D overdensities, potentially diluting or creating the reported 0.1-0.2 dex metallicity offset and factor 2-3 SFRD enhancement.

Authors: We agree that the Δz=±0.25 slice introduces significant line-of-sight projection, a standard but imperfect approach given the photometric redshift uncertainties (typically σ_z/(1+z) ≈ 0.05–0.1) in the CEERS and JADES catalogs. This depth is chosen to ensure sufficient neighbor statistics while remaining within the redshift range where the sample is complete. We will revise §3 to explicitly discuss this limitation, including a quantitative estimate of interloper fraction based on the observed redshift distribution and a comparison of Σ5 values for the subset of galaxies with spectroscopic redshifts. We will also add a statement that the observed trends represent lower limits due to potential dilution. Full end-to-end mock validation with realistic photo-z scatter is beyond the scope of the current analysis but will be noted as future work. These changes constitute a partial revision. revision: partial
Referee: The abstract and trends are presented without reported error bars on the 0.1-0.2 dex offsets, without quantified sample completeness or selection functions, and without tests for systematics in the gas-phase metallicity calibrations. The statement that 'including SFR increases the separation' is post-hoc and unaccompanied by any assessment of covariance or selection bias between SFR and environment. These omissions are load-bearing for the central claim that environment accelerates enrichment and star formation.

Authors: We accept these criticisms and will revise the manuscript accordingly. Error bars on the metallicity offsets will be added to all relevant figures and text using bootstrap resampling of the sample. A new subsection will quantify sample completeness as a function of stellar mass, redshift, and environment, based on the magnitude limits of the parent catalogs. We will include a brief assessment of metallicity calibration systematics, referencing the specific strong-line diagnostics employed and any consistency checks with alternative calibrations. The discussion of SFR inclusion will be rephrased to present it as a direct result of the analysis, accompanied by partial correlation coefficients and joint SFR–environment distributions to address covariance. These additions will be made throughout the results sections. revision: yes

Circularity Check

0 steps flagged

No circularity: purely observational data trends with no derivation reducing to inputs by construction.

full rationale

The paper compiles JWST-derived measurements (M*, SFR, gas-phase metallicity, projected Σ5 within Δz=±0.25) for 225 galaxies and reports direct empirical comparisons: metallicity offsets at fixed M*, SFRD enhancement in high-Σ5 regions, and age-metallicity trends. No equations, models, or 'predictions' are derived; all results are stated as measured trends from the data. No self-citation chain, fitted-parameter renaming, or ansatz smuggling is present. The central claim (environment accelerates enrichment and star formation) rests on the observational classification and comparison, which is independent of any internal reduction. This matches the expected non-circular outcome for a data-driven observational study.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claim rests on standard assumptions about spectroscopic metallicity indicators and the fidelity of projected nearest-neighbor density as an environment tracer; no free parameters or new entities are introduced in the abstract.

axioms (2)

domain assumption Gas-phase metallicities derived from JWST spectroscopy are reliable and comparable across different environments at z=4-10
Invoked when reporting 0.1-0.2 dex offsets between dense and field regions
domain assumption Projected fifth-nearest-neighbour surface density within a redshift slice of Δz=±0.25 traces true local environment without significant line-of-sight contamination
Used to define dense versus field samples

pith-pipeline@v0.9.0 · 5700 in / 1408 out tokens · 36622 ms · 2026-05-07T15:45:29.313214+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

1 extracted references · 1 canonical work pages

[1]

B., et al., 2017, A&A, 600, A90 Caminha G

Adachi K., Kodama T., Pérez-Martínez J. M., Suzuki T. L., Onodera M., 2025, ApJ, 994, 4 Adams N. J., et al., 2024, ApJ, 965, 169 Andrews B. H., Martini P., 2013, ApJ, 765, 140 Asplund M., Grevesse N., Sauval A. J., Scott P., 2009, ARA&A, 47, 481 Astropy Collaboration et al., 2013, A&A, 558, A33 Astropy Collaboration et al., 2018, AJ, 156, 123 Astropy Coll...

work page doi:10.5281/zenodo.6825092 2025

[1] [1]

B., et al., 2017, A&A, 600, A90 Caminha G

Adachi K., Kodama T., Pérez-Martínez J. M., Suzuki T. L., Onodera M., 2025, ApJ, 994, 4 Adams N. J., et al., 2024, ApJ, 965, 169 Andrews B. H., Martini P., 2013, ApJ, 765, 140 Asplund M., Grevesse N., Sauval A. J., Scott P., 2009, ARA&A, 47, 481 Astropy Collaboration et al., 2013, A&A, 558, A33 Astropy Collaboration et al., 2018, AJ, 156, 123 Astropy Coll...

work page doi:10.5281/zenodo.6825092 2025