The biological carbon pump, responsible for the annual transfer of 5–10 Gt of carbon to the ocean's interior, plays a crucial role in long-term carbon sequestration and deep-sea ecosystem energy supply. Its efficiency is initially determined by processes within the euphotic zone, such as aggregation and grazing, followed by substantial vertical flux attenuation in the mesopelagic zone. This attenuation significantly impacts atmospheric CO2 concentrations. Traditionally, the vertical particulate organic carbon (POC) flux attenuation has been assessed using methods like multiple-depth radionuclide sampling, sediment trap deployments, and optical particle measurements, often expressed as a power-law exponent or *b* value. The *b* value serves as a comparative metric for flux attenuation across different oceanic regions, helping to explore the influence of factors like oxygen, temperature, primary productivity, and phytoplankton composition. Despite decades of research, the mechanisms controlling POC flux attenuation remain unclear, with conflicting latitudinal trends in *b* values reported in various studies. This inconsistency highlights the complexity of the process, as POC flux attenuation results from multiple interacting factors, ranging from zooplankton flux-feeding to particle-attached microbial degradation. To improve our understanding of the biological carbon pump and refine model projections, a more mechanistic approach is necessary, separating the individual contributions of these processes. This study focuses on disentangling the roles of microbes and zooplankton in determining regional patterns of particle flux attenuation.

Existing research on the biological carbon pump and POC flux attenuation reveals a complex interplay of factors. Studies have attempted to correlate flux attenuation with various environmental parameters, such as oxygen levels, temperature, primary productivity, and phytoplankton community composition. However, these studies often yielded conflicting results, with inconsistencies in latitudinal trends of the power-law exponent (*b* value) used to describe flux attenuation. This highlights the limitations of simple empirical relationships and the need for a more mechanistic understanding that separates the contributions of different processes. Previous studies have also investigated the role of microbes in POC degradation, emphasizing their importance in remineralization, recycling, solubilization, and respiration of sinking particles. However, the relative importance of microbial processes compared to zooplankton grazing and diel vertical migration (DVM) remains debated, and quantifying their contributions accurately has proven challenging due to methodological constraints.

This study employed the C-RESPIRE (Carbon RESPIRE) dual particle interceptor and incubator, a novel in situ experimental system. Unlike conventional sediment traps, C-RESPIRE intercepts settling particles that have already undergone transformations due to zooplankton and DVM. The system then incubates these particles in situ, allowing researchers to measure microbially mediated POC flux attenuation (microbial remineralization, MR) separately from the effects of other processes. Oxygen consumption was measured using optodes, and a respiratory quotient (RQ) was used to calculate MR. The residual POC flux, representing the fraction not transformed by microbes, provides indirect information about the contribution of zooplankton-mediated processes. The C-RESPIRE was deployed across six contrasting oceanic regimes, spanning a wide range of POC fluxes and environmental conditions: highly productive (South Georgia (SG) and Benguela (BEN)), high nutrient-low chlorophyll (HNLC; Subantarctic Zone (SAZ) and northeast subarctic Pacific (PAPA)), and oligotrophic (Mediterranean (MED) and South Pacific Subtropical Gyre (SPSG)). These regions differed in mesopelagic temperatures, oxygen concentrations, and phytoplankton assemblages. The study also accounted for variation in incubation times between sites due to logistical constraints. Data analysis involved fitting power-law models to the cumulative POC flux data, allowing for comparison with the classic Martin curve. Additionally, a C-specific remineralization rate (*C*remin) was calculated by normalizing MR to incubation time and cumulative POC flux, providing a metric for comparing microbial activity across sites and depths. Sensitivity analyses were conducted to assess the impact of uncertainty in RQ values on the results.

The study's key findings reveal a complex picture of POC flux attenuation. Across all six sites, the microbial contribution to POC flux attenuation ranged from 7% to 29%, indicating that zooplankton-mediated processes (flux-feeding and DVM) play a more dominant role. Microbial remineralization rates, normalized to POC flux (*C*remin), varied substantially (20-fold) across sites and depths, with lower rates observed at sites with higher POC fluxes. Vertical profiles of microbial remineralization showed distinct regional patterns. At low-latitude sites (with strong vertical temperature gradients), *C*remin decreased with depth, potentially reflecting the influence of temperature on microbial activity. However, at mid- and high-latitude sites (with less pronounced temperature gradients), *C*remin increased with depth, suggesting that other factors, such as particle biochemistry, fragmentation, and microbial ecophysiology, are more important. The cumulative POC flux generally followed the canonical Martin curve at five out of six sites, confirming the validity of the study's reconstruction approach. The C-specific remineralization rate (*C*remin) ranged widely (0.02–0.43 d−1), providing evidence that particle-attached microbes had sufficient carbon to meet their needs in the upper mesopelagic. The lack of a clear relationship between *C*remin and cumulative POC flux across sites suggests that several factors influence the rate of microbial remineralization, including depth of the primary production zone, particle lability, and biogenic versus lithogenic particle composition. The analysis of vertical trends in *C*remin revealed a distinction between low-latitude sites (where temperature was a significant factor) and mid- and high-latitude sites (where other factors were more influential).

The findings challenge the prevailing assumption that temperature is the dominant driver of POC flux attenuation globally. The study demonstrates that the relative importance of microbial remineralization in the upper mesopelagic is generally less than that of zooplankton processes. The observed regional differences in the relationship between *C*remin and depth suggest that the mechanisms controlling POC flux attenuation are more complex than previously understood and vary depending on environmental and biological factors. At low-latitude sites, temperature gradients significantly influence microbial activity, while at mid- and high-latitude sites, factors such as particle biochemistry, fragmentation, and shifts in microbial community structure are more important. The findings highlight the need for more complex biogeochemical models that account for this regional diversity in the relative contributions of microbes and zooplankton. The study suggests parallels between the oceanographic setting and soil microbiology, where substrate lability and microbial community composition interact with temperature to influence decomposition rates. Further research should focus on exploring this analogy more deeply to improve our understanding of microbial controls on POC flux attenuation.

This study provides a novel deconstruction of the Martin curve for POC flux attenuation, revealing the distinct roles of microbes and zooplankton across diverse oceanic regimes. The less than 30% contribution of microbial remineralization to overall attenuation suggests a greater influence of zooplankton. Temperature significantly impacts microbial remineralization at low-latitude sites, while other factors become dominant at mid- and high-latitude sites. These findings highlight the need for more sophisticated biogeochemical models that incorporate regional variability and the complex interplay between microbial ecology, particle biochemistry, and zooplankton grazing in shaping the biological carbon pump. Future research should focus on further elucidating the interactions between these factors and their implications for predicting the ocean's future carbon cycle.

The study's reliance on a fixed respiratory quotient (RQ) for converting oxygen consumption to carbon remineralization introduces some uncertainty. Variations in RQ values among different particle types could influence the estimates of microbial contribution to flux attenuation. The relatively short incubation times of C-RESPIRE may not fully capture the long-term microbial processes affecting POC. The indirect estimation of zooplankton contributions based on residual POC flux also limits the precise quantification of zooplankton's role in flux attenuation. While the C-RESPIRE approach represents a significant advancement, further methodological refinements could improve accuracy and reduce uncertainties.