Abstract. Summertime mixed-layer drawdown of dissolved inorganic carbon in the absence of measurable nutrients in the ocean's subtropical gyres and non-Redfieldian oxygen : nitrate relationships in the underlying subsurface waters are two biogeochemical phenomena that have thus far eluded complete description. Many processes are thought to contribute to one or both, including lateral nutrient transport, carbon overconsumption or non-Redfield C:N:P organic matter cycling, heterotrophic nutrient uptake, and the actions of vertically migrating phytoplankton. To obtain insight into the likely magnitude of potential contributing mechanisms that can remove nitrate from the nutricline while supporting dissolved inorganic carbon (DIC) drawdown tens of meters higher in the water column, we investigated the seasonal formation rates for negative preformed nitrate (preNO3) anomalies (oxygen consumption without stoichiometric nitrate release) in the subsurface and positive preformed nitrate anomalies (oxygen production without stoichiometric nitrate drawdown) in the euphotic zone at the subtropical ocean time series stations ALOHA (A Long-Term Oligotrophic Habitat Assessment) in the North Pacific and BATS (Bermuda Atlantic Time-series Study) in the North Atlantic. Non-Redfield -O2:N stoichiometry for dissolved organic matter (DOM) remineralization accounts for up to ∼15 mmol N m−2 yr−1 of negative preNO3 anomaly formation at both stations. We present a new formulation for calculating preNO3 (residual preNO3) that includes components resulting from non-Redfield DOM cycling. Residual negative preNO3 anomalies in excess of that which can be accounted for by non-Redfield DOM cycling are found to accumulate at a rate of ∼32–46 mmol N m−2 yr−1 at Station ALOHA and ∼46–87 mmol N m−2 yr−1 at the BATS station. These negative anomaly formation rates are in approximate balance with residual positive preNO3 anomaly formation rates from the euphotic zone located immediately above the nutricline in the water column. We evaluate three mechanisms to explain these anomalies, calculating that transparent exopolymer particle (TEP) cycling and heterotrophic nitrate uptake can contribute to the formation of both residual preNO3 anomalies. However, a significant fraction, estimated at ∼50 %–95 %, is unexplained by the sum of these processes. Vertically migrating phytoplankton possess the necessary distribution, nutrient acquisition strategy, and biogeochemical signature to simultaneously remove nitrate at depth and transport it above the nutricline. Reported transport rates by known migrators equal or exceed the residual preNO3 anomaly formation rates and potentially explain both the negative and positive residual preNO3 anomalies as well as the mixed-layer DIC drawdown at the stations ALOHA and BATS within the limits of scarce detailed abundance profiles. However, the three processes examined are not independent and mutually exclusive. The model Rhizosolenia mat system (and perhaps other migrators) produces TEPs, suggesting that migration could provide accelerated vertical transport of TEPs and provide labile carbon for heterotrophic nitrate uptake. These results based on geochemical distributions suggest that, in the absence of additional mechanisms and rates, phytoplankton vertical migrators, although rare and easily overlooked, play a larger role in subtropical ocean nutrient cycling and the biological pump than generally recognized.