Characterization of phytoplankton productivity and bio-optical variability in a polar marine ecosystem

Highlights

• Primary production (PP) was linked to in-situ absorption properties in the ISSO.

• Phytoplankton absorption (a_ph) was a better predictor of PP than chlorophyll-a.

• ISSO PP models may need to account for intra-regional variation in a_ph and a*_ph.

Abstract

Knowledge of Southern Ocean carbon cycling is limited by a paucity of phytoplankton primary productivity (PP) and spectral absorption data in this globally-important region. We measured ¹³C-based PP in the Indian sector of Southern Ocean (ISSO) during austral summer 2017, examining its link with spectral absorption coefficients and phytoplankton size structure derived from an absorption-based global model. Phytoplankton productivity was assessed at both coastal (60°S-69°S) and frontal stations (40°S-60°S), characterized by silicate- replete and -deplete water masses, respectively (indicated by measured nutrient ratios) to capture a range of phytoplankton growth conditions. Bio-optical relationships were used as indicators of phytoplankton community size structure and to assess the extent of cellular pigment packaging - a phenomenon reported previously for phytoplankton in this region. Blue-Red (B/R) ratios of phytoplankton absorption (a_ph) spectra indicated that microphytoplankton (more prone to “package effects”) were the dominant size class at most sites sampled. Overall, PP was better explained by a_ph (R² = 0.85) than total chlorophyll-a (R² = 0.64) in surface waters. The a*_ph (675)-chlorophyll-a relationship explained package effects more effectively in frontal regions (R² = 0.63) than stations further south (R² = 0.30). The global absorption-based model captured smaller (pico, nano) phytoplankton size classes but failed to identify larger microphytoplankton, underscoring the need for region-specific algorithm modifications. Our findings improve existing understanding of spatio-temporal trends in PP and bio-optical variability within the Indian Sector of the Southern Ocean (ISSO) – knowledge that is essential to improve capacity to retrieve PP from satellite-based models in this region.

Introduction

Phytoplankton primary production (PP) accounts for approximately 50% of global carbon fixation and therefore plays a crucial role in regulating the climate. A significant proportion of this productivity occurs in the Southern Ocean (SO) – making this region an increasing hotspot of interest for marine scientists due to its major contribution to oceanic uptake of anthropogenic CO₂ via the “biological pump” (Gruber et al., 2019). However, as a classic high-nutrient, low-chlorophyll (HNLC) region (Moore and Abbott, 2000), productivity in the Southern Ocean is largely underpinned by intense phytoplankton blooms that are highly-variable in both space and time (Arrigo et al. 2008a). To date, this “patchy” nature of primary productivity, combined with a shortage of in-situ observations due to its remote location and extreme weather, has hindered efforts to understand and predict carbon cycling in the Southern Ocean.

Phytoplankton productivity ultimately depends on absorption and utilisation of light by photosynthetic light-harvesting pigments including Chlorophyll-a (Chl-a). As such, development of remote-sensing models based on ocean colour (Clementson et al., 2001) has permitted investigation of physical and biological control over PP with relative ease. Indeed, bio-optical measures of phytoplankton absorption are often regarded as better predictors of PP compared to in-situ measures of biomass and/or physical variables (e.g. temperature in polar oceans (Marra et al., 2007). While utilisation of remote-sensing estimation of phytoplankton absorption in the Southern Ocean is appealing, at present, ocean colour algorithms are incapable of adequately accounting for phytoplankton absorption in this region due to large variability in bio-optical properties (Ferreira et al., 2017).

Numerous studies have reported that the bio-optical characteristics of the Southern Ocean differ from those of other major oceans (Mitchell, 1992, Sullivan et al., 1993, Arrigo et al., 1998, Dallolmo et al., 2005, Ferreira et al., 2017). In addition to unique environmental conditions including persistent cloud cover, larger solar zenith angles and deeper vertical mixing compared to lower latitude regions (Mitchell and Holm-Hansen, 1991, Mitchell, 1992), the Southern Ocean waters are also characterised by low particulate, as well as chlorophyll-specific absorption (Mitchell and Holm-Hansen, 1991). The latter reflects prevalence of the physiological phenomena referred to as the “pigment package effect” (Bricaud et al., 2004) that has been widely-observed among phytoplankton populations in this region (Tripathy et al., 2014, Ferreira et al., 2017). Pigment packaging occurs in phytoplankton cells when high intracellular Chl-a concentrations reduce light absorption per unit of pigment, leading to a decrease in their corresponding absorption coefficients (Ciotti et al., 2002, Laiolo et al., 2020, Bricaud et al., 2004). This phenomenon is particularly pronounced in phytoplankton assemblages dominated by larger cells (>10 μm diameter) e.g. at the deep chlorophyll maximum (DCM) layer (Gomi et al., 2010, Kerkar et al., 2020) – where availability of light and nutrients facilitates larger cells and increased pigment synthesis (Fernand et al., 2013). As up to half of integrated primary production and export production occurs at the DCM (Weston et al., 2005, Omand et al., 2015, Baldry et al., 2020), understanding variability in how phytoplankton absorption scales to PP in both surface and DCM waters in the Southern Ocean remains a critical gap in knowledge.

Overall, the low chlorophyll-specific phytoplankton absorption (a*_ph) reported from the Southern Ocean to date results in underestimation of Chl-a by global standard ocean color algorithms in the Southern Ocean (Dierssen and Smith, 2000, Garcia et al., 2005, Szeto et al., 2011, Johnson et al., 2013). In turn, this has implications for utilising satellite-derived Chl-a as an index of PP in the Southern Ocean, compared to other regions where this approach has been widely-applied to model dynamics of phytoplankton productivity (Graham et al., 2015). In spite of this, no extensive studies have focussed on in-situ phytoplankton absorption characteristics in this region (Arrigo et al., 1998, Reynolds et al., 2001, Stambler, 2003). Although taxonomic composition, pigment constitution and cell size confer characteristic absorption features to phytoplankton assemblages (Arrigo et al., 1998, Dierssen and Smith, 2000), the environmental complexity of the Southern Ocean renders such generalisations inapplicable. The inter-regional characteristics like upwelling and differences in concentrations of nutrients (for instance, abundant silicate concentrations at the coastal region) lead to a peculiar phytoplankton community type (Mendes et al., 2015), thereby leading to differential absorption and bio-optical variations between frontal and coastal domains. Indeed, advancing remote sensing of PP estimates in this region hinges on characterising bio-optical variability and how this relates to the surrounding environment at intra-regional scales. To date, bio-optical studies in the Southern Ocean have predominantly focussed on the Australasian sector (Shooter et al., 1998, Westwood et al., 2011), together with the Antarctic coast (Mitchell and Holm-Hansen, 1991, Mitchell, 1992, Arrigo et al., 1998). Though a few studies have addressed variability of PP in the Indian Sector of the Southern Ocean (ISSO) (Westwood et al., 2010), the importance of the DCM in PP variability (Gomi et al., 2010, Tripathy et al., 2015) and effects of pigment packaging on PP (Tripathy et al., 2014), the link between PP and bio-optical properties has not been examined in detail (Hirawake et al., 2011) to date.

Our study aims to enhance understanding of PP and bio-optical variability in the spatially-complex ISSO region. Specifically, we examine PP dynamics in surface and DCM waters at both frontal and coastal stations to determine if (1) PP can be better related to in-situ absorption properties rather than phytoplankton biomass (i.e. total Chl-a) and (2) whether absorption properties of phytoplankton can be linked to their size class, and global phytoplankton absorption-based models can be used in their original formulations in the bio-optically complex waters of the ISSO. Such knowledge would improve current understanding of spatio-temporal trends in PP and bio-optical variability within the ISSO – and thus represents a step towards improved global modelling of phytoplankton carbon cycling in this region. The list of common terminologies and abbreviations used in the present study are provided in Table 1.