Ecology, phenology, and preservation of recent dinoflagellates cyst from the Northwest African upwelling region

Abstract

Marine phytoplankton play a crucial role in ocean and climate sustainability. They are responsible for fixing one-third of atmospheric carbon dioxide and turning it into biomass. Through a biological pump, a fraction of the greenhouse carbon is sequestered in the ocean sediment, while some of the carbon is released back into the atmosphere via respiration. Along with diatoms and coccolithophores, phototrophic dinoflagellates are one of the main eukaryotic phytoplankton in marine ecosystems. This plankton group is highly diverse, distributed worldwide, influenced by certain physical and chemical factors (e.g., nutrients, light, temperature, currents, and salinity), and responds to changes in those limiting factors. In addition, around 15% of dinoflagellates undergo a resting stage during their sexual reproduction cycle, forming a distinctive structure known as dinoflagellate cysts (dinocysts). Many dinocysts are made of resistant organic walls that are highly preservable in the sediment. All of these dinoflagellate aspects make them a valuable proxy for not only ecological, oceanography, and climate studies but also for reconstructing the conditions and changes in past marine ecosystems. The current fast-changing climate is expected to drive shifts in the atmospheric-oceanic interactions, which influence the bloom dynamics of dinoflagellates and the formation of their cysts. Thus, a multi-year record of in-situ dinocysts and prominent limiting factors are required to create the reconstruction. Unfortunately, such data are scarce, and the available ones cover a short period of time series. In this dissertation, we provide an 18-year time series of organic-walled dinocysts collected by a sediment trap that has been deployed in one of the four eastern boundary upwelling ecosystems (EBUEs) located on the Northwest African coast. The coastal upwelling brings nutrient-rich subsurface waters to nourish the primary producers, such as dinoflagellate, that inhabit the photic zone. The upwelling occurred throughout the year under a seasonal trend controlled by the Inter Tropical Convergence Zone (ITCZ) annual migration. The nutrient-rich waters could reach hundreds of kilometers towards the open ocean in the form of filaments and eddies due to surface water currents in the study area. In addition, Sahara dust increases productivity in this area by providing limiting nutrients and trace elements. The aerosol dust supplies sediment materials, where microfossils could be preserved. The accumulation of the sediments is relatively high, and no hiatus during the Holocene was reported in the area of trap deployment. Consequently, the sediment archive recorded a continuous history of the ocean production in a high time resolution, allowing a comparison of sediment trap dinocysts with the down-core fossils. To understand the response of dinocyst production to the interannual dynamics in their ecosystem, the dinocyst export flux over 18 years was compared with several environmental parameters such as upwelling wind strength and direction, Saharan dust input, sea surface temperature (SST), sea surface temperature anomaly (SSTa), and sea surface chlorophyll-a (Chla) (Chapter 3). The results revealed that heterotrophic dinoflagellates contributed to a significant portion of the dinocysts association (ca.94%). Their export flux was usually higher in spring - summer, coinciding with the strongest upwelling wind that blew from the northeast. A few peaks of the high export flux of this dinocysts group in winters of some years aligned with high dust input into the sediment trap area. Stronger inter-annual variability of the export flux was shown by the photo-/mixotrophic dinocysts when their annual highest flux occurred in autumn - winter. Furthermore, Canonical Correspondence Analysis (CCA) confirmed that the upwelling indicator (wind strength and direction) was the most influential parameter, and the SST was the least significant. According to this ordination technique, five dinocyst groups were identified in relation to the environmental parameters. Dinocysts taxa in group 1 (Echinidinium delicatum/granulatum, Echinidinium spp., E. transparantum/zonneveldiae, Trinovante-dinium spp., and Protoperidinium latidorsale) thrived under maximal upwelling intensity. Group 2 (Archaeperidinium spp., P. americanum, P. stellatum, and P. subinerme) consists of taxa favoured by maximal upwelling and dust input. Group 3 (Gymnodinium spp. and L. polyedra) preferred the upwelling relaxation phase. Dinocyst taxa in group 4 (Bitectatodinium spongium and Protoceratium reticulatum) occurred when the upper water column was warmer. Lastly, taxa in group 5 (Brigantedi-nium spp., E. aculeatum, Impagidinium aculeatum, P. conicum, P. monospinum, Pentapharsodinium dalei, and Spiniferites spp.) did not show a specified correlation to any environmental parameters. In 2009, the dinocyst association shifted from the abundance of various taxa to the abundance of Echinidinium, which coincided with the intensification of the dust input. Regarding cysts of potentially toxic dinoflagellates, the sediment trap collected five photo-/mixotrophic taxa of dinocysts affiliated with biotoxin-producing species that threaten the marine ecosystem and human health. Their export flux is low due to the highly mixed water column, which is unfavourable for photo-/mixotrophic dinoflagellates. To extend our knowledge about the bloom dynamics of the dinocysts and the taxa in the same ecological group that was determined in the previous project, the dinocysts and environmental parameter time series were analysed using Morlet wavelet analysis (Chapter 4). This technique allows the visualisation of the periodic or cyclic patterns and their variation throughout the time series by examining the correlation between the wave patterns of the datasets with the Morlet wavelet. Based on the warm spectra representing a high correlation, the total dinocysts time series contained four periodicities: 240-day, 480-day, 180-day interpreted as the half-year cycle, and 360-day interpreted as the annual cycle. The half-year and annual cycles of the total dinocyst export flux coincided with the same cycles in the wind speed, wind direction, aerosol dust time series, and SSTa. The dinocyst winter peak (December - February) aligned with the dust winter peak, whilst the dinocyst spring/summer peak (April - June) aligned with primary upwelling and dust summer peak. The SST wavelet spectra only indicated the annual cycle, but it did not align with months when the annual cycle of the dinocyst was observed. The cycle of total dinocysts indicated variation in three phases, which is in line with changes in the taxa composition. Phase I (2003 - 2008) was indicated by the low significance of 240-day and 480-day periodicities when the dinocyst taxa of maximal upwelling and dust and upwelling relaxation group were abundant. Phase 2 (2009 - 2012) is characterised by the half-year and annual cycles as the product of maximal upwelling group cycles. Phase 3 (2013 -2020) showed the stronger significance of the half-year and annual cycles that were also observed in the upwelling group, plus a slight contribution by the cosmopolite/no relation and upwelling and dust groups. The three phases coincided with the changes in the cycles of the aerosol dust and upwelling wind time series. The results suggested that the stepwise changes in the environments were driven by the southward movement of the ITCZ position that could strengthen the upwelling and dust emission during the time span of our investigation. The dinocyst cycles and taxa composition changes supported this interpretation, indicating that climatic-driven changes in the environment influenced the bloom dynamics and community structure of dinoflagellates. The information gathered from the sediment trap data in the previous two projects was expected to be preserved in the sediment (paleoarchive), following the long-known hypothesis of uniformitarianism. This theory has allowed microfossils, including dinocysts, to be utilised as past environmental and climatic change proxies. However, some perturbations are known to disrupt the water column signal carried by microfossils in the marine sediment. Therefore, we compared the sediment trap data with sediment core data under the same time span and resolution (Chapter 5). The age of every core sample was determined with 210Pb (lead) half-time decay, which the result stated 1.6 years per 3.1 mm of sediment. After matching the time span and resolution between the trap samples and core samples, the comparison results showed that the dinocyst association in the core resembles a coastal upwelling environment dominated by brown dinocysts mostly affiliated with heterotrophic taxa. The most abundant dinocyst taxa in the sediment trap (e.g., Archaeperidinium spp., Brigantedinium spp., Gymnodinium spp., P. americanum, L. polyedra, and Echinidinium species) are all presented in the core samples. The dinocyst accumulation rates between trap and core samples are comparable; the heterotrophic group ranged in the scale of 106, and the photo-/mixotrophic group ranged in the scale of 105. However, the relative contribution and accumulation rates of the heterotrophic dinocysts were lower in the core samples, which is the opposite of the photo-/mixotrophic dinocysts. The susceptibility of the brown (coloured) cysts that are mostly produced by the heterotrophic dinoflagellates is suggested to cause the declining rate and contribution of this group in the core samples. In contrast, some heterotrophic taxa (e.g., P. conica, P. monospinum, and P. stellatum) showed an increase in concentration and percentages, indicating foreign materials that were brought to the location of the core. Principal Component analysis (PCA) was applied to the trap samples and core samples, dissecting the trap samples into three groups and the core samples into four groups. The PCA groups of trap samples marked the dinocyst association shift in 2009 to Echinidinium abundance (maximal upwelling group) reported in the two previous projects. This shift was detected in the core samples under the same time frame. In addition, the four groups of core samples not only confirmed the association shift but also showed that the four upper samples contained significantly different compositions. These samples are suspected to be the result of a stronger impact of pre and post-depositional processes, such as bioturbation, compaction bias, resuspension, lateral transport, and aerobic microbial degradation. The knowledge derived from these three projects has enriched our knowledge about the driving factors of dinocyst production, their relation to environmental changes, and the post-depositional processes that potentially disturb the primary signal of preserved dinocysts. Some crucial points, such as inter-annual variability in dinocysts production, a significant shift in the dinocysts association, and how they were influenced by the changing environment and climate, were only made possible due to the availability of 18-year sediment trap series. The high-resolution comparison study between dinocysts produced from the water column with the embedded ones was the first investigation done, aiming to confirm the role of dinocysts as a reliable tool for reconstructing the condition and changes in the past environment and climate.