A Biomarker Reconstruction of the Past Occurrence of Harmful Algal Blooms Using Sediment from the Northern New England Coast

Primary tabs

Project Type: 
Inception Date: 
Completion Date: 
Theme Area: 
Coastal Ecosystem and Public Health


Meixun Zhao Dartmouth College Principal Investigator


In recent years the blooms of the brown tide alga Aureococcus anophagefferens have caused damage to fisheries along the New York and Southern New England coasts (Anderson, 1994). It has been detected as far north as Portsmouth, NH (Anderson et al., 1993). The blooms of the red tide alga Alexandrium tamarensis have had similar harmful effects along the north New England coasts (Anderson, 1994). Human induced changes such as excessive nutrient input and climate changes have been suggested to be responsible for this and other harmful algal blooms (HABs) (Keller and Rice, 1989; Yamaguchi et al., 1997). Although this appears to be a new problem, it is unknown whether HABs have occurred in the past.

One means to reconstruct the occurrence and indeed the frequency of past HABs is the analysis of sterol biomarkers in the sediments. Dinosterol, a compound produced by most dinoflagellate species, has been used to reconstruct the history of dinoflagellate production. More recently, species-specific sterols has been identified for the red tide algae Gymnodinium brevis and the brown tide algae Aureococcus anophagefferens (Faraldos and Giner, 1998; Giner and Boyer, 1998).

These sterols afford an opportunity of species-specific reconstruction of HABs. Sterols are useful biomarkers for paleo-algal bloom reconstruction because they are found in relatively large amounts in algae (ca. 0.5% of dry weight) and because they are very stable and can be preserved in sediments for up to millions of years.

Goals and Objectives

1) To survey sediments for evidence of the occurrence of the brown tide algae Aureococcus anophagefferens along the New England coast from Rhode Island to the Gulf of Maine.

2) One sediment core will be selected for the analysis of sterols to provide a chronology of such events and to correlate them with other environmental factors such as climate.

3) To perform an exploratory search for a biomarker for the red tide algae Alexandrium tamarensis, which have occurred in the Gulf of Maine (in collaboration with Dr. Giner).

General and Analytical Approaches

Up to 12 surface sediment samples have been collected along the New England coast from Rhode Island to the Gulf of Maine. These will be analyzed for dinosterol and Z-24- propylidenecholesterol for evidence of the occurrence of brown tide. Based on this survey, one sediment core will be obtained for sterol analysis to estimate the history of these algal blooms over the last few hundred years. We also carry out alkenone analysis to estimate past sea surface temperature history to be correlated with algal bloom history.

Sedimentary sterol and alkenone analyses: After adding the appropriate internal standards, freeze-dried sediments are Soxhlet extracted by refluxing 24 hours using a dichloromethane/methanol mixture (3:1 v/v). The obtained total lipid extract is first fractionated into three fractions using silica gel columns eluted using hexane, dichloromethane (DCM) and methanol. The DCM fraction (which contains the alkenones) is derivatized with N,0- Bis(trimethylsilyl)-trifluoroacetamide (BSTFA) and analyzed using a Varian 3400 GC and a Varian Saturn 2000 GCMS, both are fitted with an autosampler and with a 50 m capillary column (Chromapack CP-Sil 5, 0.32mm i.d. and 0.25micro m film) ( Zhao et al., 1995). Alkenones were quantified by comparing their peak integration with the internal standard (C36 n-alkane). SST is then calculated using the following equation: T(degrees C) = (Uk37-0.039)/0.034 (Prahl & Wakeman, 1988) where Uk37 =37:2/(37:2+37:3).

The methanol fraction (which contains the sterols) is normally derivatized with BSTFA and analyzed using a Varian 3400 GC and a Varian Saturn 2000 GCMS. Identification of sterols is achieved by comparison of the retention times and mass spectra of the samples with those of the authentic standards (Huang and Meinschein, 1976; Giner and Boyer, 1998). Quantification of the sterols is made by using an internal standard, cholestane, just before GC analysis. For samples with a complicated mixture of sterols and stands, a silica gel thin-layer chromatography (TLC) step can be employed to isolate and concentrate the specific sterols of interest before GC and GCMS analysis. A portion of the methanol fraction is loaded onto the 0.25 mm silica gel G25 UV254 TLC plates and developed in hexane/diethyl ether/acetic acid (50:50:1, v/v/v). Typical Rf value is 0.45 for long-chain alcohols are 0.36 for 4-methyl sterols and 0.29 for desmethyl sterols (Volkman et al., 1993).