Areas of the coastal ocean where oxygen is low or absent in bottom waters, so-called dead zones, are expanding worldwide (Diaz and Rosenberg, 2008). Increased inputs of nutrients from land are enhancing algal blooms, and the sinking of this organic matter to the seafloor and subsequent decay leads to a high oxygen demand in bottom waters. Depending on the physical characteristics of the coastal system, this may initiate periodic or permanent water column anoxia and euxinia, with the latter term implying the presence of free sulfide (Kemp et al., 2009). Global warming is expected to exacerbate the situation, through its effects on oxygen solubility and water column stratification. In many modern coastal systems, anthropogenic changes are superimposed on natural variation and lack of knowledge of such variation makes the prediction of future changes in water column oxygen challenging (e.g., Grantham et al., 2004). That natural drivers alone can be the cause of widespread coastal anoxia is evident from studies of greenhouse periods in Earth’s past, including the oceanic anoxic events of the Cretaceous and Toarcian (Jenkyns, 2010).

Sediment proxy records are essential to any reconstruction of variations in anoxia and euxinia on time scales beyond several decades to a century. A variety of biological and geochemical indicators can be used for this purpose, such as the presence of the remains of benthic and pelagic organisms, laminations, biomarkers for eukaryotes or prokaryotes, and inorganic geochemical and mineralogical signatures in the sediment, and ideally, these methods are combined. Sediments that are deposited below a euxinic water column are, for example, typically enriched in organic carbon, sulfur, iron, and trace metals such as rhenium and molybdenum (Gooday et al., 2009). Recent additions to this paleo-redox toolbox are the isotope systems of Fe and Mo (Lyons et al., 2009). Reconstruction of the temporal changes in the oxic-anoxic interface (chemocline) in the water column forms a key step in the identification of the external drivers and internal feedbacks that contribute to anoxia and euxinia in a given system. In their study of sediments from the Black Sea, Eckert et al. (2013, p. 431 in this issue of Geology), make this step by providing, for the first time, a basin-wide reconstruction of the evolution of the chemocline in this silled coastal basin over the Holocene.

Silled basins in humid areas such as Kau Bay (Indonesia), the Baltic Sea, and the Black Sea, are particularly sensitive to low oxygen conditions because of salinity stratification and associated reduced vertical mixing (Kemp et al., 2009). All these inland seas have an intriguing history and were originally coastal lakes that were transformed to marine basins due to postglacial sea-level rise. Kau Bay is only semi-euxinic, and is subject to incursions of low-oxygen non-sulfidic bottom waters that alternate with periods of anoxic, sulfidic bottom waters (Middelburg et al., 1991). The Baltic Sea also alternates between redox states: it experienced various periods of low oxygen over the Holocene, but is currently subject to a human-induced period of anoxia, with its bottom waters largely oxic around 1900 CE (Conley et al., 2009). The Black Sea is the largest euxinic basin in the world and differs in being permanently euxinic. This is the result of the strong stratification that developed after its fore-runner fresh water lake became connected to the Mediterranean Sea through the narrow, shallow Straits of the Bosporus at ca. 9 kyr B.P. Water column anoxia developed across the deep basin from ca. 7.5 kyr B.P. onward (Degens and Ross, 1974), and the chemocline is presently located at ∼100 m depth.

Strong variations in the geochemical and paleo-ecological composition and genetic signature of the sediments in the Black Sea provide testimony that the conditions in the water column have been far from constant over the past ∼7.5 kyr. Two phases of deposition are generally distinguished based on visual characteristics of the sediments. Following the onset of anoxia, a finely laminated, dark, organic-rich sediment layer formed first (Unit II), followed by deposition of alternating microlaminae of calcareous (white) and organic- and clay-rich material (black) from ca. 2.6 kyr B.P. to the present (Unit I). The shift from Unit II to Unit I was originally attributed to the invasion of the coccolithophore Emiliania huxleyi when salinity rose above 11 (Arthur and Dean, 1998). However, genetic analyses show that this calcifying haptophyte colonized the photic zone of the Black Sea shortly after the connection to the Bosporus, and the Unit I–II transition marks the moment that coccoliths began to be preserved in the sediments (Coolen et al., 2009).

The delayed appearance of Unit II on the slopes of the basin has been taken as an indicator of a slow rise of the chemocline following the onset of anoxia (Degens and Ross, 1974). The rise was fast enough, however, for the chemocline to reach the photic zone by the time of deposition of the lower part of Unit II, as indicated by the presence of biomarkers for photosynthetic green sulfur bacteria (Sinnighe Damsté et al., 1993; Repeta, 1993). Results of similar analyses for the upper part of Unit II suggested a subsequent descent of the chemocline followed by re-establishment in the photic zone during deposition of Unit I. At the time, controversy remained about the temporal and spatial variability in the position of the chemocline, and the extent to which the water column and photic zone remained euxinic throughout deposition of Units I and II. This debate was partially resolved when Wilkin et al. (1997) showed that the size of the pyrite framboids in Units I and II were in line with a continuously euxinic water column. Using a composite record of sediment Fe, Mo, and Fe-isotopes derived from data for nine sites throughout the basin, Eckert et al. (2013) now confirm the evolution of Black Sea euxinia, as suggested in these earlier studies, and provide a more consistent and basin-wide timing for the series of events.

The variation in strength of the ‘Fe shuttle’ forms the heart of their reconstruction. This term is used to describe the lateral transfer of Fe released from suboxic shelf sediments to the deep basin. The authors use their Fe/Al record as a direct indicator of the position of the chemocline, where low Fe/Al indicates a weak shuttle with a chemocline impinging on the slope. A high Fe/Al, in contrast, indicates a chemocline allowing suboxic water to spread over part of the shelf and supporting an intense Fe shuttle. The authors also make use of the fact that Mo data can be used to reconstruct the hydrography of a basin, which for the Black Sea allows an estimate of the inflow of Mediterranean seawater. Fe isotope analyses bolster the argument for the shelf-source of Fe. The emerging timeline is as follows (Unit II): (1) a gradual rise of the chemocline over a period of ∼2 kyr following the onset of anoxia at ca. 7.6 kyr B.P., (2) fully developed euxinic conditions with an ascent of the chemocline onto the shelf at ca. 5.3 kyr B.P., (3) a subsequent descent of the chemocline, and (Unit I) (4) establishment of the chemocline in its present-day position at the shelf break from 2.7 kyr B.P. onward.

But this is not the full story. Besides a good timeline for euxinia in the Black Sea, we need to understand the hydrographic and biogeochemical processes that drove these changes in redox conditions, and there much work still needs to be done. The evolution of the salinity in the basin, for example, is not well constrained. Recent qualitative reconstructions of salinity based on various proxies suggest that values of surface water salinity in the Black Sea rose until ca. 3 kyr B.P., followed by a gradual freshening to present-day values (van der Meer et al., 2008; Coolen, 2011). Possible causes for the freshening include an increase in fluvial discharge and decreased evaporation (Giosan et al., 2012). An associated increase in stratification may have contributed to the shallowing of the chemocline at the onset of the deposition of Unit I. Also, the processes leading to the increased total organic carbon (TOC) in Unit II as compared to the overlying and underlying units are not well understood. The high TOC is frequently interpreted as an indicator of enhanced nutrient availability and productivity following the inflow of Mediterranean seawater, and transition of limnic (oxic) to marine (anoxic and euxinic) conditions (also see Eckert et al., 2013). However, the sources of the nutrients fuelling this productivity have not been identified and whether, for example, phosphorus release from sediments or river water is more important is still an open question. Finally, the cause of the descent of the chemocline after ca. 5.3 kyr B.P. remains unknown. While Eckert et al. (2013) propose a decreased seawater input or increased river input as potential causes, van der Meer et al. (2008), in contrast, suggest that the absence of a shallow chemocline can be best explained by the high sea-surface salinity at the time.

Despite the open questions, the Eckert et al. (2013) study is important because it provides a more solid timeline and integrated view of the evolution of euxinia in the Black Sea, which is highly useful for assessments of climatic and other drivers of temporal change. The tools used can also be applied to better interpret sediment records from other marine systems, both modern and ancient, and can thereby aid in the assessment of the time scales of a possible decline into, and recovery from, wide-scale anoxia and euxinia. Such knowledge is important in a warming world where water column deoxygenation in the coastal zone is becoming more and more common.