Regional precipitation variability modulates Holocene fire history of Iceland

We present the first continuous Holocene fire record of Iceland from a lacustrine archive in the northeast region. We use pyrogenic PAHs (polycyclic aromatic hydrocarbons) to trace shifts in fire regimes, paired to a continuous record of n-alkanes, faecal sterols, perylene, biogenic silica, and ¹³C, as proxies for soil erosion, lake productivity, and human presence.

Paleoclimate research across Iceland provides a template for changes in climate across the northern North Atlantic. The role of orbitally driven cooling, volcanism, and human impact as triggers of local environmental changes, such as fire and soil erosion, is debated. While there are indications that human impact could have reduced environmental resilience in a context of deteriorating climatic conditions, it is still difficult to resolve to what extent human and natural factors affected Iceland landscape instability, due also to a lack of data on natural fire regime prior and during human colonisation.

Pyrogenic PAHs can be formed during the incomplete combustion of biomass initiated by humans or natural wildfires. Factors such as fire temperature, biomass typology, and source distance can strongly affect pyrogenic PAH molecular weight and spatial distribution.
Faecal sterols/stanols and their ratios have been used in archaeological and paleoclimate studies to detect human and/or livestock/herbivore waste. The absence of large herbivorous mammals and humans in Iceland prior to settlement means that increases in the occurrence of faecal sterols and bile acids over natural background values should mark the arrival of humans and associated livestock in the catchment, which could be traced regionally.

Our results indicate that the Icelandic fire regime during the Holocene followed four main phases. Among these, a very long period centred around the Holocene climatic optimum (ca 9.5 – 4.5 ka BP) was characterised by a generally low frequency fire regime, both in the lake catchment as in the whole north-eastern Iceland. This same period was also marked by relatively low background levels of faecal sterols/stanols. At 4.5 ka BP a new phase started, with a general increase of all PAHs values. According to both our PAH and sterol data, there is no apparent human signal around the 9th century C.E., where an increase in man-made fires would likely be expected in connection to the historical data of Viking colonisation of Iceland (870s C.E.), suggesting that fire regimes have primarily been controlled by natural factors.
In addition, the pyrogenic PAHs record also differs from the trend of a general stepwise climatic “deterioration” previously highlighted by other lake proxies throughout Iceland, linked to decreasing summer insolation and related cooling, as highlighted also by our other proxies.

A comparison to recent palynological data from a nearby site and to δD data from the NW region suggest shifts in NAO regimes as the main forcing behind shifting fire regimes in Iceland. Changes in precipitation regimes would have determined shifts in the composition of the regional vegetational community, increasing fuel availability and flammability with decreasing precipitation, leading to widespread low temperature fires, easily trigged by frequent volcanic episodes.