This includes quantifying the state of the environment prior to and during

a non-indigenous species invasion, and its recovery state following their eradication. This information is not generally available, particularly on oceanic islands with no long-term history of human occupation or scientific monitoring. In the absence of such information, a palaeoecological approach (the study of past environments) may be used. Palaeoecological methods have been extensively used around the world to examine the influence of humans on landscapes including lakes and rivers and their catchments. As a result, their value for providing a framework against which to assess ecosystem impacts and response and recovery is well recognised (see Bennion and Battarbee, 2007, Crutzen

see more and Stoermer, 2000, Froyd and Willis, 2008 and Smol, 2008 for examples and reviews). Palaeoecological methods have previously been applied on oceanic islands such as the Galapagos Islands, Hawai’i’ and the Azores showing that their highly diverse pre-Anthropocene landscapes were completely transformed with the arrival of humans and the introduction of non-indigenous species. This in turn caused a decline BGB324 in biodiversity and the extinction of many native species (Athens, 2009, Burney and Burney, 2007, Burney et al., 2001, Connor et al., 2012 and van Leeuwen et al., 2008). Lakes provide a particularly useful cAMP palaeoecological archive as their sediments accumulate in layers over time and integrate information from both the lake and its surrounding catchment (Smol, 2008). These layers of sediment may be dated and changes in

a lake and its surrounding environment studied over time using a range of biological and non-biological proxies. Anthropogenic impacts are often particularly well recorded (Smol and Stoermer, 2010) and lake sediments can therefore provide long-term data on the state of the catchment and lake prior to, during and after the introduction of an invasive species (Korosi et al., 2013). These data can include measures of changes in soil erosion rates, vegetation (Restrepo et al., 2012 and Sritrairat et al., 2012), and within-lake production (Bradbury et al., 2002 and Watchorn et al., 2011). This study presents a palaeoecological study of a lake in a heavily rabbit-impacted area on sub-Antarctic Macquarie Island (54°30′ S, 158°57′ E, 120 km2, Fig. 1). A sediment core collected from the bottom of Emerald Lake was analysed to assess changes in sedimentation rates, grain size distribution, geochemical properties and diatom composition over the last ca. 7200 years.