Search

Agricultural expansion drives biodiversity loss globally, but impact assessments are biased towards recent time periods. This can lead to a gross underestimation of species declines in response to habitat loss, especially when species declines are gradual and occur over long time periods. Using Cold War spy satellite images (Corona), we show that a grassland keystone species, the bobak marmot (Marmota bobak), continues to respond to agricultural expansion that happened more than 50 years ago. Although burrow densities of the bobak marmot today are highest in croplands, densities declined most strongly in areas that were persistently used as croplands since the 1960s. This response to historical agricultural conversion spans roughly eight marmot generations and suggests the longest recorded response of a mammal species to agricultural expansion. We also found evidence for remarkable philopatry: nearly half of all burrows retained their exact location since the 1960s, and this was most pronounced in grasslands. Our results stress the need for farsighted decisions, because contemporary land management will affect biodiversity decades into the future. Finally, our work pioneers the use of Corona historical Cold War spy satellite imagery for ecology. This vastly underused global remote sensing resource provides a unique opportunity to expand the time horizon of broad-scale ecological studies. [ABSTRACT FROM AUTHOR]

In densely populated fire-prone regions, interactions between global change drivers, such as land-cover changes and climate change, may increase the frequency and severity of wildfires impacting forest ecosystems, thus diminishing their capability of provisioning key ecosystem goods and services for these societies. Yet, landscape mosaics play a crucial role in fire dynamics and behaviour. Here, we argue that promoting heterogeneous agro-forest mosaics could reduce the area affected by future fires. Specifically, we evaluated 24 landscape-scale management scenarios based on agricultural conversion, i.e. the creation of new agricultural land, that also explicitly incorporated fire suppression. Scenarios differed in the annual rate of such conversion, the spatial pattern (aggregate vs. scattered), and the location of new agricultural patches. To quantify the interactions between vegetation dynamics, fires, land-cover changes, and fire suppression, we coupled two spatially explicit models: a landscape dynamic fire-succession model and a land-cover change model. When applied to the Mediterranean region of Catalonia (NE Spain), new landscape mosaics favoured firefighting extinction capacity only after 15 years (on average) of cumulative land transformations. Agricultural conversion of at least 100 km2 year−1 was required to reduce total area burnt. A conversion rate of 200 km2 year−1 substantially improved fire suppression effectiveness, but subsequent conversion increases did not. When aggregated, new agriculture patches contributed more effectively to reduction in total area burnt and decreased the edge effect on remaining forest patches. Agricultural conversion in Mediterranean landscapes opens a new window for long-term spatial planning aimed at minimizing negative impacts of wildfire on forest ecosystems. These alternative strategies could help to develop landscape management practices in other fire-prone regions. [ABSTRACT FROM AUTHOR]