In a new paper, Dr Simon Pooley from the Department of Geography argues that moving beyond historical ecology to draw on methods from environmental history is crucial for the field and allied sub-disciplines of ecology like restoration ecology and invasion biology.

As researchers turn increasingly to history as a resource for understanding the trajectory of a planet under unprecedented human pressure, we need to be clear about the methodological challenges of finding and using historical data.

Mainstream ecology recognises the importance of recovering long-term ecological records by applying rigorous methods to the study of charcoal, ice cores, pollen and tree rings. The uses of historical textual and graphic sources are seldom accorded analogous rigour.

Ecologists use scientific methods to learn from the past, drawing on resources developed by historical ecologists to source data and assemble it into time series, predicting fluctuations and trends in variables over time.

Recently, interest in the impacts of human actions on earth systems has led scientists to aggregate data on generalised human perturbations like greenhouse gas emissions or the industrial manufacture of fertilizers. They then compare these with variables tracking impacts on related earth systems like earth surface temperatures and ocean acidification.

For environmental historians, the resulting framings of relationships between human agency and earth systems are often too linear and generalised. Further, historians urge analysts to pay closer attention to the contextual factors shaping data production, rather than dip uncritically into ‘the past’ as a ‘goody bag’ of useful facts.

This paper is aimed at ecologists rather than historians; it aims to communicate some of the major considerations they should be aware of when collecting and interpreting historical data. These can be handily divided into four themes, summarised as questions:

• Why, how, and by whom were the data collected?
• How were the data measured (relating to time periods and spatial units in particular)?
• What conceptual filters shaped data collection and interpretation?
• What anomalies, exceptional events and human influences shaped data collection?

The paper notes that the purposes for which data were initially collected, and by whom they were collected, using which methods and tools, all importantly shape what is (and is not) collected and in what form. Long time-series often combine datasets shaped in different ways in accordance with all these factors. A simple example is that, beginning in the late sixteenth century, the Gregorian calendar was adopted sporadically across Europe over several centuries, and then in parts of Asia and elsewhere in the late 1800s and early 1900s. In each instance, days were won or lost as regional or national calendars were adjusted accordingly.

The reporting periods used by institutions vary (and change) and researchers compiling time series must take these changes, and ‘longer’ and ‘shorter’ years used to adjust to new reporting cycles, into account when incorporating their data.

And of course datasets should never be taken at face value: whale catch data published by Russia and Japan in the postwar period, for instance, should be treated with due scepticism.

In the 1970s, mathematical models were devised to estimate historical (baseline) saltwater crocodile populations in northern Australia, as a means for judging when conservation measures introduced early in that decade had resulted in an adequate recovery to pre-exploitation levels. However, when a series of attacks followed an unexpectedly quick population recovery, public opinion demanded a rethink of models which calculated that these populations were will still at only 2% of their historical size.

Northern Territory researchers consulted historical records and interviewed former hunters, discovering that pre-exploitation crocodile population densities had varied significantly from river system to river system. According to their revised estimates, populations were at 30-50% of pre-exploitation levels, sufficient to allow controlled sustainable use of crocodiles. This resulted in a policy shift which has proven to be a successful long-term management approach in the region.

The ways in which underlying theories and conceptual frameworks shape data collection are challenging to work out. Examples include powerful narratives about desertification and deforestation which shaped how colonial foresters and agricultural experts interpreted landscape change and management in the colonies in which they worked. Ideas about fire, forests and rainfall long shaped expert thinking about landscape change in India, Madagascar and many parts of Africa, and arguably still influence management interventions aimed at changing or preventing traditional practices by local peoples. In India, ideas about degradation through burning still guide management of savannas, which are arguably misclassified as degraded woodlands.

The four themes identified in my paper are discussed in the text in relation to a schematic figure. Of course, any such visualisation conceals considerable complexity, but it provides a useful framework for exploring some of the challenges. The paper gestures to some of these further complexities through examples from my monograph Burning Table Mountain: an environmental history of fire on the Cape Peninsula (Palgrave 2014; UCT Press 2015) and a recent paper on fire and expertise in South Africa (Environmental History 23:1, 2018). These examples show some ways in which published theories don’t always match the authors’ recommendations for management in the field, or the results of experimental science are ignored in the face of powerful narratives about ecological phenomena like fire, soil erosion and drought. Finally, they demonstrate how complex factors – social, economic, cultural – combine in diverse ways to shape whether and how scientifically informed policies are implemented.

The paper concludes by arguing that the use of historical approaches as summarised in its accompanying figure can help improve the accuracy of the data we draw on to do important conservation work like estimate the conservation status of ecosystems or species, or prioritise conservation action. At a deeper level, historical methods allow us to tease apart the many interacting factors which shape policy and management and influence their interrelations with other narratives, knowledge systems and priorities in particular contexts, over time.

The paper is available (open access) with supporting figure and references in Conservation Letters.