Due to rising demand for both food and environmental services, agriculture is increasingly required to deliver multiple outcomes. Characterising differences, across agricultural landscapes, via the identification of broad archetypal groupings, is an important step in exploring spatial patterns in the capacity of land to deliver these potentially competing functions. Creating characterisations at multiple levels, for landscape and farm management, can allow policy-makers and land managers to harmonise delivery of ecosystem services at different intervention scales. This can identify ways to increase the complementarity of public goods and the sustainability of farmed landscapes. We used data-driven machine learning to create landscape and agricultural management archetypes (1 km resolution) at three levels, defined by opportunities for adaptation. Tier 1 archetypes quantify broad differences in soil, land cover and population across Great Britain, which cannot be readily influenced by the actions of land managers; Tier 2 archetypes capture more nuanced variations within farmland-dominated landscapes of Great Britain, over which land managers may have some degree of influence. Tier 3 archetypes are built at national levels for England and Wales and focus on socioeconomic and agro-ecological characteristics within farmland-dominated landscapes, characterising differences in farm management. By using a non-nested hierarchy, we identified which types of management are restricted to certain landscape settings, and which are applicable across multiple landscape contexts. Understanding variation within and between agricultural landscapes and farming practices has implications for planning environmental sustainability and food security. It can also aid understanding of the scale at which interventions could be most effective, from incentivising changes in farmer behaviour to policy drivers of large-scale land use change.
Biological invasions are one of the major causes of biodiversity loss worldwide. In spite of human aided (anthropogenic) dispersal being the key element in the spread of invasive species, no framework published so far accounts for its peculiar characteristics, such as very rapid dispersal and independence from the existing species distribution. We present a new method for modelling biological invasions using historical spatio-temporal records. This method first discriminates between data points of anthropogenic origin and those originating from natural dispersal, then estimates the natural dispersal kernel. We use the expectation-maximisation algorithm for the first step; we then use Ripley’s K-function as a spatial similarity metric to estimate the dispersal kernel. This is done accounting for habitat suitability and providing estimates of the inference precision. Tests on simulated data show good accuracy and precision for this method, even in the presence of challenging, but realistic, limitations of data in the invasion time series, such as gaps in the survey times and low number of records. We also provide a real case application of our method using the case of Litoria frogs in New Zealand. This method is widely applicable across the field of biological invasions, epidemics and climate change induced range shifts and provides a valuable contribution to the management of such issues. Functions to implement this methodology are made available as the R package Biolinv (https://cran.r-project.org/package=Biolinv).
Archetypes of land- and socio-ecological systems, generated using unsupervised classification methods, enable the assimilation of complex environmental and socio-economic information. Such simplification has considerable potential to feed into decision support systems for sustainability planning. But, the usefulness of archetypes depends on how well they relate to sustainability criteria, such as ecosystem service (ES) delivery, that are external to the input datasets employed for archetype generation. Sensitivities in such post-hoc association analyses, and the associated utility of the archetype framework in a decision support context, remain unexplored. Here we emulated post-hoc association analysis procedures using simulated socio-ecological datasets and ES response variables. Our simulations revealed a substantial influence on analysis performance from (1) the number of variables used as inputs in archetype generation, (2) the correlation structure of input datasets, (3) the type and distribution of input variables, and (4) the functional form (linear or non-linear) characterising the relationship between ES variables and their predictors. We observed near-identical performance when archetypes were generated using K-means clustering and Self-Organising Maps (SOMs) – two commonly used archetype classification methods. Further, better archetype classifier performance did not guarantee better discrimination of ES value distributions between archetypes. Our results suggest that designing a framework to generate archetypes for sustainability planning, and the selected methodological choices therein, should place greater emphasis on what the archetypes will be used for in downstream analyses, and not focus solely on archetype classifier performance. This would better ensure the identification of archetypes adaptable to a diverse array of sustainability indicators and sufficiently robust for monitoring decision outcomes over time.