Mark Young

The baseline

Essential to any conceptual or predictive study of a 'system' is a reference  to real examples of its state and dynamics. Concepts and models should consider how the past has given rise to the present before they can can hope to predict how the present can be the basis of a sustainable future. At the beginning of this study, there was little systematic information at the scale of the field on either present or past states.

Historical trajectories

The condition of soil and farming in the past is not well documented. The major surveys of soil and vegetation provide useful background but are too infrequent. The main systematic source of data is the June Agricultural Census from which the areas grown with different crops and grass can be collated. Ten areas, defined by groups of parishes, shown circled in red in the map to the left, were selected as covering most types of farming within the arable-grass system. The Scottish Government provided data on cropped areas in each group, as electronic files from 1982 and as scanned paper documents before that. Other sources are being used to complement the information from the June census, notably surveys of pesticides and fertilizer.

Coexistence and ecological biosafety of two GM crops in Europe

The three-year, EU-funded project SIGMEA combined skills from many disciplines to examine the biological, environmental, agronomic, economic and legal issues that determine whether GM and non-GM crops can feasibly be grown in the same agricultural landscape. Its conclusions differed for the two crops that have been most widely studied.

Map of field patterning in a study of cross pollination - provided by Enric Mele, SpainSIGMEA reported mainly on coexistence, but also on ecological biosafety. Coexistence refers to the need to separate, in the food production chain, different types of crop, such as those that have been developed with or without genetic modification. While zero impurity of one type of crop in another is impossible to guarantee, the EU had set a threshold of 0.9% GM content for produce that can be labelled as GM-free.

Maths, modelling and quantitative biology

The diverse group of modellers and mathematical biologists in EPI has now reached critical mass with some 15 in-house researchers and students. They direct a range of concepts and tools to questions in systems biology, at scales of organ, individual and community, and in applications as diverse as plant-plant sensing, multi-trophic interactions, ‘industrial’ genotypes and GM coexistence policy.

Modelling and various mathematical approaches now permeate much of the science and some of the applications in EPI. A common and defining feature of the work is the exploration of 'the individual' in 'the system', in which the interactions among individual organisms, organs or cells give rise to emergent properties not predictable from the characteristics of the individuals themselves. Biologists, modellers and software developers combine their skills to address central and essential challenges in modern biology. The examples below are of current work (main funders in parenthesis).

Seedbanks of arable east Scotland 2007

The lack of a baseline on which to assess differences between farming preferences and soil types necessitated this extensive study of seedbanks in representative fields in the east of Scotland from Moray, through Aberdeenshire, Angus, Fife and the Lothians to the Borders. Soil samples were collected in 2007 from more than 100 fields. The aims were to see whether seedbanks differed in relation to soil, latitude, crop rotation and management inputs. The information will contribute as a reference and baseline along with data on soil physical and microbiological properties, vegetation, agronomy and yield as part of the RERAD workpackage on Sustainable Crop Systems.

The methodology, using the emergence method,  was similar to that used in the FSEs. Soil samples were taken by field teams from SCRI and SAC and laid out in trays in the glasshouses at SCRI in Dundee. The first flush of seedlings, mainly of spring-germinators and generalist species was measured for several months after sampling. In the autumn of 2007, the soil in each tray was remixed and emergence was again measured, this time capturing the autumn-germinators. The advantage of this substantive database is the wide range of associated measurements, much more than in the FSEs, that should enable us to quantify the seedbank's dual role of weed burden and base of the arable food web.

The FSEs (the UK's GM crops trials) 1999-2005

All seedbank measurements in the FSEs were carried out by SCRI using the emergence method applied to soil sampled from the 250 or so sites used in the experiment by field staff from CEH, Rothamsted Research and SCRI.

The method was developed in 1999 on three spring and four winter oilseed rape sites. On the basis of these initial measurements, the group estimated that differences of 1.5- to 2- fold in seedbank density between treatments would be detected from around ten samples of each one litre of soil from each treatment (half field). In the event, the estimates proved correct.

From 2000 onwards, a baseline sample was taken before the treatments were applied at each of the 250 sites, and repeat samples from the same locations 12 and 24 months later. At the height of activity in 2001, thousands of trays containing soil were spread throughout several large cubicles in glasshouses equipped with temperature control and shading. In all, SCRI's seedbank records in the FSE comprise the largest arable seedbank survey in the UK.

Clone of Sustainability Research Platform at Balruddery Farm

A new experimental research platform is being established at Balruddery Farm for long-term studies on arable sustainability.Photograph of a poppy field

The overall goal is to test whether or not potential solutions for sustainable agriculture arising from the current RERAD workpackages, actually result in improved arable biodiversity, resilience, crop productivity and yield stability at a commercial, field-scale over at least four rotation cycles (>20 years).

To do this, we will design a sustainable cropping system based on existing research at SCRI that optimises inputs, yield, biodiversity and ecosystem processes. The effect of this ‘sustainable’ system on long-term trends in yield and system health will be tested by comparison with current commercial practice.

Data mining techniques for analysing complex simulation models

Individual-based models of plant populations and communities can be highly complex, reflecting the underlying dynamics of the natural systems under study. Analysing and understanding model behaviour can be extremely challenging. To address this we are collaborating with researchers from the Department of Knowledge Technologies of the Jožef Stefan Institute, Slovenia on the application of data mining and machine learning techniques to the analysis of the IBMs we have developed.

Model analysis

Machine learning methods are being used to analyse the relationship between simulation outputs with IBM inputs (model parameters) in order to gain insight into the behaviour of the model system. Here we rely on a Monte Carlo approach in which simulations are based on IBM parameter values sampled at random from across a predefined parameter space. By applying machine learning methods, we can generalise over the specific simulations made and derive more general rules concerning the behaviour of the system.

Wild arable plants - diversity and function

The ecology and biology of wild arable plants are poorly understood. Of the more than 250 plant species to be found on arable farmland, typically five to 10, among which are wild oat, blackgrass, barren brome and cleavers, constitute the main weed burden of arable cropping. Many of the rest, particularly the broadleaf (dicotyledonous) species, support an arable food web that includes insect groups, mammals and birds. Despite the economic, ecological and aesthetic importance of wild arable plants, embarrassingly little is known about their ecology and genetic diversity.

Our first research paper in this new topic demonstrated the lack of basic information for even the common species (Hawes et al., 2005). If arable cropping systems are to be sustainable, then co-existence between crops and weeds must be managed with minimum or no herbicide application. To achieve this, fundamental and strategic research is necessary into the way wild arable plants respond to crops, weather and field management.

The persistence and spread of novel genes

A transgenic, that is genetically modified, oilseed rape (Brassica napus) provides the model system with which we have investigated the dynamics of novel genes in local populations. A stochastic, spatially explicit individual based model (IBM) was developed to simulate the dynamics of the transgene introduced into a non-transgenic population. The model combines life-history and management processes with environmental drivers to examine the effect of these on the spread and persistence of the transgene and the conferred trait.

The model has been used to explore a number of features of this system:

Gene flow patch diagramSpread and persistence in patchy populations - Plant populations, including arable weeds, typically exhibit spatial heterogeneity, that is patchiness, in their distribution. Together with the localised nature of plant to plant interactions this has the potential to affect the dynamics of a population and the spread and persistence of the genes they posses. Model simulations have shown that small scale spatial heterogeneity in the distribution of transgenic plants combined with localised pollen dispersal reduces mixing between populations and acts to limit the spread of the gene through the population.

The seedbank

Seeds from the arable seedbank - photograph by Gladys Wright/Stewart MaleckiBuried living seed - the seedbank - is central to the composition and succession of disturbed vegetation, allowing regeneration after agricultural or natural clearing of the existing plant cover. In arable-grass systems, the seedbank is the source of both the weed burden and the vegetation that supports the arable food web. Of around 250 species typically found in arable regions, only five to 10 are economically important as competitors to crops. Few of the poisonous species that were once a concern now remain in fields. Most of the other seedbank species have been reduced, in particular the broadleaf or dicotyledonous species that provide food and habitat for detritus feeders, herbivores, parasitoids, predators and pollinators. Knowing the seedbank is therefore essential for balancing the weed burden and biodiversity.

Syndicate content