Guidelines for Environmental Risk Assessment and Management
[This document refers, in a number of instances, to the then Department of the Environment, Transport and the Regions (DETR). The text of this document has not been updated since the transfer of environmental protection functions to Defra.]
Annex I
Case studies
A1.1 Risk assessment for the release of genetically modified sugar beet
This annex is adapted from an application for consent to release genetically modified (GM) sugar beet submitted under the Genetically Modified Organisms (Deliberate Release) Regulations 1992 (as amended 1995 and 1997).
The release is for research purposes only and relates to a small-scale trial to assess the field performance of GM sugar beet modified for tolerance to the herbicide glufosinate ammonium.
The GM sugar beet has two inserted genes: a bacterial gene (pat) which encodes the enzyme phosphinothricin-N-acetyltransferase enabling the sugar beet to withstand applications of the herbicide glufosinate ammonium; and a bacterial gene (neo) which encodes the enzyme neomycin phosphotransferase which confers resistance to the antibiotic kanamycin.
The GM sugar beet seeds were sown in March 1997 in a sugar beet growing area and were monitored at weekly intervals during the release to ensure that any bolting plants (plants which enter their reproductive phase) were removed and destroyed before the onset of flowering. At harvest, in October 1997, the beets were collected and any remaining plant material pulverised and ploughed into the soil. After harvest, the release site was planted with a cereal crop and monitored for one year to ensure any regrowth of the sugar beet (volunteers) was destroyed.
The risk assessment described below indicates that the release of this genetically modified organism (GMO) may pose a low risk to the environment because of the likelihood of gene transfer to wild relatives or sugar beet crops, and/or survival of the GMO. Any risk from these hazards is reduced to near zero by destroying bolters during the release and any volunteers during post-release monitoring.
Hazard identification
Capacity to survive, establish and disseminate in the environment
The GM sugar beet may have a greater tendency to overwinter, establish and invade habitats that are normally beyond sugar beet's range. It may also become an agricultural pest, ie a weed.
Potential for gene transfer between the GM sugar beet and other organisms
Three scenarios are envisaged:
- GM sugar beet cross-pollinates commercial sugar beet crops;
- GM sugar beet cross-pollinates a wild relative - B. maritima
- to produce
kanamycin-resistant glufosinate-tolerant hybrids; and - the glufosinate gene and kanamycin gene in GM sugar beet are transferred
to
micro-organisms resident in the gut of humans and animals during ingestion.
Products of expression of inserted genes
The GM sugar beet produces two novel enzymes: phosphinothricin acetyltransferase that confers herbicide tolerance and neomycin phosphotransferase that confers tolerance to the antibiotic kanamycin. Both gene products are non-toxic to plants, animals or man, but the allergenic properties of these proteins are unknown, as are the indirect effects of the expression of these foreign proteins in sugar beet.
Phenotypic and genetic instability
The loss of kanamycin resistance and herbicide tolerance would make these plants sensitive to kanamycin and glufosinate ammonium. This in itself does not represent a hazard but re-integration at a different site in the sugar beet genome could inactivate the expression of other genes and this might be harmful.
Pathogenicity to other organisms
Neither gene (neo and pat) are associated with pathogenic traits in bacteria. Neither strain of the donor organisms - Escherichia coli K12 and Streptomyces viridiochromogenes - are associated with pathogenicity. No hazard is identified.
Identification of consequences
Survival, establishment and dissemination
- The invasion of natural and semi-natural habitats and the erosion of species diversity; and
- A significant agricultural weed.
Potential for gene transfer between the GM sugar beet and other organisms
- B. maritima hybrids with glufosinate and kanamycin tolerance - the ecological consequences of this are uncertain. The traits might confer a selective advantage to these plants increasing their weediness that might result in habitat invasion.
- B. maritima hybrids may become an agricultural weed.
- Cross-pollination of sugar beet crops to yield seed that could lie dormant in locations at some distance from the release site. Resulting hybrids may be more resistant to weed management programmes to control volunteer beet in following crops.
- The transfer of the kanamycin resistance gene from sugar beet to micro-organisms which reside in the human or animal gut might reduce the effectiveness of treating bacterial infections with kanamycin.
Products of expression of inserted genes
- Neomycin phosphotransferase and phosphinothricin-N-acetyltransferase may cause an allergenic response in some people exposed to pollen expressing these two proteins.
- Interaction of these proteins with the metabolic pathways of sugar beet may give rise to toxic compounds.
Phenotypic and genetic instability
The loss of kanamycin resistance and herbicide tolerance would make these plants sensitive to kanamycin and glufosinate ammonium. This in itself does not represent a hazard but, the genes may integrate at a different genetic location disabling the expression of genes which may lead to harmful consequences.
Estimation of the magnitude of the consequences
Survival, establishment and dissemination
The proposed field trial with GM sugar beet is a small-scale test (less than 50 m2) of limited duration (one growing season). It will take place within an agricultural environment where sugar beet is regularly grown. The magnitude of the consequences for the agricultural and natural environment is estimated as: Mild.
Gene transfer between the GM sugar beet and other organisms
B. maritima hybrids and sugar beet crops with kanamycin and glufosinate tolerance:
The scale of any impact on the surrounding countryside would be limited to a maximum distance of two miles initially. If a selective advantage is conferred by these genes they will be perpetuated and their gene frequency in the B. maritima population will rise. The time taken to see an environmental effect may take several years. If they become troublesome weeds in the agricultural environment then the time to see an effect will probably be shorter. The magnitude of consequences is estimated as: Mild.
Transfer of antibiotic resistance:
The scale of the impact would be restricted to animals fed the sugar beet as fodder and humans ingesting the sugar. Kanamycin is not important in clinical and veterinary medicine. The treatment of infections by pathogenic bacteria would not be compromised by the uptake of the kanamycin resistance trait. The magnitude of consequences is estimated as: Mild.
Products of expression of inserted genes
Toxicity:
Birds and animals that visit the site will be exposed. Any sugar harvested from the trial (50 m2) will be mixed with sugar from conventional sugar beet crops thereby reducing any toxicity to people. The magnitude of consequences is estimated as: Mild.
Allergenicity:
The frequency of bolter formation is variable but normally less than 5% of the crop. Pollen production will therefore be limited, but any pollen produced can disperse significant distances and Peterborough is located 2 miles away so these people may be exposed for two months in the year. The magnitude of consequences is estimated as: Mild.
Phenotypic and genetic instability
Effects are likely to be of limited scale and duration. The magnitude of consequences is estimated as: Mild.
Probability of the consequences
Survival, establishment and dissemination
The recipient:
Sugar beet root fragments not destroyed at harvest can survive through mild winters and sprout to form volunteer plants the following year. Sugar beet is, however, a very uncompetitive plant in agricultural, natural and semi-natural habitats. It is non-invasive by nature and self-sustaining populations of sugar beet have never been reported in any habitat.
The genes inserted:
Tolerance to kanamycin and glufosinate are traits that are unlikely to increase the capacity of the GM sugar beet to survive, establish and invade natural or semi-natural habitats. Kanamycin and glufosinate are not factors which influence plant competitiveness in nature.
The GM beet:
Laboratory studies indicate that the parent and GM sugar beet are very similar in several characters including frost tolerance. Therefore, considering the characteristics of the parent, the properties of the genes inserted and preliminary data on the characteristics of the GM beet, the probability that GM sugar beet will have a greater capacity to survive, establish and invade natural habitats is estimated as: Negligible.
Survival of the GM sugar beet after the trial may allow limited survival and establishment in the agricultural environment where it may become a weed. Tolerance to glufosinate will give the GM sugar beet a selectable advantage when this herbicide is applied in weed management programmes. However, the GM sugar beet is sensitive to other herbicides used to control sugar beet volunteers and so its establishment is likely to be transient. Therefore, the probability of the modified sugar beet becoming an agricultural pest is estimated as: Low.
Gene transfer
Sugar beet is a biennial, growing vegetatively in the first year and flowering in the second; sugar beet is harvested at the end of the first year. However, a minority of individuals (2-5%) flower in the first year, produce pollen and set seed. Pollen is wind rather than insect dispersed and can travel in excess of 2 miles. The transfer of the transgenes to sugar beet crops which are located 100 metres away from the site and to B. maritima which grows at this location is therefore likely.
Beta maritima:
Glufosinate- and kanamycin-tolerant B. maritima are unlikely to become more competitive and invade other habitats including the agricultural environment as a result of inheriting these traits. Therefore, the probability of B. maritima hybrids invading new habitats and becoming an agricultural pest is estimated as: Low.
Sugar beet crops:
Hybrid sugar beet seed which emerge in following crops will be susceptible to normal agricultural practices of weed control and will not persist. Therefore, the probability of sugar beet hybrids becoming an agricultural pest is estimated as: Low.
Bacteria:
High levels of the kanamycin resistance already exist in bacterial populations. Transfer of this gene from plants to bacteria (from where it was derived) would not lead to a significant increase in background levels. Also, the frequency of horizontal gene transfer from plants to bacteria is very small. Therefore, the probability of GM sugar beet compromising the therapeutic use of kanamycin in people which consume the sugar is estimated as: Negligible.
Products of expression of inserted genes
Toxic effects:
Toxicity due to the insertion of these non-toxic genes into sugar beet is unlikely. Furthermore, the mixing of sugar derived from this small plot with sugar derived from conventional crops will dilute any toxins produced. Therefore, the probability of toxic effects is estimated as: Low.
Allergenic effects:
These are not expected because the proteins do not resemble known allergenic proteins. Also the production of pollen from the plot will be small (bolter frequency X plot size). However, people in Peterborough could be exposed for two months. Therefore, the probability of allergenicity is estimated as: Low.
Phenotypic and genetic instability
Observations from successive generations show stable transgene insertion and expression. Therefore, the probability of gene instability is estimated as: Negligible.
Evaluation of the significance of the risk
The overall risk of damage to human health and the environment is low to effectively zero as the following components of risk have been assessed as:
| Survival, establishment and dissemination | Low |
| Gene transfer to sugar beet crops | Low |
| Gene transfer to B. maritima | Low |
| Horizontal gene transfer to bacteria | Near Zero |
| Phenotypic and genetic instability | Near Zero |
| Toxicity and allergenicity of gene products | Low |
| Pathogenicity to other organisms | Near Zero |
Risk management
Survival, establishment and dissemination
At harvest the GM sugar beet material will be pulverised and ploughed into the soil. Therefore, the risk of survival, establishment and dissemination is reduced from low to near zero.
Toxicity:
No GM sugar beet will be permitted to enter the human food or animal feed chain. Therefore, the risk of toxic effects is reduced from low to near zero.
Allergenicity:
Bolters will be removed from the GM sugar beet before the onset of flowering to prevent pollen dispersal. Therefore the risk of pollen-mediated allergenic effects is reduced from low to near zero.
Gene transfer to crops and B. maritima:
Bolters will be removed from the GM sugar beet plants before the onset of flowering to ensure gene transfer to neighbouring sugar beet crops and B. maritima is prevented. Therefore the risk of habitat invasion from B. maritima hybrids is reduced from low to near zero.
Monitoring
During the trial
The GM sugar beet will be monitored at weekly intervals during the release to ensure that any bolting plants (plants which enter their reproductive phase) are destroyed before the onset of flowering.
Post trial
The release site will be inspected twice in the year following the trial for the effective control and destruction of any GM sugar beet.
A1.2 Risk assessment for road transport: a semi-quantitative methodology
This case study was undertaken at both a screening level (Tier 1) to identify the primary areas of concern, and a generic level (Tier 2) to quantify some of the risks facing the environment. The risk assessment was undertaken by the Environment Agency which, although not having a formal remit in relation to road transport, needed to take a holistic long-term view of the associated issues since they have a bearing on its ability to regulate and manage the environment effectively. The Agency also needed to be informed in discussions with Government and other organisations. A risk-based framework was imposed on the information available for a wide range of issues. At such a strategic level, it is necessary to make and record broad assumptions and understand how uncertainties in information affect the final outcome.
TIER 1 - RISK SCREENING
Hazard identification
Many authoritative studies have highlighted the severe and widespread environmental impacts of road transport. From discussions with experts in the field, the Environment Agency determined the following environmental consequences to be of particular concern:
- Raw materials
- Road construction
- Road maintenance
- Road run-off
- Accidents and spillages
- Exhaust emissions
- Waste and tyre disposal
The full range of consequences arising from road transport has yet to be fully established. From discussions with experts in the field, the following consequences were determined to be of concern:
- Climate change
- Poor air quality
- Poor soil quality
- Poor water quality
- Flooding
- Impact on water resources
- Ecological damage
- Landscape
- Property
- Human health
- Quality of life
An expert elicitation exercise was carried out to determine the priorities with respect to the consequences. This was based on three key factors: the significance of the consequences, whether the Environment Agency has a policy or formal remit in the area, and whether there is capacity to mitigate the effects.
The following consequences were deemed to be important for certain hazards: water quality, flooding and water resources, ecological quality, soil quality, air quality and climate change. Many other issues were considered to be as important but too far outside the Agency's remit.
Probability of the consequences
At the screening level, the probability was determined to be unity, as the activities were known to occur, and the consequences had been linked by scientific knowledge and professional experience to the hazards.
Evaluation of the significance of the risk
The expert elicitation process enabled the consequences for each hazard to be prioritised. As a result the following consequences for each hazard were determined to be of sufficient concern to warrant further investigation at Tier 2.
Road construction and maintenance
Ecological quality, water quality, flooding and water resources, ecological quality and/or habitat loss.
Road use
Water quality, climate change, air quality, soil quality.
TIER 2 - GENERIC QUANTITATIVE RISK ASSESSMENT
Hazard identification
The primary focus for hazard identification at this stage was the Tier 1 risk screening described above. A wide range of risks was identified and the example given here for illustration is that of road construction giving rise to water quality problems.
A range of scenarios was discussed by an expert group, and the following hazard scenario was specifically chosen for further assessment:
- The transport of sediment into watercourses during the construction of roads.
Suspended solids are responsible for the siltation of spawning grounds for migratory fish such as salmon, and for harm to habitats and aquatic macrophytes. The first of these consequences can lead to significant declines in populations of key fish species.
Estimation of the magnitude of the consequences
The magnitude of the consequences was determined through the use of event trees containing information derived from scientific literature, monitoring programmes, and expert opinion.
The consequences were deemed to be high when intense storms coincide with lower river flows in summer.
Probability of the consequences
The contribution to suspended sediment concentrations from road construction was represented in an event tree. Under average conditions most sediment (0.89) is incorporated into the ground works of the road. Around 0.02 of the sediment is removed in control structures and less than 0.05 is transported into surface water. The remainder is deposited elsewhere in the catchment from dirty vehicles or by wind erosion and deposition.
Evaluation of the significance of the risk
The risk was determined to be significant for this impact. Discharges from road construction sites may be up to 14 times greater than the Environmental Quality Standard (EQS) for suspended solids in surface waters taken for potable supply.
Risk management
The sediment trapping efficiency of control structures has been taken into account in determining the probability of the consequences. Clearly further efficiency improvements in this area would be useful. In addition, the timing of construction is important, and could contribute to risk management planning. In particular, construction during times of lower river flow may be inappropriate.
Monitoring
The Environment Agency will detect the consequences of such activities through its General Quality Assessment scheme for surface waters. The monitoring of sediment erosion by those responsible for construction will greatly assist early remediation of the hazard.
A1.3 Risk assessment of coastal flooding: a semi-quantitative methodology
This risk assessment provides an indication of the relative level of risk for various sections of developed land behind a coastal defence. The risk relates to flooding with consequential damage to property, and harm to humans, including injury and potential loss of life.
This case study was undertaken at a Tier 1 screening level to assess coastal flood risks to different types and mixes of development from a well-characterised flood hazard. Its principal use was in combining what was known quantitatively about the flood hazard with qualitative information that could reasonably be inferred regarding community vulnerability and response, to provide an overall risk screening of the area behind the existing sea defences.
TIER 1 - RISK SCREENING AND RISK PRIORITISATION
Hazard identification
Flood waters
The principal hazard is:
- a threat to human life and property from saline water overtopping or percolating through a coastal defence structure.
The hazard is initiated by a combination of wind, waves, storm surges and astronomical tides which exceed the strength of the coastal defences and lead to flooding of the land area behind. The volume, depth and velocity of flood waters are important in determining the consequential harm.
Secondary hazards
The force of water can introduce further hazards, the adverse impacts of which can be significant:
- debris including beach shingle, cars and skips may be picked up and carried by the flood waters and may also cause injury and structural damage.
Flood waters
The overriding consequence of such an event is that flood waters inundate a land area behind the coast. In the context of this study, the principal resulting consequences of this event were considered to be:
- injury to humans including potential loss of life from debris, flowing and ponded water;
- structural damage to residential and commercial properties due to debris and flowing water; and
- water damage to residential and commercial properties due to ponded water.
Importantly, the presence of flood waters can also lead to:
- restricted access to roads, disruption to travel, and delayed access by emergency services.
For this case study, the flood hazard had been previously characterised from specialist studies and the severity of various flood events assessed by considering:
- the mechanism of flooding (overtopping or percolation);
- the flood volume;
- the mode of action for moving water (direct wave impact, surging water flow, secondary impact of moving water, steady water flow or ponded water); and
- the presence of debris in areas of land behind the coast.
Flood volumes and water velocities for each land section were estimated. This assessment provided an inventory of potential consequences by land section according to whether the flood event was initiated by overtopping of, or percolation through, the coastal defence and the amount of water involved in each case.
Estimation of the magnitude of consequences
The nature of harm posed by the flood hazard and the vulnerability of receptors determine the magnitude of the consequences. Here, each land section was assessed by considering the severity of the flood event, the nature of harm that could result and the vulnerability of the development in each case; the latter with reference to the development 'mix' of each land section. The principal consequences were weighted (in italics) according to the nature of harm posed within the context of this study:
- no significant damage (1)
- minor water damage to properties (10)
- minor structural damage (50)
- major structural damage or injury (500)
- loss of life (2000)
The vulnerability of the development type in each section of land to these consequences was assigned an indicative 'value', from 1 (for low value structures) to 5 (for developments with a high residential mix).
For each land section, a product of the exceedance probability referred to below (eg 0.02) and the principal consequence type (eg 500) for each land section, weighted according to its vulnerability (eg 4), provided a relative assessment of the magnitude of the consequences. Damage scores were aggregated for floods of selected exceedance probability (and severity).
Probability of the consequences
Flooding is a natural and episodic risk. It is important to identify the consequences that may occur for any given probability (and severity) of the hazard. In this regard, flood risk is inherently distinct from most chemical risk assessments.
In this case study, the magnitude of the consequences was estimated for floods with annual exceedance probabilities of 0.1, 0.02 and 0.005 (ie with mean return intervals of 10, 50, and 200 years) as described above.
The resulting damage assessment or 'damage profile' (Table A1.1 below) represents the relative probability and magnitude of the consequences for each section of land. The higher the ranking, the higher the estimated damage. It assumes, however, that the damage sustained is independent of community response in each land section and one of the aims of the study was to factor this in accordingly. For example, the probability of harm to humans is dependent on their ability to respond to the flood in advance and whilst it occurs, and also to the ability of emergency services to respond to their needs.
To account for these issues, the generalised sections of land behind the existing defences were each scored according to their reliance on various community response factors. The main factors controlling the probability of harm being sustained were deemed to be (weighting according to relative importance in italics):
- the availability of access by emergency vehicles (5);
- the availability of easy routes of evacuation to shelters (5);
- the amount of advanced warning available (5);
- a prior knowledge of evacuation procedures (3);
- the availability of access to shelter within the property (2); and
- the existence of protection for properties (drop boards, etc.) (2).
The extent to which these factors were likely to influence community response in each land section following the issue of a red flood warning was accounted for by scoring them on a 1 (low influence) to 5 (high influence) basis. Local flood defence staff with knowledge and experience of procedure and likely response were used to elicit individual scores, the means of which were summed and taken to represent a 'response profile'.
Evaluation of the significance of the risk
Table A1.1 shows damage, response and risk (sum of damage and response) profiles for the land sections studied. To avoid over-attribution to the scores, sections were banded and assigned a qualitative risk designation. In a relative risk context, the very high designation represents a risk of substantial damage with little ability to respond on receipt of a red warning. High represents substantial or moderate damage with possible access to escape routes. Medium represents lower damage with reasonable ability to respond due to the level of emergency access, etc. These designations were interpreted in the context of flood probability that was well-characterised, assessed in detail and had accounted for storm severity, probability and sea defence performance.
Risk management
The principal risk management action determined from this assessment and other supporting data is a recommendation that development or re-development be steered away from areas where the risk is deemed to be high or very high.
In general, the case study has shown the need to consider community response factors and strengthen the public awareness of how to respond to flood hazards. It has highlighted the importance of emergency planning and the ability to provide the necessary resources including people, equipment and materials for dealing with flooding.
Monitoring
The monitoring of the risk will be undertaken within an overall flood monitoring and warning framework. In addition, it will be necessary to monitor the development of properties within the flood plain, in areas that are deemed to be at greatest risk from coastal flooding.
| Table A1.1 Damage, response and risk 'profiles' and designations for sections of land behind a coastal defence scheme | ||||
|---|---|---|---|---|
| Land section | Damage profile | Response profile | Risk profile | Designation |
| 1 | 350 | 102 | 452 | Very High |
| 2 | 125 | 102 | 227 | High |
| 3 | 125 | 97 | 222 | High |
| 4 | 100 | 98 | 198 | High |
| 5 | 105 | 87 | 192 | High |
| 6 | 100 | 92 | 192 | High |
| 7 | 105 | 84 | 189 | High |
| 8 | 105 | 69 | 174 | High |
| 9 | 48 | 53 | 101 | Medium |
| 10 | 14 | 84 | 98 | Medium |
| 11 | 3 | 87 | 90 | Medium |
| 12 | 1 | 79 | 80 | Medium |
Page published 2 August
2000;
Page last modified
19 September, 2002