Following numerous discussions of the risks associated with biological control, (see Howarth, 1991; Simberloff & Stiling 1996; Thomas & Willis 1998) literature was reviewed in order to investigate whether biological control was an environmentally friendly or a risky business. Although a lack of firm evidence suggests that risks may be ‘perceived’ rather than ‘real’, the release of the biological control agent Harmonia axyridis by countries lacking in regulation has severely damaged biological control’s reputation and ecosystems all over Europe. Biological control is the most sustainable, cost efficient and natural method of pest management and therefore it should be used to its full potential. Harmonized regulation is required to improve biological control’s reputation by preventing the release of ‘risky’ agents in the future. Regulation should facilitate rather than restrict the use of biological control below its potential. Regulation should be specifically designed for biological control and should enforce the use of an environmental risk assessment (ERA). Scientifically based methodologies are required to ensure an efficient ERA is conducted for potential biological control agents. An efficient ERA should identify unsuitable agents as early as possible to reduce cost and time requirements. This will allow the continued growth of the biological control industry. Biological control should be utilised as part of Integrated Pest Management to ensure the most efficient control of each pest.
Insects are the foundations of ecosystems, vectors of disease and agricultural pests around the world (Gassmann et al. 2009). Table 1 shows that as agricultural pests, insects cause economic losses of billions every year.
The economic damage caused by insect pests (see Table 1) and the increased consumer demand for blemish free produce has led to the utilisation of different approaches to pest management (Castle et al. 2009). For example, modern pesticides have been used since their development in the 1940s and it has recently been estimated that 8000 metric tons of insecticide (FAO, 2009) are used around the world at an approximate cost of $40 billion every year (Akhabuhaya et al. 2003).
The advantages of pesticide use include the short time between application and effect, the eradication of the pest in the area of application and the predictability of success (Bale et al. 2008). The speed and assumed efficiency of pesticides led to their great popularity up to the 1970s when concerns arose about their effects on health and the environmental (see Table 2).
The rise in public concern and increased evidence of the negative effects of pesticides (see Table 2) led to the reduction of their use in the 1970s (Chiu & Blair, 2009). Pesticides associated with the more serious risks were made illegal, such as DDT in 1984 (Attaran & Maharaj, 2000). The great reduction in pesticide use over the last 50 years has allowed other pest management techniques, such as biological control, to be further developed and utilised (Suckling & Brockerhoff, 2010).
Biological control has received great support due to its natural mechanisms. For example, van Lenteren (2005) estimates that 95% of native insects are already controlled through natural biological control. In addition, a continuous increase in international trade and travel has led to increased insect dispersal between countries (Waage & Mumford, 2008). For example, 62,000 pests were reported following an Animal and Plant Health Inspection Service (APHIS) studycarried out on airplane and boat passengers in the USA (Dunn, 1999). Also, there has been a recent increase in the number of crops grown in glasshouses across Europe. Glasshouse conditions are much more suited to invasive insects, so this has allowed increased establishment (Hunt et al. 2008). The movement to reduce pesticide use, popularity of natural control, increased levels of insect invasions and the use of glasshouses to grow crops greatly increased the demand for biological control in the 1980s (Sheppard et al. 2003).
Biological control is the use of living organisms to actively reduce the population density of a pest species. A biological control scheme is deemed a success if the pest population densities are lowered to the extent that they are no longer considered an economic or environmental threat (van Klinken & Raghu, 2006).
Biological control can be further classified as classical, augmentative (inundative) or conservation. Classical biological control is the permanent reduction in the population of an exotic pest species through introduction of its exotic natural enemy. The introduced control agent is required to establish as it is meant for self sustaining control of the pest (Eilenberg et al. 2001). A classical biological control scheme that has reached great success is the use of Rodolia cardinalis against the invasive scale insect Icerya purchasi. Following its accidental introduction into California, I. purchasi was threatening to ruin the Californian citrus industry. R. cardinalis was selected as a monophagous natural enemy and 128 individuals were introduced to California. Populations of I. purchasi were controlled within a year (Frank & McCoy, 2007). Classical biological control schemes that only reached partial success, i.e. pest population densities were reduced but the agent did not fully establish, led to the development and use of augmentative biological control.
Augmentative biological control is the release of natural enemies in an inundative or seasonal inoculative manner (van Lenteren, 2005). Inundative biological control is the mass release of biological control agents to quickly reduce a pest population density (Eilenberg et al. 2001). Inundative control agents are not meant to establish so agents may require reintroduction. An example of this is the mass release of the parasitoid Trichogramma brassicae to control the European corn borer (Ostrinia nubilalis) (Bigler, 1986). Seasonal inoculative biological control is the release of a natural enemy species with the aim that they will reproduce, survive and control pests throughout a crops growing season (van Lenteren & Woets, 1988).
Conservation biological control is the alteration of the environment towards one more suited to the pest’s natural enemy. For example, the provision of extra host plants (Anethum graveolens and Coriandrum sativum) for the natural enemies (Edovum puttleri and Pediobius foveolatus) of the Colorado potato beetle (Leptinotarsa decemlineata) (Patt et al. 1997). The aim is a long term increase in natural enemy populations resulting in increased control of pests (Landis et al. 2000).
Until the mid 1980s, the introduction of over 2000 natural enemy species and the successful control of over 165 invasive pest species, led to the belief that biological control was an environmentally safe and cost effective alternative to pesticides and GM organisms (van Lenteren et al. 2006a). However, Howarth’s (1991) argument that there were serious risks associated with biological control was followed by a flood of papers discussing evidence of similar risks (for example, Simberloff & Stiling, 1996; Louda et al. 2003). It was recognised that an unsuitable biological control agent may cause the problems associated with an invasive insect.
The potential risks of biological control include the possibility that the exotic agent could be poisonous, allergenic or the vector of a disease that is dangerous to humans (Howarth, 1991). Introduced species could become essential crops pests or they could indirectly cause an increase in other crop pest populations (Howarth, 1991). For example, the reduction in target pest species may allow previously outcompeted insects to increase population size to pest densities (Kenis et al. 2009). Biological control agents may kill a plant that other insects rely on for food or shelter (Simberloff & Stiling, 1996). For example, the destruction of ash by the Chinese buprestid Agrilus planipennis has threatened the whole Frazinus genus of leptidoptera (Kenis et al. 2009). Further-more, biological control agents may predate or outcompete insects involved in plant in tri-trophic interactions or they may kill plant essential pollinators (Simberloff & Stiling, 1996).
The greatest risks of biological control are those that impact on the environment. These risks include non target effects (Hokkanen, 2003). For example, the generalist biological control agent Compsilura concinnata has threatened the extinction of six non target Lepidoptera species in North America (Boettner et al. 2000). The effect of a biological control agent on non target organisms may be direct, such as the parasitisation of a non target host when the target is unavailable, or the preference of exotic prey over the target (Simberloff & Stiling, 1996; Kriticos et al. 2009). For example, Cotesia glomerata parasitised the non target butterfly Pieris oleracea which is now at risk of extinction (Van Driesche et al. 2003). A reduction in non target population size may reduce their genetic diversity and therefore ability to adapt to future environmental changes (Kenis et al. 2009). Introduced agents may hybridise with native species or be a vector of a disease to which native invertebrates have no resistance (NRC, 2002).
The possible indirect effects of biological control include resource competition (Delfosse, 2005). For example, the introduced parasitoid C. concinnata appears to have outcompeted the native silk moth parasitoid (Lespesia frenchii) in New England (Parry, 2009). Biological control agents may share predators with a native herbivore. This may result in the disruption of natural biological control: reduced predation of the native herbivore may allow its population to increase to pest densities. Severe alterations to the ecosystem may occur if the introduced species affects an ecosystem’s keystone species or becomes a keystone species (Wagner & Van Driesche, 2010). This would alter natural co-evolved relationships (Strong & Pemberton, 2000) inducing evolutionary changes (Kenis et al. 2009). Finally, biological control agents may disperse from their area of introduction. This means the risks described are relevant to any neighbouring habitats and countries (Howarth, 1991).
The increased discussion of these risks has led to demand for regulation implementing a thorough risk assessment to ensure that only ‘safe’ biological control agents are released (Delfosse, 2005). Numerous publications have been released by organisations and countries (such as IPPC, 1997; EPPO, 1999; EPPO, 2001; EPPO, 2002; EU-funded ERBIC, 1998-2002; OECD, 2004; IOBC/WPRS, 2003; IPPC, 2005; REBECA, 2007). These publications provide useful regulatory guidelines but they are not legally binding, they are also too vague as they do not state appropriate Environmental Risk Assessment (ERA) methods (Kuhlmann et al. 2006). Many countries have not produced regulations or do not actively utilise them and this has resulted in extremely patchy regulation across the globe.
Advocates of biological control argue that the discussions of the potential risks do not provide adequate evidence that observed effects were due to biological control (Lynch et al. 2001). Also, insect invasions occur accidently all the time with little evidence of any harm and therefore, an increase in regulation is not required.
To answer the question posed (is biological control an environmentally friendly or risky business?) this review will address the following questions: Are the perceived risks of biological control founded on relevant evidence? What and where are the current biological control regulatory systems? Are ERA methods described and if so are they efficient ortoo strict, expensiveor vague? Do they acknowledge the differences between classical and inundative biological control and are they applicable to both? What should an efficient ERA comprise of? Is biological control compatible with other pestmanagement schemes?
The initial literature search was conducted to identify the scope of the topic: Web of Knowledge was used because this search engine has a database holding a wide range of journals. The use of Web of Knowledge also has the advantage of being able to read the abstract before downloading the paper and the search can be restricted to ‘Science Citation Index Expanded’ to increase the relevance of results. In order to study the full history of biological control, there was no restriction placed on the year of publishing and a range of broad key words were used including “‘insect biological control’ AND history”.
Following the initial search and study of primary papers, key areas of interest were identified where further research was required in order to answer the question posed. Papers of interest were found using article reference lists and topic specific searches. These searches were conducted using key words for each area that required more detailed research. For example, papers on the problems associated with pesticide use were searched for using PubMed. PubMed is a biomedical database so was a more relevant search engine for this particular topic. Key words used included “pesticide limit* AND human health”. Once found, citation searches were used on key papers to help establish their importance and accuracy.
Boolean operators were used to combine keywords in the Topic search. An asterisk (*) was typed at the end of words that could have various endings. This allowed a wider search including titles with singular and plural word forms. The ‘OR’ operator was used between possible key words to allow for variation in terminology. When a search found too many results (over 100), ‘AND’ or ‘NOT’ were used between words and more specific key words were identified to help make the results more relevant. More specific keywords were identified using terminology that was common in the titles of interesting papers. When a search resulted in less than 100 papers all abstracts were read. If the abstract suggested the paper might provide evidence towards answering the question posed, the full paper was read. This search strategy allowed the efficient search of specific papers relevant to each area of interest.
Effects of a Biological Control Introduction on Three Non-target Native Species of Saturniid Moths
Boettner et al. (2000) Conservation Biology, 14, 1798-1806.
To answer the question posed, (is biological control an environmentally friendly or risky business?) this review needs to consider whether or not the risks discussed for biological control are founded on relevant evidence. Examples used to demonstrate non target effects are often criticized because they do not account for native predation causing non target mortality (Lynch et al. 2001). This study is pioneering as it is the first to directly assess the non target effects of the classical biological control agent Compsilura concinnata and compares these effects to native predators.
The effects of C. concinnata on the non targets Hyalophora cecropia, Callosamia promethean and the state endangered Hemileuca maia maia were studied. This experiment was conducted following observations that these non target species populations had declined since the introduction of C. concinnata. Cohorts of 100 H. cecropia larvae, densities of 1 – 100 C. promethean larvae and wild H. maia maia eggs were observed in the field. The percentage mortality of each species that was due to C. concinnata was calculated.
Boettner et al. (2000) found that 81% of H. cecropia mortality was due to C. concinnata (see Table 5). 67.5% of C. promethean larvae and 36% of H. maia maia mortality were also found to be due to C. concinnata.
Boettner et al. (2000) found that C. concinnata was responsible for the majority of non target deaths and that the numbers of individuals surviving may be less than the minimum viable population size for each species. Biological control should never result in a loss of biodiversity (Kuris, 2003).
Methods utilised were supported by previous studies and were conducted in realistic conditions. This is important because host selection is effected by physiological conditions including the availability of hosts (van Lenteren et al. 2006b). However, the species were reared in a laboratory before and after exposure to parasitoids. This is undesirable as larvae were reared in unnatural conditions which could alter the parasitoid’s host selection (van Lenteren et al. 2006b). In addition, repeats should have been conducted for each experiment to allow for natural variation in host selection (Bigler et al. 2005).
Although this paper accounts for mortality due to native predators, it is still limited by the assumption that the observed reduction in saturniid moth populations was due to increased levels of predation. Other possible reasons for non target population declines and the parasitisation rate prior to the introduction of C. concinnata require consideration. Van Lenteren et al. (2006b) states that firm evidence non target population declines are due to biological control is often lacking. Therefore, it may be argued that this study does not provide substantial evidence that C. concinnata has caused the observed decline in non target populations.
Overall, Boettner et al. (2000) provide evidence that C. concinnata parasitises non target species. Since its initial release in 1906, C. concinnata has been observed parasitizing over 180 native North American species. In combination with other evidence of non target effects and with the knowledge that non target studies are rarely conducted following introductions, this study assists in the argument that non target effects are a reality (Louda & Stiling, 2004). Therefore, biological control has the potential to be environmentally risky.
Changes in a lady beetle community following the establishment of three alien species
Alyokhin & Sewell (2004) Biological Invasions, 6, 463-471.
The successful introduction of Rodolia cardinalis was followed by the introductions of numerous coccinellids without a thorough risk assessment (van Lenteren, 2005). As a result, many indirect effects have been recorded. However, numerous experiments that appear to provide evidence for indirect effects have been criticized because they took place over such a short time scale. This means that limited conclusions can be drawn because they do not allow for natural variation in species abundances (Alyokhin & Sewell, 2004). Long term research is required in order to provide adequate evidence for the indirect effects of biological control. This is particularly relevant to coccinellids as they are known for population fluctuations (Alyokhin & Sewell, 2004).
This paper provides evidence of the biological control agents Harmonia axyridis, Coccinella septempunctata and Propylea quatordecimpunctata competitively displacing native coccinellids. This paper is pioneering as the change in coccinellid populations was observed over a 31 year period so it allows for natural variation.
Alyokhin & Sewell (2004) found that prior to 1980 the majority of coccinellid species recorded were native. Following the establishment of C. septempunctata in 1980, native species were outcompeted; the abundance of C. septempunctata increased from 6.1% in 1980 to 100% in 1994 (see Figure 1). In 1993 and 1995 P. quatordecimpunctata and H. axyridis established respectively (see Figure 1). Alyokhin & Sewell (2004) concluded that the increase in exotic coccinellid establishment was strongly correlated with a statistically significant decline in native coccinellid populations.
This study provides evidence for the indirect effects of biological control. The methodology allows for natural population fluctuations and both methods and results were supported by previous studies (such as Brown & Miller 1998; Elliott et al. 1996). However, controls were obtained from an archive, this is undesirable as it does not ensure the use of the same protocol. Experiments should always include appropriate positive and negative controls to enable the drawing of accurate conclusions (van Lenteren et al. 2006b). In addition, this study does not consider other factors that might have affected native species populations such as temperature and other native species.
The establishment of exotic coccinellids did not result in the total displacement of native species; native species were present throughout the study in reduced abundance. This may indicate that although competition took place, it was not substantial enough to place the native coccinellids at risk of extinction. Therefore, it may be argued that the benefits of aphid control are worth a reduction in native coccinellid populations (Pearson & Callaway, 2005).
In addition, this study is further limited as it took place on a potato field and potato is exotic to the area. Therefore, this experiment may not reflect the effects of an introduction exotic insect to a naturally evolved ecosystem. For example, potato and native coccinellids did not evolve together and this may have provided exotic species with a competitive advantage (Strong & Pemberton, 2000).
Despite the limitations discussed, this study provides evidence of habitat displacement in biological control. Alyokhin & Sewell (2004) utilised appropriate statistical tests to provide valuable insight into the change in native species populations following biological control agent establishment. The regulations and assessments under which biological control agents such as H. axyridis and C. septempunctata were released needs to be reassessed to ensure biological control is environmentally safe.
Harmonia axyridis in Great Britain: analysis of the spread and distribution of a non-native coccinellid
Brown et al. (2008) BioControl, 53, 55-67.
Harmonia axyridis has been released to control aphids and coccids across Europe (for example, Ukraine in 1964, Belarus in 1968, France in 1982, Portugal in 1984, Italy in 1990s, Greece in 1994, Spain in 1995, Netherlands in 1996, Belgium in 1997, Germany in 1997, Switzerland for a short period in the 1990s before it was deemed too risky and finally, Czech republic in 2003). Since its introduction into these countries, H. axyridis has also been observed in Austria, Denmark, the UK, Liechtenstein, Luxembourg, Norway and Sweden (Brown et al. 2007). This paper provides evidence of H. axyridis dispersal into Great Britain, where it has never intentionally been released. This paper was selected as unlike other countries, Great Britain has monitored the spread of H. axyridis since its initial arrival in 2004 (Majerus et al. 2006).
Brown et al. (2008) utilised a web based survey to follow the dispersal of H. axyridis across Great Britain. Between 2004 and 2006, the analysis of 4117 H. axyridis recordings indicated that H. axyridis dispersed an average of 58 km north, 144.5 km west and 94.3 km north-west per year. The increased western dispersal rate is suggested to be due to multiple invasions from the European mainland. H. axyridis recordings increased by an average of 2.9 fold each year and the mean number of adults per recording increased from 2.9 in 2004 to 6.2 in 2006.
The results from this study indicate that H. axyridis has invaded Great Britain on multiple occasions and through multiple methods. For example, a single northern population of H. axyridis was recorded in Derby. This indicates that this population must have arisen from a separate invasion than those populations spreading across the UK from the East.
Public recordings were verified before inclusion in the analysis. Although this would have increased the accuracy of results, 4316 recordings were not verified so were not included. Some of the non verified recordings were likely to be H. axyridis but verification was not possible. Therefore, the analysis in this paper could be a huge underestimate of the actual dispersal and abundance of H. axyridis across the Great Britain. This data set is also limited due to the uneven spread of human populations across Great Britain. This would have resulted in a variation in the frequency of recordings in different areas. Therefore, these results may not accurately represent the species abundance.
This paper demonstrates that the currently inconsistent regulation for biological control across Europe is not adequate. The release of a biological control agent in one country will inevitably affect neighbouring countries. For example, H. axyridis has never been intentionally released in the UK but it has been estimated that since its invasion, H. axyridis could negatively affect 1, 000 of Great Britain’s native species (Majerus et al. 2006). The release of H. axyridis provides evidence that patchy regulation is a risk of biological control in itself.
Review of invertebrate biological control agent regulation in Australia, New Zealand, Canada and the USA: recommendations for a harmonized European system
Hunt et al. (2008) Journal of Applied Entomology, 132, 89-123.
Whilst the potential risks of biological control have only recently been acknowledged in Europe, they have been recognised and regulations have been implemented to avoid them for over forty years in Australia, New Zealand, Canada and the USA. Following a thorough and pioneering review of current regulation, Hunt et al. (2008) have discussed the adaptation of some concepts for Europe.
Hunt et al. (2008) found that although most European countries have regulation in place, only eight countries utilise them. Therefore, like Australia, New Zealand, Canada and the USA, Europe requires the passing of legislations to enforce the safe use of biological control. Australia is the only country to have a governing body specifically for biological control. Regulations in New Zealand, Canada and the USA fall under plant, conservational, environmental or endangered species Acts (Hoddle, 2004). Europe requires an EU level body and regulation specifically for insect biological control. This body should cover both environmental and agricultural issues and should be composed of experts representing each country. The EU body should implement regulations across Europe and should make decisions for the release of biological control agents. Like Canada, the USA, Australia and New Zealand a group of scientific experts should be utilised to review applications and recommend decisions to the EU body. This will ensure the decision for each introduction is based on the opinion of experts covering a broad range of expertise.
Following the establishment of an EU wide body and the passing of legislation, scientifically based ERA procedures need to be developed. In both Australia and the USA, approval is sought for the non target list prior to host specificity testing, however, this may restrict the ideally flexible nature of host specificity testing where species should be added or removed when appropriate (Kuhlmann et al. 2005). Hunt et al. (2008) suggest European regulation should follow New Zealand by involving discussions with experts. This will ensure the consideration of all risks, costs, benefits and the use of a scientifically based ERA. Discussion with experts will also reduce costs and time wasted on projects that do not have potential or are not being completed in an efficient manner.
This paper uses examples from the USA and Canada to demonstrate that a regulatory body over the whole of Europe is possible. It also emphasises the importance of utilising previous experiences of regulated countries to implement effective regulation in Europe. However, Messing (2005) argues that the USA has unresolved legislative problems between their federal and state governing boards. For example, Hawaii has such strict ERA regulations that the use of biological control is hindered and the federal ERA regulations are insufficient as they do not involve adequate application review. In addition, Cameron et al. (1993) argues that only 24% of biological control projects in New Zealand have been a success. Goldson et al. (2010) adds that Australian and New Zealand legislations are too strict. For example, in order to receive approval for release, evidence is required to prove agents do not pose any risks but this is often impossible due to time and cost constraints.
Care is required when reviewing the regulation of biological control in other countries. The presence of regulation does not necessarily mean it is enforced and information from government employees may be susceptible to political issues. Europe wide legislation is required but time and cost constraints need to be taken into account. In conclusion, regulation is needed to enforce the environmental safety of biological control but it should not restrict its effective use.
Establishment potential of the predatory mirid Dicyphus hesperus in northern Europe
Hatherly et al. (2008) BioControl, 53, 589-601.
Many guidelines have been released for an ERA (such as EPPO, 2001; NAPPO, 2001; IPPC, 2005) but none state a clear and effective methodology to test for establishment. As a result of this, climate matching has been widely accepted as an efficient predictor of establishment (for example, Messenger & van den Bosch, 1971; Stiling, 1993). However, the augmentative biological control agent, Neoseiulus caliginosus has proved its inadequacy as individuals with diapause ability were released unintentionally (Jolly, 2000). McClay & Hughes’ (1995) use of a degree-day model to predict establishment potential has also been criticized due to its labour intensive nature (McClay, 1996). In addition, the numerous methods utilised to determine developmental thresholds have led to differing conclusions for the establishment potential of the same insect (Hart et al. 2002). Hatherly et al. (2008) utilise a clear and scientifically based methodology for a test for establishment that should be used as an alternative to climate matching and day degree models.
Each experiment involved treatments of fed and unfed first instar nymphs, adults and diapause induced adults. Supercooling points (SCP), Lower lethal times (see Figure 2) and temperatures were determined. Field experiments were completed to study the effects of naturally fluctuating temperatures and a control experiment was conducted to ensure experimental conditions did not damage the mirids.
Statistical tests (one way ANOVA and Tukey’s HSD test) found no significant differences between the SCP (-20oC) for different life cycles and Ltemp90 was found to be -20.4oC for diapausing insects. After 140 days in the field, 5% of fed nymphs and 50% of fed diapausing adults were alive. After 148 days, 15% of fed non diapausing adults were alive. Following transfer to the lab, the survivor adults were observed laying viable eggs.
Overall, it was concluded that D. hersperus were able to diapause and individuals from each life cycle were able to survive outdoors in the UK. Feeding increased survival times and the polyphagous nature of D. hersperus meant it was likely to find food.
Laboratory methods to test the establishment potential of possible biological control agents need to be environmentally relevant (Hoelmer & Kirk, 2005). To determine SCP, the rate of temperature decrease was 0.5oCmin-1, this could be reduced to make it more realistic. Mortalities for lower lethal temperatures were recorded after 24 and 48 hours, however, winter lasts for four to six months. In this case, this was appropriate as 90% mortality was reached at each temperature exposure within the timescale. To make this study more realistic, it was ensured that D. hersperus was experimented on in the condition received by commercial buyers. To ensure that the results did not occur by chance, lower lethal temperatures and time were determined in addition to SCPs (Bale, 2005).
To determine establishment potential, both biotic and abiotic factors need to be taken into account (van Lenteren et al. 2003). Hatherly et al. (2008) included diapause ability, each life cycle stage, sub lethal injury and food availability in their analysis. The use of fed and non fed individuals is essential so methods are applicable for insects using freeze avoidant and freeze tolerant strategies. Also, the addition of a field experiment enables an essential comparison between laboratory and field survival results (van Lenteren et al. 2006a).
Bale (2005) supports this method and its adequacy for all invertebrates that may be considered for biological control. Therefore, a method such as this should become compulsory for ERAs around the world, to stop the reliance on climate matching which will increase the success rate and safety of biological control.
Assessing host specificity of a classical biological control agent against western corn rootworm with a recently developed testing protocol Toepfer et al. (2009) Biological Control, 51, 26-33.
The acknowledgment of non target effects in insect biological control is relatively recent compared to weed biological control. As a result, there are far fewer examples of host specificity tests and as of yet, a realistic, practical and efficient protocol has not been established for use in global ERAs (Van Driesche & Hoddle, 1997). Field studies are too labour intensive due to difficulties observing predator-prey relations and electrophoretic analysis of the gut is effective but, prior to agent release its use is limited to the country of origin (Bigler et al. 2005). Toepfer et al. (2009) utilise a novel method of non target selection (after Kuhlmann et al. 2006) and host specificity testing (after van Lenteren et al. 2006b) that are efficient, realistic and science based.
Toepfer et al. (2009) selected nine non target species to study the host range of the potential biological control agent Celatoria compressa. Small area no choice and small area sequential no choice tests were conducted (see Figure 3). Each test used a sample size of 10 adult beetles and was repeated approximately twenty times. Suitable positive and negative controls (30 replicates) and statistical tests were used (Turkey HSD post hoc tests following ANOVA and paired t-tests).
Toepfer et al. (2009) concluded that if released, C. compressa would have a narrow host range. This result corresponds with C. compressa’s native, ecological host range and with host range studies for other Celatoria species. In addition to the methods proposed by van Lenteren et al. (2006b), Toepfer et al. (2009) completed small area choice and sequential choice tests. This allowed the evaluation of van Lenteren et al.’s (2006b) methodology.
Conditions in the lab were fully controlled. This is supported by Keller (1999) who states that abiotic factors all have an effect on host selection. Van Lenteren et al. (2003) support Toepfer et al.’s (2009) inclusion of the target species as this ensures conditions are favourable. Experimental conditions were based on reproductive information for C. compressa (Zhang et al. 2004). This is important as host selection varies depending on physiological conditions (van Lenteren et al. 2006b). Laboratory studies are often limited as conditions are unrealistic (Kuhlmann et al. 2006; Sands and Van Driesche, 2000). Toepfer et al. (2009) offered hosts in natural conditions over a natural duration of exposure. This helps to ensure the experiment as realistic as possible (Bigler et al. 2005).
Careful identification of all species and field collection was conducted. Field collection is advantageous as lab rearing has been suggested to disrupt natural insect behaviours. For example, Ferguson et al. (1999) found host selection altered depending on the food the parasitoid was reared on. The use of replicates allowed for some natural variation in host selection (Bigler et al. 2005). Also, Boyd & Hoddle (2007) supports the use of a negative control to measure natural mortality rates. However, each test was conducted in a small area, a larger area would have been desirable. Boyd & Hoddle (2007) found that the results of host specificity tests varied in different area sizes.
In conclusion, Toepfer et al. (2009) provide evidence that the non target selection (after Kuhlmann et al. 2006) and the novel step by step method proposed by van Lenteren et al. (2006b) to determine host range are effective. However, further studies are required to give a more reliable estimate of the effectiveness of these methods. This paper is the first step in finding a supported method for host specificity that should be included in a globally enforced ERA. The establishment of such an ERA will help to avoid non target effects and it will therefore help to ensure biological control is environmentally safe.
An approach for post-market monitoring of potential environmental effects of Bt-maize expressing Cry1Ab on natural enemies
Sanvido et al. (2009) Journal of Applied Entomology, 133, 236-248.
In order to determine biological control’s position in Integrated Pest Management (IPM), its effectiveness alongside other pest management methods needs to be considered. Genetically modified (GM) crops have become an increasingly popular alternative to pesticide application and Bt crops are now grown in 110, 000 hectares across Europe (James, 2007). However, concerns have been expressed for GM crops. These include suggestions that they are incompatible with other pest management methods (Ferry et al. 2006). For example, GM crops may alter the availability or quality of food and shelter for biological control agents. In contrast, the two schemes may work well together, biological control agents could reduce the density of pests not targeted by GM crops and could decrease pest resistance rates (Lundgren et al. 2009).
Previous studies to investigate the effects of GM crops on biological control have been conducted in laboratories or in complex field studies. These studies have been criticized as they are extremely unrealistic, time consuming and expensive (Hilbeck et al. 1998). For example, Poppy and Sutherland (2004) criticise Hilbeck et al. (1998) study as natural enemies were forced to consume prey that had been fed Bt crops. Other studies have only taken one species of natural enemy into account. Sanvido et al. (2009) argue that a loss in species richness may not alter an ecosystem as many species carry out the same role. Therefore, Sanvido et al. (2009) pioneer a novel approach to investigate the effect of Bacillus thuringiensis (Bt) maize expressing Cry1Ab proteins on all natural enemies over cultivation. This experiment was completed following previous findings that Cry1Ab proteins were too specific for direct effects on natural enemies (Ferry et al. 2006) so an indirect protocol was utilised instead. Individual natural enemies were not identified, instead questionnaires were sent to farmers who recorded herbivore outbreaks. Herbivore outbreaks can be used to assess the levels of biological control.
The methodology described by Sanvido et al. (2009) has numerous advantages over other methods. It is less time consuming and more accurate as it covers all natural enemies within a field. Questionnaires would be easy to complete since farmers are already attentive to herbivore populations reaching economic injury levels. This method is applicable for all variations of GM crops and provides an effective method that can be utilised prior to full GM crop availability.
Questionnaires could be statistically analysed as a comparison could be made between GM and non GM maize. However, this method does not identify the direct cause of any detrimental effect observed. Following the observation of a reduction in biological control, further experiments would be required to prove the effect was due to GM crops rather than other factors (Poppy & Sutherland, 2004).
To conclude, this paper provides a novel methodology to study the effect of GM crops on biological control. IPM is extremely important to ensure the utilisation of the most efficient management for every pest situation. A reliance on biological control may lead to the release of unsuitable agents and therefore a combination of biological control with other pest management methods helps to keep it environmentally safe.
Following a review of the literature available, it is evident that negative effects are often assumed to be caused by biological control agents before other possibilities are considered. For example, both Boettner et al. (2000) and Alyokhin and Sewell (2004) lack explanations why other factors that alter species populations could be disregarded. Lynch et al. (2001) state that in only 1.5% of over 5000 classical biological control introductions have non target effects occurred and that less than 10% of these cases have resulted in a reduction in non target population size. It appears that the risks discussed for biological control are not founded on relevant evidence and therefore it may be argued that they are ‘perceived’ rather than ‘real’. Despite this, the damage Harmonia axyridis has caused to the ecosystems it has dispersed into (see Brown et al. 2007; Brown et al. 2008) provides clear evidence that non target effects have occurred in biological control. H. axyridis demonstrates that non target effects can be extremely environmentally damaging.
Biological control agents frequently disperse into countries neighbouring their release site. This statement is particularly applicable to mainland Europe, but agents have also been found to disperse across seas. For example, Petit et al. (2009) recorded the spread of Gonatocerus ashmeadi to islands 1400 km away from its release site. Therefore, if a lack of regulation has allowed the release of an unsuitable biological control agent in one country, neighbouring countries will be at environmental risk, regardless of their own regulation. This was seen with the release of H. axyridis by countries lacking in efficient regulation. It may be that H. axyridis has received disproportionate amounts of attention considering biological control has been used to successfully control over 165 pests. None the less, H. axyridis has damaged biological control’s reputation to such an extent that the current patchy regulation across Europe must be addressed (Wagner & Van Driesche, 2010).
The regular discussion of the potential risks of biological control and the publicity received by H. axyridis has slowed the use of biological control in the last decade or so. For example, between 1900 and 1980 an average of 4 insect agents were introduced to Hawaii every year. Since 1990, an average of one agent is introduced every two years (Follett et al. 2000). Biological control is the most natural, sustainable and economic method of pest management (van Lenteren et al. 2003) and therefore, biological control’s reputation must be improved so it can be utilised to its full potential (Castle et al. 2009). Biological control’s reputation can only be improved by the harmonization of regulation to ensure that unsuitable agents (such as H. axyridis) will not be released in the future (Thomas & Willis, 1998). Regulation should be implemented in order to facilitate rather than restrict biological control (Mason et al. 2005).
Biological control is currently regulated in 20 countries but to very different extents (Loomans, 2007). Figure 4 shows that only eight countries in Europe utilise regulation (Austria, Czech Republic, Denmark, Hungary, Norway, Sweden, Switzerland and the UK) (Bigler et al. 2005). As established by Hunt et al. (2008) only Australia has a governing body specifically for biological control. In other countries, legislation falls under Plant Protection or Conservation Acts. In addition, Goldson et al. (2010) state that in both Australia and New Zealand, the regulation of biological control falls under multiple legislations. This requires the cooperation of different governing bodies to approve releases which can be quite time consuming.
Each country requires a governing body specifically for biological control to ensure regulation is appropriate. In addition, regulation is meaningless unless it provides a method of assessing the risk posed by potential biological control agents (Hunt et al. 2008). For example, if the ISPM3 is made legally binding, it will offer information on environmental risks for classical and inundative biological control but it will not provide methodologies to assess potential agents (van Lenteren, 2005). Without scientific assessments, there is no way of predicting how successful or safe an agent will be (Hoelmer & Kirk, 2005).
In most regulated European countries (see Figure 4), introductory decisions are based on information provided in a set application regardless of the inundative or classical nature of the project. In countries where regulation does not fall under Environmental Acts, applications for release are too vague as they only require information related to the possible effects on humans and plants. Where regulation is conservation based (Norway, Netherlands, Switzerland & UK), applications include an environmental risk assessment (ERA). These ERAs vary, but they all require a minimum of information on establishment and host specificity (Loomans, 2007). In New Zealand, the ERA of potential agents is extremely strict as insects are declined release unless it can be proved that they do not pose any risk at all (Barratt & Moeed, 2005). As previously noted, regulation should be designed to facilitate biological control, the implementation of a too strict or inappropriate ERA will restrict it (Blum et al. 2003). This situation has occurred for microbial biological control agents as their regulation (EU Directive 91/414/EEC) was originally designed for synthetic pesticides. The ERA requirements were unrealistic for microbes and as a result were so cost and time consuming that they restricted the number of microbial agents available (Blum et al. 2003). However, even if an appropriate ERA is used for insect biological control, it will still increase cost and time requirements. Cock (2003) argues that this will restrict biological control since the decision of whether or not to release an agent is often dependent on financial factors.
A lack of regulation implementing ERAs across Europe has allowed the biological control industry vast success and growth in the last ten years. For example, regardless of their lack of regulation (see Figure 4), Belgium, France and Italy all commercially rear and sell inundative natural enemies (van Lenteren, 2005). There are approximately 85 companies and hundreds of small scale farmer or state run facilities who mass rear and sell natural enemies around the world and the European biological control industry has an annual turnover of approximately 200 million Euros (van Lenteren, 2005). The inevitable increase in costs due to the implementation of an ERA could limit these commercial producers. Due to the permanent nature of classical biological control, there is no commercial value in agent production. Therefore, money from governments would be required to cover classical ERA costs. This would dramatically decrease the impressive data demonstrating the cost-effective nature of biological control over other pest management techniques.
Overall, it must be concluded that a slight increase in cost and time requirements is unavoidable if regulation is to prevent the release of ‘risky’ biological control agents (van Lenteren et al. 2006b).
Throughout the world, the methods utilised to conduct ERAs vary greatly. For example, APHIS use a low to high scoring system and Biosecurity Australia utilise matrices (Loomans & van Lenteren, 2005). Early ERA procedures suggested for the harmonization of Europe involved the sum of all the products of magnitude and likelihood values for establishment, host range, dispersal, direct and indirect non target effects (van Lenteren et al. 2003). However, this procedure was criticized as it would be too cost and time consuming to complete for every potential agent. An efficient ERA should identify unsuitable biological control agents as early as possible to reduce costs and time. Taking this into account, van Lenteren et al. (2006a) have proposed a step by step ERA procedure (see Figure 5).
For inundative biological control, no long term environmental effects can occur if the agent cannot establish. Therefore, no further ERA tests would be required if this is concluded. In contrast, continuation of the ERA is only required if a potential classical agent can establish (see Figure 5). McClay (1996) states that classical biological control projects often fail due to the agent’s inability to establish. It follows that the first step in an ERA should be a test for establishment as this will quickly identify agents that are not suitable for release (see Figure 5). Hatherly et al. (2008) utilise a scientifically based test for establishment, combining diapause ability, SCP, lethal time and temperature thresholds, field experiments and sub lethal injury.
Following establishment potential, host range tests are required. The use of Wapshere’s (1974) centrifugal phylogenetic method to select non target species for host range tests is insufficient for insect biological control because there are too many environmental influences on host selection (Messing, 2001). As a result of this, Kuhlmann et al. (2006) provided a novel technique to select non target species. This needs to be conducted by an expert to ensure the correct taxonomic identification of both the potential agent and non targets (Gitau et al. 2009). Van Lenteren et al. (2006b) provide a novel method to test host specificity. Emphasis is placed on the importance of replicates, controls and statistical tests. The framework is open to adaptation for the specific natural enemy and information already available. This flexibility is essential for a globally applicable ERA (van Lenteren et al. 2003).
Hatherly et al. (2008) and Toepter et al. (2009) provide evidence for the successful use of the test for establishment and host specificity tests provided by Bigler et al. (2006). Further experiments are required to fully support the efficiency of these relatively novel protocols but they are examples of methods that should be implemented by world-wide regulation. The final decision for the release of a biological control agent should be based on a comparison with other pest management techniques using a risk-cost-benefit assessment completed by a governing body of experts (OECD, 2004). This will ensure the most appropriate and effective control of the pest (van Lenteren et al. 2003).
Integrated farming should be utilised to ensure the best control of any pest whilst minimising risks to the environment or human health. It must be ensured that pest management is not detrimental to the natural biological control of 95% of native insects (van Lenteren, 2005). The movement to reduce pesticide use (for reasons see Table 2) in addition to evidence that pesticides do not work well alongside biological control (for example, Duso et al. (2009) found that the use of pesticides disrupted the natural biological control of phytophagous mites in European apple orchards. Also, Stark et al. (2003) found that the biological control agent C. septempunctata was more susceptible to pesticides than the aphid pest (Acyrthosiphon pisum)) has increased interest in the compatibility of GM crops and biological control. Recent studies have found biological control and GM work well together (see Poppy & Sutherland, 2004; Ferry et al. 2006) and Sanvido et al. (2009) provide an easy approach to study the effects of GM crops on natural enemies. This combination could well be a key component in the future of pest management. However, biological control is considered the ‘greenest’ option for pest management so it is likely that advocates of biological control will not correlate with advocates of GM. This needs to be addressed to ensure the most efficient form of IPM is utilised.
Firm evidence for the risks discussed for biological control is limited, one might conclude that biological control is fairly environmentally friendly. However, non target effects have occurred following the introduction of unsuitable agents. It appears biological control is neither environmentally friendly or a risky business but H. axyridis has damaged biological control’s reputation to such an extent that it enforces the requirement of global regulation. Regulation is meaningless unless it enforces an efficient ERA. To allow biological control’s full potential, ERA methodologies need to minimise costs and time requirements and be applicable to both inundative and classical biological control whilst acknowledging their differences. A suitable ERA step by step procedure and methodologies for each step are provided by van Lenteren et al. (2006a) and Bigler et al. (2006) respectively. The legal binding of which will ensure inappropriate releases such as H. axyridis will not occur in the future. The natural mechanisms and cost effectiveness of biological control are so advantageous over other pest control techniques that it should be utilised to its full potential, as safely as possible.
Suggestions for further study
Bacigalupe (2009) suggests that a lack of genetic diversity in invasive species often prevents their establishment. A lack of genetic diversity may be the reason why many classical biological control agents fail to establish, on the other hand, it may help to prevent inundative biological control agents from establishing. Manipulating the genetic diversity in the sample of natural enemies released could help to ensure biological control is environmentally safe. A suitable classical biological control agent that has previously failed to establish should be rereleased in a sample of high genetic diversity and a sample of low genetic diversity. The two samples should be monitored and their establishment should be compared. Specific regulation and ERAs are necessary, but if another mechanism to ensure the safety of biological control agents was found, ERAs could be relaxed and biological control would not be at risk of restriction.
Studydriver writers will make clear, mistake-free work for you!Get help with your assigment
Please check your inbox