IMPACT AND EFFECTIVENESS OF Phytophthora TREATMENT ON HEATHLAND AND OTHER SEMI-NATURAL HABITATS

BIODIVERSITY MethodOLOGY

January 2013

Simon Conyers

Naomi Jones

Food and Environment Research Agency

Sand Hutton, York, YO41 1LZ

Contents

1.Introduction

1.1Phytophthora on Vaccinium

1.1.1Pathogens involved

1.1.2Symptoms and epidemiology

1.2Distribution and characteristics of infected sites

1.3Importance of the lowland heath habitat

1.4Pathogen control options

1.5Potential impacts of treatment

1.6Biodiversity monitoring issues

1.7Aims of the study

2.Review of field monitoring methods

2.1Vegetation

2.1.1Community composition

2.1.2Quadrat size

2.1.3Sampling design

2.1.4Structure

2.2Invertebrates

3.Proposed Experimental Design and Monitoring Methods

3.1Treatments and experimental design

3.1.1Treatments

3.1.2Sites

3.1.3Sampling design

3.2Vegetation monitoring

3.3Invertebrate monitoring

3.4Other assessments

3.5Statistical analysis

4.Bibliography

Summary

Background

  1. Two exotic species of Phytophthora (P. kernoviae and P. ramorum) have, in the last decade, been found to infect, and kill, a range of native and non-native woody species. Initially, infection was largely restricted to nurseries and Rhododendron ponticum in the wild. However, Vaccinium myrtillushas been recorded infected with Phytophthora in open heathland(predominantly P. kernoviae) and woodland (predominantly P. ramorum). Treatment of infected V. myrtillus is important because this species supports high levels of sporulation of Phytophthora, therefore has the potential to spread rapidly through a habitat.
  2. Sites infected with Phytophthora are in the south west of the UK, mostly in Cornwall. Affected heathland sites are generally areas of degraded heathland, which have received little if any management in recent years. However, lowland heathland is a Priority Habitat under the UK Biodiversity Action Plan which aims to improve the condition of existing heathland.
  3. Methods for treating infected Vaccinium have been proposed after research based on small scale plots. Only herbicide application will control the pathogen by removing the host species from the immediate area and herbicide application is recommended on small sites where the extent of infection is limited. However, it is difficult to accurately identify symptoms on Vaccinium particularly in a varied vegetation mosaic. A two stage treatment has therefore been recommended comprising burning or cutting followed by targeted herbicide application of infected vegetation.
  4. The aim of this study is to monitor effects of large-scale treatment on vegetation and invertebrate communities. Impacts on other groups may be inferred from these data. The purpose of this review is to develop suitable methodologies for monitoring.

Review of monitoring methods

  1. Vegetation community composition can be assessed by a range of methods including presence/absence, density and cover. The value of presence/absence assessments is limited by the size of quadrat chosen, however a series of quadrats of increasing size, used in the monitoring of Environmentally Sensitive Areas, allows the optimum scale for different species to be assessed. This system is a rapid assessment at a range of scales. Monitoring of plant density generates more detailed data, however it is time-consuming and it can be very difficult to identify individuals, particularly of perennial species. Cover is a subjective assessment, unless point measurements are used but this is a very time-consuming technique. Cover can also change significantly in a short period of time.
  2. Suitable quadrat size for vegetation monitoring encompasses a wide range. However large quadrats are recommended for characterising vegetation (NVC recommends 2 m x 2 m for dwarf shrub heaths) whereas detection of changes over time will be more accurately assessed by smaller quadrats. Quadrats can be placed completely at random or in a stratified random design. Monitoring studies which aim to detect long term changes are usually based on fixed quadrats. Vegetation structure is an important ecological aspect of the habitat. Structure can be measured by maximum height (using a sward stick or drop disc), height when a pin is touched or for Calluna vulgaris, by assessment against four life stages.
  3. Impacts of treatment on the vegetation will vary depending on treatment (and how it is applied) and habitat characteristics (community composition, recent management). However, very little information is available regarding the impacts of herbicide application in these habitats.
  4. Heathland may support a diverse range of invertebrates including some species specifically associated with the dwarf shrub component, but also less specialist species. A wide range of invertebrate monitoring methods can be used, but all will only sample a specific element of the invertebrate community in terms of functional type and size. Commonly used methods include sweep netting, suction sampling and pitfall trapping. Many other methods are very time consuming. Sweep netting is an effective method particularly for flying insects (e.g. Araneae, Orthoptera and Homoptera). Suction sampling is a complementary method which can give higher numbers and diversity for some invertebrates than sweep netting and is particularly effective for Diptera, Hymenoptera and smaller invertebrates. Pitfall trapping samples ground dwelling arthropods and does sample nocturnal species. However, this method gives a measure of activity/density which could be significantly affected by treatment and is time consuming in terms of preparation, placement and retrieval of traps.

Proposed methodology

  1. Herbicide application is the most effective treatment for infected Vaccinium because it removes infected plants without spreading the disease. At sites where limited infection exists, herbicide will be applied over the whole area plus an additional buffer zone. On larger sites it is difficult to identify infected plants within the varied mosaic therefore initial burning or cutting will be followed by targeted herbicide application of plants with infected regrowth. Detailed assessments (of both the burnt/cut treatment and the additional herbicide application) will be undertaken at larger sites, focussing on the heathland habitats.
  2. Sites will be monitored pre-treatment to establish baseline communities and subsequently at 6 to 12 month intervals, in late spring and early autumn, until the end of the study. Permanent quadrats will be randomly located within a stratified sampling design.
  3. All plant species will be recorded as present/absent in a series of nested quadrats within a 1 m x 1 m quadrat (importance and frequency scores will be calculated). In addition, percent cover will be recorded for all species in 1 m2 and 4 m2 quadrats. Bracken abundance will be assessed by a stem count in the 1 m2 quadrat. Height will be recorded using a drop disc and the predominant vegetation layer in which each species occurs will be recorded.
  4. Invertebrate populations will be monitored using a combination of sweep netting and suction sampling in order to best assess the phytophagous species associated with the vegetation communities. Suction samples will be taken from the 1 m2 quadrats assessed for the vegetation. Sweep net samples will necessarily encompass a much larger area, but samples will be centred on the 1 m2 quadrat. Invertebrates will be identified to family level.
  5. Other assessments will include physical parameters such as altitude, aspect, slope, soil type and soil pH. Burn temperatures will be monitored using a range of paint temperatures, however burning is likely to be done using hand held torches which may result in highly variable burn temperatures.
  6. Changes in vegetation (cover, abundance and importance scores) and invertebrate numbers will be analysed for each site individually. A range of relevant groupings (e.g. plants – negative indicators of condition; invertebrates – bird food species) will also be analysed. Ordination techniques will be applied to vegetation and invertebrate community data.

1.Introduction

1.1Phytophthora on Vaccinium

1.1.1Pathogens involved

Phytophthora is a genus of plant-damaging water moulds, which are responsible for a range of economically damaging diseases including potato blight. Two species, Phytophthora ramorum and Phytophthora kernoviae are exotic species, which infect a range of native and non-native woody species. They are notifiable pathogens under European legislation, which were first confirmed present in the UK in 2002 and 2003 respectively. P. ramorum has a very wide host range (Fera, 2010a), whereas that of P. kernoviae is much smaller (Fera, 2010b).

Native hosts have been found to include ericaceous heath species. P. kernoviae was first recorded on Vaccinium myrtillus in the natural environment in woodland in Cornwall in 2007 (Turner et al., 2009) and P. ramorum infection has since been recorded on V. myrtillus in woodland (2009) and in November 2010 in lowland heathland (Fera, 2010a). P. ramorum infection of V. myrtillus is predominantly found in woodland, whereas P. kernoviae is most common on V. myrtillus in heathland.

1.1.2Symptoms and epidemiology

Symptoms of Phytophthora infection of Vaccinium initially include stem and leaf blackening followed by petiole necrosis and stem dieback (Beales et al., 2009). Early symptoms on V. myrtillus occur towards the top of the plant and then spread down the plant, at least until stem junctions. Often a banding effect of alternating infected and healthy tissue along stems develops where multiple infection points have been initiated along the stem (Turner et al., 2009). Both species produce reproductive structures known as sporangia which release mobile zoospores that enter plant mainly through wounds. The pathogen can spread by: aerial spore dispersal, through waterways, by direct plant to plant contact and by movement of contaminated soil and debris. The pathogens do not sporulate on most native host species, however, infected V. myrtillus has been found to support very high levels of sporulation, therefore there is the potential for the disease to spread rapidly through a habitat.

1.2Distribution and characteristics of infected sites

Sites with V. myrtillus which are infected with Phytophthora are largely confined to the south west of the UK on dry lowland heathland and in woodland, where V. myrtillus can form the dominant understorey species. However, characteristics of the sites are variable. Generally P. kernoviae is present in heathland sites and P. ramorum in woodlands. Only six sites outside Cornwall have been recorded as infected, all with P. ramorum in woodland. Four of these sites have been treated and are apparently clear of infection.

Affected heathland sites are mostly areas of dry heath, mainly on isolated small hills where there has been a lack of management of any kind for many years, in a matrix of mixed farmland. Heaths are characterised by low plant species diversity with the cover of ericaceous shrubs partially correlated with the area of heathland (Webb and Hopkins, 1984). The dominance of these shrubs decreased on smaller heaths which results in changes in structure and diversity (Webb, 1989), and therefore the vegetation becomes less representative of classic heathland (Webb and Hopkins, 1984).

A pilot survey of the vegetation of two infected heathland sites in Cornwall in March 2012 (Appendix 3), has shown that these are degraded heathland areas with extensive areas of Pteridium aquilinum and Rubus fruticosus agg., and also grasses such as Molinia caerulea. Both sites are outside but adjacent to designated areas of the BAP Priority habitats for Lowland Heathland[1]. On the sites surveyed, Vaccinium myrtillus itself occurs generally as small patches within the vegetation and it maybe that it has survived due to its ability to tolerate the low light level under P. aquilinum canopy.

1.3Importance of the lowland heath habitat

Dry heathland in unexposed areas (below 300 m and not coastal) is semi natural, being derived from woodland with a long history of burning and grazing (JNCC 2009), and succession will continue in the absence of use or management (Webb, 1989). The ideal condition for lowland heath is a mosaic of mixed-aged vegetation dominated by ericoids or Ulex species. This is achieved by rotational management that maintains the dwarf shrub heath species at a pre-climax successional stage. These lowland heath communities are associated with thin acidic, nutrient poor soils and broadly restricted to elevations <300 m. On more freely-draining soils along with Calluna vulgaris, Erica cinerea, Ulex gallii and V. myrtillus may also occur. Lowland heathland is a Priority Habitat for conservation under the UK Biodiversity Action Plan (BAP) (Hewins et al., 2007). This Action Plan aims to arrest loss of lowland heathland habitat, improve the condition of existing heathlands and to create new areas of lowland heathland.

1.4Pathogen control options

Previous research has investigated the efficacy of different control methods on the pathogen. Recommended management actions will vary according to the specifics of the site, but potential treatments include burning, cutting and herbicide application. Only herbicide application will control the pathogen, by removing the host species from the immediate area. Regrowth after burning or cutting will be susceptible to reinfection from spores present in the soil. Herbicide application is recommended on small sites where the extent of disease can be effectively surveyed and spraying is practical. However, on large sites, it can be difficult to reliably identify all infected plants due to issues of variable mosaics, V. myrtillus growing beneath the canopy and variability of symptom expression on mature plants. It is therefore recommended that large areas are burnt or cut on these sites, with a follow up herbicide application applied to infected regrowth, which will be much more accurately identified.

Treatments will therefore be a herbicide application or a combination of cutting or burning followed by targeted herbicide application of infected regrowth.

1.5Potential impacts of treatment

Some species may benefit from the use of herbicides. Grass species may recolonise more quickly than ericaceous species; Milligan et al. (2003) found that Deschampsia flexuosa increased when Calluna cover was reduced. There is equivocal evidence with regard to the effect of glyphosate on bryophytes. Newmaster et al. (1999) in a study of its effects on species in a Canadian boreal mixedwood ecosystem has shown that their abundance and species richness decreased for two years after the treatment, whereas bryophytes were reported to increase by Marrs et al. (2004) and Todd et al. (2000). Positive effects on other species were presumably a result of protection of the understorey from herbicide application and a subsequent reduction in competition from the canopy species.

Burning and cutting will not control the disease or affect the viability of Vaccinium rhizomes. In fact severe fires or repeated burning are thought to favour V. myrtillus over Calluna in the first years after fire before heather reasserts dominance (Ratcliffe, 1959). Vegetative regeneration of dwarf shrub heath and Ulex species seems to be unaffected by the fire temperature, but is affected by the age of the vegetation burnt (Allchin et al., 1996) although evidence for this effect in species other than Calluna is equivocal.

A study of the after-effects of fire on Thursley Common reported that burning can destroy all above-ground bryophytes (Duckett et al., 2008), however the temperature of a fire is likely to have an impact on their survival. Cooler fires, which may result from controlled burning, may have less impact on bryophytes. Species with rhizoid systems that survive in the peaty substratum will recover quickly while others will colonise by spores blown in from elsewhere and may not have been present beforehand. Bryophytes may become the predominant vegetation in the first year.

Cutting is a less severe form of management than burning and will have a much lesser impact on low-growing species such as bryophytes. Cutting of heathland sites does not reduce the dwarf shrub cover but maintains it at a stage of vigorous growth, has a more diverse grass element and prevents the dominance of invasive scrub such as bracken and gorse. Regeneration times have been reported to be very similar to those found from burning, though at some sites a visible reduction in Calluna height and bare ground have been apparent six years after cutting.

All forms of management can open up the vegetation, exposing and disturbing bare ground and so increasing the likelihood of undesirable plants such as bracken and gorse scrub becoming established.

Controlled burns have been shown to have little effect on various invertebrate populations (Coulson, 1988; Gardner & Usher, 1989). Recolonisation by Lepidoptera in particular happens rapidly after Calluna begins to grow after burning (2-3 years), provided that mature stands are retained near the burned site (Haysom and Coulson, 1998). However for some heathland-indicator species such as Hemiptera, colonisation rates of a treated area declined exponentially with distance from an untreated source (Littlewood, 2006c).

1.6Biodiversity monitoring issues

The study aims to monitor effects on the vegetation and invertebrate communities affected by the disease and its treatment. Resource constraints and, at some sites, the scale of treatment, mean that monitoring of other groups is not appropriate, however it will be possible to extrapolate from the vegetation and invertebrate information to draw some conclusions about the likely impacts on other groups.

The purpose of this document is to review the sampling methodology for vegetation and invertebrates and to develop suitable methodologies appropriate for monitoring these groups at infected sites. The review (Section 2) is not intended as an exhaustive literature review, but to highlight the range of possible approaches relevant to heathland habitats. Section 3 describes the proposed methodology developed specifically in relation to this study. Factors considered here include the distribution of sampling points and the relative position of vegetation and invertebrate samples. Methods of vegetation sampling are discussed including the size of quadrats, unit of measurement (e.g. cover, presence/absence, density) and structure. Invertebrates are considered in terms of the sampling method and the level of identification appropriate for the study.

1.7Aims of the study

This study aims to:

  1. monitor the impact of disease control methods on vegetation and invertebrate communities on sites where V. myrtillus is infected with Phytophthora and;
  2. further develop the understanding of the effectiveness of Phytophthora control using different treatments.

Although there is considerable knowledge of the impact of burning and cutting on heathland vegetation, there is little similarity between the infected heathland habitats and those managed heathlands where past research has been undertaken. Also there is little information available on the impact of herbicide applications. A replicated trial is not appropriate because of the potential for reinfection within a site. This study will therefore be aligned with an eradication exercise.