Does light pollution affect the ability of the male glow-worm, Lampyris noctiluca, to find females through phototaxis?

Stephanie Bird 2008, BSc Biology
Current email address: stephaniebird@rhs.org.uk
Supervised by Dr. J. Parker

Abstract

In this project the effect of light pollution regarding the ability of male Lampyris noctiluca (glow-worms) to find females, was explored. Light pollution had already been considered as a potential factor in the reported decline of the species, as it has the potential to interfere with reproductive behaviour as the males find the females through phototaxis.

Imitation females (LEDs) were used to lessen the impact on the local population. These were set out under two different conditions; one light polluted and one control, with the numbers of males attracted to each being recorded. The distance away from the light source was then increased, decreasing the level of illumination (lux) each LED was subject to. A correlation was found between the level of light pollution falling on the LEDs and the difference in numbers of males attracted depending on light pollution condition. As the lux level decreased more males were attracted to the LEDs, with a lux value of 0.065 being predicted as a critical value, below which the light would no-longer have an effect. The implications of these results are discussed, with the focus on conservation.

Regarding future projects in this field it was concluded that it would be easier to conduct them on the female L. noctiluca as the duration of their activity is longer and less dependant on the weather. This project also provides background research into and evaluates the feasibility of future studies. The distribution of females across the hillside was recorded using a global positioning system and found to be clumped, however work could be done to confirm this and investigate possible reasons. The abundance of females over time, in conjunction with the first male sighting was also noted so that for future years a more informed decision of the best time to conduct trials can be made, for the site at Brush Hill (Princes Risborough).

A Summary for the lay person

Lampyris noctiluca is a species of glow-worm, incidentally a beetle and not a worm. It is wide ranging and found in Great Britain, however these numbers are believed to have declined in recent years. It has been thought that light pollution from street lamps has played a role in this process as it could negatively interfere with their reproductive behaviour. To attract males, females climb to a prominent position and emit a constant green light to which the male then flies towards.

This project used imitation females, made from light emitting diodes (LEDs). These were placed under two lighting conditions; polluted (with light) and unpolluted (without), the distance from the light source was then varied. A white light producing camping torch was used to mimic a street lamp. It was not possible however to reproduce the angles that light pollution comes from as positioning the lamp above the mock females was not viable. The number of males attracted to each set of LEDs was noted and then compared to the light pollution level faced graphically.

It was found that the light pollution did significantly affect the males’ ability to find the females, although further research would have to be undertaken at other sites to confirm this. A correlation between the level of light pollution falling on the LED and the difference in numbers of males attracted depending on light pollution condition was found. As the illumination (lux) decreased more males were attracted to the LEDs, with a lux value of 0.065 being predicted as a critical value, below which the light would no-longer have an effect. This result has implications for the future conservation of the glow-worm. If it is the case that street lamps have the same effect as the light source used then new lighting plans for areas surrounding or containing glow-worms should be carefully researched and current lighting in such areas re-evaluated for any detrimental ecological impacts, this is discussed further in the project.

From this work it was also concluded that it is more feasible to work on females than males, as they are easier to observe, emerge earlier and are more tolerant of adverse weather conditions. It was also possible to count the females each night and note when the males were first observed, whilst recording temperature, wind and rain. In the future this data could be compared to that of other sites or to the same one over the next few years. From preliminary background work it was also found that the positioning of the females within the field was clumped, although the cause of this is unknown and studies at other sites would need to be conducted. The effects of the imitation females on real females and the height they climbed up plant structures was also briefly examined, to provide a greater background understanding for this project and a starting point for the work of future students in this field.


Contents

Abstract     
                                                                                                   
A summary for the Lay person
                                                              
1. Review of the literature                                                                          
          1.1. The Glow-worm; L. noctiluca   
          1.2. Bioluminescence   
          1.3. Light Pollution  

2. Aims of the Project

3. Materials and Methods                                                                         
          3.1. Materials     
          3.2. Preliminary experiments       
          3.3. Main experiment   

4. Results

5. Discussion

6. Acknowledgements  
  
7. References   
 
8. Appendix A   

9. Appendix B 

10. Appendix C  

11. Appendix D (Risk assessment) 
    

1. Review of the literature

1.1. The glow-worm; L. noctiluca
The Common European glow worm, Lampyris noctiluca, is a coleopteran of the family Lampyridae, which also includes fireflies. It was first defined as a species by Linnaeus in 1758. The defining feature of many members of this family is their ability to ‘glow’, or more correctly ‘bioluminesce’, this is apparent from their names; noctiluca comes from the latin for night light (nox/noctis; night, and lucere meaning to shine). Female L. noctiluca produce a continuous constant cold light via a series of chemical reactions to attract males, this is in contrast to fireflies that can modulate the light emitted resulting in species-specific flash patterns (Branham & Wenzel, 2003).

L. noctiluca is one of only two Lampyridae species to be found in the United Kingdom, with the other being Phosphaenus hemipterus (Tyler, 1986), the distribution of which is thought to be restricted to regions of Sussex and Hampshire, though sitings are rarely reported (Scagell, 2008). The populations of L. noctiluca are found throughout Britain however they are more abundant in the South of England (Tyler, 1994). They are a widespread species, with a range that currently covers Europe and extends from Portugal to China (Fig. 1.1.)(Tyler, 1986 & 1994). Brush Hill, the location of this study, is found close to the village of Princes Risborough in Buckinghamshire.

 
(Fig 1.1 Approximate distribution of L. noctiluca from Tyler, 1986.)

Glow-worm distribution tends to be patchy and limited to areas of unimproved grassland (Gardiner & Gardiner, 2005); Tyler (1986) analysed data contributed by members of the British Naturalists Association, finding that out of 255 L. noctiluca sightings 58% were grassland habitats. In 2007, Gardiner notes that colonies are found on small sites, again preferring grassland habitats or hedgerows. Tyler (1994) also mentions that glow-worms appear to be particularly abundant on chalk containing, calcareous soil, whilst Alexander (1992), notes that they tend to be found on neutral to base-rich soils. As suggested by Tyler, this could be due to the mineral content of the ground as the larval stages feed on gastropods, predominantly snails (Tyler, 1986, Alexander, 1992 & Gardiner, 2007(c)), which require calcium, as it is calcium carbonate crystals that ultimately provide approximately 95-99% of the mass of a snail shell (Zhang & Zhang, 2006). It has been shown that there is a correlation between the distribution of L. noctiluca and species density of snails (Tyler, 1986). However Gardiner et. al. (2003) found no such link between pH and distribution of L. noctiluca at 16 sites in Essex.

It is the immature larvae of L. noctiluca that predate the snails, although it is unknown whether they do this by encountering them through chance or by following the trails of mucus they produce (Tyler, 1986, cited Wootton, 1971). On discovering a prey item the larvae inject it with poison that has paralytic effects (Schwalb, 1961). They live as larvae for up to three years, (Schwalb, 1961, Tyler, 1986) undergoing five moults, before they pupate. The length of this process varies according to gender, with the females taking eight days compared to the males eleven (Tyler, 1986, cited Vogel, 1927).

L. noctiluca are sexually dimorphic as adults, with the females being larger, flightless and capable of greater bioluminescence. This glow produced by the females is predominantly achieved by two large light-emitting organs, located on the ventral surface, in the fourth and fifth abdomen segments (Sala-newby et. al.,1996), although light is also emitted from the terminal segment of the tail (Tyler, 1986).  

Though the adult males are not capable of the level of bioluminescence exhibited by the females, they do still have two small glowing organs located at the tip of the abdomen (Tyler, 1992). However, unlike the females they are able to fly using fragile wings, encased by ‘leathery’ elytra (Tyler, 1992) (Fig. 1.1.). Due to their delicate nature they are susceptible to rain and wind, they do therefore have a lower tolerance of these conditions and can abandon flights if they are too adverse (Tyler, 1992).

The vision of the males is superior to that of the female. They possess compound eyes of on average 3412 facets (Schwalb, 1961), compared to the females of only 300 (Tyler, 1992). This presumably improves their vision, enabling them to more easily spot the glow of the females they do not aid in foraging for food because as adults they do not eat. They subsist entirely on the energy stores they built up as larvae, so their lifespan is therefore shorter in the adult stage, with no individual usually lasting longer than two weeks (Tyler, 1986). They consequently need to reproduce as quickly as possible, however both sexes die shortly afterwards (Tyler, 1986).

It is difficult to estimate the abundance of L. noctiluca due to this short adult lifespan however there is a consensus among those involved in the field (Tyler, 1986, 1994, 2004, Alexander, 1992 & Gardiner 2007, (a) & (c)) that there has been a decrease in the number and size of populations in Britain. Tyler (1986) suggests that there has been a marked decline during the last 50 years, however, Gardiner (2007(c)) suggests that they have been declining in areas such as Epping Forest since the 1800’s, linking it to linking it to the industrial revolution. Although the available information is limited, it indicates that the decline is not restricted to one area or habitat (Tyler, 1986).

The conservation status of the glow-worm varies across different counties. Those of Essex and Sussex are thought to currently be stable following surveys in 2001 and 2004 respectively (Gardiner, et. al., 2001) (Howard et. al., 2004), although the Sussex sites are due for review in 2010. Appendix 1 of the biodiversity action plan of Buckinghamshire County council, the authority that governs the area where this study will be conducted, lists L. noctiluca as a species of concern. The local distribution is ranked as scarce, meaning that it is currently found in 0.6 - 4.0% of tetrads in the biodiversity action plan area, whilst its decline rate is estimated at 25-49% over the last 25 years (Buckinghamshire & Milton Keynes Nature Conservation Forum, 2000).

1.2. Bioluminescence
Light is produced by L. noctiluca at all stages of the life cycle, including when the potential larva are still encapsulated by the egg. This luminescence begins a couple of days after they have been layed (Sala-newby et. al., 1996). As larvae, both genders produce light, from photophores in the tip of the abdomen (Tyler, 1992) upon being disturbed (De Cock & Matthysen, 2003). It has been proposed that this operates as an aposematic signal mechanism, to warn potential predators of distastefulness (De Cock & Matthysen, 1999, De Cock & Matthysen, 2003). Schwalb (1961) observed that mature L. noctiluca had few predators presumably because of this, especially when coupled with their short lifespan as any predator could not feed solely on them, they would have to be a generalist tolerant of the toxin. It has been further suggested recently by the work of Tyler et. al. (2008) that aposematism was the original function of the bioluminescence with the reproductive signalling behaviour evolving at a later date. It is now however the adult female L. noctiluca that is primarily associated with bioluminescence (fig. 1.3). Here it serves the function of an attractant for reproduction, to call attention to the females’ location and guide the males towards them; it induces positive phototaxis. As such it is a vital mechanism for reproduction and plays a role in sexual appetency behaviour (Schwalb, 1961). This bioluminescence lasts for several hours (Sala-newby et. al., 1996) and is usually visible to the human eye from several metres under normal conditions (Tyler, 2002) and up to 50m when they are favourable (Tyler, 1986).

The physiological pathways that result in bioluminescence have been investigated scientifically. Carlson and Jalenak, (1986) found that the neurotransmitter octopamine, present in adult lantern segments as well as larval lantern tissue, was capable of inducing bioluminescence. Although this study was conducted on fireflies, Photuris versicolor, De Cock and Matthysen (2003) were able to stimulate luminescence in larval L. noctiluca using this hormone.

The cold light is produced by the enzyme luciferase, comprised of 547 amino acids, which catalyses the oxidative decarboxylation of luciferin, a benzothiazole (Sala-newby et. al., 1996). This reaction also requires adenine triphosphate (ATP), mg+2 and oxygen, producing oxyluciferin. Schwalb (1961), measured the wavelength of light produced at 570 nm, a yellow light, although more recent studies (Sala-newby et. al., 1996) state that the maximum frequency emitted is 550nm, which produces a green light.As Schwalb (1961) also found that the response of males is greatest to yellow light, with scant attention being paid to red, blue or green wavelengths, it could be that the ‘yellow’ light used In this case was actually green. This preference for green light frequencies over yellow, is reinforced by Booth et. al. (2004) who aver that they are most attracted to light of 555nm. This information will have an impact on the experimental design of any imitation females required.

This bioluminescence only occurs at night, when the females are active, as they remain covered during daylight to avoid predation (Gardiner, 2007(c)). They appear and glow each night of their adult phase, which lasts from late May until early August, depending on latitude (Tyler, 1986), although their presence is most noticeable in July (Sala-newby et. al., 1996 & Gardiner, 2007(b)). During this period the females climb up to the top of a grass stalk or other suitable prominent structure to become more visible, before twisting to expose their underside, orientating their luminous organs upwards (Schwalb, 1961). Towards the end of their lifespan if fertilisation has not occurred the females may rotate their abdomen, moving it in a circular motion whilst protruding their external genital appendages (Schwalb, 1961), presumably to further increase the chances of being seen. This occurs until fertilisation (Schwalb, 1961). Dreisig (1978) found that the females glowed on average for 138 minutes, with the intensity of the bioluminescence decreasing with time until no light was emitted. This however would only be the case in unfertilised females, as during copulation the bioluminescence is either reduced or ceased completely (Schwalb, 1961).

It was also found by Dreisig (1975) through studies on L. noctiluca in Denmark, that this bioluminescent process starts after the light levels have decreased below a critical threshold, typically during or just after twilight. However this process is also reliant upon a natural circadian rhythm of sensitization, during which the female becomes more receptive to ambient light levels (Dreisig,1975).

Schwalb (1961) found that the length of time the females glow for is unaffected by temperature, this was subsequently supported when it was found to be the case if temperature remained constant. However, when the temperature is raised the period of bioluminescence is reduced, whilst if it is lowered the length of time is decreased (Dreisig, 1978). Females also commence glowing earlier, with natural variations that decrease daylight intensity, including cloud cover and changes in onset of darkness through the progression of seasons (Schwalb, 1961). They start glowing later as day length increases before the summer solstice (21/07/07), after which they would steadily begin to glow earlier, in keeping with the shortening day length. This activity was noted in both males and females (Schwalb, 1961).

Another question, important to this study, is whether the positive phototaxis is the only method employed by female L. noctiluca to attract males? Many arthropods emit pheromones, including some Lampyridae species. Lucidota atra , a firefly, relies solely on the production of pheromones to attract a mate (Branham & Wenzel, 2003, cited Lloyd 1972). It was Schwalb (1961) that first noted that female L. noctiluca exude a ‘somewhat cabbage-like odour’ that could act as a pheromone when detected by the male.

To determine whether the presence of this odour alone induced a response from the male, studies were conducted using deceased, non-glowing, female L. noctiluca (Schwalb, 1961). On encountering the female the males exhibited sexual behaviour, which resulted in copulation of a normal duration. The arrangement of female entrails, head, thorax and abdomen in a circular pattern also elicited a sexual response from the males. These results show that the bioluminescence is not required for copulation, the odour is sufficient stimulation, however, for this to occur the male would have to already be in close proximity to the female. It also shows that the female body shape is secondary for sexual behaviour.

Schwalb (1961) came to the conclusion that the female odour did not act as a remote stimulus to the male, so flight towards it would not be possible, it only intensified the sexual reaction. The bioluminescence is therefore essential for the females to achieve reproduction as without it the chances of encountering a female would be significantly reduced if not minute. From a distance the cold light reproductive signalling by the female is the only attractant.

1.3. Light pollution; lighting where it is not required or where it negatively impacts nature.
Several theories have been offered as explanations for the aforementioned decline in L. noctiluca, these include pollution, climate change, habitat fragmentation and changes in grazing pressures (Tyler, 1986). However it has been proposed that light pollution is also a contributing factor (Tyler, 1986).It has already been established scientifically that light pollution does have a negative impact on many nocturnally active species. These include other arthropods such as moths as well as larger animals like bats, fish, turtles and birds. The darkness also improves avoidance of predation and maintains biological rhythms. As all species of bat and their habitats have been protected since 1981 by the Wildlife and Countryside Act, light pollution is already of some concern as it can disturb roosting and foraging behaviour (Boldogh et. al., 2007).

The yellow light emitting low-pressure sodium bulbs, which are currently being phased out and replaced with Mercury & Ceramic discharge Metal halide (white lights), which emit UV light, which many insects are attracted to, including glow worms (Alexander, 1992). This change is due to levels of crime and fear of crime (Adams, pers. comm.). The newer white light producing bulbs make it possible to distinguish between colours and more easily recognise individuals (The Institute of lighting engineers, 2005).

In this project light pollution will be measured by the amount of light passing a given point, for a given period of time (luminous flux), upon a set surface area, with the scientific unit being lux (lx) (The Institute of lighting engineers, 2005). Lighting engineers seek to provide an even level of lighting along a road surface. The lux at ground level produced by a street lamp depends on the type of bulb and its wattage and position within the lamp, as well as the spacing between lamp columns and their height (Adams, pers. comm.). The level to which paths and roads are lit to depends on their usage, how reflective the surface is, crime levels and fear of crime (Adams, pers. comm.). It is the horizontal light pollution emitted from these sources that the female L. noctiluca have to contend with, along with the transient light of car head lamps if they are situated along the verge of a road.

The real effects of light pollution on L. noctiluca are currently unknown however work has been done to determine the intensities of light to which the males are attracted. It has been shown however that males react positively to light intensities up to about 200 lux, whereas at higher intensities, approximately 1000 lux they either do not respond, or react negatively (Schwalb, 1961).

Tyler (1986, 1994) mentions that males despite being able to distinguish between females and artificial lights may become distracted if they mistake them for other celestial bodies and become deflected by their light. This could be caused by household lighting in heavily populated areas or by road and footpath lighting; however Tyler (1986) believes that this is only a minor factor in the overall decline of the species.

It could be that even if the males are not attracted to the light pollution source that it would still interrupt reproductive behaviour. Lloyd (2006) wrote that both male and female fireflies, which also use bioluminescence for reproductive signalling, may be attracted to artificial lights as they could falsely indicate a suitable habitat. The males would perceive it as a colony of brightly glowing females, whilst would assume it was an appropriate site for oviposition. It was also added by Longcore and Rich in 2007 that stray light could ‘wash out the visual signals between males and female fireflies’. It could be that L. noctiluca may also be unable to perceive the less intense natural light of the females against that of more powerful streetlamps.

There is not a vast amount of literature regarding the effects of light pollution on females. Schwalb (1961) established that the sexual appetency behaviour of the females was disturbed at light intensities over 80 lux and with overhead lighting of 500 lux the females retreated and ceased glowing. Although Dreisig (1975) found that some females did not commence their reproductive behaviour with light over 1 lux, also stating that there was no activity when light levels remained above 10 lux. This figure falls in the middle of the range of recommended lighting levels for subsidiary roads and pedestrian areas (The Institute of lighting engineers, 2005), so it may not be unusual for individuals to encounter these levels.

Tyler (1994) also notes that during a glow-worm survey conducted during 1992 most of the sites were situated ‘well away from any form of artificial lighting’, whilst it was also found during a survey of L. noctiluca in Essex that the majority of sites (78%) were found in areas of no artificial lighting, including nature reserves (Gardiner et. al, 2001). However without visiting or conducting furthers studies in the area it is impossible to tell whether this was due to the lack of extra lighting or the result of other factors; levels of disturbance. It is still an indication that lack of light pollution may be an important constituent for a suitable L. noctiluca habitat.

None of the aforementioned research examines the impact that the artificial light has on the ability of female L. noctiluca to attract males. It could be that under increased background lighting the female bioluminescence is no-longer visible. It is the aim of this investigation to provide preliminary research into this field and evaluating the feasibility of conducting future studies. The implications of any results obtained will be discussed and interpreted in the context of glow-worm conservation.

2. Aims of the project

The aim of this investigation was to determine the effects of light pollution on the glow-worm, L. noctiluca, in terms of its reproduction. As the males are attracted to the females through phototaxis, light pollution from street lamps and other night lights could cause background lighting sufficient to drown out the females signal. The implications of the results obtained will be discussed and relevant conclusions will be drawn regarding the conservation of L. noctiluca in the future.

Counts will also be made each night to determine the numbers of females present at the site, to contribute to background knowledge, to aid in future studies. These figures will also be provided to the UK glow-worm survey, the long-term aim of which is to enable better estimations of L. noctiluca populations and to establish whether the numbers at specific sites are decreasing through year by year comparisons. Future projects will also be suggested and their feasibility discussed.

3. Materials and Methods

This project was conducted over six weeks of the summer of 2007. One week was spent searching for an appropriate site with a large enough population of L. noctiluca was found. A nature reserve of unimproved chalk grassland was found: Brush Hill, near Princes Risborough in Buckinghamshire. Studies were then conducted every night (42 nights) from the 22nd of June until the 2nd of August 2007.

3.1. Materials
3.1.1. The imitation female
Real female glow-worms could not have been used to attract males in this experiment because their size and hence the intensity of their bioluminescence varies naturally and they do not all stay lit for the same length of time. The females are receptive to changes in temperature, altering their length of glowing time accordingly (Dreisig, 1978). They also cease to bioluminesce once they have attracted a male (Schwalb, 1961), meaning that if real females were used they would have to of been monitored closely and replaced as soon as they had attracted a mate. Therefore imitation females comprised of LEDs will be used in this project, as advised by Tyler (pers. comm.).

Conducting this investigation with LEDs instead of females also lessens any possible negative impacts on the female glow-worms. They are small, delicate creatures which may be harmed with too much handling and moving them could cause distress or harm their reproductive success.

Schwalb (1961) found that male L. noctiluca were attracted to an almost completely drained flash-light, however the rate of attraction to decoy females was much higher.
Schwalb (1961) investigated the effect of decoys of different light intensities and found intensity-dependent behaviour. He discovered that if the imitation females were of too high an intensity compared to the light emitted by natural bioluminescence then the males either reduced or ceased activity completely. Therefore the LEDs used could not be too bright. After reviewing the literature concerning the frequency of light emitted by females and the wavelengths that males are attracted the LEDs used were green.

Imitation females were constructed (fig. 2.1.) using green 5mm light emitting diodes (LEDs) (make; WL28). These were soldered to a circuit board 2cm by 1.5cm, fitted with a mini resistor (make; M270R). This was attached to a 9V battery to create a circuit. The LEDs, circuit board, resistor and connection to the battery were all purchased from the branch of Maplins in Southampton. Garden canes were cut to a height of 30cm to simulate grass stalks, then the LEDs were attached to the top and bound on with duct tape.

It was assumed that the bioluminescence was the only guidance the males had to the presence of the female, as the sexual appetency inducing odour emitted by female L. noctiluca is not detectable over a great enough distance to act as a pheromone (Schwalb, 1961). Therefore the imitation LEDs do not need to be coated in this odour.

A tape measure was used to enable correct measuring of the distance from the light source, to ensure that the LEDs were equidistantly positioned. A torch was also taken to read the tape measure and to fulfil the risk assessment. All the other equipment listed in the risk assessment was also taken along with spare torch and LED batteries so that they could be replaced if lights started to dim.

3.1.2. Light pollution source
A ‘white light’ pollution source was used as this most closely resembles the mercury or ceramic metal halide (CMH) lights that are being used to replace the older low pressure sodium bulbs (Adams, pers. comm.). After preliminary tests a light source was chosen that emitted light of an intensity of 0.3 lux at a distance of 0.5 metres, (measured using a council light metre) this lux level decreased with increasing distance from the lamp (Appendix A, table 1.). This light pollution level was the highest obtainable from the light sources available, however it still falls between that of natural starlight; 0.0009 (Austin et. al. 1976) and the council’s recommended lighting level for footpaths of 1.5 lux (Adams, pers. comm.). It is therefore an appropriate level to use, as if the males are affected by this lower intensity artificial light then they are more likely to be affected by the 1.5 lux emitted by footpath lamps. When the vertical light pollution the light source also produced was blocked more horizontal light pollution was produced (Appendix A, table 1), however due to interrupted internal reflection the light emitted was further decreased at greater distances and was therefore unsuitable.

It was assumed that aside from the light emitted from the light pollution source it would have no other effect on the male L. noctiluca, i.e. heat energy. This meant that a secondary light pollution source did not have to be purchased, switched on with the lights covered and placed in the centre of the control.


3.2. Preliminary Experiments
The aims of the preliminary experiments were to firstly determine whether the fake females would attract males and secondly to test the design of the male catching trap.

For the preliminary experiment the LED part of the circuit described above was mounted in the centre of transparent circular platforms of radius 2.5cm (fig. 2.3.) were cut to provide a surface on which to apply a water soluble wax gel, the purpose of which was to stick and thus render immobile any male attracted to the LED in such a way that it was possible for the individual to be counted and then released. As this substance was water soluble it enables the wax to be rinsed off with a water spray and the males freed alive. This would have lessened what would have otherwise been a negative impact on the local glow-worm population as other insect collecting surfaces suggested were sticky permanently, so the insects could not be removed alive. This method would also have removed observer effects as the LEDs would not need to have been checked until the end of the trial.

The plastic discs were transparent so that if viewed from underneath the LED could still be seen, this design made it possible for males that had sighted the ‘female’ from above to still see their target once they had landed nearby, enabling them to crawl onto the surface of the disc. The whole structure was mounted onto the aforementioned garden cane (fig. 2.2.).

Three sets of six canes were set out, six as described above (fig. 2. 2) these consisted of the presence of both the LED and the leg wax, to see if the wax could indeed hold a male glow worm still, it had already been established that it was capable of slowing the movement of the ant; Pogonomyrmex. Another six had exactly the same plastic disc surrounding the LED except that it was not coated in the wax, acting as a control, making sure that the wax neither deterred or attracted more males (this one had to be monitored carefully as there would be nothing to prevent the males from leaving). Whilst the last six had had the LED switched off, acting as a control for the background number of males, to see if the LEDs were capable of attraction.

This was conducted for 6 days; however, modifications needed to be undertaken due to the impact of the adverse weather conditions; it rained for half the trails. This caused either the raindrops or the damp grass to dissolve the soluble wax, rendering the traps useless. This design could still have work in a climate with little rainfall. The design was altered to a plastic cup, with the LED inside, creating a pitfall-like trap (fig. 2.3.). Small holes for drainage were made in the bottom to prevent males from drowning in the water that would otherwise have accumulated there.

A second series of trials had then to be conducted to ensure that the modified design would work (Appendix A., table 3.). In these it was found that the LEDs did attract the male glow worms and that the cups functioned well at retaining them. As soon as this was established, the main experiment was set up, due to the time restrictions imposed by the short lifespan of the adult L. noctiluca (Tyler, 1986).

The preliminary experiments were also used to determine the depth to which the garden canes should be driven into the ground. This was done by randomly measuring the distance up the stems of grass plants that 12 randomly selected L. noctiluca females had climbed. The average was found to be approximately 9cm, so that will be the height that the LEDs will be placed above the ground.

3.2. Main Experiment
The focus of this investigation was into the effect of light pollution on the ability of male glow worms (L. noctiluca) to find females. However during preliminary experiments background notes were taken for students who may consider this area for a dissertation in the future.

If light pollution does interfere with the ability of male L. noctiluca to find the females, then it would be expected that fewer males would be found on the imitation females subject to the light pollution. It could also be predicted that with increasing distance from the light source any observed effect would decrease.


3.2.1. Methods
Two positions on the hillside of Brush hill were determined at random, although there was sufficient distance between them to safely assume that light pollution from one site would not affect the other. At one a light pollution source was set up from which six imitation females were placed equidistantly and at the other the six LEDs were placed in the same layout only without the lantern; as a control. The next night the positions were switched to remove any effect that the different positions could have on results. The LED’s were set at four increasing distances from the light pollution source: 0.5m, 1m, 1.5m and 2m. Increasing the distance away from the light source like this effectively reduces the amount of lux received by each LED (Appendix A, table 1.). However the control (no light source) is needed each night as abiotic factors may not be the same on subsequent nights. This control was also used each night to ensure that males were still present.

As daylight intensity affects the activity of both female and male glow-worms (Schwalb, 1961) the experiments therefore needed to be conducted at the same light intensity rather than the same time each night. As the readings were taken after the summer solstice, the light intensity required to induce bioluminescence would gradually be reached earlier, due to decreasing day length (Gardiner, 2006(b)). Consequently the experiment was started between 10.45p.m. and 11.30p.m. depending on light intensity. Each night the experiment was left running for the same length of time: 2 hours.

Due to the alterations in male-trap design following the preliminary trials the LED’s had to be monitored periodically. A plastic tub was used to collect the males as they were attracted to the LED’s. This was done because it was unlikely that they would remain by the fake females for the duration of each trial, as the imitation females could not produce the chemicals that induce sexual behaviour, they might attempt escape although the pitfall like nature would hinder this. The males were also caught because they could obscure the light of the LED they crawled onto from other passing males. They were not released until the end of each trial as otherwise pseudoreplication could occur (Hurlbert, 1984).

Results were noted in a field book as they were obtained to avoid reliance on memory. As abiotic factors, such as temperature, wind direction and rainfall could not be controlled, they were instead measured and also recorded to enable a more informed discussion of any anomalous results.

A 2-way analysis of variance (ANOVA) will be performed on the results, if the data meets all assumptions, with the two factors being presence/absence of light pollution and distance from the light source, to determine if any results are statistically significant. Graphs of the residuals will first be plotted to ensure that the data meets all the assumptions.  

3.2.2. Additional complementary experiments
As already mentioned in the aims of the project there was an intention to aid in establishing a better knowledge of L. noctiluca population sizes to add to the UK glow worm survey run by Robin Scagell. This data would serve a duel function as it could also be useful in interpreting any results and would provide background information for students wishing to work on glow-worms in the future. Each night the same transects were walked across a set area of the hill-side at approximately the same light intensity with the numbers of females observed being noted. This ensured that same area was counted each night to enable valid comparisons to be made between nights, although it will have to assumed that the individuals counting will not improve with time. The females were also counted before males became active as during copulation or directly afterwards they cease to be bioluminescent (Schwalb, 1961). The dates where males first started to appear were also noted to aid in later interpretation.

For consecutive evenings a global positioning system (GPS) was borrowed, this meant that when the females were counted their positions could be recorded. This enabled analysis to determine whether or not the distribution of females was clumped or random.

During the preliminary trials it was noticed that the females in the vicinity of the LED traps continued to glow after the females in the rest of the field had stopped. Whilst still testing the improved apparatus experiments were also conducted to see whether the females did glow for significantly longer. LEDs were placed by 12 females, with another 12 had only garden canes beside them. The glow-worms were then monitored by several people on a circuit and how soon they stopped glowing was noted.

4. Results

4.1. Experiments on light pollution.
Males were only attracted to and caught by the LED traps over a period of 20 days, although this activity span of the males was short, they were still not caught on each occasion.

Distance (metres)	Average of the total number of males found at both light polluted positions.	Average of the total number of males found at both non light polluted positions.
0.5	0	33
1	0	33
1.5	6	26
2	4	6

Table 4.1. The total number of males found on the LEDs under light polluted and unpolluted conditions at increasing distance from the light source.


Out of a total 108 male L. noctiluca attracted to imitation females at varying distances from the light pollution source or centre of the circle, only 10 of them were found on the LEDs under light polluted conditions. No males were attracted to the imitation females at 0.5 metres or 1metre and as distance from the light pollution source increased; lux falling on the LEDs decreased, more males were caught (Table 4. 1.) (Appendix B. 3.1.). To see whether this was a statistically significant difference a parametric two-way analysis of variance (ANOVA) was going to be performed, with the light pollution and distance from it as factors. However the data obtained did not fulfil all the assumptions of ANOVA, as it was not normally distributed (P<0.005)(figure 4. 1).

 

The percentage values of the total number of glow-worms caught at each distance/lux level were used to draw the graph. The raw differences could not be used because the total number of males caught at each distance were different.  For example if the difference had been two out of a total of ten, then this would have been a bigger difference than two out of a total of 100.

Matlab was used to fit several models to the data, before a custom model was generated; f(x)=a*exp(-b*x)+100, depicted graphically (figure 4. 2.) with 95% confidence bounds. This shows the correlation between light level and number of males attracted. The light pollution interfered with the males ability to find the females, as it increased the percentage difference between the males attracted to the control and the light polluted LEDs increased.

 
Figure 4.2. The percentage difference between the numbers of males attracted to LED’s, under four different lux levels, and control conditions. 95% confidence limits, upper=green, lower=red

The curve crosses the x-axis at 0.065 lux, predicting that at this light intensity the reproductive signalling is no-longer affected. No difference in the numbers of males attracted between the control LEDs and ones subject to this level of lux would be expected. This is consistent with the results obtained at a distance of 2.5 metres from the light source: 0.045 lux. At this lux level there was no difference in male numbers, this fits the above model, however results were not obtained for the second position due to a lack of males, so it could not be included in the graph.

4.2. Count of females
Females were first observed on the 22nd June at Brush Hill and data was collected until the 2nd August, as by this time no males were being caught (Appendix C, table 1.). As is shown graphically (figure 4.2.), it was found that the numbers of females producing bioluminescence showed an upwards trend until day fourteen, to a maximum of 331 individuals, with the only major departure coming on day 5. After this observed numbers decline for the next two days before rising again on day seventeen. Day seventeen was also the day when the presence of males was first noted, more than two weeks later than the females. After this female numbers decrease again, although, more do appear towards the end of the glowing season but do not reach a third of the number seen at the peak (day fourteen).

 
Figure 4.3. Numbers of bioluminescent females found over a set area of Brush Hill over time, red line indicates first appearance of males

4.3. Distribution of females
A grid was superimposed onto the GPS data (fig. 4. 3.), enabling the squares which were more than 50% covered including line of sight, as it was possible to see glow-worms from at least 5 metres away, to be shaded green. To see whether the observed distribution of females was clumped, random or over-dispersed the coefficient of dispersion (CD) was calculated by dividing the standard deviation by the mean number of females per square on the grid (because the mean equals the variance for a Poisson distribution) (see Appendix B, table 3.2. for the values used). As the CD was greater than one; 1.429, this suggests that the data was clumped.

The standard error was calculated for the data giving a value of 0.236. As the difference between the coefficient of dispersion and one is greater than that of the standard error, the divergence from unity can therefore be considered significant.
 
Figure. 4.3. Path walked across Brush Hill with GPS system, females are marked with enlarged points and numbers

4.4. Effect of imitation females on real females
A two sample T-test was performed on the data (Appendix B, table 3.3) obtained from comparisons between the duration of female bioluminescence when either an LED or a control garden cane had been placed beside them. The difference between the lengths of glowing times was not statistically different; (T22=1.66, P>0.05), the females did not glow significantly longer without an LED beside them however there was a trend in that direction.

4.5. Other observations
Females did not tend to move between nights, they were seen in the same place night after night. It also appeared that they were found predominantly on Poaceae.

5. Discussion

5.1. Light pollution.
5.1.1. Analysis.
From the results a correlation with lux level and ability of the imitation females to attract males was established, using Matlab a model was produced: f(x)=a*exp(-b*x)+100 (figure 4. 2.). This gives the difference in number of males that would be expected for a given level of light pollution in lux.

It was at a distance of 1.5 metres that the imitation females subject to light pollution at a level of 0.09 lux began to attract males, this increased at 2 metre, 0.07 lux although it was still not the equal to the number found under the control conditions. No difference was found between the numbers males attracted to the LEDs 2.5m from the light pollution source at 0.045 lux (Appendix B,  table 3.1.), however not enough males were present the next night to do the second position, so it could not be included in the data used to produce the graph. It is however consistent with the Matlab model as at this lower intensity no difference was predicted.

More sites need to be looked at to ensure the findings of this site are not anomalous, so that the results can be applied to other populations. Female L. noctiluca are only active over a couple of months during the summer, whilst the lifespan of each individual adult is limited to approximately two weeks (Tyler, 1996) with the males emerging later than the females. The period of male activity was further limited by reliance on the weather, as the males are not very tolerant of adverse conditions (Tyler, 1992). The weather during the study was poor and on windy nights or ones on which it rained few males were caught (Appendix C., Table 1.).

These factors imposed a severe time constraint on the study, which meant that it was not possible to conduct it at more than one site or further increase the distance from the light source. The results were collected until there were no-longer any males being caught. Ideally the experiment would have been run simultaneously at several sites, however this was not possible as it was difficult to find a site with a sufficient population. During experimental development other sites including Aston Rowant, Winter Hill, Christmas Common and various ones around Flackwell Heath were investigated, but the numbers of females at Brush Hill were by far the greatest found. This will have to be considered by students before selecting this topic for an undergraduate project. If a second suitable site had been found it would also have to have been close enough to the first to enable travel between them unless there had been enough volunteers to stay at both, although each participant would have to have been accompanied to fulfil the risk assessment at it is probably neither safe nor wise to conduct such an investigation alone in a dark field.

Further studies would also be able to determine whether the results of this investigation really are comparable to the effects of streetlamps, as the angle of lighting would be different. There would have been no point setting this series of experiments up under a real street lamp, as during the two weeks spent looking for a substantial colony of glow-worms, before the population at Brush Hill was discovered, no females were found underneath or in close proximity to street lamps. Without the presence of females it would not be possible to know whether there was a population of L. noctiluca, making it unlikely that males would be present. The only way it could have been conducted would be by transplantation of a street lamp into the centre of a glow worm population, although, the disruption alone could have unknown or negative effects. One population may have to be sacrificed in order for the true results of light pollution to be discovered. This is neither possible, nor desirable, so instead a camping light had to be used. 0.3 lux was the highest obtainable light pollution from all the light sources available, which is less than the average light pollution emitted from lighting on a foot path (Adams, pers. comm.). The light source did emit light pollution vertically, which might have affected the results, although this should not matter as some poorly designed street lamps do this to a greater extent (Adams, pers. comm.). If the source had been laid on its side to emit the light purely horizontally then the lighting would not have been uniform for all the LEDs, unless more light sources of the same make were purchased and a far greater number of positions within the field used. During the trials no male L. noctiluca were attracted to the actual torch, presumably because it did not emit the same frequency as the females, however a number of Lepidoptera, earwigs and other arthropods were attracted.

The light from cars passing along the road, although brighter than the street lamps, is transient, so they could have less of an effect than the constant light from street lamps. Populations of glow-worms that are found in light polluted areas, such as road verges, could be the result of females nearby, from the same population, who are far enough away from the street lamps for their reproductive success to not be affected. This was also suggested by Stewart in a draft version of a submission to the Royal Commission on Environmental Pollution (2008). The dispersal to the light polluted areas could occur during the larval stages, however as adults they may be unsuccessful under these conditions.

It could also be that the males at these sites could have eyes that are better adapted to dealing with light pollution compared to the ones at Brush Hill. There was no light pollution at the site at Brush Hill so there would have been no selection pressure for eyes that are more tolerant of or that can filter out light pollution. This may have to be considered if trying to reinforce populations, translocation from the site at Brush Hill to another in a more light polluted area could result in outbreeding depression.

5.1.2. Implications
As has already been mentioned the strength of light pollution used was less than that normally used to light the average footpath, however it interfered with the males ability to find the imitation females. As glow worm females rely entirely on phototaxis and not on pheromones to achieve reproduction (Schwalb, 1961), this observed effect of light pollution has serious implications. Real females may be able to move away from a light pollution source, to reduce the levels of light pollution they are subject to. However if they are able to do this then it could still deplete their energy stores; potentially reducing the energy they can spend on bioluminescence, lowering their chances of reproductive success.

Populations of L. noctiluca although widespread are patchy (Tyler, 1986). Increased light pollution could fragment and further isolate these glow-worm populations, given that the females are sessile by nature, rarely moving more than one metre from one night to the next (Tyler, 1986). The larvae are also flightless, which further hinders dispersal and would presumably also affect the species ability to recolonise an area, even once the cause of a local extinction has been remedied (Alexander, 1992).  

Street lamps along roads in an area containing L. noctiluca could result in habitat fragmentation. The light pollution would prevent the glow-worms from occupying habitats that would otherwise be suitable, or from accessing suitable habitats with intervening light polluted ground. If higher light intensities do effect the reproductive behaviour of L. noctiluca then as the females emerge from where the larvae dispersed then only the ones under natural lighting conditions will be found by the males and reproduce. The individual populations or patches on which they were found would become islands more vulnerable to extinction through stochastic processes.

If the number and size of L. noctiluca populations are declining, then the light pollution in areas surrounding their habitats needs to be re-evaluated. A number of councils including those of Buckinghamshire and Hampshire have plans to switch off or dim some road lighting between the hours of midnight and 5.30 (Adams, pers. comm.), (Gray, 2007). This is due to a growing concern for the energy shortage that will be faced in the future (Gray, 2007) Dimming of lights by 15-25% is the method likely to be employed in more built up areas, whereas turning lights off completely would probably be used along safer rural roads and paths (Adams, pers. comm.). These plans could benefit L. noctiluca populations as at the Brush Hill site females were observed glowing until past midnight, the latest one recorded stopped producing bioluminescence at 2.17 am.

If councils are already planning to turn lights off in some areas from midnight, then they may be persuaded to do so in areas surrounding glow-worm populations. This could be carried out between the end of May until the beginning of August to ensure the entire adult reproductive season covered, although an earlier start, maybe at 10.30-11.00pm should be considered as during this study the females were generally active by then. In work not yet published Gardiner draws the same conclusions regarding similar plans proposed by Essex County Council, highlighting the fact that it would not be glow-worms solely that would stand to benefit from such schemes.

However all the effects of switching the lights off should be considered. If lighting is switched off on the surrounding roads at night this could have detrimental impacts on other nocturnally active species as well as potentially fascilitating crime (Adams, pers. comm.) In the same communication it was also mentioned that most crime takes place before 11-12.00 however on the Thorn Hill Housing Estate, near Harefield where this ‘lights out’ strategy was tested three cars were vandalised on the first night of the trial.

It may also be that species avoid crossing roads because of the lighting there, without it they will only have the sound of the car and the lighting of the approaching headlamps as a warning, they could be more likely to cross road where they would be more likely to be hit by passing cars., however there is no research into this field.

The results found in this project also raise interesting questions regarding the sensitivity of the males eye and why it falls off with increasing light intensity. It could just be that the light pollution lights the females background to the same intensity of the light she produces. Then as the lux she is subject to decreases the contrast between the light produced by the natural bioluminescence and the darker surroundings increases, making the females presence more apparent to passing males. Another possibility is that the males simply do not search for females in light polluted areas.


5.2. Female abundance
The number of females at the site on Brush Hill was recorded for 42 days and at its peak the total number of females counted was 331 on the 5th July. Overall the female numbers increased until this point, see figure 4. 2., indicating that all the larvae do not pupate at the same time and so they do not all hatch together. On the 5th day however there were only 19 females observed, this could be explained by the weather data also collected (Appendix C., table 1.). The fifth night was the coldest of the whole trial, with temperatures only reaching 9°C, the conditions were also remarkably bright due to no cloud cover and the approaching full moon (second of the month).

On the 9th day (30th of June) there was also a dip in the number of females, it did rain that night which could have affected them and it was also the night of the blue moon. Gardiner (2006(a)) found that a full moon increased the difficulty humans had in counting glow-worms, although this search was thorough it could have been the case that a few may have been missed along the transects. Alternatively, there could have been fewer females glowing if their critical light intensity threshold, a concept mentioned by Dreisig (1975), had not been passed.

There was also a decrease in the number of females counted 2 days before the first males were observed (day 17/8th June), however there could actually have been males present two to three days earlier. These results concur with the work of Tyler (1986); he stated that the males appear after the females although in this case it was approximately two weeks after and not a couple of days although it could be that the poor weather delayed their emergence. Table 1. of Appendix C. provides After the males presence was first noted the numbers of bioluminescent females decrease, presumably because females were being fertilised, negating the need to glow.

A considerable number of female L. noctiluca were observed, indicating a population of considerable size. However it should be mentioned that this number could have been due to the weather conditions and may be a bad omen for numbers of L. noctiluca at this site in future years. Gardiner (2006b) found counts at a site in Essex, were higher on nights with rain or drizzle, than without. As bioluminescence ceases when fertilised there was probably a large number of unfertilised females, because the weather would have been too poor on several occasions for the males to fly (Appendix C, table1.). Tyler (1986) writes that ‘bad weather during the critical period of adult maturity can have serious consequences for reproductive success’, this may mean that in three years time (2010) there could be fewer females at this site; a project for the future?

In the future weather-dependant species like the glow-worm could be under increasing threat from climate change, as their reproductive stage is limited to approximately two weeks; 1-2% of their lifespan (Tyler, 1986). Glow worms may actually be doing better when you can’t see that many glowing, as it may mean they have successfully reproduced. However this does therefore makes it hard to establish population sizes. Although the L. noctiluca population appears to be under no imminent threat at Brush Hill it still ought to be monitored.

It was assumed that ability to observe females remained constant. The weather however did vary between nights, altering the visibility; this could not be controlled however the paths walked across the field were tight enough that it should not have had an impact on the results.


5.3. Female distribution
Upon analysis of the GPS data it was found that as the coefficient of dispersion was still greater than one (1.429) with the standard error, so the distribution of females can be considered clustered. The coefficient of dispersion could only be calculated for one of the GPS sets of data as analysing the second would have resulted in pseudoreplication, as it could not be assumed that all the females from the previous nights had attracted a mate.

The females could have been clumped within the site due to a local distribution of snails, their prey, as the adults emerge where they pupated which would probably be in the same area as where they fed. Tyler found a significant correlation between L. noctiluca and the species density of snails, although this was on a larger scale (Tyler, 1986, cited Tyler 1979). In a future study the distribution of snails should be recorded at the time L. noctiluca pupate; approximately 8-11 days before the start of the glowing season, to see if this is the case.

 L. noctiluca usually spend three years as larvae, so it is unlikely that individuals from the same clutch would remain in such close proximity, therefore not accounting for the distribution observed. Another possibility is that they become clumped as adults to increase the chances of attracting mates; males could be more likely to see a group from a distance than an individual.

To draw any firm conclusions this distribution study would have to be carried out at a number of sites. If comparing GPS data between sites the weather would have to be taken into account in the experimental design. It could be that during certain weather conditions only females in certain areas of a site would emerge, i.e. more sheltered ones on windy nights ( though this was not a problem at this site). The data could be recorded at one site then a second site could be travelled to the next night and visited until the conditions there were the same as at the first. However there would be no guarantee of a repetition of conditions, so this would not be an effective use of time. The progression of the glowing season could also represent an additional factor. For example it could be that the behaviour of females changes during the glowing season, they could begin the season more dispersed and gradually become clumped. A project could be conducted exploring this although the emergence of new females during the study could affect results.

5.4. Effect of LEDs on the length of female glowing period.
From initial observations made during the preliminary experiments, females in the vicinity of the LEDs appeared to glow for longer. Upon testing, however a trend was found in the opposite direction. The females’ ceased bioluminescence sooner with the LEDs placed directly beside them, although this difference was not statistically significant.

It could be the case that the initial females found continued to glow longer entirely by chance. However there may be other explanations. It could be that when females detect lights at a distance, they continue to glow, as genes for continuation of bioluminescence in the presence of other females could be widespread throughout the population as aggregation could attract more mates, increasing the chances of reproductive success. However when the LED is closer then they could retire earlier as it would appear much brighter; it could be that the intensity of the LED at such a close proximity is great enough to inhibit bioluminescence, as it was also marginally brighter than the average female. As the adult females cannot eat they have a limited energy supply (Tyler, 1986), so if too strong a competition is faced by an individual then it could do best to conserve energy for a time of less competition.

5.5. Plant species that the females are found on
Whilst counting the females one night the plant species the females found on was also recorded. It was not possible to do this at a second site though as there was not another nearby with a large enough population close enough to enable the light pollution experiment (the focus of this project) to also be run. Out of 64 females, 59 were found on grass. This could be due to its height or its structural stability, the fact that when the positions of the females were marked with garden canes that the next night they were found on the canes also suggests this. A future student could look at the composition of the flora at Brush Hill, identify the species and find out if the L. noctiluca were found more frequently than would be expected by chance on a plant species given the percentage of ground covered by that species. The morphological differences between these species could then be discussed.

5.6. Other trials and observations
Other shorter experiments were also conducted out of interest in the subject. Whilst waiting for the emergence of the males, trials were conducted into the feasibility of looking at the distance moved by females each night (both vertical and horizontal). This type of study could provide information into a possible trade-off between the energy expended climbing and the length of time (energy) spent glowing. Although it would be expected that weather would be an important factor, as there would be no point in glowing or at least not climbing so high to glow under conditions that males are unlikely to or do not fly in.

Garden canes with coloured pegs were used to mark the position of and distinguish between the individuals. After these trials however, this method is not recommended, though it could work better at a smaller less used site. It was difficult to find the canes in the dark, it was easily possible for them to be knocked by larger wildlife and on one occasion it was clear that they had been moved deliberately. Only using the canes did not allow whether a female had actually moved or if a second female had emerged to be known.

It may be better to mark the females with UV pens in conjunction with the canes. This would ensure that the same female was measured on subsequent nights. It would however first need to be determined that the UV marking technique does not affect survival.

During this trial it was observed that some females were visible one night and not the next. It can be concluded that either it was the same female both times; it had not come out one night, or that a new one had emerged in the same place after the first had reproduced. In captivity L. noctiluca larvae have been observed feeding collectively on the same snail (Tyler, 1986, cited Wootton, 1971), it could be that this also occurs occasionally in the wild if two larvae come across the same prey item or its trail before pupating, as it can take up to 12 hours for a larva to complete feeding (Tyler, 1986). This could explain two different females emerging in the same place. Marking females with UV pens would help to shed light on this.

Whilst collecting data on one occasion bats were observed catching flying male L. noctiluca, this contradicts the findings of Schwalb (1961) who found that this species has few predators, this is probably due to its short lifespan (no specialists) and aposematic nature. It could be that the bats were unable to distinguish between the glow-worms which were unpalatable to them and the moths when both species were airborne. Or it could be that the energy reserves of the bats had been depleted as subsequent to these observations the weather quality had been poor (Appendix C, table 1.), as Rowe and colleagues (2007) found that predators were increasingly likely to eat more defended prey when their diet was restricted; energy levels affect foraging behaviour.

5.7. Additional studies for the future
There is still a vast deal that remains unknown about this species, so there is plenty of scope for future projects, several potential topics have already been mentioned in the text above. After the experience in this field provided by last summer it is the advice of this author that future undergraduate projects are dependant on the activity of the female. The primary reason being that it is easier to tell when they are active and that they appear to be active for longer periods. In contrast male L. noctiluca, emerge later and it is often difficult to tell when they are active unless you find them on a female. Male activity is also weather dependent, restricting their already short periods of activity further, limiting data collection. The conditions experienced this summer were awful (Appendix C, table 1.) which was reflected in the numbers of males on some nights. It could be useful to find out precisely what conditions male L. noctiluca will tolerate flying in, as this could also have implications on their conservation, regarding any future changes in climate. Imitation females could be set out at several sites and abiotic factors such as rainfall, wind speed, temperature and also potentially moonlight (depends on phase of moon and cloud cover) recorded. The experiment would have to be designed carefully though to ensure statistics could be performed.

Dreisig (1978) noted that the females glowed for 138 minutes on average, however whilst making my observations I noted that on most occasions the females glowed for longer, this could have been due to the prolonged glowing associated with a decrease in temperature mentioned in the same paper. The factors determining the length of this bioluminescent period could be investigated by a future student, with the other potential trade-offs examined.

It is also worth noting that numbers of glow-worms found at established sites can still vary hugely from year to year, as has been found in the literature through the work of Tyler (1986) who looked at the data also collected from a site in Buckinghamshire, discovering figures over three years that ranged from 5 to 200 females. This observation was supported by the fact that at Aston Rowant, a site which had previously been found to sustain a large glow-worm population (Scagell, pers. comm.) relatively few were seen.

To conclude, this project suggests that light pollution may indeed be a contributing factor to any decline of L. noctiluca in this country. If further studies at other sites reinforce these findings then street lighting surrounding or in between areas containing glow-worms will need to be re-evaluated. Although the glow-worm is not a keystone species, no-one can deny that the British countryside would be a darker place without it.

Acknowledgements

I would like to thank the following people: firstly Joel Parker (supervisor) & Lex Kraaijeveld for advice regarding my dissertation, in particular the statistics, John Tyler, Alan Stewart and Robin Scagell for initial advice regarding the feasibility of studies on L. noctiluca, Robin Scagell again, for offering information on glow-worm sites in Buckinghamshire and for the use of his GPS system, both John Tyler and Tim Gardiner for sending me some of their articles and research papers.

I would also like to thank all the people who came with me to the site at Brush Hill, providing transportation and aiding in counting the females, whilst ensuring risk assessment was fulfilled, these include: Helen, Charlotte & Neville Bird, Emily Taylor, Bronwen Taylor, Martin Udall, Fliss Evans, Alix Whiteway, Martha White & Duncan Lofty. I would also finally like to thank Mike Adams (Southampton city council lighting manager) for information regarding council lighting policies and luminescence (lux) emitted by street lamps.


References

Adams, M., 2008. [Discussion on council lighting practises] (Personal communication)

Alexander, K. N. A., 1992. The glow worm, Lampyris noctiluca (L.), in Gloucestershire and its conservation. The Gloucestershire Naturalist. 5, p.1-5

Austin, R. H., Phillips, B. F. & Webb, D. J., 1976. A Method for Calculating Moonlight Illuminance at the Earth's Surface. The Journal of Applied Ecology. 13(3), p741-748.

Boldogh, S., Dobrosi, D. and Samu, P., 2007. The effects of the illumination of buildings on house-dwelling bats and its conservation consequences. Acta Chiropterologica. 9(2), p.527-534.

Booth, D., Stewart, A. J. A. & Osorio, D., 2004. Colour vision in the glow-worm Lampyris noctiluca (L.) (Coleoptera: Lampyridae): Evidence for a green-blue chromatic mechanism. 207, p.2373-2378.

Branham, M. A. and Wenzel, J. W., 2003. The origin of photic behaviour and the evolution of sexual communication in fireflies (Coleoptera: Lampyridae). Cladistics. 19, p.1-22.

Carlson, A. D. and Jalenak, M., 1986. Release of octopamine from the photomotor neurones of the larval firefly lanterns. The Journal of Experimental Biology. 122, p.453-457.

De Cock, R. & Matthysen, E., 1999. Aposematism and bioluminescence: Experimental evidence from glow-worm larvae (Coleoptera: Lampyridae). Evolutionary Ecology. 13, p.619-639.

De Cock, R. & Matthysen, E., 2003. Glow-worm larvae bioluminescence (Coleoptera: Lampyridae) operates as an aposematic signal upon toads (Bufo bufo). Behavioural Ecology. 14(1), p.103-108.

Driesig, H., 1975. Environmental control of the daily onset of luminescent activity in glowworms and fireflies (Coleoptera: Lampyridae). Oecologia. 18(2), p.85-99.

Driesig, H., 1978. The circadian rhythm of bioluminescence in the glowworm, Lampyris noctiluca L. (Coleoptera, Lampyridae). Behavioural Ecology and Sociobiology. 3(1), p.1-18.

Fowler, J., Cohen, L. and Jarvis, P., 1998. Practical statistics for field biology, second edition. Somerset: John Wiley and Sons Ltd.

Gardiner, T., Pye, M. & Field, R., 2002. Glow-worms Lampyris noctiluca L. (Coleoptera: Lampyridae) in Essex: results of the 2001 Essex Glow-worm Survey. Essex naturalist. 19, p.151-159.

Gardiner, T., Pye, M. & Field, R., 2003. The influence of soil pH on the abundance of the Glow-worm Lampyris noctiluca (Linnaeus) (Lampyridae) in Essex. The Coleopterist. 12, p.121-123.

Gardiner, T. & Gardiner, M., 2005. Monitoring Glow-worm Lampyris noctiluca L. (Col.: Lampyridae) populations in grazed and mown grasslands. Entomologists’ Record & Journal of Variation 117, p.263-266.

Gardiner, T., 2006(a). Does the phase of the moon affect Glow-worm Lampyris noctiluca L. (Lampyridae) counts? The Coleopterist. 15, p.92.
 
Gardiner, T., 2006(b). Effect of survey start time on counts of the Glow-worm Lampyris noctiluca (Col.: Lampyridae). Entomologist’s Record & Journal of Variation. 118, p.184-185.

Gardiner, T., 2007(a). Short-term changes (2001-2005) in Glow-worm Lampyris noctiluca L. (Coleoptera: Lampyridae) abundance in Essex.  British Journal of Entomology & Natural History. 20, p.1-7.

Gardiner, T., 2007(b). Optimum survey dates for detecting Glow-worms Lampyris noctulica Linnaeus (Lampyridae) in Essex. The Coleopterist. 16, p.12-13.

Gardiner, T., 2007(c). The decline of the Glow-worm Lampyris noctiluca in Epping Forest. Essex naturalist. 24, p.90-92.

Gardiner, T., soon to be published. Lights out for glow-worms. Entomologist’s Record & Journal of Variation.

Gray, R., 2007. Dimmer switches to be tested on street lights. Telegraph [internet]. 19 February. Available at: http://www.telegraph.co.uk/news/main.jhtml?xml=/news/2007/02/18/nlight18.xml [accessed 1 march 2008].

Howard. R., Howard. J. & Platt. B., 2004. Species Action Plan for Sussex: Glow-worm, Lampyris noctiluca. Sussex wildlife Trust.

Hurlbert, S. J., 1984. Psedoreplication and the design of ecological field experiments. Ecological Monographs. 54(2), p.187-211.

Linnæus, C. 1758. Systema naturæ per regna tria naturæ, secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locis. Tomus I. Editio decima, reformata. - pp. [1-4], 1-824. Holmiæ. (Laurentii Salvii).

Longcore, T. & Rich, C., 2004. Ecological light pollution. Frontiers in Ecology and the Environment. 2(4), p.191-198

Lloyd, J. E., 2006. Stray light, fire flies and fireflyers. In C. Rich & T. Loncore. (eds.). Ecological Consequences of Artificial night lighting. Island Press, Washington, D. C. p.345-364.

McKillup, S., 2006. Statistics explained. An Introductory Guide for Life Scientists. Cambridge: Cambridge University Press.

Phillips, K., 2004. Glowing attraction. The Journal of Experimental Biology. 207(14), p.i-a.

Rowe, C. & Guilford, T., 2000. Aposematism: To be Red or Dead. Trends in Ecology and Evolution. 15, p.261-262.

Sala-Newby, G. B., Thomson, C. M. & Campbell, A. K., 1996. Sequence and biochemical similarities between the luciferases of the glow-worm Lampyris noctiluca and the firefly Photinus pyralis. Biochem. J. 313, p.761-767.

Scagell, R., 2007. [E-mail on feasibility of glow-worm studies] (Personal communication, 1 May 2007).

Schwalb, H. H., 1961. Contributions to the domestic lampyridae Lampyris noctiluca GEOFFR. and Phausis Splendidula LEC. and experimental analysis of their predatory and sexual behaviour (Translation). Zoologisches Jahrbuch (Zoological annual). 88, p.399-550.

Stewart, A., 2008. [E-mail correspondence. Draft submission to: the Royal Commission on Environmental pollution in response to the call for information on artificial light in the environment: The potential impact of Artificial Light on Glow worms, 16 Jan 2008].

The Institute of Lighting Engineers, 2005. The Outdoor Lighting Guide. Abingdon: Taylor & Francis.

Tyler, J., 1986. The ecology and conservation of the glow-worm, Lampyris noctiluca (L.) in Britain. Atala. 10/12, p.17-19.

Tyler, J., 1994. Glow-worms. Stratford-upon-Avon, UK: Herald Press

Tyler, J., 2002. The glow-worm. Sevenoaks. Kent: Lakeside printing Ltd.

Tyler, J., 2004. Luminescence on the world wide web: the UK Glow-worm
Survey website. Luminescence. 19(6), p.358-358.

Tyler, J., 2007. [E-mail on feasibility of glow-worm studies] (Personal communication, 1 May 2007).

Tyler, J., Mckinnon, W., Lord, G. A., & Hilton, P. J., 2008. A defensive steroidal pyrone in the Glow-worm Lampyris noctiluca L. (Coleoptera: Lampyridae). Physiological Entomology. doi:10.1111/j.1365-3032.2007.00610.x.

Zhang, C. and Zhang R. Q., 2006. Matrix proteins in the outer shells of molluscs. Marine biotechnology. 8, p.572-586.

Appendix A

Table 1. Light pollution emitted by chosen light source in different positions and at different distances

Distance from light source (metres)	Lux emitted horizontally	Lux emitted vertically	Lux emitted horizontally with top covered
0.5	0.3	27.3	2.2
1.0	0.18	6.0	0.04
1.5	0.1	2.4	0.035
2.0	0.07	1.2	0.03
2.5	0.045	0.9	0.02

Table 2. Preliminary data to test improved experimental design of glow-worm trap

	Trap with LED	Trap without LED
Number of males found	4	0

Table 3. Preliminary data testing the light source

Day number	Number of males on light polluted LEDs	Number of males on non light polluted LEDs
Day 1	0	4
Day 2	1	9
Day 3	0	10

 
Appendix B

Table 3.1.  Number of males attracted to LEDs under the two conditions: with light pollution and without

Distance from light source if present (metres)	Illuminance at that distance if light source present (lux)	)    Total numbers of males attracted (light polluted)	Total number of males attracted (no light pollution)
0.5	0.3	0	33
1.0	0.18	0	33
1.5	0.09	6	26
2.0	0.07	4	6

Table 3.2.  Statistics used to analyse the GPS data collected from Brush Hill

Mean	0.703
Standard deviation	1.004
Coefficient of dispersion	1.429

Table 3.3. Number of minutes after LED/cane was placed beside a female glow-worm that they ceased bioluminescence.

Female number	with LED	without LED
1	5	5
2	5	18
3	19	20
4	22	21
5	26	23
6	26	38
7	30	45
8	33	55
9	34	56
10	36	61
11	45	97
12	62	122
Total	343	561
Average	28.58	46.75

Appendix C
Table 1.  Raw data table containing abiotic factors recorded and numbers of L. noctiluca thoughout the period of study

Date	Night	Females	Males	Temp.	Rainfall?	Wind (Beaufort scale)	Other conditions
22 June	1	61	0	12	Earlier on	0	misty
23	2	58	0	12	Earlier on	1
24	3	86	0	12	Drizzle	1
25	4	92	0	11.5	Heavy rain	4
26	5	19	0	9	0	0	very light
27	6	154	0	10.5	Drizzle	1
28	7	179	0	13.5	0	2
29	8	202	0	13	Earlier on	1
30	9	185	0	14.5	Rain	1	full moon
1 July	10	271	0	13	Earlier on	2
2	11	229	0	11	Drizzle	3
3	12	283	0	13	Earlier on	1
4	13	300	0	13	Drizzle	4
5	14	331	0	13	Earlier on	7
6	15	314	0	13	0	2
7	16	232	0	13	Earlier on	1
8	17	257	5	14	0	0
9	18	248	0	14	0	2
10	19	214	4	14	0	1
11	20	93	10	13	0	1
12	21	68	10	13	Earlier on	2
13	22	18	2	16	Rain	1
14	23	23	14	15	0	1
15	24	13	19	18	Earlier on	1	new moon
16	25	27	22	14	0	0	misty
17	26	26	11	12.5	Earlier on	2
18	27	28	20	16.5	0	1
19	28	59	12	15	Earlier on	3	lightning
20	29	36	0	15	Rain	6	too windy, not even months around light source, difficult walking
21	30	53	0	12.5	Rain	1	misty
22	31	66	0	12.5	0	2
23	32	64	1	13	0	2
24	33	49	3	14	0	2
25	34	55	0	19.5	Earlier on	4
26	35	56	0	14	Earlier on	4
27	36	55	7	14	0	1
28	37	20	0	14.5	Rain	0
29	38	1	0	9.5	0	1	full moon
30	39	15	0	10.5	0	1
31	40	27	2	14	0	1
1 August	41	9	0	16	0	0
2	42	4	0	14	0	0

Appendix D: Risk assessment
Risks
1.    During preparation of the LED circuits there will be a risk of being burned by the soldering iron.
2.    The experiments are going to be conducted in the field at night, so there is also a risk of attack and of becoming lost/disorientated.
3.    It will be dark (human cone cells cease working at a light intensity of 0.001Wm-2, whilst rod cells are functional in lower light levels, until approximately 0.0000001Wm-2), so there will be an added danger of tripping over unnoticed objects, standing on sharp items (glass, metal spikes) or spraining an ankle in an unseen rabbit hole.
4.    Stinging plants like nettles or plants like thistles and brambles may cause injury if undetected. As could insects that bite or sting (centipedes, ticks, hymenoptera etc.)
5.    The study cannot be conducted in the rain, however it could still be cold, despite being summer and there is always the possibility of a sudden downpour.

Precautions
1.    Due care will be taken when soldering and a lab coat worn to protect the arms
2.    It will always take at least one person with me and I will inform another where I am going & at what time I should be expected back. A map will also be taken along with a mobile phone.
3.    I will carry both a torch and a basic first aid kit. The mobile phone already mentioned can be used contact people in case of emergency.
4.    The first aid kit will also include piriton which contains an antihistamine to combat allergic reactions, antiseptic wipes and plasters.
5.    Appropriate clothing will be taken; a jumper/coat as well as waterproofs incase it suddenly rains.
6.    I will also check the sites for permission of use with the authorities in question; family friends/national trust/farmers.