Pendleton, T.A.. Pendleton, D.T. and Dale, M.G. "Some notes and observations on the bioluminescence exhibited by the larva of the Glow Worm (Lampyris noctiluca) in captivity" www.eakringbirds.com; December 2012.

In August 2012, following the end of another adult glowing season, we decided to engage in two research projects, both of which produced results directly linking the two projects together in several aspects of research.

It was anticipated at the outset, that both research projects would probably take several years to complete before enough data had been collected and from which any firm conclusions could be derived. This page details our current findings.

Project 1: Phase 1. 2012 was conducted by Trevor and Dilys Pendleton (TDP), based at Market Warsop, Nottinghamshire and Martin Dale (MD), based at Mapperley in Nottingham. The primary aims of this research were to try and determine the following aspects of bioluminescent activity in larvae of the Glow Worm Lampyris noctiluca.

• To determine consistent patterns in the bioluminescent glows produced by hunting/active larvae
• To determine if there is any form of bioluminescent interaction between larvae, other invertebrates and prey location etc
• To determine conditions conducive to larval bioluminescence (seasonal, temperature, wet or dry substrate, hunger etc)
• To explore the theory that larvae which habitually climb plant stems and glow for long periods, eventually produce females
• To determine if any other specific larval behaviour patterns, could eventually be attributed to their resulting sex

At the start of each research project, we were unaware of any similar research projects, either in progress, or already completed and the results published. The different aspects of each project were something which we had wanted to discover for ourselves, hopefully providing some answers to many questions we had asked during the previous years spent surveying adults at a well known Sherwood Forest Glow Worm colony.

The two projects were to look into the number and frequency of bioluminescent glows produced by captive Glow Worm larvae on a nightly basis (Project 1: Phase 1. 2012) and to determine the size ratios of larvae in the wild during late Summer and Autumn (Project 2: Phase 1. 2012).

1.1 ... Captive setups

11 larvae were collected from two Glow Worm colonies near Edwinstowe in Nottinghamshire, during late August 2012. The larvae were of varying sizes, with the smallest measuring 9mm and the largest being 23mm. All larvae were separated, numbered from L1-L11 and housed in large plastic containers of two types, either containing a sand/soil mix or stone substrate with a naturalised layout to provide daytime shelter etc.

All 11 larvae were kept indoors, in an unheated room and near a natural light source. The use of artificial lights was prohibited during the research period and all setups were exposed to the sun at some part of the day. Throughout the period, only dull red illumination was used to avoid disturbing the larvae and depleting the observers dark adapted vision. Water or moisture was provided by occasional spraying when the used substrate dried out and spraying was restricted to during the day when required.

The two basic setups are shown in the following three photographs below. The top two photographs show the typical setup devised by MD for his nine larvae. The setups were covered with perspex sheeting during the day, which was removed when in full sun. The 230mm pots contained approximately 40mm of coarse gravel for drainage and 40mm of sandy soil substrate, filling the pots to approximately half full. In order to simulate a natural habitat, stones, loose bark, moss and grass were added to each pot, together with a stick offering a permanent structure to observe any climbing behaviour.

Another behavioural pattern noted from some of MD's larvae, was the distinct preference for hunting just along one section of the pot wall. This was observed as a 180° circling of the container, before turning round and retracing the same track back until reaching it's original start point, then turning back once again. Why some did this is not clear, though it could possibly have been due to sunlight illumination during the day, warming only one side of the pot. The same problem was not believed to have been evident in the hunting movements of L1 and L2, which were housed in rectangular containers with TDP. The lower right hand photograph shows the stone substrate setups TDP used for L1 and L2.

The methodology used for recording the data obtained during the observation periods was a simple process. We (TDP) were fortunate to be able to have one person watching larvae 1 and 2, then relaying the duration of bioluminescent glows produced by each larva to the other person, who acted solely as the recorder, noting the times, length of glows and recording the information on paper. Because of our involvement with Project 2, it was decided that we would monitor just two captive larvae, leaving MD to monitor the remaining nine.

To help timing accuracy, MD used an old talking clock that could be operated by pressing the button with his feet, leaving hands free to record the times on paper. It was a technique which worked well, although the occasionally high number of glows from a number of larvae, did mean that data transfer onto paper became quite a hectic procedure.

It was estimated that a minimum of two weeks monitoring would be required to provide us with enough data and so a single, two hour observation period on consecutive evenings was pre-selected. This was to run from 22:00h - 00:00h, with the start tme allowing for complete darkness by the beginning of each observation period.

The observation periods ran from 29/08/12 - 22/09/12 for Larvae 1 and 2 (L1and L2); from 01/09/12 - 22/09/12 for L3, L4, L5 and L6 and from 05/09/12 - 22/09/12 for L7, L8, L9, L10 and L11.

2.0 ... Larval bioluminescence

Observations to record larval bioluminescence over a two hour period ended on 22/09/12. By then, we believed that we had gathered enough data to be able to satisfy our own criteria and that the data would prove accurate enough for later publishing. MD continued monitoring his nine larvae on a more casual basis well into October and this served a useful purpose in confirming the information all observers had gathered from earlier results.

2.1 ... The bioluminescent rates produced by hunting/active larvae

The results obtained were to surprise all three observers, with some quite staggering counts of larval bioluminesence (GPN) produced within the selected two hour observation periods.

We found that not all larvae produced high counts of bioluminescent activity and certainly for five larvae used in the research, it was considerably less. MD also noted that L3 and L7, actually appeared to glow brighter than the other larvae in his charge. We were not able to differentiate between the bioluminescent brightness of L1 and L2, but obviously had a much smaller sample of larvae to enable any comparison.

When the resulting figures in the Table 01 are examined, an obvious difference in the amount of bioluminesent activity produced by L1, 2, 3, 4, 6 and 7, is evident from that produced by L5, 8, 9, 10 and 11.

The number of nights before larvae resumed glowing after completing feeding, was also found to vary between larvae.

L1, 2, 6 and 7 were all recorded to glow the night following feeding, on at least one occasion and it is interesting to note that these four larvae also produced considerably higher total GPN rates than all the others.

Larvae could suddenly become active after a period of dormancy/inactivity during the selected observation period, or opposingly, could suddenly cease all activity and become dormant. Occasionally, individual larval activity did not begin until well after midnight, or larvae would again become active well into the early morning hours following earlier activity.

The average GPN (glows per night) recorded for each larvae, over all observation periods, are shown in Table 01 and again splits the 11 larvae into two categories. The average GPN between the six larvae producing highest levels of bioluminescence (L1, 2, 3, 4, 6 and 7) and those producing the lowest rates (L5, 8, 9, 10 and 11) show a limited amount of variation in the GPN figures produced for both groups. Those larvae with the higher rate of activity produced a GPN range of 26.95 - 42.55 GPN, whilst larvae with a lower activity rate, produced a range of 1.83 - 9.16 GPN, excluding the returned 0.00 GPN showed by L8.

If we look at the GPM (glows per minute) rates of the 11 captive larvae and group them according to their highest GPM rates, then we have virtually identical groupings of larvae, as based on the highest levels of bioluminescence just discussed. The only difference would be that L5, is now included in the group producing the highest GPM (L1, 2, 3, 4, 5, 6 and 7) and missing from the group producing the lowest GPM (L8, 9, 10 and 11).

Wild larvae were primarily being studied to determine the size range of larvae in the wild during late Summer and Autumn, but in order to locate larvae for data collection, we first had to find them. This was done after dark, when we looked for the bioluminescent glows produced by larvae at a Sherwood Forest site.

Sudden reduction in bioluminescent activity was noted from both wild and captive larvae around mid-September. This was actually found to coincide nicely at the point where both day and night become of equal length. The combined GPN activity profile of all larvae in Figure 01 shows this decline.

One of the aspects of larval bioluminescence we wanted to research, was variation in glow duration of hunting/mobile larvae. To do this, we timed the duration of each glow from L1 and L2, then calculated the average across each two hour observation period. Occasionally, larvae glow for extended periods of over ten seconds, so two sets of figures are given in the results. Six individual glows during a two hour observation period was believed to be the lowest total high enough, from which to derive an accurate average glowing length. A total number of glows below six, is signified in Table 04 by "insufficient data".

Table 04 shows the data obtained from L1 and L2 on a total of 24 dates between 29/08/12 and 22/09/12. No data is available for 20/09/12, as we were recording larvae in Sherwood Forest as part of Project 2: Phase 1. 2012.

The total average glow duration on nights when bioluminescent activity was produced, shows remarkable similarity for both larva. Table 05, provides a summary of the combined average glow duration data for L1 and L2.

Long duration bioluminescent glows of 10+ seconds were extremely unusual in hunting or mobile larvae in captivity. Adding the total GPN for L1 and L2 from Table 01, we get a total of 1,591/48 GPN (48 representing the total number of observation periods). Out of this total, the number of 10+ second bioluminescent glows recorded was just 16, which equates to a rate of about 1.00%.

Although the data gathered from the two wild larvae is limited, we believe that there is enough available data to use as a direct comparison with captive larvae, if we use the averages for L1 and L2. A comparison between the average glow durations of L1 and L2 against the two wild larvae, are given in Table 06.

A wild larva found on 15/09/12 was recorded glowing eight times within a ten minute observation period. Two bioluminescent glows lasted for 10 seconds, leading to an average glow duration of 7.33sec/6 if the 10+sec glows are taken out of the calculations. Add in the 10+sec duration glows and an average glow duration of 8.00sec/8 is reached. Based on this single observation, the larva had a maximum GPM rate of 2.00.

The bioluminescent data for the second wild larvae found on 19/10/12, gives averages of 7.88sec/4, if the 10+sec glows are taken out of the calculations and 12.46sec/13 with the inclusion of 10+sec duration glows. What was interesting about this larva, was the high number of long duration glows, producing nine long glows out of a total of 13 in a 15 minute observing period, including single glows of 20sec and 26sec duration. This larva also produced a maximum GPM rate of 2.00.

Observations revealed that hunting larvae were capable of producing an extraordinary number of glows and differed in the number and frequency of glows produced, but a definite pattern of peaks and troughs within the glowing activity was found. This was most evident on evenings when larvae were extremely active, often producing a continuous series of glows throughout the two hour observation periods, when visual detection of both the end and start points of separate bioluminescent glows became almost impossible. The 'y' axis scale on the left represents the number of glows per minute.

On 13 occasions, the number of glows produced by a single larva was recorded at well over 100 times per two hour observation period. The highest GPN (glows per night) rate recorded, was a staggering 379 by L3 on 02/09/12, averaging 3.15 GPM (glows per minute). This is illustrated in Figure 02, but two other good examples of high GPN rates are illustrated in Figure 03 and Figure 04. These two graphs show two different high bioluminescent activity profiles and illustrate very well, how even high levels of larval activity can vary.

It was noticed after studying the resulting data from the observing periods, that we could split our captive larvae into two groups based on their low level activity profiles, i.e. their low GPN and GPM rates. Larval size was not a factor we believed that we had to consider, as all 11 larvae were still at least several months away from being full grown.

Our present theory, is that larvae with low GPM and GPN rates, eventually develop into adult males.

2.4 ... The occurrence of long duration bioluminescent glows from captive and wild larvae

That there could possibly be a way to separate male and female larvae based on an aspect (or aspects) of larval bioluminesence, was something we had discussed many months before the start of our Project 2: Phase 1. 2012 research in July and August 2012. The origin of our theory was not based on larval GPN or GPM rates, but as a potential explanation for the very long duration glows often recorded from larvae in the wild.

Section 2.2 dealt with variations found within the duration of larval bioluminescence exhibited by hunting/mobile larvae, where a glow of over ten seconds was the criteria required to assign a glow as being of long duration. But this referred only to larvae which were actively hunting or mobile and we felt it neccessary to have a different criteria for long duration bioluminescence from stationary larvae.

As the captive larvae were being observed without the use of bright artificial light, we had to ensure that the criteria used in determining long duration bioluminesence between hunting and stationary larvae was not misinterpreted and decided that a higher minimum glow length would help differentiate between stationary and hunting/mobile larvae long duration bioluminesence. This was established at the start of both research projects, with 20 seconds decided as being the most appropriate minimum length of time.

Table 07 lists all the long duration (20+sec) bioluminesent glows produced by captive larvae 1 - 11. Bioluminesence longer than 20 seconds from active larvae was extremely rare, occurring just once out of a total of 4,431 larval glows recorded between 29/08/12 and 22/09/12. Yet within a few days of ending this area of our larval bioluminescence research, three long duration glows were recorded from L1 and L11, occurring after normal hunting/mobile bioluminescence had declined.

Two glows produced by L1 on 01/10/12, were the only glows we recorded from both L1 and L2 after 22/09/12. MD also noted the sudden decrease in the number of hunting/mobile glows from larvae 3 - 11, but did record a single glow of over two minutes by L11 on 06/10/12.

We believe that we are almost able to rule out temperature change as being a cause for this drop in GPN rates. In captivity, L1 and L2 (TDP) were kept in an unheated, naturally warm room at a constant temperature of 18 - 20°C, with larvae 3 - 11 (MD) also in an unheated room, but subject to greater temperature fluctuations of 12 - 17°C.

Outdoor temperatures recorded from the Sherwood Forest site used for our wild larvae research, ranged from 6 - 17°C between August and October. Despite the constantly warm temperatures L1 and L2 were kept at, the decline and eventual cessation of larval bioluminescence was in accordance with captive larvae kept in lower temperatures and with larvae in the wild. It seems that the contributing factor had to be the change in day/night length and around September 25th, when night lengths finally exceed 12 hours.

Table 07 also lists instances of extremely long duration bioluminescent glows produced on several occasions by feeding larvae. We later termed these "feeding glows", which were characterised by their extreme longevity and faintness, although they could be interrupted by short periods of non-bioluminescence.

If prey was caught and feeding commenced during daylight, feeding glows were not observed when it subsequently became dark, even if larvae were still continuing to feed.

As with other types of larval bioluminescence, we recorded a general decline in the number of feeding glows observed during September and although feeding did continue, feeding glows were rarely recorded after early October.

Another potentially interesting observation was that feeding glows were not produced by all larvae that caught prey during September's observing periods.

Larval hunting was not something we had actually discussed researching before we engaged in Project 1: Phase 1. 2012, but we feel it useful to publicise our observations to a wider readership here. Initially, we were more interested in the bioluminescent activity produced by larvae upon the detection of prey, and any bioluminescent reaction produced when larvae came into direct contact with other invertebrates and potential predators.

The clear advantage of having a number of larvae in captivity, meant that larval hunting and the detection of and immobilisation of prey items, could be observed under more controlled conditions than are possible through field observations. Although we do have first hand experience of larval hunting in the wild, a captive environment was much more suitable for research purposes.

3.1 ... Conditions conducive to larval activity and bioluminescence

Determining the suitability of conditions leading to general, or increased larval activity, proved difficult and there is plenty of scope for a more detailed long-term study into determining what particular conditions (if any), are most conducive to larval activity and bioluminescence associated with hunting/mobility. From our own research and observations, larval activity is dependant on moisture, humidity, temperature, nocturnal feeding, post-feeding dormancy and moult preparation and/or recovery.

The two substrate types and average room temperatures used in both our own (TDP) and MD's larval observation setups, allowed for varying levels of moisture, humidity and temperature levels to be tested.

The peat and soil substrate favoured by MD for larvae 3 - 11 was kept generally damp throughout the larval observations of August and September, but did have to be watered occasionally. The stony substrate used by TDP, dried out considerably quicker and although perhaps more representative of the substrate conditions found locally in Sherwood Forest, substrate was subsequently shown not to be the important factor.

But we were unable to determine if temperature had any definitive effect on larval activity in a captive environment, with little temperature variation (17 - 20°C between TDP's and MD's setups) during August and September. Even when provided with a consistantly warm setup, larval activity still showed a greater amount of variation than we had expected. But this does not mean that temperature is not altogether a conducive factor towards larval activity, because on-going research as part of Project 2: Phase 1. 2012 has found evidence to suggest that larvae do respond to temperature increases in the wild, but will also be dependant on other contributing factors listed at the beginning of section 3.1.

Post-feeding dormancy was another contributing factor towards larval activity. While most larvae would not become active for several evenings after feeding on substantial prey items, it was not unknown (within the short period of our captive observations) for larvae to resume bioluminescent activity the evening following completion of feeding.

The whole aspect of larval hunting and prey detection has always been open to debate. There are certainly two theories, one suggesting of deliberate (intelligent?) larval hunting techniques and the other that both hunting and prey detection are just completely random processes. Either way, regular larval bioluminescence would probably be a characteristic of both suggestions, even though it would have no significance in hunting methods. But we were interested in seeing if bioluminescence was perhaps representative as an indicator of larval awareness of prey items, or was increased at the moment of prey location and capture.

Larval bioluminescence at the moment of prey location, was one of the main aspects of Project 1: Phase 1. 2012 and observing larvae in captivity was the only practical method to observe this, but we were unable to devote the required amount of hours to this aspect of our research in 2012 for anything conclusive to show. We were particularly interested in looking for any obvious changes in the larval glow rate, or in the bioluminescent activity pattern produced at the time of prey location and how long bioluminescence was maintained after capture.

Observations revealed that hunting was almost certainly, a totally random process and prey items seemed to be located purely by chance. A larva could quite easily miss a meal, just by walking the wrong side of a stone or other obect on the substrate. We looked for indications that larvae deliberately followed the mucus trails left by prey, but were unable to find any really conclusive evidence to support the theory that they actually did. When larvae were particularly active, bioluminsecent activity did appear to increase where larvae were recorded spending more time on areas of the substrate where prey had been present previously. This may have been down to larvae coming up against a natural obstacle on the substrate, or in the case of L1 and L2, a corner of the container and then being indecisive about which way to go next. Again, we could find no firm evidence to suggest that larvae deliberately tracked intended prey by following mucus trails.

We also found no real evidence to show that larvae glowed more frequently when prey items were available. Even when larvae were active during one of the larval observation periods, the amount of bioluminescent activity (GPN and GPM) was only once obviously higher than an averaged activity level. This is shown by L1 between 23:16h and 23:33h in Figures 04 and 04a. However, we did find some evidence to show that larvae do glow well at the moment of finding and biting food, but unfortunately not well enough to be consistantly discernable above general bioluminescent activity preceding prey detection.

It soon became obvious that Snails were less likely to be lost by the larva after the initial bite. Although both Slugs and Snails recover quite quickly from a single bite, Slugs would often get 'lost' by the larva as they move away, despite the Glow Worm larva circling them. Snails however (when bitten initially) retreat into the shell and in all the instances of Snail capture we observed, the larva climbed straight onto the shell and remained in situ before inflicting the first bite and remained on the shell until the prey was dead.

Figure 06 shows the bioluminesent activity profile produced by larva 2 after attacking a sub-adult Garden Snail (Cornu aspersum) prior to the start of a two hour observation period at 21:35h. Our notes made at the end of the observation period, stated that the Snail was still moving slightly at 00:00h, with the larva still on the shell but had not commenced feeding.

The number of prey items taken by each captive larva shown in Table 09, are thought to be accurate, although there may have been times when feeding on small prey items took place out of sight of the observer. Even taking into account the differences in larval size, it became apparent that some larvae were less able to hunt and kill as effectively as others. As a large larva, L8 only managed to kill 2 prey items and remained inactive and dormant for much of the time, but eventually went on to complete a lengthy moult. Conversely, L1, 2 and 3 all appeared to have a large appetites, often feeding for extended periods and swelling to more than twice their original size.

Feeding by all larvae reduced sharply towards the end of September/early October, with L5 the last larvae recorded to feed on the 23/11/12. Despite by now being in an obvious state of diapause, L1 and 2 were easily encouraged to take food on 06/11/12, although when prey was placed adjacent to them, it did take a minute or so for full activity to commence. Once they were aware of it, prey was despatched in the usual manner.

It would also be expected that nocturnal hunters would be encouraged to hunt for longer as the nights lengthen, even when the temperature was as conducive to larval activity in late October/early November, as it was in early late August/early September. But it appears that the increase in night length is equally an important factor towards reducing hunting activity, as much as it is a factor in reducing larval bioluminescence.

MD noted that larvae occasionally returned to feed on prey, sometimes after having become dormant for more than 24 hours after feeding and even when being at some distance from the kill. L3, 5, 7 and 8 all did this at various times, returning to finish off what was left, and often disappearing entirely inside the remains of the snail shell. Glow Worm larvae, must also be opportunitic feeders, something which would account for several larvae observed feeding on the same prey, noted before all MD's larvae were separated.

Another interesting aspect of feeding behaviour recorded by MD, was the dragging of prey items under cover. L3 and 5 both clearly exhibited this, with L5 visibly moving a 75mm bark chip some 10mm in the process on 24/11/12. We have also recorded wild larvae moving prey on two occasions, but L1 and 2, would also move prey to a more safer environment before consuming it in captivity.

In order to monitor any growth, all larvae were measured at suitable (monthly) intervals to record their relative sizes throughout the period of captive observations (September-December 2012). Larvae were measured lengthwise in a manner we have used for several years, while the larvae were flat to the surface (or hand) they were on. The first opportunity to carry this out, was when the larvae were initially separated and set up in early September. Subsequent measuring then usually took place during habitat maintenance in October, November and December 2012.

MD found larval shrinkage and size reduction alarming initially. However, it became clear that individual larvae could increase or decrease in size, markedly over short periods of just a few days following feeding. This was probably attributed to larvae rapidly taking on water during feeding. Some may be down to measuring difficulties where the larva was less inclined to 'open up' during dormancy, although it was clear that some larvae had lost significant body mass from the start of the season. Larval size will need to be monitored more closely during the 2013 phase of this research project.

Only two larvae moulted their skins whilst in captivity. L11 moulted on the 17/09/12 (within 24 hours) and became active again shortly after and L8 eventually completed it's moult on the 20/09/12, after having remained dormant for much of its time after being transferred to its own container.

The direct reaction exhibited by some invertebrates to sound, is a relatively well known phenomenon and perhaps the best example is that displayed by moths avoiding capture from Bats. Moths are able to detect the echo location produced by Bats and react accordingly by literally dropping out of the air to avoid capture.

The possibility that Glow Worm larvae may exhibit some reaction to sound or vibration in the form of bioluminescence, was something that had interested us and the following results and findings have been brought together from both current research projects. Before proceeding further, it is perhaps worth listing the most typical forms of external physical influences we employed to try and trigger a direct response.

Under a captive environment, external physical influences could be attributed to the following list

• isolated knocks or jarring
• vibration through movement
• sudden lound noise
• sudden bright light

Our observations of larvae kept in captivity, produced only limited direct reactions to external physical influences. All physical influences tended to have a negative effect on larval bioluminescence, or movement displayed by an actively hunting larva, when both regular bioluminescence and movement would temporarily cease, before quickly resuming again.

However, we did find that physical influences would initate a positive response from a feeding larva, already producing the faint bioluminescence typical of feeding glows. In these cases, it was found that larval reaction would consist of a sudden increase in brightness for a short period of time, before the level of bioluminescence gradually decreased to become just visible to the naked eye. A positive reaction to sudden (sometimes accidental) knocks on the container was also recorded from larvae which were dormant or inactive (post-feeding) at the time, but these were found to be occasional and with no reliable expectancy.

Under field conditions, external physical influences could be attributed to

• vibration caused by motor vehicles
• walking
• trampling through grass
• direct disturbance of vegetation
• sudden bright light

Although all larvae will glow for a period when directly handled, our observations showed completely negative responses to any of the listed physical influences in the wild. In the field, physical interaction either had a direct negative effect, or failed to promote any bioluminescent response when purposefully looked or tested for.

If Glow Worm larvae were to show any form of bioluminescent response or reaction to vibration, then we believed that a vehicle moving just a metre away from vegetation known to contain larvae, would trigger and confirm any larval response. But despite numerous attempts, with one observer walking directly behind a vehicle travelling along forest tracks, we found no evidence to suggest that even the vibration from motor vehicles would trigger any form of bioluminescent response.

Cessation of long duration larval bioluminescence was regularly recorded when attempting to locate stationary larvae glowing on plant stems or leaf litter, for the purpose of capturing them to measure their size. Bioluminescence was found to cease almost immediately vegetation close to larvae was disturbed, or especially when the actual stem the larva was on at the time was touched or knocked.

Sudden bright light from a hand held torch used during the latter stages of larval location in the field, would also act as the trigger for any larval bioluminescence to stop.

3.6 ... Bioluminescent interaction between Glow Worm larvae and other invertebrates

After collecting the first of the nine larvae to be used in our captive research, they remained temporarily housed for several days in a clear plastic, rectangular container with dimensions of 20cm x 15cm. These larvae later became the nine used by MD in his controlled observational research. At night, they were watched collectively for signs of any obvious larval interaction on meeting, but despite much activity from several larvae at any one time, we recorded no bioluminescent glows or any series of glows that could be attributed to direct larval interaction upon two larvae making contact.

Finding larvae glowing in close proximity to adult females in the wild, had prompted the question of whether larvae were habitually attracted to the bioluminescence emitted by adult female Glow Worms. This was something that we had wanted to test in captivity, so we took the opportunity to trial this by using an artificial lure designed and constructed by MD, which had proved extremely successful in attracting adult males during trials at Sherwood Forest in 2011. The lure was placed in the container, to see if it would successfully initiate any form of larval response, but returned results all proved negative and we observed no evidence to suggest that there was any larval attraction to bioluminescence, and that larvae observed glowing in close proximity to adult females in the wild, had been purely coincidental.

Research on the bioluminescent reactions produced by larvae when coming into contact with other invertebrates was more successful. We had proposed to conduct some trial experiments using L1 and L2, but these never took place as we felt that enough evidence had been obtained from wild larvae as part of Project 2: Phase 1. 2012 based in Sherwood Forest.

Between August and October 2012, we recorded a total of nine instances where wild larvae were found glowing within approximately 1.00cm of other invertebrates. Bioluminescence from the stationary larvae, suggested that both parties had just come into immediate contact and that the larval bioluminescence was clearly a responsive and/or a defensive reaction to the encounter.

Among the selection of invertebrates were two predatory species. The very common Wolf Spider Pardosa amentata (Lycosidae) was recorded with (but not attacking) larvae on two occasions and the small Ground Beetle Notiophilus biguttatus (Carabidae) once. Invertebrates offering no threat, but appearing to initiate a bioluminescent reaction, were the Common Rough Woodlouse Porcellio scaber (Porcellionidae), Common Shiny Woodlouse Oniscus asellus (Oniscidae) and the Pill Millipede Glomeris marginata (Glomeridae) which was recorded on four occasions.

The larvae under the careful observation of MD, showed a further decline in activity towards the end of November, with individual larvae remaining dormant for extended periods as diapause commenced. The smaller larvae (L9, 10 & 11) showed a tendency to seek out cracks and crevices in the soil substrate and remained well out of sight. As an example, L11 disappeared on 27/11/12 for a total of 26 days, after having crawled 30mm down the side of a gap between the stick and the soil substrate. L9 and 10 were often to be located deep within Sphagnum moss on the substrate surface, or in the grass root stocks. The larger larvae (L3, 4, 5, 6, 7 & 8) more frequently choose to remain under the bark, either on the substrate, but most often hanging upside down from the bark.

The decrease and eventual cesation of larval activity shown by MD's larvae, coincided nicely with the inactivity of L1 and 2 with TDP. L1 and 2 also spent most of the time immobile and clinging upside down under moss or loose stones. Both tended to remain inactive, although occasional movements, characterised by either larva transferring from under one stone to another, were noted regularly until late December.

In late October, the internal temperature of MD's unheated room fell to reach a night-time low of just 3.5°C. In order to see if a rise in temperature would bring some of the larvae out of hibernation, L3, 4, 5 and 6 were moved to a heated part of the house on 08/11/12. The rest of the larvae remained in the unheated room, provided the temperature did not dip below zero.

With an average night-time temperature not lower than 14°C, there was an increase in larval activity from the larvae that were moved the following week. L3 was observed actively hunting, glowing and killing a snail on 09/11/12 with L4, 5, and 6 also active around the same time. L6 was recorded attacking a Snail, but never managed to actually kill it. Despite the increase in temperature, within a week, all four larvae were dormant again, and showed little difference in activity to the larvae remaining in the unheated area of the house.

The choice of location for diapause shown by captive larvae, leads us to believe that similar sites would be chosen in the wild, depending on the site. At our Sherwood Forest study site, we believe that larvae simply retreat under stones, moss, within leaf litter and deep within grass tussocks and probably rarely remain in one place all Winter, moving when mild conditions allow.

Phase 2 of our research will see the continuation of work started as part of Phase 1. As these projects tend to throw up potentially new avenues of research (as did Project 1: Phase 1. 2012) it is always difficult to be able to know exactly where our research will take us. However, the following basic points, are certainly among our list of research priorities for 2013.

• To rear the 11 larvae used in our captive research to adult. If the number of males and females emerge as we expect, then we will be considerably closer to answering at least one of our original research aspects.

• To collect a further sample of wild larvae, seperable by their high or low GPN and GPM rates in the field and rear to adult in captivity, enabling accurate comparison with data collected from captive larvae in 2012.

• To monitor the GPN activity profile of captive larvae during the Autumn, to see if returned results are similar to those of 2012 and to confirm the equal day/night length theory in association with the cessation of larval bioluminescence.

• To determine the height at which male Glow Worms fly using artificial lures on posts of different heights. This work will probably run in conjunction with Project 2: Phase 2. 2013.

• To observe males in captivity and gather data on the bioluminescent flashes they produce.

Pendleton, T.A.. Pendleton, D.T. and Dale, M.G. "Some notes and observations on the bioluminescence exhibited by the larva of the Glow Worm (Lampyris noctiluca) in captivity" www.eakringbirds.com; December 2012.