The Lamp Test Results

• Total UVB output
• "photobiological" activity
• spectra
• UV Index gradients
• The effect of reflectors and domes
• UVC

What is different about these lamps?

It was difficult, at first, to understand the reason for the specific problems with these particular lamps.

Total UVB output at the manufacturer's recommended basking distances

The ZooMed Reptisun 10.0 Compact Lamp is described as " ideal for all desert and basking reptiles" and the 5.0 Compact Lamp as "ideal for all tropical basking reptiles". No information is available on the box or on the ZooMed website regarding any recommended basking distances. ZooMed are now including a comprehensive information sheet in the packaging of all new stock, but this was not inside any of the new boxes we obtained back in April. Hence it seemed, at first, that maybe the lamps were just being placed too close. However, we measured the output of their lamps with a broadband UVB meter (Solarmeter 6.2) and found total UVB levels not unlike those found in natural sunlight, at distances which caused eye problems in reptiles. For example, a brand new ZooMed Reptisun 10.0 Compact Lamp gave a reading of 306 µW/cm² at 6 inches. Reptiles have been recorded basking and foraging in sunlight in the wild at similar levels (up to 322 µW/cm²). The reading at 10 inches was 121 µW/cm².

The Big Apple Herpetological Mystic lamps are claimed to be "Rated #1 most powerful UVB emitting fluorescent reptile light" and to "produce as much UVB as the highest quality UVB mercury vapor bulbs".
The manufacturers recommend a minimum basking distance of 10 inches if used without a screen between lamp and reptile. We recorded total UVB levels from a brand new Mystic Compact Lamp as 128 µW/cm² at 10 inches, and from a brand new Mystic 24" linear tube as 142 µW/cm² at 10 inches.

The new R-Zilla Desert 50 Series are described as "ideal for these desert-dwelling reptiles: African fat-tailed gecko, leopard geckos, Schneider's skinks, bearded dragons, nile monitors, tegus, blue-tongue skinks, red- footed tortoises, uromastyx, emerald swifts and savannah monitors".
No minimum basking distance is specified regarding the T8-size linear tube. We recorded total UVB levels from a 15 watt tube that had been burned-in for 150 hours as 69 µW/cm² at 10 inches.
Included with the T5-size "Low Profile Single Bulb Fixture" (a tube inside a ballasted fixture complete with a UVB-transmitting acrylic cover sleeve) is a detailed instruction sheet stating that the lamp unit is "for use over covered terrariums only" - presumably the cover would be a screen. The recommended tank sizes are all of maximum height 12 inches, so this is presumably the suggested maximum distance from the reptile; a minimum distance is not specified. We recorded total UVB levels from a brand new T5-size "Low Profile Single Bulb Fixture" with its acrylic sleeve in place, as 137 µW/cm² at 10 inches.

Thus, the total UVB output of all these lamps, although high for fluorescent lamps, does not seem alarming. Natural sunlight can have a much higher total UVB output (measured in µW/cm²) than these lamps at reasonable basking distances, and yet reptiles do not develop photo-kerato-conjunctivitis in natural sunlight. In addition, there are other types of UVB lamp available, such as some UVB mercury vapour lamps, which do not appear to cause photo-kerato-conjunctivitis at distances where a high, or higher, total UVB level is recorded.

The "photobiological activity" of a lamp: measuring the UV Index

The lower UV wavelengths of sunlight (around and below 300nm) are much more biologically active than the higher wavelengths (up to 320nm). It is these lower wavelengths which enable vitamin D3 synthesis, and in general, the lower the wavelength of UV light, the greater the potential for damage to skin and eyes, too.

Could these lamps have a higher proportion of their UVB in these lower wavelengths, making their UV light more photobiologically active - i.e. "stronger" - than sunlight?

Broadband UVB meters, such as the Solarmeter 6.2, measure the entire UVB range (280 – 320nm) and give a readout of the total UVB irradiance in µW/cm². They cannot give any indication of what percentage of the total UVB is in the more biologically active, lower wavelengths. Therefore, a simple comparison of the readings from a broadband meter from the sun and from a lamp with "stronger" UV light could be misleading as regards the potential of the lamp to cause cell damage, photo- kerato-conjunctivitis, or, for that matter, to enable the synthesis of vitamin D3.

We therefore needed to find out how much of the UVB from these lamps was in the more biologically active, lower wavelengths, and how the output of the lamp at different distances would compare to natural sunlight. This can be done by measuring the UV Index (UVI) of light from the lamp at various distances.

The UV Index is a universal, unitless measure indicating the intensity of UVB radiation in the biologically active range of wavelengths - specifically, the wavelengths that enable skin synthesis of vitamin D3, cause damage to the DNA of living cells, and produce erythema (sunburn) in human skin. It is widely used in assessment of solar radiation -and even announced on weather forecasts. (1)(2)

We have just begun to use the Solarmeter 6.5 UV Index Meter to reveal the "photobiological activity" of different lamps. This meter gives a direct readout of the UV Index. It has a sensor response which follows the Diffey Erythemal Action Spectrum effective irradiance (Eeff), which closely follows the vitamin D action spectrum (Deff) in the UVB range (280 - 320nm).(3) In simple terms, the meter "weights" the readings from different wavelengths, giving a "higher score" to the more biologically active wavelengths. The final, total "score" is the UV Index.

When readings were taken from sample lamps from the three brands described above, and compared with readings from other lamps and with solar recordings, the answer was revealed: there were, indeed, considerable differences.

Figure 6 shows the UV Index for the sun and for a selection of UVB reptile lamps, recorded with the UV Index Meter at the distance at which the Solarmeter 6.2 reads exactly 100 µW/cm² of total UVB.

The proportion of low-wavelength UVB in sunlight (shown at the top of the chart) increases with solar altitude, because when the sun is lower in the sky, the rays must pass through a thicker layer of atmosphere. Since low-wavelength radiation is more effectively absorbed by the atmosphere than the higher wavelength light, a higher proportion of this is removed. This accounts for the difference in the UV Index between paired solar readings with the total UVB reading of 100 µW/cm² taken with a clear early morning sky, and taken with an overcast sky close to mid-day. (There's more about the solar UV Index later...)

The "photobiological activity" of the first five lamps in the table - a mixture of linear and compact fluorescents - can be seen to be relatively similar to that of the sun. The next two - the Reptisun 10.0 linear tube and the Exoterra 8.0 linear tube - have slightly higher relative UV Index recordings suggesting that their spectra may have a slightly higher percentage of lower wavelength UV light.

However, the final six lamps in the table are markedly different. The UV Index recordings are between four and eight times higher than the solar UV Index for the same amount of total UVB (in µW/cm²). Light from these lamps would therefore appear to be between four and eight times as photobiologically active as light from the sun.

The spectra of the fluorescent UVB lamps

The next step in evaluating these lamps was to examine their spectra. What wavelengths of light do they emit, to produce such high photobiological activity?

Our spectrometer is an Ocean Optics Inc. USB2000 spectral radiometer with a UV-B compatible fibre-optic sensor with cosine adaptor, calibrated for absolute intensity. The full spectrum (UV and visible light) was recorded for each lamp, and the UV portion compared with solar recordings, and those of other lamps.

Fig. 7, left, shows an example of a solar UV spectrum (blue line). was taken in full sunlight close to mid-day in Wales, UK, with the sun at its highest point, close to the summer solstice. Solar elevation was 60.8 degrees. The solar spectrum begins at around 300nm and the irradiance rises fairly steadily through the UVB wavelengths and continues to rise up through UVA to visible light. Reptiles can actually see it from about 350nm upwards.

The black curve is the Pre-Vitamin D3 Action Spectrum (CIE 174:2006) (4), which shows the potential of a given wavelength to enable vitamin D3 synthesis. As can be seen, this overlaps with the solar spectrum up to about 315 - 320nm. Wavelengths at around 298nm (where the action spectrum peaks) are most efficient, but the amounts of this in sunlight are very small indeed. Higher wavelengths are progressively less efficient, until wavelengths above about 315nm can be seen to have very little effect indeed. Having said that, there is vastly more light at these higher wavelengths, so their effect may be just as important as the tiny but more potent amounts at around 300nm.

The red curve is an action spectrum for DNA damage (adapted from Setlow 1974)(5). This shows the potential of a given wavelength to cause injury and death to cells. From this graph, it can be seen that the lowest wavelengths of sunlight, around 300 - 310nm, do overlap this action spectrum.... no strong sunlight is perfectly "safe" as we know... but most of the solar UV spectrum lies outside of this "danger zone". The action spectrum of photo-kerato-conjunctivitis (for which we were unable to obtain data sufficient to plot on a graph) in several species of mammal, including man, peaks at around 270nm (6).This suggests that it is not dissimilar to the action spectrum for DNA damage which we have shown. Although no studies appear to have been done on reptiles, there is no reason to suppose their response would be markedly different.

Figure 8 shows the UV spectra of a small selection of reptile UVB lamps which as far as we know, have not been associated with any photo-kerato-conjunctivitis or other health problems. A quick comparison of the absolute irradiance of these lamp spectra (all recorded at the very close range of 10cm.) with that of the sunlight shows how feeble this irradiance (even at 10cm) actually is; but it is not the total amount of light we are concerned with, but rather the spectral power distribution (SPD); the proportion of light at each wavelength, i.e., the "shape" of the spectrum compared to that of the sun.

All have a similar threshold wavelength around 290 - 295nm, followed by a steady rise up towards UVA. The shape of the spectra is similar to that of sunlight up to about 340nm, but then the UV output falls again and there is very little higher wavelength UVA except for the mercury emission peak at 365nm.

With the exception of the ExoTerra lamps, these all have UV spectra very similar to the lamp known as UVA-340.(7) The typical spectrum of one of these lamps may be viewed here (external link; opens in new window; pdf file - takes short time to load) This is widely used in photobiological experiments and other tests requiring artificial sunlight, since it is claimed to have the closest possible resemblance to solar UV in the UVB portion of the solar spectrum.(8)(9) The spectrum is also very similar to that of human tanning lamps with a high UVB content.(10) Slight differences in the phosphor blends and in the composition of the lamp glass cause some variations in the spectral pattern, and affect the total amount of radiation produced, as does the wattage of the lamp and its type (compact or linear tube). The ExoTerra ReptiGlo lamps have a similar spectrum to the others in the UVB region, but significantly more UVA, especially higher-wavelength UVA. This would probably make the lamp more visible to a reptile (with visual threshold from 350nm)(11) than other types.

The spectra of the fluorescent lamps which have been connected in some way with cases of photo-kerato-conjunctivitis are shown in Figure 9. All of these appear to utilise a very different phosphor, which is generating low wavelength UVB from a threshold as low as 275 - 280 nm.

This is the typical UV spectrum of the phosphor used in so-called "FS" lamps.(12)(13) The typical spectrum of one of these lamps may be viewed here (external link; opens in new window; pdf file - takes short time to load).

FS lamps are used for testing the deterioration under UVB of resistant materials such as roofing and car bodywork,(12) and in older-style human clinical phototherapy lamps.(14) Eye protection is always worn when any of these lamps are used in clinical work. These lamps are never used as tanning lamps or for cosmetic purposes.

Lamps which contain a high proportion of this "phototherapy" phosphor, and which have glass which does not block UVC, emit radiation as low as 270nm. However, the lower wavelengths are apparently attenuated by reducing the proportion of phosphor in the blend and by use of different glass. The Reptisun 10.0 lamp, for example, does not emit UVB below 280-285nm; the Big Apple Herpetological Mystic tube and the R-Zilla Desert 50 Series T8 lamp emit UVB down to 275nm.

UV Index Gradients: How much radiation do reptiles receive from these lamps?

At the height of summer, in the UK, full mid-day sunshine can produce a UV Index (UVI) of about 9. In the tropics, mid-day sunlight may reach levels of 10 - 14, but few animals are out and about under such strong sunlight. A UVI above 16 -17 is not naturally found in the most intense sunlight on the face of the earth.

A small number of recordings taken in the field, alongside basking lizards in the tropics -usually seen in early to mid-morning or from mid-afternoon onwards - suggest that reptiles which bask in the sun choose times of day when UV levels are fairly low - the highest recorded UVI in this study, for example, was only 7.6.

Reptiles which live in the shade, (in rainforest, for example) may never expose themselves to as much as this; the UV Index in rainforest shade may be less than 1.0 even at mid-day.

There is insufficient evidence to make any definite recommendations regarding an "ideal" UV Index gradient at a basking spot. If the highest UVI in which a sun-loving lizard has been seen basking is 7.6, then a maximum of 6 - 8 might seem appropriate for that one species. No more can really be said; but in the absence of any other evidence, it seems reasonable to use figures like this as a temporary guide. Certainly it would seem advisable to ensure no animal is ever exposed to abnormal levels of UV radiation, which could not occur anywhere in the animal's natural habitat.

Sets of recordings were made from all the lamps on test, with the UV Index Meter (Solarmeter 6.5) to construct charts of the UV Index gradient created by each lamp. Measurements were all taken at right angles to the lamp axis, halfway down its length.

Figures 10 and 11 show the results for the Big Apple Herpetological Mystic linear fluorescent tube, and compact lamp, respectively, both before and after burning in.

It can readily be seen that at close range, both these lamps are emitting extremely hazardous levels of radiation. At 6 inches, even after burning in, both lamps yield a UVI higher than any sunlight on the face of the earth. At the manufacturer's recommended minimum basking distance (10 inches) the UVI with the new linear tube we tested was 16.6 (that of the most extreme equatorial mid-day sun) and after burning in, it was still 11.2 (that of tropical late morning sun.) At 10 inches the UVI reading of the new compact lamp we tested was 13.9 (equivalent to intense tropical mid-day sun.)

If a maximum UVI of 6.0 is required, a new linear tube must be positioned at least 20 inches from the reptile, and a new compact lamp needs to be at least 16 inches away.

Figure 12 shows the recordings for the R-Zilla Desert 50 Series lamps on test.

With these lamps, too, at close range the radiation is very hazardous. The brand new T5 lamp, with its removable acrylic cover taken off, yields a UV Index of 81.6 at 4 inches. Animals have been placed at this distance from the lamp by unsuspecting owners. Even at 12 inches, a distance often recommended as a maximum for other fluorescent UVB lamps, the new unshielded T5 lamp is hazardous with an unearthly UV Index of 20.7. Even at 18 inches, the level is still above a tropical 10.0. Burning in does reduce the output considerably, of course, but both the T8 lamp (at 150 hours) and the T5 lamp with the acrylic shield in place (at 105 hours) still emit radiation exceeding UVI 20 at 6 inches and UVI 10 at 10 inches.

Figure 13 is a complex chart as it plots the UVI gradients of all the ZooMed Reptisun 10.0 lamps which were on trial. The first 3 lamps shown in the key were brand new samples; these, not surprisingly, have the highest output. Lamps BW3, 4 and 5 were all lamps known to have caused photo-kerato-conjunctivitis in reptiles. The final two lamps (BZC1 and MB2) were Reptisun 10.0 compact lamps of an earlier design; these had a much lower UVB output even when brand new.

Once again, even well burned-in lamps produce unearthly levels of UV radiation further than six inches from the side of the lamp. The new lamp with the highest output would need to be placed 16 inches above a reptile for the UVI to be below 6. Interestingly, the manufacturer's minimum recommended distance for a new Reptisun 10.0 lamp (positioned horizontally, like this, above the reptile with no mesh or reflector) is 23 - 25 inches, according to their new information sheet. This, according to our recordings (not shown on the chart) would yield a UVI of about 2.3.

The last chart in this series is that for the ZooMed Reptisun 5.0 Compact lamps. (Figure 14)

The first lamp, BJP1, caused photo-kerato-conjunctivitis in a small group of hatchling lizards which could approach to about 4 inches. Although the output of this lamp is far lower than that of a Reptisun 10.0, it it not insignificant, and the UVI climbs rapidly to hazardous levels at ditances closer than 4 inches.
The manufacturer's minimum recommended distance for a new Reptisun 5.0 lamp set up in this way is 17 - 19 inches. This would give a UVI of only about 0.8 with lamp BJP1.

The second lamp (BZC4) was of the older design, and had a lower output although it has also been in use for considerably longer, making a comparison difficult.

It is tempting to assume that all that is necessary, to enable these lamps to be used safely, is to calculate "safe" minimum basking distances with what seems to be a suitable UV Index, and ensure that these are always used. Although this approach should reduce the irradiation to a level where photo-kerato-conjunctivitis is unlikely, it may not be a long-term solution.

It is important to remember that the UV light which is producing these high UV Indices contains non-solar wavelengths. The effect of non-solar radiation on living cells is different to that of sunlight. Even using the UV Index as a guide for a safe basking distance may not be appropriate with these wavelengths in the beam. This will be considered further in the discussion.

The Effect of Reflectors and Domes

In the vivarium, the shape of the UVB beam from a lamp, and the UVB gradient which it creates, is very important. The beam must be wide enough for the reptile to place most, if not all of its body within the zone of effective UVB radiation at a suitable basking distance.

The shape of the beam from a fluorescent lamp, which is a relatively diffuse light source, is determined by the size and shape of the lamp and whether or not a reflector of any type is fitted.

Aluminium is an extremely good reflector of UV light. Our previous studies have demonstrated that simple strip reflectors fitted behind linear fluorescent tubes can double the effective amount of UVB beneath the lamp. The use of a strip reflector or aluminium lamp housing behind any of these lamps (compact or linear tube) might therefore be problematical.

An even more powerful increase is seen if an aluminium dome reflector is used. We first investigated the effect of placing compact lamps in dome reflectors of different types in 2005. Our first tests showed that the output was more than six times greater beneath a compact lamp, if either a polished aluminium or a brushed aluminium dome reflector was fitted. Obviously, an increase of that magnitude with a lamp that was already emitting very high levels of UVB could be dangerous.

Iso-irradiance charts ("spread charts") for both total UVB output (µW/cm²) and UV Index were therefore constructed according to the method previously devised by FB, for one recently burned-in new-style Reptisun 10.0 lamp (ref. BW2), with no reflector, and also when fitted with four different types of dome reflector. The type of reflector has a profound effect upon the shape of the beam and the intensity of the UVB under the lamp.

With no reflector, the beam is, roughly speaking, an extension of the shape of the lamp into the air around it, forming a flattened sphere with a UVB gradient increasing sharply as the lamp is approached.

Fig. 15(a) and (b): ZooMed Reptisun 10.0 Compact Lamp (no dome)

When the lamp is inside a white painted porcelain dome such as the ZooMed 10.5inch brooder dome, there is little change in the shape of the beam. The white surface does not reflect UVB light. The blocking of the upper parts of the lamp from view actually reduces the overall output slightly, below and to the sides of the lamp. (Fig. 16(b) )

 

Fig. 16(a) and (b): ZooMed Reptisun 10.0 Compact Lamp in ZooMed 10.5in dome with white porcelain interior

With the aluminium domes, the situation is very different. (figs. 3 - 5)

Fig. 17(a) and (b): ZooMed Reptisun 10.0 Compact Lamp in deep 8in. brushed aluminium brooder dome
Fig. 18(a) and (b): ZooMed Reptisun 10.0 Compact Lamp in extra large 10.5in aluminium dome
Fig. 19(a) and (b): ZooMed Reptisun 10.0 Compact Lamp in Pifco 8.5in aluminium dome

Compact fluorescent lamps have a comparatively large surface area. When most of the light from all sides of such a lamp is gathered and reflected downwards, as it is in a dome reflector, the resulting beam may be extremely intense. When the dome is shaped like a parabola, and the aluminium is polished, as with the Pifco 8.5inch dome, the effect is even more pronounced as a directional beam is created beneath the dome. (Fig. 19(b))

Measurement of the UV Index gradient directly beneath the lamp in each situation reveals the increase in intensity produced by the aluminium reflectors. From this data, the percentage increase in UV index produced by each dome may be calculated (Figure 20).

The different shape of each aluminium reflector produces a different effect at any given distance, but 12 inches below the lamp, the parabolic shape of the Pifco dome amplifies the UVI to 762% of the UVI of the bulb with no dome, measured at the same distance below the lamp; the large brushed aluminium dome gives a 627% increase, and the deep aluminium dome increases it by 388%.

These enormous increases are not something that most people would expect. The use of the popular "clamp lamp" type of fixtures, which typically hold the lamp about six inches above the edge of the tank to which they are fixed, rarely raise the lamp more than 8 - 10 inches above an elevated basking spot, such as a log or rock, in the vivarium below.

Any domes with aluminium interiors should only be used with extreme caution, with high-output lamps. The effect varies widely with the exact size and shape of the dome, so it is difficult to offer specific advice about basking distances.

UVC

Recordings with the Solarmeter 8.0 broadband UVC radiometer (240 - 280nm) taken from every lamp in the trial gave the results shown in Figure 21 (below).

Some of these lamps do emit traces of UVC, which correspond to the high-wavelength UVC (from 275nm - 280nm) seen on the spectrograms. However, this UVC output is low and at more than a few inches from the lamp, seems unlikely to be very significant.

Continue to: discussion
Return to: beginning of report