OPEN ACCESS Provision of ultraviolet basking lights to indoor housed tropical birds and their effect on suspected vitamin D3 deficiency

Abstract

Vitamin D deficiency (measured as 25(OH)D3) can occur if birds are fed a vitamin D deficient diet and do not have access to ultraviolet B light (UVB). This can result in eggs with deficient yolks and ultimately to metabolic bone disease (MBD) in chicks. In this study, hypovitaminosis D was suspected in 31 adult birds, from five orders, housed indoors long-term without prior access to UVB light. The study aimed to assess the effect of providing UVB basking lights on vitamin D status and the incidence of MBD in chicks. It also aimed to assess whether the birds would access the UVB provided. Breeding and pathology records were analysed, and birds were blood tested for 25(OH)D3 before, and 12 months after, being provided with access to UVB basking lights. The area of perching with UVB irradiance was filmed before and after the UVB basking lights were switched on. There was a significant increase in 25(OH)D3 after 12 months of UVB provision from a mean of 9.3 nmol/L to 14.2 nmol/L (P=0.001, CI=2.35 to 9.47). Annual incidence of metabolic bone disease in chicks dropped from an average of 14.3% over the three years prior to UVB provision to 3.2% in the two years afterwards. Birds appeared to actively seek the basking spots and significantly increased the proportion of time spent in the area of UVB irradiance (P=0.02). No correlation was found between the total amount, or change, in time spent in the UVB area and the final, or change, in number of individual birds with circulating 25(OH)D3 levels. These results show that indoor-housed birds will bask in UVB light if provided and that this radiation can increase vitamin D levels of the birds, which may prevent MBD in their offspring.

Full text

OPEN ACCESS JZAR Research arcle
Journal of Zoo and Aquarium Research 5(4) 2017 1
OPEN ACCESS
Research arcle
Provision of ultraviolet basking lights to indoor housed tropical birds
and their eect on suspected vitamin D3 deciency
Gabby J. Drake1*, Robyn L. Shea2, Andrea Fidge3, Javier Lopez1 and Robert M. Christley4
1North of England Zoological Society, Caughall road, Upton, Chester, CH2 1LH, U.K.
2Sandwell and West Birmingham Hospital NHS Trust, Vitamin Laboratory, Sandwell General Hospital, Lyndon, West Bromwich, B71 4HJ, U.K.
3San Diego Zoo Global, PO BOX 120551, San Diego, CA 92112-0551, United States
4Instute of Infec on and Global Health, University of Liverpool, Leahurst Campus, CH64 7TE, U.K.
*Correspondence: Dr. Gabby Drake; g.drake@chesterzoo.org
Keywords: avian, MBD, osteodystrophy,
tropical birds, ultraviolet light, UVB,
vitamin D
Arcle history:
Received: 16 Dec 2016
Accepted: 9 Aug 2017
Published online: 31 Oct 2017
Abstract
Vitamin D deciency (measured as 25(OH)D3) can occur if birds are fed a vitamin D decient diet and
do not have access to ultraviolet B light (UVB). This can result in eggs with decient yolks and ulmately
to metabolic bone disease (MBD) in chicks. In this study, hypovitaminosis D was suspected in 31 adult
birds, from ve orders, housed indoors long-term without prior access to UVB light. The study aimed
to assess the eect of providing UVB basking lights on vitamin D status and the incidence of MBD
in chicks. It also aimed to assess whether the birds would access the UVB provided. Breeding and
pathology records were analysed, and birds were blood tested for 25(OH)D3 before, and 12 months
aer, being provided with access to UVB basking lights. The area of perching with UVB irradiance was
lmed before and aer the UVB basking lights were switched on. There was a signicant increase
in 25(OH)D3 aer 12 months of UVB provision from a mean of 9.3 nmol/L to 14.2 nmol/L (P=0.001,
CI=2.35 to 9.47). Annual incidence of metabolic bone disease in chicks dropped from an average of
14.3% over the three years prior to UVB provision to 3.2% in the two years aerwards. Birds appeared
to acvely seek the basking spots and signicantly increased the proporon of me spent in the area of
UVB irradiance (P=0.02). No correlaon was found between the total amount, or change, in me spent
in the UVB area and the nal, or change, in number of individual birds with circulang 25(OH)D3 levels.
These results show that indoor-housed birds will bask in UVB light if provided and that this radiaon
can increase vitamin D levels of the birds, which may prevent MBD in their ospring.
Introducon
Vitamin D is a steroid hormone essenal for calcium and bone
metabolism in most vertebrates, including birds (Bauck, 1995,
Watson, 2014). Adequate vitamin D, calcium and phosphorus
are required for normal growth. Vitamin D deciency can
cause metabolic bone disease (MBD), characterised by low
bone density, bowing of long bones and pathological fractures
(Adkesson and Langan 2007; Cousquer et al. 2007; Stanford
2006; Tangredi and Krook 1999; Van Wyk 1993).
Vitamin D precursors are either obtained through the diet
or endogenously produced in the skin or uropygial gland
secreons through photo-conversion by ultraviolet B (UVB)
light found in sunlight, wavelengths 290–315 nm. The major
circulang vitamin D metabolite is 25-hydroxycholecalciferol
(25(OH)D3) which represents the sum of dietary and sunlight/
UVB sources (Lupu and Robins 2013; Watson 2014). 25(OH)
D3 has the longest half-life of the metabolites, and serum or
plasma levels provide the most stable indicator of vitamin
D status (Bar 2008); thus, it is the most commonly used
metabolite for measuring vitamin D. 25(OH)D3 is converted
to calcitriol (1,25-dihydroxyvitamin D), the acve vitamin D
metabolite, in the kidneys.
Diet is a signicant source of cholecalciferol (vitamin D3)
in carnivorous or piscivorous birds. Ergocalciferol (vitamin
D2), an alternave pre-cursor of calcitriol, found in plant
material, may be ulised by some omnivore or herbivore
species. However, it is unknown which bird species, if any, have
the metabolic pathways required to process this pre-cursor
(Watson 2014). Birds generally have high renal clearance and
a poor metabolism of dietary vitamin D2 so it is unlikely to
constute a substanal source of vitamin D (Stanford 2006;
Bar 2008; Watson 2014). Furthermore, the diet of granivorous
and frugivorous species (Lupu and Robins, 2013) is commonly
decient in vitamin D3. In such cases, therefore, if there is a
prolonged lack of exposure to UVB light, vitamin D deciency
can result. Parental deciency of calcium or vitamin D3
can cause poor quality yolk with lile or no reserves for the
developing chick (Coto et al. 2010a; Mala et al. 2004). UVB
dependent vitamin D deciency and MBD is well documented
in callitrichid monkeys (Thornton 2002), reples (Calvert 2004)
and birds, such as African grey parrots (Psiacus e. erithacus)

Journal of Zoo and Aquarium Research 5(4) 20172
Drake et al.
(Stanford 2005). No published data could be found relang to the
species invesgated in this study.
Producon of 25(OH)D3 via the UVB pathway is self-regulang
due to build-up of inert metabolites in the cus (Lupu and Robins
2013). However, oral supplementaon of 25(OH)D3 is poorly
regulated via negave feedback (Bar 2008), and can result in an
excess, which could be harmful to the bird (Watson 2014; Sco
Echols 2006). For this reason, exposure to UVB light is a safer
alternave to dietary supplementaon (Stanford 2006; Lupu and
Robins 2013), especially if species requirements are unknown.
Domesc poultry given UVB via natural sunlight signicantly
increased the 25(OH)D3 content of their egg yolks (Kuhn 2014),
suggesng that the provision of UVB light in sucient quanes
to the hens in this study should be a way to increase yolk vitamin
D3 to provide for the chick in the rst three weeks of life.
At Chester Zoo, osteodystrophy and pathological fractures
consistent with metabolic bone disease were idened in the
chicks of mulple tropical bird species breeding in an indoor aviary.
The aected chicks and their parents had no access to either
sunlight or UVB radiaon in the wavelengths 290–315 nm. Since
dietary calcium and calcium to phosphorus raos were known to
be adequate, vitamin D deciency in parents was hypothesised
to be the cause of the MBD in chicks. This study was designed
in response to these clinical issues and addressed the following
objecves: to invesgate whether this populaon of birds would
ulise arcial UVB basking spots; and to assess the eect of this
method of UVB light supplementaon on the circulang vitamin
D levels (measured as 25(OH)D3) of the adult birds and on the
incidence of MBD in chicks.
Materials and methods
This study was opportunisc as UVB lighng was being installed
in the study aviaries as part of ongoing husbandry improvements.
Study design was longitudinal, each bird was its own control, thus
ensuring no bird was denied UVB light. The study was approved
by the Ethical Commiee of both Chester Zoo and the University
of Liverpool. Samples were taken during clinical invesgaon and
followed up by a qualied veterinary surgeon.
Animals and husbandry
Birds were housed in ve large aviaries (minimum 15 m2 to
maximum 25 m2) within a tropical house. Thirty-one adult birds
of ve orders (Table 1) had been living in these aviaries for more
than one year at the start of the study in January 2012, and were
thus eligible for inclusion. A calculaon using Altman’s nomogram
revealed a sample size of 25 birds should be sucient to idenfy
biologically signicant changes in vitamin D (Sabin and Petrie
2011). Aviary groups remained stable, with the excepon of
birds added through breeding (n=25), or removed by death (n=3
adults) or removed as young stock (fully grown) once independent
from their parents (n=16). A singe male was briey moved to
the neighbouring study aviary due to aggression. Pathology
and breeding records for all birds housed in these aviaries were
reviewed for breeding acvity and cases of metabolic bone
disease between 2009 and 2014 inclusive.
Birds were fed species-appropriate diets formulated by a
nutrionist, comprising a combinaon of commercial pellets, fresh
chopped fruit and live invertebrates, and blends made with pellets,
crushed egg shell, vegetables and cooked egg as appropriate. Diets
had a mean calcium content of 1.36% dry maer. Diets contained
suitable calcium to phosphorus raos (mean 5.7:1) and remained
unchanged throughout the study period.
Husbandry alteraons
Two UVB spot-lights (Arcadia 160W Basking lamps, Arcadia
Products plc, Redhill, U.K.), producing light in the wave lengths
290–315 nm, were installed in each aviary. In each aviary, one light
was posioned high, above the branches used for perching, to
give tree dwelling birds UVB access—subsequently referred to as
the high resource area; and the second was located at a height to
provide ground dwelling birds with an area of UVB, subsequently
referred to as the low resource area. Lights were situated a mean
distance of 66.3 cm (range 60–90 cm) from the perch or ground.
Lights and perches were ed for at least three weeks prior to light
illuminaon, to allow the birds to acclimase to the new aviary
furniture. Basking lights were then switched on (Day 0 for study
mings) from 10:00 to 14:00 every day from Day 0 to Month 4,
then from 06:00 to 14:00 for the rest of the study (Months 5–12).
The output of the lights was monitored throughout the
Order Species (Common name) Scienc name Number in study MBD seen
Galliformes Congo peafowl Afropavo congensis 1Yes
Cuculiformes White crested turaco Tauraco leucolophus 3Yes
Passeriformes Snowy headed robin chat Cossypha niveicapilla 2Ye s
White-rumped shama Copsychus malabaricus 2Yes
Chestnut-backed thrush Zoothera dohertyi 1
Montserrat oriole Icterus oberi 2Yes
Fairy bluebird Irena puella 2
Brazilian tanager Ramphocelus bresilius 2
Columbiformes Mindanao bleeding heart dove Gallicolumba criniger 4Yes
Luzon bleeding heart dove Gallicolumba luzonica 1Yes
Socorro dove Zenaida macroura graysoni 4
Green naped pheasant pigeon Odiphaps nobilis nobilis 2Yes
White naped pheasant pigeon Odiphaps nobilis aruensis 3Ye s
Piciformes Brown-breasted barbet Lybius melanopterus 2
Total 31
Table 1. Species, number and history of metabolic bone disease (MBD) in birds

Journal of Zoo and Aquarium Research 5(4) 2017 3
Vitamin D and ultraviolet light in indoor birds
study using a UVB meter (Solarmeter® Model 6.2, SOLAR LIGHT
COMPANY INC. Glenside, PA, USA, measuring µw/cm2 UVB light).
Bulbs were tested at 30 cm and gave readings between 90 and 134
µw/cm2 when new. They were replaced if their output dropped
by more than 50% from the inial reading. There was no other
source of UVB in the study aviaries.
Health assessment and vitamin D3 sampling
All birds were examined prior to the main breeding season
in 2012, which was prior to Day 0. The birds were weighed,
clinically examined, radiographed and blood sampled. Where
sucient sample was obtained, blood biochemistry was checked
as well as vitamin D (measured as 25(OH)D3). The birds were re-
examined and blood sampled again 12 months aer illuminaon
of the basking lights. At the inial examinaon, radiographic bone
density and conformaon was considered normal for all birds
and plasma ionised calcium, where measured, was within normal
avian limits (>1.00 mmol/L) (data not shown), so these tests were
not repeated.
Blood samples for 25(OH)D3 analysis were taken from the
jugular, ulnar or metatarsal vein (as appropriate) and placed in
heparinised blood tubes (Microvee 500 lithium heparin, Sarstedt,
Numbrecht, Germany). Sample volume never exceeded 1% of body
weight as per avian guidelines (Best 2008). Heparinised samples
were centrifuged at 11,800 rpm for three minutes and separated
plasma samples were placed in Eppendorf tubes and either sent
directly to the laboratory or frozen at –20°C unl batches could
be sent for analysis. The drops remaining in the syringe hub were
placed on lter paper (Whatman 903 lter paper, GE Healthcare,
Whatman Plc, Maidstone, UK) and air-dried for blood spot analysis.
In some birds, small sample volume allowed only one technique to
be used (blood spot, n=4; plasma, n=5).
At Sandwell & West Birmingham Hospitals NHS Trust Vitamin
Laboratory, UK, blood spots were analysed for 25(OH)D3 as three
millimetre punches from the lter paper on a Waters TQS Tandem
Mass spectrometer using an electro-spray ionisaon interface
with a Water’s i-Class UPLC, following liquid/liquid extracon with
a derivasaon step enhancing small volume analysis.
Plasma 25(OH)D3 was assayed using a Waters ACQUITY
UPLC and Quaro Premier XE MS/MS mass spectrometer
(Herordshire, England) with an electro-spray ionisaon interface
following liquid/liquid extracon of the sample. Both methods are
linear to 1,100 nmol/L with a limit of quantaon of 7.5 nmol/L
for 25(OH)D3. The plasma assay is accredited by the Vitamin D
External Quality Assessment Scheme (DEQAS) and the laboratory
is accredited by the Clinical Pathology Accreditaon (CPA) UK Ltd.
Using human samples, the blood spot assay had been aligned
with the plasma assay so that blood spot and plasma results were
comparable.
UVB resource use
UVB meter readings were used to map out the area of UVB
radiaon produced by each bulb (high and low) and the intensity
of radiaon at bird level. In each aviary, a “UVB resource area” was
dened as the basking areas at bird level where >10 µw/cm2 UVB
irradiance was detected. This constuted a radius of between 60
and 80 cm in each case. The peak UVB reading at bird level was
also recorded, with an average of 31 µW/cm2 (24–65 µW/cm2).
A webcam (Logitech Webcam HD C270, Logitech Europe S.A.,
Lausanne, CH) was used to lm the UVB resource areas in each
aviary, as the presence of an observer could alter bird behaviour.
Footage was recorded on a laptop outside the aviary using
movement recognion soware (iSpy freeware, hp://www.
iSpyconnect.com). Each resource area was lmed in each aviary
(n=6) for ve to seven consecuve days between 06:00 and
14:00 during the six-week period prior to UVB light illuminaon
(baseline data). Filming was repeated at least four weeks aer
light illuminaon (UVB data).
Mpeg video clips were reviewed by a single observer using
QuickTime Player (QuickTime 7.0, Apple Inc., Cuperno, CA) and
the me stamp was used to record the me and duraon of a visit
by any bird accessing the UVB resource during the observaons.
Birds were idened to species level in all cases and individually
where possible. Where two indisnguishable birds were present
in an aviary their UVB resource use was recorded as the average of
the total me one or both birds were visible in the UVB resource.
Aviaries were not lmed if birds were breeding, as brooding could
restrict access to the UVB resource.
Stascs
All data analyses were performed using Microso Excel (2010,
Microso Corporaon, Redmond, WA) and R (R core team, 2013).
Annual incidence of MBD in chicks before and aer provision
of UVB supplementaon was compared using Fisher’s exact
test due to small sample size in the post treatment group. This
test was performed excluding the rst treatment year (2012)
as iniaon of supplementaon coincided with the start of the
breeding season, thus there would have been insucient me for
circulang 25(OH)D3 levels in the hen birds to change.
25(OH)D3 levels obtained by each method (blood spot and
plasma) were averaged between the methods for each sampling
point to minimise missing data, as some samples only provided
enough volume for one method. The lower limit of the assay was
7.5 nmol/L; thus, values reported as <7.5 nmol/L were given a
nominal value of 7.5 nmol/L for stascal analysis. A Wilcoxon-
signed ranks test for non-parametric paired data was used to
compare values from pre-supplementaon and 12 months aer
UVB supplementaon.
Behavioural data were only available from three of the six
aviaries holding a total of 15 birds due to issues with the video
footage. Data were analysed for me spent in the resource area
for baseline and UVB access. For each bird, the total me, in
seconds, spent in the resource area was calculated. Total me
observed was calculated. As not all observaon periods were the
same length, the proporon of me spent in the resource area
during observaon was used for comparison.
Year Number chicks
bred
Number chicks
with MBD
Annual incidence of MBD
in chicks at risk (%)
No UVB provision
2009 23 4 17.4
2010 27 3 11.1
2011 41 614.6
Total 91 13 14.3
Paral UVB provision
2012 25 4 16.0
UVB provision
2013 13 0 0.0
2014 18 15.6
Total 31 1 3.2
Table 2. Number of birds bred and cases of metabolic bone disease (MBD)
idened per year.
UVB, ultraviolet B light. Paral UVB provision refers to supplementaon
being for insucient duraon prior to the breeding season for data to be
interpreted.

Journal of Zoo and Aquarium Research 5(4) 20174
Drake et al.
UVB radiaon doses (mJ/cm2) were calculated from the
irradiance received by each bird (mW/cm2) mulplied by the
exposure me (seconds) (Lupu and Robins 2013).
It was noted during video analysis that some birds roosted
overnight on the UVB resource perches. As light intensity is the
trigger for birds to rise in the mornings, birds were frequently
sll roosng during the rst hour of data collecon during the
baseline data collecon period, but roused immediately once
the UVB basking lights came on during the UVB data collecon
period. In order to allow a comparison of baseline and UVB data,
only acvity between 10:00 and 14:00 was compared to avoid
this “roosng eect”. A Wilcoxon-signed ranks test was used to
compare the proporon of me spent in the UVB resource pre-
and post-UVB supplementaon.
Spearman’s rank correlaon tests were used for all correlaon
assessments. The full data set from 06:00 to 14:00 was used to
compare me spent accessing the UVB basking light with nal
25(OH)D3 concentraon and changes in 25(OH)D3, as baseline
observaon data were not required. The data from 10:00 to
14:00 were used to assess for correlaon between change in UVB
resource use and the nal 25(OH)D3 and changes in 25(OH)D3
concentraons.
Results
MBD cases
Annual incidence of MBD appeared to decrease from a range of
11.1–17.4% before UVB supplementaon to a range of 0–5.6%
in the years aer UVB supplementaon was iniated (Table 2).
However, this dierence was not stascally signicant (P=0.19).
Vitamin D3
There was a signicant increase in circulang 25(OH)D3 levels aer
12 months of UVB supplementaon (P=0.001, CI=2.35 to 9.47;
Figure 1, full data set Appendix 1). However, there was individual
variaon with some birds showing a marked increase, while lile
change was observed in others. The mean pre-supplementaon
circulang 25(OH)D3 was 9.3 nmol/L and the mean aer 12
months was 14.2 nmol/L, thus exhibing a mean increase of 4.9
nmol/L across all birds. The pre-supplementaon range was 7.5 to
27.7 nmol/L, interquarle range (IQR) 7.5 to 8.25 nmol/L. Twelve
months aer UVB supplementaon the range was 7.5 to 48.1
nmol/L, IQR 7.5 to 15.2 nmol/L.
25(OH)D3 levels were reported as <7.5 nmol/L in a number of
birds since this was the lower limit of quantaon of the assay.
As a result, actual levels in these birds may have been anywhere
between 0 and 7.4 nmol/L. Therefore, actual increases could be
larger than idened for 12 of the birds whose baseline values
were at 7.5 nmol/L and increased above this threshold aer 12
months, which could mean a biologically signicant increase for
the individual. Six of the birds remained below 7.5 nmol/L, so it is
unknown whether their levels increased or decreased between 0
and 7.4 nmol/L. The measures of three out of four birds with levels
that dropped, went below 7.5 nmol/L, so there may actually have
been larger decreases than could be detected by the assay.
Seven hens bred in 2012, and three of these were birds with
circulang 25(OH)D3 levels which dropped aer supplementaon.
Table 3 shows the mean dierence between blood spot and
plasma methods, range and standard deviaons for these results
by species.
Use of UVB resource area
Baseline observaons compared to UVB provision
The proporon of me spent in the UVB resource area increased
signicantly aer the basking lights were illuminated (P=0.02;
point esmate of mean 0.0272 ± 0.0204; Figure 2, full data set
Appendix 1). This corresponds to an average increase of 6 min 32
sec per bird, per four hour observed window per day (CI 1 min 39
sec to 11 min 25 sec) in the UVB resource area.
The pre-supplementaon minimum proporon was 0.000360,
corresponding to 5 seconds, median 0.005008 (1 min 12 sec),
IQR 0.002644 to 0.017090 (38 sec to 4 min 6 sec) and maximum
Species (common name) Average dierence between blood
spot and plasma D3 (nmol/L)
Range (nmol/L) Standard deviaon (nmol/L)
Congo peafowl -2.4 -2.7–2.0 0.5
White crested turaco -0.1 -2.1–1.8 1.4
Snowy headed robin chat 1.6 0–5.5 2.6
White-rumped shama -5.7 -13.8–0.1 7.3
Chestnut-backed thrush n/a n/a n/a
Montserrat oriole 0.2 0–0.3 0.2
Fairy bluebird -0.4 -2.8–1.1 1.7
Brazilian tanager 0 0–0 0
Mindanao bleeding heart dove 0.4 -6.6–4.9 3.5
Luzon bleeding heart dove 0.2 -0.2–0.6 0.6
Socorro dove -1.4 -7.2–0 3.2
Green naped pheasant pigeon 0.3 -1.2–2.6 1.6
White naped pheasant pigeon 0 0–0 0
Brown-breasted barbet n/a n/a n/a
Table 3. Comparison of blood spot and plasma 25(OH)D3 results grouped by species.

Journal of Zoo and Aquarium Research 5(4) 2017 5
Vitamin D and ultraviolet light in indoor birds
light to adults and increased 25(OH)D3 in adults to be consistent
with other more-studied avian species. Sub-clinical vitamin D
deciency in chickens can result in deciency in their egg yolks and
in the developing chicks, leading to MBD, which can be resolved by
addressing the vitamin D3 levels of the adults (Coto et al. 2010b).
No clinical issues were found in the adult birds in this study,
despite hypothesised vitamin D deciency. This is consistent with
work on domesc poultry where hens with vitamin D deciency,
but ample calcium, show no clinical signs of hypovitaminosis D
(Nascimento et al. 2014). The diet provided was adequate for
calcium. Although oered zoo diets are formulated complete, it
is well documented from intake studies there may be dierences
in nutrional balance compared to the actual diet eaten (Fidge
and Robert 1993). Therefore, we cannot be certain of nutrient
bioavailability from the data available and cannot rule out a
dietary component of the MBD seen in chicks. However, the diets
oered did not change during the study and we have no reason to
suspect the birds altered their intake during this me, therefore
diet is unlikely to be the cause of the reducon in MBD.
Arcial provision of UVB basking lights had a posive eect
on vitamin D3 levels in the birds in this study, with a stascally
signicant increase in circulang 25(OH)D3 following 12 months
of eight hours per day ad libitum UVB basking light access. As
there is believed to be eecve feedback to prevent excess levels
of vitamin D3 accumulang from UVB exposure (Lupu and Robins
2013), this suggests that some of the birds could have been sub-
clinically decient at the start of the study, as their levels were
physiologically able to rise.
Average pre-supplementaon circulang 25(OH)D3
concentraon was 9.3 nmol/L in this study, which appears low
when compared to some other avian taxa. In experimental work
with African grey parrots (Psiacus e. erithacus), Stanford (2005)
reported a mean pre-UVB plasma 25(OH)D3 concentraon of
23.1 nmol/L in capve birds. In budgerigars (Melopsiacus
undulatus) given no UVB access, 25(OH)D3 levels of 7.93–16.03
nmol/L have been reported (Lupu and Robins 2013). In other avian
orders, cases of metabolic bone disease have been reported in
growing wild birds with vitamin D3 levels this low. In American
crows (Corvus b. brachyrhynchos), plasma 25(OH)D3 values of
8.0 (± 2.65) nmol/L were found in young wild birds with MBD
compared to 20.0 (± 6.04) nmol/L in unaected birds (Tangredi
and Krook 1999). In Columbiformes, values of 3.90 nmol/L were
reported for three clinical MBD cases in young wild collared doves
0.043700 (10 min 29 sec) per bird per four hour observed window
per day. Aer 12 months of UVB supplementaon this had
increased to a minimum proporon of 0.01010 (2 min 25 sec),
median 0.03360 (8 min 4 sec), IQR 0.01360 to 0.05805 (3 min 15
sec to 13 min 56 sec) and maximum 0.11950 (28 min 41 sec) per
bird per four hour observed window per day. However, within
each of the aviaries, some birds dramacally increased their usage
of the UVB resource area while others reduced their usage.
Esmated UVB doses
When the lamp was in use, the daily average me spent in the
UVB resource area (>10 uW/cm2) was 18 min 35 sec (range 2 min
46 sec to 40 min 17 sec). Minimum average daily UVB doses were
therefore 11.15 mJ/cm2 (range 1.66 to 24.17 mJ/cm2). Maximum
average calculated daily doses based on the average maximum
irradiance of 31 µW/cm2 were 34.57 mJ/cm2 (5.15 to 74.93 mJ/
cm2) if the birds remained in the peak area of UVB emied by the
basking light.
Correlaons between UVB resource use and average 25(OH)D3
Final, and change in, circulang 25(OH)D3 concentraon aer
12 months was not associated with the proporon of me spent
in the UVB resource (rho=0.08, P=0.8 and rho=0.17, P=0.59,
respecvely). The nal, and change in, circulang 25(OH)D3
concentraon was not associated with the change in proporon
of me spent in the UVB resource before and aer the basking
lights were illuminated (rho=0.12, P=0.7 and rho=0.23, P=0.47,
respecvely).
Discussion
The incidence of MBD in chicks appeared to reduce following
UVB basking light provision to the adults from 14.3 % to 3.2 %
subsequently. However, the small numbers involved, combined
with the limited number of years of follow up, may have contributed
to the lack of stascal signicance. Fewer chicks were bred in
2013 and 2014, compared to previous years. However, three of
four birds which bred in 2013, and four of ve birds which bred
in 2014, had previously produced chicks with MBD, supporng
the possibility of a true reducon following UVB provision. The
annual incidence of 5.6% in 2014 represents a single case of MBD
from the snowy headed robin chats. Physiologically, we nd the
reducon in MBD in ospring following provision of UVB basking
Figure 1. Box and whisker plot of vitamin D (25(OH)D3) concentraons
before and aer 12 months of UVB basking light provision.
Figure 2. Box and whisker plot of the proporon of me spent in the UVB
basking spot between 10:00 and 14:00 before and aer UVB basking lights
were illuminated.

Journal of Zoo and Aquarium Research 5(4) 20176
Drake et al.
(Streptopelia decaocto) and a range of 10.0–25.0 nmol/L is quoted
as normal, although this is unreferenced (Cousquer et al. 2007).
If this ‘normal’ range is accurate, then 83% (25/30) of the birds in
this study were decient in vitamin D3 at the start, falling to 44%
(12/27) by the end of the study.
Average 25(OH)D3 concentraon aer 12 months of UVB
supplementaon in this study was 14.2 nmol/L; an average
increase of 4.9 nmol/L. These are considerably lower levels and
smaller changes than those idened in African grey parrots by
Stanford (2005) who reported a mean increase in plasma 25(OH)
D3 concentraon of 77.9 nmol/L aer one year of 12 hours per
day of direct UVB light. However, the same study also found
mean concentraons of 33.68 nmol/L (95% CI 8.00–59.63) in
19 wild African grey parrots, suggesng such high 25(OH)D3
concentraons, although not detrimental to the birds, may not
be physiologically required. In budgerigars, increases in 25(OH)
D3 of 1.61–11.31 nmol/L, to 9.54–27.34 nmol/L, occurred aer
short duraon UVB exposure experiments (Lupu and Robins
2013); changes which are comparable with those observed in
the present study. Despite there being considerable variability in
response to UVB provision in this study, the overall tendency was
for circulang 25(OH)D3 levels to increase with UVB basking light
provision. This is especially signicant as birds were not forced to
remain under the UVB basking lights, unlike in the experimental
work discussed, but were free to access the resource, suggesng
provision of UVB basking lights for indoor housed birds does not
have to be ubiquitous to have an eect. In addion, the spectrum
of the lights used in this study is more akin to natural sunlight
than those used by Lupu and Robins (2013), which contained
proporonally higher levels of very short wavelength UVB light
which can cause damage to the skin and eyes of birds.
Blood spot and plasma results were averaged to minimise
missing data where only one method was possible in this study.
The blood spot assay was aligned with the plasma assay using
human samples, not avian samples. There is evidence that
these two methods may not be comparable in reples, likely
due to issues with lower haematocrits signicantly altering the
spreading characteriscs of replian blood compared to mammals
(Whitehead and Fogge 2016); however, bird haematocrits are
more similar to those of mammals and our data suggest that the
two methods are acceptably similar. This is concurred by Michaels
(2015) who states that results for these two methods should be
comparable within the same study, as applies here.
Mulple species were used in this study, and despite the
mechanism for producon of vitamin D3 appearing to be well
conserved throughout the taxonomic groups (Dacke 2000; Bar
2008), species variaon, both physiologically and in feeding and
basking behaviours, may have aected conversion eciency and
therefore inuenced these results. In a zoological collecon, it
is rare to have large numbers of the same species to study, and
therefore when an opportunity arises to perform a study such as
this, even with mulple species, the generated data are valuable.
Some of the birds’ circulang 25(OH)D3 levels did not change
detectably post-supplementaon (n=6) and some actually
reduced (n=4). Some birds may not have received adequate
doses because in a mixed aviary, there may be compeon for
the UVB resource. In addion, birds were given the freedom to
voluntarily access the resource, and the UVB provided by a bulb
is weak compared to sunlight. Doses of 180 mJ/cm2 (0.012 mW/
cm2 over 6 hr for 5 d) have been reported to produce a signicant
increase in plasma 25(OH)D3 in budgerigars, while doses of 65 mJ/
cm2 (0.012 mW/cm2 for 2 hr) resulted in no signicant increase
(Lupu and Robins 2013). These are considerably larger doses and
of shorter wavelengths—due to dierences in the lamps used—
than the birds in this study received, but this perhaps suggests
that the UVB dose could have been inadequate for some study
birds. However, the experiments of Lupu and Robins (2013)
lasted only ve days, unlike the present study in which long-term
exposure was provided. This may be why, in the majority of birds,
circulang 25(OH)D3 increased through bioaccumulaon, despite
the low doses. In chickens, circulang 25(OH)D3 has a half-life
of approximately 21 days (Coto et al. 2010a) and changes in
circulang 25(OH)D3 may occur more slowly than accounted for
by Lupu and Robins (2013).
Three of the four birds whose circulang 25(OH)D3 levels
decreased aer 12 months of UVB provision were breeding
hens, raising the possibility that their starng levels were either
arcially elevated as serum stores transferred to yolk at the
me of rst sampling, or were depleted by egg laying by the
me of the second sample. One of these hens was the snowy
headed robin chat hen, which produced a MBD-aected chick
following provision of UVB light. In domesc poultry, increases
and decreases in yolk vitamin D have been shown to occur in line
with both circulang vitamin D levels and UVB access (Narbaitz et
al. 1987; Kuhn 2014). Values from wild birds would also aid in the
understanding of normal ranges for each species, to truly dene
deciency in these hens.
The dierences in eect may also be due to interspecic
variaon in the eciency of conversion of 7-dehydrocholesterol
to pre-vitamin D3 in the skin. For example, canopy-dwelling
species, where UVB exposure should be high in the wild, might
have received insucient doses compared to naturally-reclusive
species which may be beer adapted to photoconvert at lower
light levels. This hypothesis is consistent with the poor responses
seen in this study in the Brazilian tanagers, fairy bluebirds and
white crested turacos, all of which are canopy species; this eect
warrants further invesgaon.
The indoor housed tropical birds in this study chose to access
UVB basking lights when given the opportunity, with a stascally
signicant increase in the proporon, and amount, of me spent in
the UVB resource area observed aer the lights were illuminated.
This nding is consistent with another study in which various birds
showed a preference for an area of ultraviolet spectrum lighng
(Ross 2013). Not all birds showed increased usage in each aviary.
The variaon in use could represent compeon from dominant
species prevenng access by the shyer ones. For instance, the
snowy headed robin chats were observed compeng with the
white crested turacos (which are much larger and more aggressive
birds) for access, and the proporon of me spent in the resource
by the robin chats decreased as that of the turacos increased. This
compeon eect would be worth invesgang further, as it could
have a signicant impact on the eecveness of UVB basking light
provision in mixed species exhibits.
It is unclear why the birds chose to spend more me under
the lights. They may be insncvely basking in the “sunlight”;
indeed, many birds were observed preening while sing under
the lights. Equally, the birds may be seeking the heat produced
by the lights, or the UVA light emied by the bulbs, as birds can
see into this spectrum (Rajchard 2009). Addional experiments
substung the UVB bulbs with basking lamps which produce
dierent combinaons of heat, UVA and UVB would be required
to disentabgle these possible inuences and determine the best
condions to aract birds to a therapeuc source of UVB.
There was no signicant correlaon between the nal, or
change in, circulang 25(OH)D3 levels in the birds and the total
or change in proporon of me spent in the UVB resource
area. However, data for both circulang 25(OH)D3 levels and
behavioural observaons were only available from 13 birds for this
comparison. Hence, the sample size is small in these analyses, and
a weak correlaon may have been missed. The results could also
have been inuenced by interspecies variaon, the species mixes
and the necessity to average data between birds which could

Journal of Zoo and Aquarium Research 5(4) 2017 7
Vitamin D and ultraviolet light in indoor birds
not be individually idened. Some females in this comparison
bred in the 2012 season and depleon to yolk may have aected
circulang 25(OH)D3 levels. Finally, the behavioural data collecon
period constuted one week, which equates to only a fracon of
the study me, and so may not truly represent the birds’ acvity.
In conclusion, provision of UVB basking light in the wavelengths
290–315 nm to indoor housed tropical birds on an ad libitum basis
can result in increases in circulang vitamin D3, levels which may
be biologically signicant. Therefore, UVB basking light access
could be used to reduce vitamin D3 deciency in a variety of
species without restricng aviary sizes. Importantly, the incidence
of metabolic bone disease in chicks appeared to be reduced
following UVB provision for the adult birds. Birds appeared to
acvely choose to access the ultraviolet basking spots and showed
a signicant increase in the proporon of me spent in an area
where this resource was provided. Whether this preference was
due to the UVB, UVA or increased heat from the lamps could not
be determined from the data available. Parcular care should be
taken when providing UV to mixed species aviaries, as interspecies
compeon may prevent some species from accessing the
resource.
Acknowledgements
The authors would like to thank the Bird Team of Chester Zoo, with
special acknowledgment to Andrew Owen, Wayne Mcleod and
Clare Wylie for their support of this work and the veterinary team,
with special thanks to Alison Kelsall and Tanya Grubb for help with
the diagnosc sampling.
References
Adkesson M.J., Langan J.N. (2007) Metabolic bone disease in juvenile
Humboldt penguins (Spheniscus humbold): Invesgaon of
ionized calcium, parathyroid hormone, and vitamin D3 as diagnosc
parameters. Journal of Zoo and Wildlife Medicine 38(1): 85–92.
Bar A. (2008) Review: ‘Calcium homeostasis and vitamin D metabolism
and expression in strongly calcifying laying birds. Comparave
Biochemistry and Physiology Part A 151: 477–490.
Bauck L. (1995) Nutrional Problems in Pet Birds. Seminars in Avian and
Exoc Pet Medicine 4(1): 3–8.
Best J.R. (2008) Passerine birds: approach to the sick individual. In: Chiy, J.
and Liertz, M. (eds). BSAVA Manual of Raptors, Pigeons and Passerine
birds, Gloucester, UK: Brish Small Animal Veterinary Associaon,
356–364.
Calvert I. (2004) Nutrional problems. In: Girling, S.J. and Rai, P. (eds).
BSAVA Manual of Reples (2nd Edion), Gloucester, UK: Brish Small
Animal Veterinary Associaon, 289–308.
Coto C., Cerate S., Wang Z., Yan F., Min Y., Costa F.P., Waldroup P.W. (2010a)
Eect of source and level of vitamin D on the performance of breeder
hens and the carryover to the progeny. Internaonal Journal of Poultry
Science 9: 623–633.
Coto C., Cerate S., Wang Z., Yan F., Waldroup P.W. (2010b) Eect of source
and level of maternal vitamin D on carryover to newly hatched chicks.
Internaonal Journal of Poultry Science 9: 613–622.
Cousquer G.O., Dankoski E.J., Paerson-Kane J.C. (2007) Metabolic bone
disease in wild collared doves (Streptopelia decaocto). Veterinary
Record 160: 78–84.
Dacke C.G. (2000) Chapter 18 – The Parathyroids, Calcitonin, and Vitamin
D In: Causey Whiow G. (ed). Sturkie’s Avian Physiology (5th Edion),
Cambridge, Massachuses: Academic Press, 473–488.
Fidge, A.L. and Robert, J.N. (1993) An invesgaon into nutrion and
mortality in capve St. Lucia parrots (Amazona versicolor). Dodo,
Journal of the Wildlife Preservaon Trusts 29: 103–125.
Lupu C. and Robins S. (2013) Determinaon of a safe and eecve
ultraviolet B radiant dose in budgerigars (Melopsiacus undulatus):
A pilot study. Journal of Avian Medicine and Surgery 27(4): 269–279.
Mala P., Valaja J., Rossow L., Venalainen E., Tupasela T. (2004) Eect
of vitamin D2- and D3-enriched diets on egg vitamin D content,
producon, and bird condion during an enre producon period.
Poultry Science 83(3): 433–440.
Michaels C.J., Antwis R.E., Preziosi R.F. (2015) Impacts of UVB provision
and dietary calcium content on serum vitamin D3, growth rates,
skeletal structure and coloraon in capve oriental re-bellied
toads (Bombina orientalis). Journal of Animal Physiology and Animal
Nutrion 99: 391–403.
Narbaitz R., Tsang C.P.W., Grunder A.A. (1987) Eects of vitamin D
deciency in the chicken embryo. Calcied Tissue Internaonal 40:
109–113.
Nascimento G.R., Murakami A.E.I., Guerra A.F.Q.M.I., Ospinas-Rojas I.C.I,
Ferreira M.F.Z.I., Fanhani J.C.I. (2014) Eect of dierent vitamin D
sources and calcium levels in the diet of layers in the second laying
cycle. Brazilian Journal of Poultry Science 16(2): 37–42.
R Core Team (2013). R: A language and environment for stascal
compung. R Foundaon for Stascal Compung, Vienna, Austria.
[online] ISBN 3-900051-07-0. Available from hp://www.R-project.
org/.
Ross M.R., Gillespie K.L., Hopper L.M., Bloomsmith M.A., Maple T.L. (2013)
Dierenal preference for ultraviolet light among capve birds from
three ecological habitats. Applied Animal Behaviour Science 147:
278–285.
Sabin A., Petrie C. (eds) (2011) Medical Stascs at a Glance (3rd Edion),
Hoboken, NJ: Wiley Blackwell.
Sco Echols, M. (2006) Evaluang and treang the kidneys. In: Harrison
G.J. and Lighoot T.L. (eds). Clinical Avian Medicine: Volume 2. Palm
Beach, Florida: Spix Publishing Inc., 451–492.
Stanford M.D. (2005) Calcium metabolism in grey parrots: the eects of
husbandry. Fellowship Thesis London, UK: Royal College of Veterinary
Surgeons.
Stanford M.D. (2006) Calcium Metabolism. In: Harrison G.J. and Lighoot
T.L. (eds). Clinical Avian Medicine: Volume 1. Palm Beach, Florida: Spix
Publishing Inc., 141–151.
Tangredi B.P and Krook L.P. (1999) Nutrional secondary
hyperparathyroidism in free-living edgling American crows (Corvus
brachyrhynchos brachyrhynchos) Journal of Zoo and Wildlife Medicine
30(1): 94–99.
Thornton S.M. (2002) Primates. In: Meredith A. and Johnson Delaney C.
BSAVA Manual of Exoc Pets (4th Edion), Gloucester, UK: Brish
Small Animal Veterinary Associaon, 127–137.
Van Wyk E., Van der Bank F.H., Verdoorn G.H. (1993) Blood plasma calcium
concentraons in capve and wild individuals of the cape grion
vulture (Gyps coprotheres). Comparave Biochemistry and Physiology
Part A 104(3): 555–559.
Whitehead M.L. and Fogge S. (2016) UV lighng, dietary vitamin D3, and
serum 25(OH)D in crocodilians and iguanas. Proceedings of the Brish
Veterinary Zoological Society, September 2016, Oxford: 20.

Journal of Zoo and Aquarium Research 5(4) 20178
Drake et al.

Journal of Zoo and Aquarium Research 5(4) 2017 9
Vitamin D and ultraviolet light in indoor birds

Journal of Zoo and Aquarium Research 5(4) 201710
Drake et al.

Journal of Zoo and Aquarium Research 5(4) 2017 11
Vitamin D and ultraviolet light in indoor birds

Journal of Zoo and Aquarium Research 5(4) 201712
Drake et al.