J. Avian Biol. 39: 539!545, 2008 doi: 10.1111/j.2008.0908-8857.04413.x

J. Avian Biol. 39: 539!545, 2008 doi: 10.1111/j.2008.0908-8857.04413.x
J. Avian Biol. 39: 539!545, 2008
doi: 10.1111/j.2008.0908-8857.04413.x
# 2008 The Authors. J. Compilation # 2008 J. Avian Biol.
Received 25 October 2007, accepted 21 January 2008
Resistance of melanized feathers to bacterial degradation: is it really
so black and white?
Alex R. Gunderson, Alicia M. Frame, John P. Swaddle and Mark H. Forsyth
A.R. Gunderson (correspondence), A.M. Frame, J.P. Swaddle, and Mark H. Forsyth, Institute for Integrative Bird Behavior Studies,
Department of Biology, The College of William and Mary, Virginia, USA. ! Present address of ARG: Department of Biology, Duke University,
Durham, NC, USA. E-mail: [email protected] ! Present address of AMF: Department of Biology, University of North Carolina,
Chapel Hill, NC, USA.
Melanins are common feather pigments that contribute to signaling and crypsis. Melanins may also help feathers resist
feather-degrading bacteria (FDB). Two recent studies (Goldstein et al. 2004, Grande et al. 2004) tested the resistance of
melanized versus unmelanized feathers to FDB using in vitro experiments, but draw opposite conclusions. Goldstein et al.
(2004) concluded that melanized feathers resist FDB more than unmelanized feathers, while Grande et al. (2004)
concluded that unmelanized feathers resist FDB more than melanized feathers. To resolve this conflict in the literature,
we replicated previous studies but included additional tests not previously used. We inoculated melanized and
unmelanized feathers of domestic geese Anser anser domesticus, with the FDB Bacillus licheniformis and measured bacterial
activity every two days over two weeks. Three metrics of bacterial activity on feathers were measured: soluble protein
content around feathers in solution, bacterial growth on feathers, and loss of feather mass. The latter two metrics were not
considered in the aforementioned studies, which indirectly measured bacterial activity. We conducted two trials, one in
which feathers were sterilized by autoclaving before inoculation (Goldstein et al. 2004, Grande et al. 2004), and a second
in which feathers were sterilized by ethylene oxide gas. This allowed us to test whether autoclaving, done in previous
studies, influences bacterial activity on feathers and could confound results. In both trials, unmelanized feathers degraded
earlier, supported greater bacterial growth, and lost more mass than melanized feathers. These results support the findings
of Goldstein et al. (2004); melanized feathers are more resistant to FDB than unmelanized feathers. Thus, using direct
metrics of bacterial activity, we resolve a current conflict in the literature. We also found that autoclaving feathers
influences FDB activity on them, and thus autoclaving should be avoided in future studies.
Melanins are the most common pigments of feathers and
are responsible for the majority of black, brown, and earthtoned colours of avian plumage (McGraw 2006). The
putative functions of melanized feathers are many, ranging
from thermoregulation (Walsberg 1983) to crypsis (Zink
and Remsen 1986) to social signaling (Griffith et al. 2006).
Placement of melanins into feathers appears to be costly
and subject to environmental influences, much like carotenoid pigmentation (Griffith et al. 2006, also reviewed in
McGraw 2006). Variation in feather melanization may
influence feather condition, as melanized feathers are harder
and more resistant to physical abrasion than unmelanized
(white) feathers (Bonser 1995, Burtt 1979, 1986). Feather
wear can influence flight performance (Swaddle and Witter
1997, Williams and Swaddle 2003) and thermoregulatory
ability (Booth et al. 1993), and thus melanization could
reduce costs associated with feather wear.
The physical properties of melanized feathers compared
to unmelanized feathers have led to the suggestion that
melanized feathers are resistant to feather-feeding ectoparasites such as lice, and there is evidence consistent with
this hypothesis (Kose et al. 1999, but see Bush et al.
2006). Similarly, melanized feathers may be more resistant
than unmelanized feathers to damage by feather-degrading
bacteria (FDB) (Goldstein et al. 2004, but see Grande
et al. 2004). FDB are a polyphyletic assemblage of
microbes that produce proteolytic keratinase enzymes
that hydrolyze b-keratin (Gupta and Ramnani 2006,
Onifade et al. 1998), the primary constituent of feathers.
FDB occur within the plumage of several birds (Burtt and
Ichida 1999, Whitaker et al. 2005), and, for some species
at least, are a ubiquitous feature of plumage (e.g. FDB
detected on 66/67 (99%) of eastern bluebirds Sialia sialis,
in southern Virginia, U.S.A.; A.R. Gunderson unpubl.
data). Despite the potential for FDB to be a potent
selective force shaping feather evolution (Burtt and Ichida
1999, Clayton 1999), little is known about how these
microbial symbionts function on live birds. FDB were first
isolated from wild bird feathers less than 10 years ago
(Burtt and Ichida 1999), and in the short time since only
a handful of studies have addressed interactions between
birds and FDB (e.g. Burtt and Ichida 2004, Cristol et al.
2005, Shawkey et al. 2007, Shawkey et al. 2003, reviewed
in Gunderson 2008).
To date, two studies (Goldstein et al. 2004, Grande et al.
2004) have tested the hypothesis that melanized feathers are
more resistant to FDB than unmelanized (white) feathers;
however, these studies reached opposing conclusions. Goldstein et al. (2004) suspended unmelanized (white) and
melanized (black) secondary flight feathers of domestic
chickens Gallus gallus, in a buffer media (referred to as
‘‘feather solution’’) and inoculated them with the FDB
Bacillus licheniformis. Feather degradation rates were determined by measuring the soluble protein concentration in
feather solution over the course of six days. This method
assumes that proteins in feather solution are the result of
bacterial activity on feathers. Feather solution had higher
soluble protein content with unmelanized than melanized
feathers (Goldstein et al. 2004), suggesting feather melanization inhibits B. licheniformis. On the contrary, Grande et al.
(2004) found that melanized feathers degrade more quickly
than unmelanized feathers. In their study, melanized and
unmelanized feathers from two (Ciconia ciconia, Corvus
corax) and three (Eudocimus ruber, Egretta garzetta, Ciconia
ciconia) species, respectively, were placed in feather solution
and inoculated with B. licheniformis. Feather degradation was
scored subjectively by observing each flask and visually
estimating feather damage. Melanized feathers showed signs
of degradation before unmelanized feathers, and melanized
feathers were more damaged than unmelanized feathers at the
conclusion of the experiment (Grande et al. 2004). This
study also included feathers coloured with carotenoids,
which also appeared to be more resistant to bacterial
degradation than melanized feathers (Grande et al. 2004).
The results of these two studies have received considerable
attention (see Bortolotti 2006, McGraw 2006, Shawkey and
Hill 2004) because they suggest feather pigmentation may
play a role in the resistance of feathers to parasites. Thus,
those studying avian colouration may have overlooked an
important mechanism that modulates colour expression. The
question remains, however: do melanins impart bacterial
resistance to feathers or do they make feathers more
susceptible to bacterial degradation? Based on the two studies
discussed above, there is no clear answer to this question.
Furthermore, we felt there were aspects of the Goldstein et al
(2004) and Grande et al. (2004) studies that suggested their
results be regarded cautiously. Goldstein et al. (2004) do not
present data on replicate experimental units for each feather
type. Their results include no measures of error and were not
subjected to statistical scrutiny. Furthermore, the authors
included an uninoculated control flask for unmelanized
feathers only, but not melanized feathers. This omission of a
control treatment could confound their results because, as the
authors state themselves and as is discussed above, melanized
feathers are more resistant to physical damage than unmelanized feathers. Thus, the discrepancy in degradation
rates could have been the result of their initial feather
sterilization method (autoclaving at 1218 C for 15 min) and/
or their incubation conditions (agitation at 120 rpm at
378 C) weakening unmelanized feathers more than melanized feathers, causing them to be less resistant to bacterial
degradation. Interpretation of the Goldstein et al. (2004)
data is further complicated by the fact that they do not
present or discuss the results for the unmelanized control that
they prepared. Grande et al. (2004) did have replication in
their study; however, they determined feather degradation
rates subjectively on a scale of 1!5 based upon visual
inspection of the feathers in the flasks, using a criterion
determined a posteriori (Grande et al. 2004). Grande et al.
(2004) also did not have controls for all bird species used in
their study and did not present data on controls.
In this study, we investigated whether melanized feathers
are more resistant to bacterial degradation than unmelanized feathers, while addressing our concerns about the
studies discussed above. Because of the physical characteristics of melanized feathers, we hypothesized that melanized
feathers would be more resistant to FDB than unmelanized
feathers. We thus predicted that FDB would degrade
melanized feathers more slowly than unmelanized feathers.
To test this, we quantified peptides released into solution
around inoculated feathers (Goldstein et al. 2004), as well
as bacterial growth and loss of feather mass (neither of
which were quantified in the aforementioned studies).
Before feather degradation trials can be conducted,
feathers must be sterilized to ensure that bacteria initially
present on the feathers do not contribute to feather
degradation or alter the activity of bacterial inocula. Feather
degradation trials were conducted using two different
feather sterilization methods, autoclaving and ethylene
oxide gas. Goldstein et al. (2004) and Grande et al.
(2004) both sterilized feathers by autoclaving them at
high temperature and pressure. High temperature and
pressure conditions can influence the structure of feathers,
and such conditions are often used to pre-treat feathers
when they are being processed into agricultural foodstuffs
(Kim and Patterson 2000). Thus, autoclaving may make
feathers more readily degradable, leading to ecologically
questionable results in bacterial feather degradation trials.
Furthermore, because melanized feathers are thought to be
more robust than unmelanized feathers, autoclaving may
differentially influence feather structure depending on
pigmentation, which would confound the experiment.
Specifically, melanized feather structure might be less
affected by autoclaving than unmelanized feathers. Ethylene
oxide gas sterilization does not expose feathers to high
temperature and pressure conditions (see Materials and
methods), and thus should not influence feather structure.
Materials and methods
We used unmelanized (white) and melanized (dark brown)
retrices of domestic geese Anser anser domesticus, obtained
from a commercial breeder. In each degradation trial,
300 mg of feathers (consisting of several small feather pieces
from different feathers and, probably, different individuals)
was placed in each of twelve 125 ml Erlenmeyer flasks; six
flasks contained unmelanized feathers and six flasks contained melanized feathers. We used only the distal
(approximately) 10 cm of feathers to ensure that most of
the feather mass consisted of barbs and barbules. The rachis
constitutes most of the mass of the proximal region and is
not usually degraded by B. licheniformis (Ramnani et al.
We conducted degradation trials with feathers sterilized
with autoclaving or ethylene oxide gas. In trials in which
feathers were sterilized by autoclaving, feathers were placed
dry in flasks and autoclaved for 15 min at 1218 C
(unmelanized feathers) (Goldstein et al. 2004) or 18 min at
1218 (melanized feathers). Melanized feathers had to be
autoclaved for 18 min, as 15 min of autoclaving repeatedly
failed to sterilize flasks containing melanized feathers
(N "30 flasks). After autoclaving, 50 ml of sterile phosphate
buffered saline (PBS), pH"7.25, was added to each flask.
We did not autoclave the feathers in PBS because in
preliminary trials this appeared to contribute to feather
dissolution. For ethylene oxide gas sterilization, feathers were
incubated in a chamber at 548 C with 170 g of ethylene oxide
gas, followed by 12 hours of aeration. Feathers were sterilized
dry and then placed in flasks containing 50 ml sterile PBS.
For both sterilization methods, flasks were kept at 358 C for
24 h after the addition of PBS to allow soluble proteins
associated with the feathers to enter solution and to allow any
potential contaminants to grow. We then sampled the feather
media to take an initial reading of soluble protein concentration before experimental bacterial inoculation. From these
samples, we plated 20 ml of the solution on trypticase soy agar
(TSA), a general growth medium, to ensure that the flasks
were not contaminated. Three flasks of unmelanized feathers
and three of melanized feathers were then inoculated with
FDB, resulting in a total of three treatment (inoculated with
bacteria) and three control (not inoculated with bacteria)
replicate flasks per feather pigment type, per trial.
Treatment flasks were inoculated with pure cultures of
B. licheniformis strain OWU 138B, a FDB originally
isolated from the plumage of a willow flycatcher, Empidonax traillii (Goldstein et al. 2004). B. licheniformis is a wellstudied FDB (e.g. Burtt and Ichida 1999, Ichida et al.
2001, Shawkey et al. 2003, Whitaker et al. 2005) that
appears to be common within the plumage of many species
(Burtt and Ichida 1999, Whitaker et al. 2005). Approximately 500,000 cells of B. licheniformis were administered
to each treatment flask. To accomplish this, the FDB were
streaked on TSA 24 hours before feathers were inoculated,
and were incubated at 358 C. Pure isolated colonies were
removed from the plates with a sterile cotton swab and used
to inoculate a solution of sterile PBS. The concentration of
bacteria in the solution was determined by measuring the
optical density at 600 nm with a Bio-Rad Smart Spec 3000
(Bio-Rad), and this concentration was used to calculate the
volume required to inoculate the flasks with the target
number of bacteria. Among trials the volume of bacterial
solution added to treatment flasks ranged from 1!4 ml, and,
thus, did not significantly affect the volume of solution in
treatment flasks. After inoculation, all flasks were placed in
a 378 C incubator with agitation at 120 rpm.
Protein sampling and concentration assay
Every 48 h, over a 12-d period, we withdrew 500 ml
aliquots of feather solution from each flask and all aliquots
were stored at #208 C. Upon thawing for protein
concentration assay, samples were centrifuged at 10,000 g
for 5 min to remove particulate feather matter and bacteria
from the supernatant. The soluble protein content of each
sample was determined by measuring the absorbance of the
supernatant at 280 nm (Lucas et al. 2003) using a
Nanodrop spectrophotometer (Nanodrop Technologies,
Inc., Wilmington, DE). This method assumes that proteins
in solution are derived from bacterial feather degradation,
but does not specifically measure the concentration of
oligopeptides produced by keratin cleavage. Other proteins
present in feathers or produced by bacteria most likely
contribute to the soluble protein content as well.
Viable cell counts
Bacteria were quantified every 48 h, at the time of protein
sampling, using the plate count method (Hattori 1982). For
each flask, 20 ml of feather solution were plated on TSA
from the following dilutions: 1, 10 #1, 10 #2, 10 #3, 10 #4,
10#5, and 10#6. Plates were incubated at 358 C for 24 h,
after which all colonies were counted. Plate counting also
allowed us to determine if flasks had become contaminated
with bacteria other than B. licheniformis over the course of
the trials. No contamination was detected.
Change in feather mass
Before sterilization, the feathers to be placed in each flask
were dried for 48 h at 558 C and weighed to the nearest
0.1 mg on an analytical balance (Ohaus AS60). After trials,
the feathers were again dried, this time for 72 h (we found
that this increase in drying time was necessary to remove all
moisture after the feathers had been suspended in aqueous
solution during trials) and weighed to the nearest 0.1 mg.
Data analysis
We used repeated measures analysis of variance (ANOVA)
for all analyses unless otherwise noted. Analyses performed
on bacterial growth data were log transformed to improve
normality and reduce disparities in variance among groups.
All statistical tests were performed using the R statistical
programming package (v. 2.4.1) employing two-tailed tests
of probability.
Autoclave sterilization trial
Feather media significantly increased in soluble protein
content for inoculated unpigmented (time by treatment
interaction, F7,28 "10.20, P B0.001; Fig. 1A) and melanized (time by treatment interaction, F7,28 "2.87, P "
0.02; Fig. 1B) feathers relative to controls. However, we
did not detect a degradation difference between inoculated
melanized and unmelanized feathers (time by pigment by
treatment interaction, F7,56 "0.59, P "0.764). As a coarse
metric of when degradation began for feathers of each
pigmentation type, we used the time at which the 95%
confidence interval of the mean of neither treatments nor
controls included the mean of the other group. Initial
protein content at the beginning of the experiment
(i.e., time "0) was standardized to zero for each flask
by subtracting each flask’s initial protein concentration
from all subsequent measurements. Using this criterion,
We did not detect a difference between inoculated
melanized and unmelanized feathers (time by pigment by
treatment interaction, F7,56 "1.10, P "0.376). Based on
the criterion for the appearance of degradation discussed
above, unmelanized feathers showed clear signs of degradation at 240 h (Fig. 3A), while there was no clear sign of
when degradation began for melanized feathers within the
time frame of our experiment (Fig. 3B).
Bacteria reached higher densities on unmelanized feathers
compared to melanized feathers (time by pigment interaction, F5,20 "4.92, P"0.001; Fig. 4A). Unmelanized inoculated feathers lost more mass than unmelanized controls
(time by treatment interaction, F1,4 "23.00, P "0.009) and
either inoculated or control group melanized feathers (time
by pigment by treatment interaction, F1,8 "19.61, P "
0.002; Fig. 4B). There was no difference in mass loss between
control and inoculated melanized feathers (time by treatment
interaction, F1,4 "0.69, P "0.451; Fig. 4B).
Resistance of melanized feathers to bacterial
Figure 1. Mean soluble protein content of media surrounding
autoclaved domestic goose feathers inoculated with B. licheniformis. (A) Unmelanized feathers. (B) Melanized feathers. Circles,
feathers inoculated with B. licheniformis; triangles, uninoculated
controls. Bars indicate 95% confidence intervals.
Irrespective of the feather sterilization method or bacterial
activity metric used, unmelanized feathers degraded more
unmelanized feathers demonstrated clear signs of degradation after 192 hours (Fig. 1A), while melanized feathers
showed clear signs of degradation after 240 hours (Fig. 1B).
B. licheniformis grew to higher densities on unmelanized
feathers than on melanized feathers (time by pigment
interaction, F5,20 "9.72, P B0.001; Fig. 2A). Plate counts
from 192 h were removed from this analysis because the
counts were performed incorrectly, as were plate counts
from 336 hours because one flask could not be counted,
which led to statistical imbalance. Inoculated unmelanized
feathers lost more mass than unmelanized controls (time by
treatment interaction, F1,4 "23.49, P "0.008). Similarly,
inoculated melanized feathers trended towards more mass
loss than control melanized feathers (time by treatment
interaction, F1,2 "17.31, P"0.053; Fig. 2B). One melanized control flask and one melanized treatment flask had
to be removed from the mass data set because an error
occurred in their measurement, so we were unable to
statistically compare feather mass loss across pigments due
to statistical imbalance. However, on average, unmelanized
feathers lost an order of magnitude more mass than
melanized feathers.
Ethylene oxide sterilization trial
Soluble protein content around feathers inoculated with
bacteria increased significantly relative to controls for
unmelanized feathers (time by treatment interaction,
F7,28 "12.99, P B0.001; Fig. 3A), but only trended
towards significance for melanized feathers (time by
treatment interaction, F7,28 "12.99, P "0.052; Fig. 3B).
Figure 2. (A) Bacterial growth on domesticated goose feathers
sterilized by autoclaving. White circles, growth on unmelanized
feathers; grey circles, growth on melanized feathers. (B) Loss of
feather mass of domestic goose feathers sterilized by autoclaving.
White bars, unmelanized feathers; grey bars, melanized feathers.
Bars in both (A) and (B) represent one standard error. (.) 0.10 !
P!0.05, (***) PB0.001.
Figure 3. Mean soluble protein content of media surrounding
domestic goose feathers sterilized with ethylene oxide gas and
inoculated with B. licheniformis. (A) Unmelanized feathers. (B)
Melanized feathers. Circles, feathers inoculated with B. licheniformis; triangles, uninoculated controls. Bars indicate 95% confidence
readily than melanized feathers. In both trials, B. licheniformis grew to higher densities, removed more feather mass,
and showed signs of degradation activity earlier on
unmelanized relative to melanized feathers. These results
support the conclusion of Goldstein et al. (2004) but
contradict those of Grande et al. (2004). Melanized feathers
are more resistant than unmelanized feathers to the
degrading action of B. licheniformis. Hence, our study helps
to resolve a current controversy in the literature and firmly
indicates that melanins can render a feather more resistant
to FDB.
Affects of sterilization on bacterial feather
We show evidence that feather sterilization method can
influence bacterial activity on feathers. Based on soluble
protein content in feather solution, unmelanized feathers
showed initial signs of degradation 48 hours earlier in the
autoclave compared to the ethylene oxide trial (compare
Fig. 1A and 3A). Melanized feathers showed initial signs of
degradation at 240 hours in the autoclaved trial, but this
never occurred in the ethylene oxide trial (compare Fig. 1B
and 3B). Interestingly, in the autoclaved trial, bacterial
populations crashed on unmelanized feathers (Fig. 2A), but
this crash was not seen in the ethylene oxide trial (Fig. 4A).
Furthermore, inoculated melanized feathers trended to-
Figure 4. (A) Bacterial growth on domesticated goose feathers
sterilized with ethylene oxide gas. White circles, growth on
unmelanized feathers; grey circles, growth on melanized feathers.
(B) Loss of feather mass of domestic goose feathers sterilized with
ethylene oxide gas. White bars, unmelanized feathers; grey bars,
melanized feathers. Bars in both (A) and (B) represent one
standard error. (NS) Not Significant, (***) P B0.001.
wards losing more mass (P"0.053) than control melanized
feathers in the autoclave trial (Fig. 2B), but there was no
difference in mass loss between these groups (P "0.451) in
the ethylene oxide trial (Fig. 4B). These effects did not
influence the qualitative interpretation of our data. In both
the autoclave and ethylene oxide trials, unmelanized
feathers were clearly more susceptible to bacterial degradation than melanized feathers. This is likely because the
difference in degradation rates between these two feather
types is large. However, if differences in degradation rate
between two feather types is more subtle, the influence of
feather sterilization method on bacterial activity could
qualitatively influence the interpretation of results. We
recommend that future studies comparing bacterial degradation rates of different feather types avoid autoclaving and
employ feather sterilization techniques unlikely to influence
feather structure, such as ethylene oxide gas (this study) or
g-rays (Shawkey et al. 2007).
We also found evidence that incubation conditions can
differentially affect different feather types. Soluble protein
content around control melanized feathers increased steadily over time in both trials (Fig. 1B and 3B), but did not
occur with unmelanized controls (Fig. 1A and 3B). Thus,
we strongly recommend that controls be prepared for all
feather types, as the influence of FDB on feather condition
must be measured relative to uninoculated feathers with the
same properties.
The mechanism(s) of FDB resistance of melanized feathers
is unknown (Goldstein et al. 2004). Goldstein et al. (2004)
suggest that the incorporation of melanin granules into a
feathers’ keratin matrix may force keratin rods into close
proximity with one another, thus catalyzing the production
of more disulfide bonds between adjacent keratin molecules. Intermolecular disulfide bonds must be broken for
keratins to become nutritionally available to FDB, and FDB
produce specific enzymes and chemicals to reduce these
bonds (Gupta and Ramnani 2006). More disulfide bonds
could impede feather degradation. However, there is no
evidence that melanins facilitate disulfide bonding among
keratin molecules. At present, not enough is known about
the biochemistry of melanized versus unmelanized feathers
to determine if this is a reasonable hypothesis.
Melanins could also inhibit FDB by binding to bacterial
proteolytic keratinases that hydrolyze keratins, a model for
which there is experimental support. Melanized fungal
mycelia resist microbial enzymatic digestion and, to test the
mechanism of enzymatic resistance, Kuo and Alexander
(1967) measured the efficiency of several microbial enzymes
in the presence of melanin. Melanins reduced the efficiency
of all enzymes, one of which was a protease of the bacterium
Bacillus subtilis (Suh and Lee 2001), a FDB closely related
to B. licheniformis. The activity of the B. subtilis protease
was reduced by 49% to 69% in the presence of various
concentrations of melanin and melanin itself was found to
be resistant to microbial degradation (Kuo and Alexander
1967). Thus, it seems reasonable to suggest that one way in
which melanins protect feathers is by binding to and
inactivating keratinases.
The next step is to determine whether or not the
resistance of melanized feathers to FDB is relevant to
plumage under natural conditions. Burtt and Ichida (2004)
began to address this question by comparing the prevalence
(number of individuals infected) and feather-degradation
rates of B. licheniformis from light and dark subspecies of
song sparrow (Melospiza melodia fallax and Melospiza
melodia morphna, respectively). The darker birds were
assumed to have higher melanin concentrations in their
feathers, although changes in the relative quantities of
eumelanin and phaeomelanin could also play a role
(McGraw 2006). There was no difference in the prevalence
of FDB between the populations, but the data indicated
FDB on the darker sparrow population degrade unmelanized chicken feathers more quickly than FDB isolated
from the lighter sparrow population (Burtt and Ichida
2004). The authors interpret this as evidence that more
damaging bacteria are selecting for more melanin deposition in feathers. However, they quantified bacterial degradation subjectively in a manner similar to Grande et al.
(2004); no protein measurements, bacterial counts, or
measurements of feathers-mass change were made. They
also note that autoclaving the chicken feathers before
inoculation damaged the feathers, and to different degrees
in different flasks, which was not accounted for in their
analysis. Furthermore, it is unclear how FDB degradation of
unmelanized chicken feathers relates to the degradation of
melanized sparrow feathers. If feather melanization puts
selective pressure on FDB, then FDB should adapt to
degrade feathers more readily in the presence of melanin,
which, depending on the mechanism of melanic inhibition,
would not necessarily influence the rate at which bacteria
degrade unmelanized feathers. Roulin (2007) found that
the rate of preening in barn owl Tyto alba nestlings
decreased with plumage melanization. Preen oil can inhibit
the growth of FDB (Shawkey et al. 2003) and the physical
act of preening may dislodge or damage bacteria (Clayton
1999). If this is the case, then more melanized individuals
may not need to preen as regularly to combat FDB. At
present, well-controlled field experiments are needed to test
the hypothesis that melanized feathers resist degradation by
FDB. However, our series of laboratory tests, and those by
Goldstein et al. (2004), indicate that melanin deposition
can make feathers relatively more resistant to degradation
by FDB. Therefore, it is possible that FDB can impose
selection pressures on the evolution of avian feather
colouration and the physical abilities of feathers to resist
abrasion and damage.
Acknowledgements ! We thank Edward H. Burtt, Jr. for kindly
supplying the Bacillus licheniformis strain used in this study and for
comments on the manuscript, Caitlin Kight for technical expertise,
Dan Cristol and George Gilchrist for helpful discussions on the
project and manuscript, and Manuel Leal for helpful comments on
the manuscript. This research was funded by graduate research
grants from the College of William and Mary to ARG, a student
research grant funded by a Howard Hughes Medical Institute
grant to the Biological Sciences Education Program at William
and Mary to AMF, National Science Foundation awards IOB0133795 and EF-0436318 to JPS, and National Institutes of
Health (R15-AI53062) to MHF.
Bonser, R. H. C. 1995. Melanin and the abrasion resistance of
feathers. ! Condor 97: 590!591.
Booth, D. T., Clayton, D. H. and Block, B. A. 1993. Experimental demonstration of the energetic cost of parasitism in
free-ranging hosts. ! Proc. R. Soc. B 253: 125!129.
Bortolotti, G. R. 2006. Natural selection and coloration: Protection, concealment, advertisement, or deception? ! In: Hill,
G. E. and McGraw, K. J. (eds). Bird coloration. Volume 2.
Function and evolution. Harvard University Press, Cambridge, pp. 3!35.
Burtt, E. H. 1979. Tips on wings and other things. ! In: Burtt,
E. H. (ed.). The behavioral significance of color. Garland
STMP Press, New York, pp. 75!110.
Burtt, E. H. 1986. An analysis of physical, physiological optical
aspects of avian coloration with emphasis on Wood-Warblers.
! Ornithol. Mon. 38: 1!126.
Burtt, E. H. and Ichida, J. M. 1999. Occurrence of featherdegrading bacilli in the plumage of birds. ! Auk 116:
Burtt, E. H. and Ichida, J. M. 2004. Gloger’s rule, featherdegrading bacteria, and color variation among song sparrows.
! Condor 106: 681!686.
Bush, S. E., Kim, D., Moyer, B. R., Lever, J. and Clayton, D. H.
2006. Is melanin a defense against feather-feeding lice? ! Auk
123: 153!161.
Clayton, D. H. 1999. Feather-busting bacteria. ! Auk 116:
Cristol, D. A., Armstrong, J. L., Whitaker, J. M. and Forsyth, M.
H. 2005. Feather-degrading bacteria do not affect feathers on
captive birds. ! Auk 122: 222!230.
Goldstein, G., Flory, K. R., Browne, B. A., Majid, S., Ichida, J. M.
and Burtt, E. H. 2004. Bacterial degradation of black and
white feathers. ! Auk 121: 656!659.
Grande, J. M., Negro, J. J. and Torres, M. J. 2004. The evolution
of bird plumage colouration: a role for feather-degrading
bacteria? ! Ardeola 51: 375!383.
Griffith, S. C., Parker, T. H. and Olson, V. A. 2006. Melaninversus carotenoid-based sexual signals: is the difference really
so black and red? ! Anim. Behav. 71: 749!763.
Gunderson, A.R. 2008. Feather-degrading bacteria: a new frontier
in avian and host-parasite research? ! Auk in press.
Gupta, R. and Ramnani, P. 2006. Microbial keratinases and their
prospective applications: an overview. ! Appl. Microbiol.
Biotechnol. 70: 21!33.
Hattori, T. 1982. Analysis of plate count data of bacteria in
natural environments. ! J. Gen. Appl. Microbiol. 28: 13!22.
Ichida, J. M., Krizova, L., LeFevre, C. A., Keener, H. M., Elwell,
D. L. and Burtt, E. H. 2001. Bacterial inoculum enhances
keratin degradation and biofilm formation in poultry compost.
! J. Microbiol. Meth. 47: 199!208.
Kim, W. K. and Patterson, P. H. 2000. Nutritional value of
enzyme- or sodium hydroxide-treated feathers from dead hens.
! Poultry Sci. 79: 528!534.
Kose, M., Mand, R. and Moller, A. P. 1999. Sexual selection for
white tail spots in the barn swallow in relation to habitat
choice by feather lice. ! Anim. Behav. 58: 1201!1205.
Kuo, M.-J. and Alexander, M. 1967. Inhibition of the lysis of
fungi by melanins. ! J. Bacteriol. 94: 624!629.
Lucas, F. S., Broennimann, O., Febbraro, I. and Heeb, P. 2003.
High diversity among feather-degrading bacteria from a dry
meadow soil. ! Microb. Ecol. 45:282!290.
McGraw, K. J. 2006. Mechanics of melanin-based coloration.
! In: Hill, G. E. and McGraw, K. J. (eds). Bird coloration Vol.
1. Mechanisms and measurements. Vol. 1. Harvard University
Press, Cambridge, pp. 243!294.
Onifade, A. A., Al-Sane, N. A., Al-Musallam, A. A. and AlZarban, S. 1998. A review: potentials for biotechnological
applications of keratin-degrading microorganisms and their
enzymes for nutritional improvement of feathers and other
keratins as livestock feed resources. ! Biores. Technol. 66:
Ramnani, P., Singh, R. and Gupta, R. 2005. Keratinolytic
potential of Bacillus licheniformis RG1: structural and biochemical mechanism of feather degradation. ! Can. J.
Microbiol. 51: 191!196.
Roulin, A. 2007. Melanin pigmentation negatively correlates with
plumage preening effort in barn owls. ! Funct. Ecol. 21:
Shawkey, M. D. and Hill, G. E. 2004. Feathers at a fine scale.
! Auk 121: 652!655.
Shawkey, M. D., Pillai, S. R. and Hill, G. E. 2003. Chemical
warfare? Effects of uropygial oil on feather-degrading bacteria.
! J. Avian Biol. 34: 345!349.
Shawkey, M. D., Pillai, S. R., Hill, G. E., Siefferman, L. M. and
Roberts, S. R. 2007. Bacteria as an agent for change in
structural plumage color: correlational and experimental
evidence. ! Am. Nat. 169: S112!S121.
Suh, H.J. and Lee, H.K. 2001. Characterization of a keratinolytic
serine protease from Bacillus subtilis KS-1. ! J. Prot. Chem. 20:
Swaddle, J. P. and Witter, M. S. 1997. The effects of molt on the
flight performance, body mass, and behavior of European
starlings (Sturnus vulgaris): an experimental approach. ! Can.
J. Zool. 75: 1135!1146.
Walsberg, G. E. 1983. Avian ecological energetics. ! In: Farner,
D. S., King, J. R. and Parkes, K. C. (eds). Avian biology. 7.
Academic Press, New York, pp. 161!220.
Whitaker, J. M., Cristol, D. A. and Forsyth, M. H. 2005.
Prevalence and genetic diversity of Bacillus licheniformis in
avian plumage. ! J. Field Ornithol. 76: 264!270.
Williams, E. V. and Swaddle, J. P. 2003. Moult, flight
performance and wingbeat kinematics during take-off in
European starlings Sturnus vulgaris. ! J. Avian Biol. 34:
Zink, R. M. and Remsen, J. V. 1986. Evolutionary processes and
patterns of geographic variation in birds. ! Curr. Ornithol. 4:
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF