Experiment 2 –Timeline for Diet Selection Experiment
To investigate whether birds alter feeding behavior to optimize
responses to infection we conducted a second experiment, in which we
simulated an infection in established zebra finch pairs using the
bacterial endotoxin lipopolysaccharide (LPS), and quantified feeding
behavior in immune challenged and control individuals, as well as birds
housed near either a control pair (no immune threat), or birds housed
near a pair given an immune challenge with LPS (social cue of heightened
infection risk). To investigate how an immune challenge and perceived
immune threat shape macronutrient preference, birds were provided with
two isocaloric diets with varied lipid and protein ratios and their
consumption was recorded. Birds were kept on a 14 L: 10 D light cycle
and housed in 24”x16”x16” cages that were divided down the center
into two separate 12”x16”x16” cage sections that each housed one pair
of birds. Birds were housed in previously established pairs with one
female and one male per cage section. Each cage section had two perches,
a water dish, and two food dishes in which birds were fed ad
libitum . Birds were given a choice of diet where each cage contained
one high lipid food block and one high protein food block, whose
composition was the same as in Experiment 1. Diet type was randomly
assigned to either the left or right food dish to avoid confounding
effects of cage side preference. To assess whether an immune threat
altered the amount and type of diet that birds consumed, birds were
provided with a choice in diet for seven days prior to experimental
treatment and for five days following experimental treatment. Diets were
weighed daily and replaced every other day. Three desiccation controls
for each diet type were also weighed daily and the average desiccation
values for each diet type (high lipid or high protein) were subtracted
from feeding values to account for daily changes in food mass due to
desiccation.
Pairs housed on one side of the cage were injected with either LPS or
Saline (Figure 1; LPS-injected: N=24 birds or Saline-injected: N=24
birds), whereas pairs housed on the other side of the cage were
unmanipulated (focal pairs; LPS-focal: N=24 birds or Saline-focal: N=24
birds). Injected pairs provided social cues to the focal pair. Solid
opaque dividers were placed on both sides of each 24 x 16 x 16 cage to
ensure that birds housed in each double cage could only see one another.
To assess how an immune challenge and social cue of heightened infection
risk shape feeding behavior, we injected previously established pairs
with either lipopolysaccharide (LPS), a non-replicating antigen that
activates the immune system and induces sickness behaviors, or a saline
solution (control). Specifically, we injected stimulus birds
intra-abdominally with either 50 μL of 2 mg/kg LPS (Sigma-Aldrich
#L7261, Salmonella enterica serotype typhimurium) or 50 μL of
phosphate-buffered saline (sham control, Sigma-Aldrich #P3813). Body
mass and fat score data were collected on all stimulus and focal birds
prior to treatment and on days 1, 2, and 5 post-treatment.
To assess whether heightened infection risk altered complement activity
or baseline corticosterone concentrations, we collected blood samples
from all focal birds 3 days prior to stimulus bird injections and 1, 2,
and 5 days following stimulus bird injections. All blood samples for
baseline corticosterone were collected within 3 minutes of entering the
room. We collected additional blood samples from all focal males to
assess changes in plasma testosterone concentrations in response to a
heightened cue of infection 4 days prior to and 3 days following
injection of the stimulus pairs. Immediately after collection, blood
samples were centrifuged and blood plasma was separated and frozen at
-20°C. To determine how treatment affected zebra finch gut microbial
diversity and composition, cloacal swab samples were collected from all
birds prior to treatment (day 0) and on day 5 following stimulus bird
injections. Cloacal swabs were placed into 300 µl of RNAlater
(Invitrogen, Thermo Fisher Scientific) and frozen at -80℃. All research
protocols were approved by the Oklahoma State University Institutional
Animal Care and Use Committee.