Storage Protein Content as a Functional Marker for Colony-Founding
Strategies: A Comparative Study within the Harvester
Ant Genus Pogonomyrmex
Daniel A. Hahnl1*
Robert A. Johnson2'*
Norman A. Buck3'*
Diana E. Wheeler3'
1Interdisciplinary Graduate Program in Insect Science,
University of Arizona, Tucson, Arizona 85721; 2Department
of Biology, Arizona State University, Tempe, Arizona 85287;
3Department of Entomology, University of Arizona, Tucson,
Arizona 85721
Accepted 7/30/03
ABSTRACT
Claustral colony founding, in which new queens rear their first
clutch of workers solely from internal reserves, is common in
the higher ant subfamilies and is believed to represent a major
innovation in ant life histories. The ability to store large
amounts of amino acids contained in storage proteins is an
essential physiological trait for claustral colony founding by ant
queens. To determine whether there is an association between
storage protein content and colony-founding strategy, we iden-
tified and quantified two major storage proteins in queens of
five harvester ant species in the genus Pogonomyrmex that differ
in colony-founding strategy. Queens of the fully claustral non-
foraging species Pogonomyrmex rugosus and Pogonomyrmex
maricopa contained the greatest amount of these proteins. Fac-
ultatively foraging semiclaustral Pogonomyrmex occidentalis
queens contained an intermediate amount. Obligately foraging
semiclaustral Pogonomyrmex californicus queens from two dif-
ferent populations contained significantly less storage protein
than the other independent-founding species. Queens of the
dependent-founding social parasite Pogonomyrmex anergismus
also contained little storage protein. Our results suggest that
*Corresponding author. Present address: Department of Entomology, Ohio
State University, 1735 Neil Avenue, Columbus, Ohio 43210-1220; e mail:
hahn.144@osu.edu.
E mail: robert.johnson4@asu.edu.
'E-mail: nbuck@ag.arizona.edu.
E mail: dewsants@ag.arizona.edu.
Physiological and Biochemical Zoology 77(1):100-108. 2004. O 2004 by The
University of Chicago. All rights reserved. 1522-2152/2004/7701-3009$15.00
storage protein content has evolved in concert with colony-
founding strategies in the genus Pogonomyrme and provides a
good functional marker for colony-founding strategy.
Introduction
Colony founding is the most vulnerable stage in the ant colony
life cycle (Holldobler and Wilson 1990; Bourke and Franks
1995). In response, many strategies to enhance founding success
have evolved (Holldobler and Wilson 1990; Bourke and Franks
1995; Rtppell and Heinze 1999; Brown and Bonhoeffer 2003).
Ants display two major types of colony founding. First, in
dependent founding, queens start a new nest with a group of
workers. Social parasites, in which queens invade mature host
colonies and species that adopt queens directly into mature
nests, are also considered dependent founders. Second, in in-
dependent founding, queens start new colonies without the aid
of workers.
Independent founding strategies also fall into two types. First,
in semiclaustral founding, queens leave the nest and forage to
meet the nutritional needs of themselves and their brood. Sec-
ond, in fully claustral founding, queens seal themselves in the
nest after mating and raise the first clutch of workers entirely
from body reserves. Fully claustral colony founding is thought
to be a major innovation in higher ants because it eliminates
the need for queens to leave the nest to forage, reducing their
exposure to predators, desiccation, and other sources of mor-
tality (H6lldobler and Wilson 1990; Bourke and Franks 1995;
Brown and Bonhoeffer 2003). The potential costs of queen
foraging make it notable that several species in the higher ant
subfamilies Myrmicinae and Formicinae have returned to semi-
claustral founding when most of their congeners remain fully
claustral (H6lldobler and Wilson 1990; Brown and Bonhoeffer
2003).
The diversity of founding strategies in ants is well docu-
mented, but the physiological correlates of these strategies are
poorly known. Stille (1996) showed an association between the
relative size of queens compared with their workers and found-
ing. Independent single claustral foundresses were proportion-
ally the largest, followed by independent foundresses that start
colonies in multiple queen groups, followed by dependent social
parasites. No semiclaustral species were considered. Keller and
Storage Protein and Colony Founding 101
Table 1: Collection locales (state, county, locale) and sample size for species of Pogonomyrmex that were analyzed in this
study
Species Latitude Longitude Elevation (m) Sample Size" Year
Pogonomyrmex californicus complex:
P. californicus (Buckley); Arizona, Maricopa, Salt
River Recreation Area, Blue Point Bridge 33033'N 11134'W 425 14 (9) 1999
12 (6) 2001
P. californicus (Buckley); California, San Diego, Cameron
Guard Station 3243'N 116028'W 990 22 (11) 2001
Pogonomyrmex maricopa Wheeler; Arizona, Pinal, 2.0
km north of Superior 3318'N 111008'W 825 14 (5) 1999
6 (2) 2000
Pogonomyrmex occidentalis complex:
P. occidentalis (Cresson); Arizona, Yavapai, Ash Fork 3513'N 11230'W 1,555 9 (3) 1999
6 (3) 2001
P. occidentalis (Cresson); Arizona, Yavapai, Chino Valley 3447'N 11227'W 1,410 10 (4) 2001
Pogonomyrmex barbatus complex:
Pogonomyrmex rugosus Emery; Arizona, Maricopa,
1.8 km southwest of McCartney Road and 1-10 3256'N 111042'W 430 14 (8) 1999
11 (4) 2001
Pogonomyrmex anergismus Cole; New Mexico, Grant, 6
km east of Separ 3210'N 108022'W 1,375 5 (1) 1998
Note. Taxonomy follows Bolton (1995). Voucher specimens are deposited in the R. A. Johnson collection, Tempe, Arizona.
Number of alate queens (number of colonies) used to measure storage proteins.
Passera (1989) and Johnson et al. (1996) both showed a strong
association between founding strategy and fat reserves. Inde-
pendent fully claustral foundresses had the greatest relative fat
stores, followed by independent semiclaustral foundresses, fol-
lowed by dependent foundresses. However, dependent-found-
ing queens still contained significant fat reserves. Dependent
and semiclaustral queens do not require fat for brood provi-
sioning, but they do require significant energy reserves to sup-
port mating, dispersal, and either digging nests or entering host
colonies.
Fat reserves support multiple metabolic functions in queens
besides provisioning brood, and body size is related to nu-
merous functions, such as increased survival or dispersal ability
(Nylin and Gotthard 1998; Brown and Bonhoeffer 2003).
Therefore, protein reserves may be more closely linked to the
colony-founding method and may provide a better predictor
than either of these two traits. Ant queens require significant
amounts of amino acids to rear their first brood (Wheeler and
Buck 1995, 1996). Semiclaustral foundresses can obtain amino
acids through foraging, and dependent foundresses can obtain
them from their hosts, but fully claustral foundresses are limited
to internal amino acid reserves accumulated during maturation
in their natal colony. These amino acids come from two sources,
histolysis of the flight muscles in the thorax and storage proteins
in the hemolymph and fat body. Histolysis of the flight muscles
begins immediately after queens locate a nest site, and the
amino acids released are used to produce brood (Janet 1907;
Wheeler and Buck 1996).
Storage proteins are large molecular-weight proteins asso-
ciated with metamorphosis and reproduction in numerous in-
sect species (Telfer and Kunkel 1991; Wheeler and Martinez
1995; Seo et al. 1998; Pan and Telfer 1999, 2001). In ants, these
proteins occur in larvae, where they are used during meta-
morphosis; in queens, these proteins are used for egg produc-
tion and brood provisioning during colony founding and oc-
casionally in workers (Wheeler and Buck 1995; Wheeler and
Martinez 1995). Wheeler and Buck (1996) demonstrated that
storage protein reserves equal or surpass the flight muscles as
an amino acid source during founding. Therefore, storage pro-
teins are essential for a fully claustral queen to rear her first
brood.
We tested the hypothesis that storage protein content is as-
sociated with colony-founding strategy by quantifying storage
proteins in alate queens of five harvester ant species differing
in founding strategies (Table 1). We predicted that storage pro-
tein content would increase as the method of colony founding
changes from dependent to obligately semiclaustral to facul-
tatively semiclaustral to fully claustral. Queens of Pogonomyr-
mex rugosus Emery and Pogonomyrmex maricopa Wheeler are
fully claustral, producing their first brood of workers solely
from internal reserves (Johnson 2002). Pogonomyrmex occiden-
talis (Cresson) queens are facultatively semiclaustral; queens
102 D. A. Hahn, R. A. Johnson, N. A. Buck, and D. E. Wheeler
Table 2: Summary of ecological and physiological data related to nest founding for the species of Pogonomyrmex
used in this study
Method of Nest Queen Queen Dry Queen : Worker Fat
Species Founding Foraging Mass (mg) Mass Ratio Content (%)
Pogonomyrmex rugosus Independent No 24.9 + .3 4.6 + .2 40.5 + .8
Pogonomyrmex maricopa Independent No 16.4 + .4 5.1 + .2 41.8 + .7
Pogonomyrmex occidentalis Independent Yes, facultative 13.6 + .3 5.2 + .3 45.4 + .6
Pogonomyrmex californicus:
Arizona population Independent Yes, obligate 5.7 + .3 2.8 + .1 31.8 + .8
California population Independent Yes, obligate 6.2 + .4 3.8 + .2 39.0 + .9
Pogonomyrmex anergismus Dependent No 4.3 + .8 NA 27.2 + 1.9
Note. Data are presented as means + 1 SE. From Johnson et al. (1996), Johnson (2002), and R. A. Johnson, unpublished data. NA = not applicable
because P. anergismus is workerless.
forage in the field but are capable of rearing the first brood
solely from stored reserves (Billick et al. 2001; R. A. Johnson,
unpublished data). Pogonomyrmex californicus (Buckley)
queens are obligately semiclaustral. In Arizona populations, P.
californicus queens found colonies singly and must forage to
produce brood. Conversely, in a Southern California popula-
tion, queens found colonies in groups where one or several
individuals forage (Rissing et al. 2000; R. A. Johnson, unpub-
lished data). We examined queens from both populations to
examine potential intraspecific variation associated with dif-
ferences in founding. As an example of a dependent founding
species, we examined queens of the workerless social parasite
Pogonomyrmex anergismus Cole, which found colonies by en-
tering a mature colony of P rugosus or Pogonomyrmex barbatus
and duping the workers into rearing their brood (Cole 1968;
Johnson 1994).
Material and Methods
Specimen (..//II ,;.;
Mature alate queens of five species of Pogonomyrmex were col-
lected from their natal nests during the mating flight season in
1999 and 2001 (Table 2). Because of rarity, our analyses in-
cluded queens from only one colony of Pogonomyrmex aner-
gismus; all other species were collected from multiple colonies
at multiple sites. Individuals from each nest were placed in 50-
mL centrifuge tubes with a piece of moistened paper towel to
prevent desiccation, taken to the laboratory, and frozen indi-
vidually at -70C within 36 h.
Sample Preparation and Electrophoresis
Queens were removed from the freezer, freeze-dried to constant
mass, weighed to the nearest microgram, and returned to
-70C until analysis. For electrophoresis, each queen was ho-
mogenized in 0.3 mL of Tris-buffered saline (20 mM Tris, 150
mM NaC1, 5 mM EDTA, pH 7.5) containing the following
protease inhibitors: leupeptin, antipain, chymostatin, aprotinin
(all at 17 gg/mL), 1.7 jg/mL pepstatin A, and 1 mM 4-(2-
Aminoethyl) benzenesulfonyl fluoride (AEBSF, an irreversible
serine protease inhibitor). Samples were homogenized in 1.5-
mL plastic microcentrifuge tubes with a plastic pestle attached
to a rotating shaft driven by a variable-speed motor. Each sam-
ple was ground for 60-90 s at 250 rpm and then centrifuged
at 12,000 g for 20 min at 40C. This procedure isolated soluble
proteins in the supernatant.
Storage protein content differed greatly among queens of the
five species. Consequently, we diluted the supernatant for each
species to bring it into a quantifiable range. Samples were di-
luted using the Tris-buffered saline protease inhibitor solution
as follows: 1 : 10 for Pogonomyrmex rugosus and Pogonomyrmex
maricopa, and 1 : 5 for Pogonomyrmex occidentalis. No dilution
was necessary for P. anergismus or Pogonomyrmex californicus
queens from Arizona or California. From these dilutions, 10
ML aliquots were taken, mixed with 20 ML of sample loading
buffer, and loaded onto gels.
Sodium dodecyl sulfate polyacrylamide gel electrophoresis
(SDS-PAGE) was used to denature native hexamerins into their
subunits for quantification. SDS-PAGE was performed follow-
ing Laemmli (1970) and adapted to 6%-15% gradient slab gels.
Gels were run at 20 mA constant current and stained with 0.1%
Coomassie Brilliant Blue R 250 dissolved in a 5: 4: 1 solution
of methanol, water, and acetic acid. Gels were destined in the
same solution without Coomassie. To quantify storage proteins,
gels were scanned at 633 nm using a laser densitometer (LKB
Ultrascan XL). Standard curves were generated using known
quantities of bovine serum albumin (BSA) that ranged from
0.2 to 4.0 jg. Each gel contained internal standards of 1.0 and
3.0 jg BSA to correct for gel-to-gel variation. Our focal proteins
separated poorly on the 10 x 8-cm SDS gels (Hoefer Mighty
Small II). Consequently, we estimated the amount of storage
proteins in each species by summing the densities of the two
storage protein bands.
Storage Protein and Colony Founding 103
VHDL-* M
GLX ---
e -
Hex ta
- 200
-- 116
-97
e 67
a
-- 45
as
Figure 1. A 6%-15% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) gradient gel, showing the two major storage
hexamerins in Pogonomyrmex queens. Both populations of Pogonomyrmex californicus and Pogonomyrmex anergismus contain significantly less
storage hexamerins than the fully claustral or facultatively foraging semiclaustral species. Each sample was extracted in 300 -ML buffer, and 10
gg of extract was loaded. Hex = storage hexamerins; GLX = high glutamine/glutamic acid storage protein; VHDL = very high-density lipo
protein; CA = California; AZ = Arizona.
Statistical Analyses of Among-Species Di;, 1.- .
We analyzed the data using two separate nested ANOVAs, one
for total storage protein content and one for percent storage
protein. Independent variables were nested in this order: spe-
cies, year, locale, and colony. An a posteriori Tukey's honestly
significant difference (HSD) test was used to assess species dif-
ferences. No transformations of the data were necessary to meet
the assumption of homogeneity of variances. Populations of P.
californicus from Arizona and California were considered sep-
arately in all cases.
Molecular Weights
Samples were run on large-format 20 x 16-cm SDS-PAGE 6%
15% gradient gels to obtain adequate separation of the subunits
of the proteins to estimate their molecular weights. We esti-
mated molecular weights of proteins in SDS-PAGE using stan-
dards in the high molecular weight calibration kit (Bio-Rad)
containing the following: myosin (200,000 kDa), galactosidase
(116,250 kDa), phosphorylase B (97,400 kDa), BSA (66,200
kDa), and ovalbumin (45,000 kDa).
Large format 20 x 16-cm 4%-12% native-PAGE gradient gels
were used to assess the molecular weights of the proteins in their
native (undenatured) state. For native-PAGE we used the buffer
system of Laemmli (1970), with SDS and beta-mercaptoethanol
omitted. We estimated molecular weights of native proteins using
the following standards obtained from Pharmacia: thyroglobulin
(669,000 kDa), ferritin (440,000), catalase (232,000 kDa), lactate
dehydrogenase (140,000 kDa), and albumin (67,000 kDa).
104 D. A. Hahn, R. A. Johnson, N. A. Buck, and D. E. Wheeler
Table 3: Amino acid composition of Pogonomyrmex storage hexamerins compared with values for the
average ant storage hexamerin and the average protein in the SWISS-PROT database
Average
Pogonomyrmex Pogonomyrmex Pogonomyrmex Average Protein
rugosus maricopa occidentalis An SWISS
Ant SWISS-
Amino Acid (mol %) Hex 1 Hex 2 Hex 1 Hex 2 Hex 1 Hex 2 Hexamerina PROT
Asparagine/aspartic acid 11.44 10.71 11.14 10.61 9.76 10.28 13.70 9.61
Glutamine/glutamic acid 8.84 9.23 8.88 9.24 9.19 10.45 9.20 10.40
Serine 5.64 6.42 6.36 5.09 5.21 5.57 6.10 7.08
Glycine 6.19 9.74 8.36 9.28 9.36 7.26 8.00 6.85
Histidine 5.27 2.15 4.83 1.50 2.51 5.22 2.60 2.24
Arginine 4.13 4.03 3.84 3.86 3.98 4.31 4.10 5.19
Threonine 5.22 4.54 4.99 4.6 4.53 5.22 4.50 5.58
Alanine 5.41 6.33 5.58 6.59 6.49 5.15 5.50 7.61
Proline 6.48 6.39 6.77 5.27 5.11 6.23 5.60 4.80
Lysine 6.03 5.98 5.39 5.91 5.79 5.31 6.00 5.97
Valine 8.49 5.42 7.90 6.78 7.54 7.98 6.20 6.61
Isoleucine 5.12 4.88 4.91 6.17 5.92 5.19 5.20 5.85
Leucine 9.23 9.73 9.47 10.73 11.15 10.22 9.20 9.53
Phenylalanine 5.70 6.81 4.88 7.54 6.90 5.29 6.60 4.10
Tyrosine 6.32 6.68 6.13 6.26 5.92 5.91 6.30 3.16
Values were taken from Wheeler and Buck (1995).
Amino Acid Composition
Purification of storage protein subunits was performed using
20 x 16-cm large-format 6%-15% SDS-PAGE gradient gels,
which facilitated the separation of the bands representing the
subunits of our two storage proteins. Proteins were electro-
phoretically transferred to a PVDF (polyvinylidene difluoride)
membrane (Problott). After staining with 0.1% Coomassie bril-
liant blue R 250, bands of interest were cut out. Amino acid
analysis was performed at the University of Arizona Biotech-
nology Core Facility using a dedicated Applied Biosystems
Model 420A Amino Acid Analyzer with automatic hydrolysis
(vapor phase at 160C for 100 min using 6M HCL) and pre-
column phenylthiocarbamyl-derivatization.
Statistical Analyses of Amino Acid Composition
Insect storage hexamerins generally contain higher than average
amounts of the aromatic amino acids phenylalanine and ty-
rosine (Telfer and Kunkel 1991). We assessed whether the pu-
tative Pogonomyrmex storage hexamerins contained higher than
average amounts of these two amino acids by plotting phen-
ylanine content against tyrosine content for each Pogonomyr-
mex protein. These values were contrasted with those of the
average ant hexamerin derived from Wheeler and Buck (1995)
and the average composition of all proteins entered in the
SWISS-PROT database (http://us.expasy.org/tools/pscale/
A.A.SWISS-PROT.html). The amino acid composition of most
proteins is astonishingly similar across numerous prokaryotic
and eukaryotic taxa (King and Jukes 1969). Therefore, we as-
sumed that the average amino acid composition across all taxa
in the SWISS-PROT database was representative of the average
amino acid composition of nonstorage hexamerin proteins in
harvester ants. Similarity was assessed graphically by generating
a 99% confidence interval ellipse around the Pogonomyrmex
data. Values falling outside of the 99% confidence interval el-
lipse were considered significantly different from the putative
Pogonomyrmex storage proteins. If the Pogonomyrmex proteins
were ant storage hexamerins, we would predict that the average
ant hexamerin would fall within the ellipse, whereas the average
protein from the SWISS-PROT database would fall outside the
ellipse. All statistical analyses were performed using the JMP
IN statistical package (SAS Institute 1996).
Results
Protein Identification
Known ant storage proteins fall into three classes: (1) hex-
amerins, which contain moderately high amounts of aromatic
amino acids (range = 7.3%-9.4%), (2) proteins that contain a
high amount of glutamine/glutamic acid (high GLX,
range = 20.3%-22.8%), and (3) very high density lipoproteins
(VHDL; Wheeler and Buck 1995; Wheeler and Martinez 1995).
The two major storage proteins found in Pogonomyrmex queens
were hexamerins. Queens of the fully claustral species Pogono-
myrmex rugosus and Pogonomyrmex maricopa and the facul-
tatively semiclaustral species Pogonomyrmex occidentalis had the
highest amounts of these two storage proteins (Fig. 1). In con-
Storage Protein and Colony Founding 105
2 3 4 5 6 7 8
% Composition Tyrosine
Figure 2. Composition of the aromatic amino acids, phenylalanine and
tyrosine, for the storage hexamerins from Pogonomyrmex rugosus, Po
gonomyrmex maricopa, and Pogonomyrmex occidentalis (filled circles);
the average ant hexamerin (x) from Wheeler and Buck (1995); and
the average protein in the SWISS-PROT database (plus sign). The black
ellipse is the 99% confidence interval for the Pogonomyrmex data;
points represent each of the two bands from the three species.
trast, both storage hexamerins were much less abundant in both
populations of the obligately semiclaustral Pogonomyrmex cal-
ifornicus, and the dependent-founding social-parasite Pogono-
myrmex anergismus. High GLX and VHDL storage proteins
occurred very sporadically in queens of these five species, al-
though apoproteins of the appropriate size for each of these
classes of proteins can be observed in the P. rugosus queen in
Figure 1. Therefore, we conclude that these classes of storage
proteins are unlikely to be significant factors in the physiology
of colony founding in these ants and are not considered further.
Molecular Weights
Mobility of the two focal proteins relative to known standards
on denaturing SDS-PAGE gels yielded an estimated molecular
weight for the apoproteins of 79 and 80 kDa, whereas the native
proteins were both approximately 500 kDa, typical sizes for
both subunits and native insect hexamerins, respectively (Telfer
and Kunkel 1991; Wheeler and Buck 1995).
Amino Acid Composition
The mol percentage amino acid compositions for the putative
storage hexamerins from three of our five species P rugosus, P
maricopa, and P occidentalis are listed in Table 3. Pogonomyrmex
californicus and P anergismus queens contained insufficient
amounts of storage proteins for compositional analysis. The
average protein fell well outside of the 99% confidence ellipse
for our Pogonomyrmex storage hexamerins, with the latter con-
taining a much higher tyrosine content and a slightly higher
phenylalanine content (Fig. 2). In contrast, the aromatic amino
acid compositions of our Pogonomyrmex storage hexamerins
encompassed the average value for phenylalanine and tyrosine
content of the known ant hexamerins (Fig. 2). Therefore, the
aromatic amino acid compositions of our Pogonomyrmex pro-
teins are consistent with the storage hexamerins of other ant
species.
Protein Quantification across Species
The full-model nested ANOVAs were highly significant and
explained >95% of the variance in both total and percent stor-
age protein content (Table 4). Species was the largest contrib-
utor to both models, with species differences strongly associated
with the colony-founding method. Storage protein content was
lowest in the dependent-founding parasite P. anergismus and
in both populations of the obligately foraging semiclaustral P
californicus, intermediate in the facultatively foraging semi-
claustral P occidentalis, and highest in the fully claustral P
rugosus and P maricopa (Fig. 3). While P rugosus and P. mar-
icopa did not differ in percentage dry weight storage protein,
the larger queens of P. rugosus contained significantly higher
total storage hexamerin content than P. maricopa (Fig. 3). Col-
ony was also a significant effect in both analyses, although it
Table 4: Results of nested ANOVAs for total storage protein and percent storage protein per queen
Total Storage Protein per Queen Queen Dry Mass Storage Protein (%)
Source df F P Whole Model (%) df F P Whole Model (%)
Whole model 71 41.4 <.001 100 71 28.0 <.001 100
Species 5 240.5 <.001 77.8 5 186.4 <.001 75.0
Year (species) 4 7.2 <.001 1.8 4 .7 .599 .2
Locale (species, year) 3 .9 .406 .2 3 4.0 .014 1.0
Colony (species, year, locale) 59 5.3 <.001 20.2 59 5.0 <.001 23.8
Error 46 46
Whole model, Whole model,
Total 117 r2 = .99 117 r2 = .98
106 D. A. Hahn, R. A. Johnson, N. A. Buck, and D. E. Wheeler
A A
6 0
3b.
C C
3- B
2-
A *
E U
1 l 7
Species
Figure 3. Total mass (a) and percentage dry mass (b) of storage hex
amerins in Pogonomyrmex queens (adjusted means from the nested
ANOVAs 1 SE). Error bars are subsumed within the points for some
groups. Significant differences among groups are denoted by the letters
A-D: D > C > B > A. Groupings are based on a nested ANOVA fol
lowed by Tukey's HSD test. See Table 1 for sample sizes. Mode of
colony founding is given beside the name of each species. CA =
California; AZ = Arizona.
explained far less of the full model variance than did species
(Table 4). Year and collection locale contributed little to either
whole model, and their effects differed between the two analyses
(Table 4).
Discussion
S",, Protein and Mode of Colony Founding among
Pogonomyrmex Species
Ours is the first study to quantify storage protein content in
ant queens in relation to colony-founding strategy. Queens of
the five harvester ant species in this study contained two major
hexameric storage proteins. Species and colony of origin had
large significant effects on storage protein content, accounting
for roughly three-quarters and one-quarter of the whole model
variance.
In agreement with our predictions, the fully claustral Po-
gonomyrmex rugosus and Pogonomyrmex maricopa queens con-
tained the highest amounts of storage proteins. Facultatively
foraging semiclaustral Pogonomyrmex occidentalis queens con-
tained an intermediate amount of storage proteins, and obli-
gately foraging Pogonomyrmex californicus queens from both
the single foundress Arizona population and the multiple
foundress California population and the dependent-founding
social parasite Pogonomyrmex anergismus contained the lowest
amounts. The low storage protein levels in P californicus and
P. anergismus queens maybe indicative of the minimum protein
reserves necessary for queens to produce eggs, when feeding
larvae from internal reserves is not necessary.
Johnson (2002) found that P. californicus queens contained
less fat reserves than claustral-founding P. maricopa and P ru-
gosus queens or facultative semiclaustral P occidentalis queens
(Table 1). Although both fat and storage protein content cor-
relate with colony-founding mode, queens of all founding types
store significant fat reserves, probably for use as fuel in flight
and nest excavation. Therefore, storage protein content is a
better predictor of colony-founding mode in Pogonomyrmex
(Table 1; Fig. 3).
In addition to containing less stores, P. californicus and P
anergismus queens are small compared with their claustral-
founding congeners, and P. californicus has a significantly
smaller queen-to-worker mass ratio than the other Pogono-
myrmex species in this study. Pogonomyrmex californicus and
P. anergismus are in separate species groups within the genus,
indicating that small body size and poor queen provisioning
have evolved at least twice in association with nonclaustral
founding. Interestingly, small queen size and poor provisioning
are associated with divergent life histories in these two species;
one is an obligate forager and the other a social parasite.
Potential Advantages of Nonclaustral Founding
Large queens rich in nutritional stores may enjoy increases in
survival in the perilous founding period by avoiding foraging-
associated mortality. However, trade-offs may exist between
increased founding survival and the costs of producing large
well-provisioned queens. Dependent or independent semiclaus-
tral founding species can produce smaller sexual provisioned
with less fat and protein. Therefore, colonies could produce
more sexual for a given amount of resources. The colony-level
reproductive advantage of producing more queens could have
contributed to the evolution of small body size and minimal
stores, followed by queen foraging in P. californicus.
Another potential selective advantage of queen foraging is
the ability to break the size/number trade-off inherent to queens
producing their first brood from limited internal reserves. In
claustral-founding species, individuals in the first cohort are
usually significantly smaller than later cohorts (Porter and
Tschinkel 1986; H6lldobler and Wilson 1990). If some mini-
mum number of workers is necessary for successful colony
Storage Protein and Colony Founding 107
founding, queens with fixed reserves will have to trade off
worker number and size. While P. californicus queens are
smaller and contain less fat and protein reserves than all the
other species, when provided access to food, they produced
more and proportionally heavier brood than their fully claustral
congeners, breaking the size/number trade-off inherent in
claustral founding (Johnson 2002). If this ability to break the
size/number trade-off was selected for, foraging could release
P. californicus queens from requiring large nutrient reserves and
a large body to contain them, allowing the evolution of smaller
body size and scant stores.
Alternatively, the evolution of small body size and scant re-
serves in P. californicus could have been the product of some
other selective force on founding strategy, such as low mortality
during the queen foraging period (Brown and Bonhoeffer
2003). In Arizona, P. californicus queens generally initiate mat-
ing flights and begin founding 6-8 wk earlier than their con-
geners in the same habitats (R. A. Johnson, unpublished data).
This difference in timing may be associated with differences in
foraging mortality that would contribute to the evolution of
semiclaustral founding in P californicus but not its later found-
ing congeners (Brown and Bonhoeffer 2003).
Other Factors ,\ ,. Si -. i,, Protein Content
The significant effect of colony on storage protein content
showed there is notable colony-level variation in queen pro-
visioning, which could have implications for the evolution of
reproductive strategies both within and between populations.
Alternatively, colony-level effects could be due to differences
in the maturation of queens between colonies. Nutrient reserves
are accumulated by queens after eclosion and before flying
(Boomsma and Isaaks 1985; Nielsen et al. 1985). All the alates
collected in this study appeared to be mature, and mating flights
had begun in the populations at the time of collection. How-
ever, asynchrony in mating flight schedules between ant col-
onies within a population is common, and it is possible that
some of the colonies in the study had not completed provi-
sioning their sexual (H6lldobler and Wilson 1990). In addi-
tion, year and collection locale had relatively small effects in
our model, but because each factor had a small but significant
effect on storage protein content, these factors also merit fur-
ther study.
Acknowledgments
The authors would like to thank Allen Gibbs and two reviewers
for commenting on earlier versions of the manuscript. D.A.H.
would like to thank Jennifer Weeks for providing critical logistic
support during this study. During this work, D.A.H. was sup-
ported by a National Science Foundation Research Training
Group in the Analysis of Biological Diversification grant to the
University of Arizona (DBI-9602246).
Literature Cited
Billick I., D.C. Wiernasz, and B.J. Cole. 2001. Recruitment in
the harvester ant, Pogonomyrmex occidentalis: effects of ex-
perimental removal. Oecologia 129:228-233.
Bolton B. 1995. A New General Catalogue of the Ants of the
World. Harvard University Press, Cambridge, Mass.
Boomsma J.J. and J.A. Isaaks. 1985. Energy investment and
respiration in queens and males of Lasius niger. Behav Ecol
Sociobiol 18:19-27.
Bourke A.FG. and N.R. Franks. 1995. Social Evolution in Ants.
Princeton University Press, Princeton, N.J.
Brown M.J.. and S. Bonhoeffer. 2003. On the origin of claustral
colony founding in ants. Evol Ecol Res 5:305-313.
Cole A.C. 1968. Pogonomyrmex Harvester Ants: A Study of the
Genus in North America. University of Tennessee Press,
Knoxville.
Holldobler B. and E.O. Wilson. 1990. The Ants. Harvard Uni-
versity Press, Cambridge, Mass.
Janet C. 1907. Histolyse, sans phagocytose, des muscles vibra-
teurs du vol, chez les reines des Fourmis. C R Acad Sci 144:
393-396.
Johnson R.A. 1994. Distribution and natural history of the
workerless inquiline ant Pogonomyrmex anergismus Cole
(Hymenoptera: Formicidae). Psyche 101:257-262.
2002. Semi-claustral colony founding in the seed-
harvester ant Pogonomyrmex californicus: a comparative anal-
ysis of colony founding strategies. Oecologia 132:60-67.
Johnson R.A., J.D. Parker, and S.W. Rissing. 1996. Rediscovery
of the workerless inquiline ant Pogonomyrmex colei and ad-
ditional notes on natural history. Insectes Soc 43:69-76.
Keller L. and L. Passera. 1989. Size and fat content of gynes in
relation to the mode of colony founding in ants. Oecologia
80:236-240.
King T. and J. Jukes. 1969. Non-Darwinian evolution. Science
164:7888-7898.
Laemmli U.K. 1970. Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature 227:680
685.
Nielsen M.G., N. Skyberg, and G. Peakin. 1985. Respiration in
the sexual of the ant Lasius flavus. Physiol Entomol 10:199
204.
Nylin S. and K. Gotthard. 1998. Plasticity in life history traits.
Annu Rev Entomol 43:63-83.
Pan M.L. and W.H. Telfer. 1999. Equivalence of riboflavin-
binding hexamerin and arylphorin as reserves for adult de-
velopment in two Saturniid moths. Arch Insect Biochem
Physiol 42:138-146.
2001. Storage hexamerin utilization in two lepidop-
108 D. A. Hahn, R. A. Johnson, N. A. Buck, and D. E. Wheeler
terans: differences correlated with the timing of egg for-
mation. J Insect Sci 1:1-9.
Porter S.D. and W.R. Tschinkel. 1986. Adaptive value ofnanitic
workers in newly founded red imported fire ant colonies.
Ann Entomol Soc Am 79:723-726.
Rissing S.W., R.A. Johnson, and J.W. Martin. 2000. Colony
founding behavior of some desert ants: geographic variation
in metrosis. Psyche 103:95-101.
Rippell 0. and J. Heinze. 1999. Alternative reproductive tactics
in females: the case of size polymorphism in winged ant
queens. Insectes Soc 46:6-17.
SAS Institute. 1996. JMP IN. Version 3.2.1. SAS Institute, Cary,
N.C.
Seo S.J., Y.J. Kang, H.M. Cheon, and H.R. Kim. 1998. Distri-
bution and accumulation of storage protein-1 in ovary of
Hyphantria cunea Drury. Arch Insect Biochem Physiol 37:
115-128.
Stille M. 1996. Queen/worker thorax volume ratios and nest-
founding strategies in ants. Oecologia 105:87-93.
Telfer W.H. and J.G. Kunkel. 1991. The function and evolution
of insect storage hexamers. Annu Rev Entomol 36:205-228.
Wheeler D.E. and N.A. Buck. 1995. Storage proteins in ants
during development and colony founding. J Insect Physiol
41:885-894.
1996. Depletion of reserves in ant queens during claus-
tral colony founding. Insectes Soc 43:297-302.
Wheeler D.E. and T. Martinez. 1995. Storage proteins in ants.
Comp Biochem Physiol 112B:15-19.
|