Behavioral Ecology Vol. 11 No. 4: 357-366
© 2000 International Society for Behavioral Ecology
Timing of oviposition and reproductive skew in cobreeding female burying beetles (Nicrophorus vespilloides)
a Zoologisches Institut (Biologie I) der Universität, Hauptstr. 1, D-79104 Freiburg, Germany, and b Department of Biological Sciences, Behavior, Ecology, Evolution and Systematics Section, Illinois State University, Normal, IL 61790-4120, USA
Address correspondence to A.-K. Eggert, Department of Biological Sciences, Illinois State University, Normal, IL 61790-4120, USA. E-mail : aegger@ilstu.edu
Received 18 April 1999; revised 4 October 1999; accepted 20 October 1999.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Burying beetles (genus Nicrophorus) are known for their elaborate parental care. Two or more conspecific females may reproduce on the same carcass, especially when the carcass is large. Here we present the results of experiments in which we observed patterns of larval hatching and parental care in unmanipulated cobreeders, manipulated hatching synchrony between cobreeders, and compared patterns of oviposition in cobreeding and single females. Our results show that in these cobreeding associations, one of the females may or may not monopolize the carcass during the period of larval hatching. We present evidence that in either case, infanticide based on temporal cues constitutes an important proximate mechanism underlying the observed reduction in average reproductive success in cobreeding females. Females with higher synchrony (i.e., greater overlap between their oviposition patterns) produce larger broods with lower reproductive skew. Cobreeding females oviposit later and less synchronously than single breeders. Such delayed oviposition may reduce the risk that a female's larvae fall victim to cannibalistic acts committed by her cobreeder or maximize her own opportunity to selectively kill her cobreeder's larvae.
Key words: burying beetles, cannibalism, communal breeding, infanticide, kin recognition, Nicrophorus, parental care, reproductive competition, reproductive skew.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Reproductive competition is ubiquitous among communally breeding animals. When several conspecific females share a breeding site, as is the case in some vertebrates ("joint nesting" or "communal breeding") as well as in some primitively social or eusocial insects (polygyny), young in the communal nest may compete for food. In this situation, females may increase their individual reproductive success by destroying their cobreeders' eggs or young while sparing their own. Such selective infanticide requires the perpetrator to discriminate between her own offspring and those of cobreeders. One simple mechanism that can enable such discrimination is a temporal switch (Elwood, 1994
), sometimes
referred to as a "change in parental state"
(Rosenblatt and Siegel, 1983),
which is possibly mediated by hormonal changes triggered by important
reproductive events ("state-dependent cues";
Elwood, 1991). Selective
infanticide may result if the individual shows infanticidal behavior toward
young at a time at which it is likely to encounter unrelated offspring, and
parental, non-infanticidal behavior at a time when it is likely to encounter
its own (Elwood, 1994).
Examples include many bird species with intraspecific brood parasitism or
communal breeding, in which females remove eggs from the nest throughout the
prelaying period (i.e., until they have laid their first egg) (e.g.,
Elmen and Wrege, 1986 ;
Macedo and Bianchi, 1997 ;
Møller, 1987 ;
Mumme et al., 1983 ;
Stouffer et al., 1987 ;
Vehrencamp, 1977), and males
of many mammalian species that exhibit infanticidal behavior toward the young
of females with whom they have not mated or cohabited long enough for the
young to be the males' own offspring (see
Hrdy and Hausfater, 1984, for
examples). Here we report the reproductive consequences of one such temporal
switch in cobreeding pairs of female burying beetles Nicrophorus
vespilloides.
Natural history of burying beetles
Burying beetles (genus Nicrophorus) are known for their elaborate
parental care which involves direct interactions between the adult beetles and
their developing young (Pukowski,
1933). In N. vespilloides, the focus of this study,
adults fly in the afternoon and evening in search of the carcasses of small
vertebrates
(Müller and
Eggert, 1987). Upon locating a carcass, single females proceed to
bury it without further delay, while lone males typically emit a pheromone
that attracts conspecific females (Eggert
and Müller, 1989 ;
Pukowski, 1933). Usually, a
male and a female participate in the burial of the carcass and the subsequent
care of larvae. Females may also perform these activities without the
assistance of a male, using sperm stored in their spermatheca to fertilize
their eggs (Eggert, 1992). The
adults prepare the buried carcass by removing fur or feathers, rolling it up
into a ball, and smearing the surface of the carrion with anal excretions
(Pukowski, 1933). During this
time, the female's ovaries grow rapidly, doubling or tripling their
prebreeding mass in 24 h or less (N. orbicollis :
Trumbo et al., 1995 ;
Wilson and Knollenberg, 1984 ;
N. vespilloides : Müller JK,
unpublished data). Female N. vespilloides lay their eggs scattered in
the soil around the carcass. After hatching, the larvae crawl to the carcass ;
the adults chew a small hole into the carrion ball within which the young
reside and where they are fed regurgitated carrion by their parents. The
adults defend the carcass and their young against predators or other burying
beetles attempting to take over the carcass
(Pukowski, 1933 ;
Scott, 1990 ; Trumbo,
1990a,
b).
Conspecific female N. vespilloides arriving on the same unburied
carcass fight repeatedly and violently. The larger female wins most of these
fights (Bartlett and Ashworth,
1988 ;
Müller et
al., 1990a ; Otronen,
1988) and repels her competitor from the carcass whenever
possible, thus severely curtailing access to the carcass by the smaller female
(Müller et
al., 1990a). In this situation, smaller females typically leave
the carcass soon after oviposition and do not provide any care for the young
(Eggert and
Müller, 1992 ;
Müller et
al., 1990a).
Joint breeding in burying beetles
Recently, several studies have documented that interactions between
conspecific Nicrophorus females are less hostile when the carcass is
large for the respective species. Fights are less frequent or absent, and two
cobreeders remain with the carcass for more than 5 days
(Eggert and
Müller, 1992 ;
Scott and Williams, 1993 ;
Trumbo, 1992 ;
Trumbo and Wilson, 1993), well
after the time that larvae typically start to hatch and appear on the carcass
(Eggert and
Müller, 1992). Females are
frequently found on the carcass together
(Eggert and
Müller, 1992 ;
Trumbo, 1992 ;
Trumbo and Wilson, 1993),
occasionally feeding the joint brood side-by-side
Eggert and
Müller, 1992 ;
Trumbo and Wilson, 1993).
Nonetheless, there is evidence for reproductive competition in these
apparently peaceful coalitions of females. In a laboratory study of European
N. vespilloides (Eggert and
Müller, 1992) and two field studies
of North American N. defodiens
(Trumbo and Fiore, 1994) and
N. tomentosus (Scott,
1994), two females cobreeding on large carcasses consistently
produced fewer offspring per female than did single females. On medium-sized
carcasses, joint broods were often skewed in favor of the larger cobreeder
(Müller et
al., 1990a ; Scott and
Williams, 1993), and, likewise, the reduction in reproductive
success in the field may have affected the smaller cobreeder exclusively
(Scott, 1994). In N.
vespilloides, however, broods on large carcasses were not significantly
skewed in favor of the larger female's offspring
(Eggert and
Müller, 1992), indicating that body
size is not the sole determinant of relative reproductive success on large
carcasses.
Based on our earlier work, we hypothesized that females may at least
occasionally commit selective infanticide of their cobreeder's offspring.
Female N. vespilloides reproducing on a carcass by themselves or with
a male accept larvae only during a limited window of time, which coincides
approximately with the period during which the female's own larvae hatch
("parental phase"). Outside this time window, they cannibalize any
larvae they encounter
(Müller and
Eggert, 1990). When females encounter larvae more than 20 h before
their own larvae hatch, they invariably kill these larvae, and some females
cannibalize larvae hatching 9-12 h before their own
(Müller and
Eggert, 1990). Based on these findings, we predicted that an
opportunity for temporally based cannibalism of larvae exists if cobreeding
females differ significantly in their oviposition times. Here we present the
results of laboratory experiments in which we (1) determined the timing of
larval hatching and observed the disappearance of larvae in cobreeding pairs
of females, (2) manipulated hatching synchrony to determine its effects on
various aspects of larval survival and female reproductive success, and (3)
compared the oviposition patterns of singly breeding and cobreeding
females.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Rearing conditions and egg dyes
Experimental beetles were laboratory reared, sexually mature adults aged 20-50 days. Unless otherwise mentioned, they were F1 offspring of beetles collected in the field at Freiburg, Germany (Mooswald, 48°00' N, 7°51'E). We kept beetles in small plastic containers half-filled with moist peat and fed them dead mealworms twice a week ; up to six same-sex siblings shared the same container. Beetles were kept at 20°C on a 16 h:8 h light : dark cycle. For 2-3 weeks before an experiment, females were fed small amounts of ground beef mixed with a fat-soluble dye. We used either 200 mg Sudan Red 7B or 400 mg Solvent Blue (available from BASF Germany, or Aldrich Chemicals) per 20 g of ground beef. These dyes are incorporated into the eggs at oviposition such that females produce bright pink and blue to turquoise eggs, respectively. This enabled us to identify the eggs of individual females in experiments. Preliminary experiments indicated that the dyes did not affect the timing of oviposition, female fecundity, hatching rate, or larval survival.
General experimental procedure
We kept females with unrelated males for 48 h before the start of
experiments to ensure acquisition of a store of viable sperm (see
M
ler and
Eggert, 1989). Immediately before an experiment was initiated, we
measured each female's pronotum width under a stereomicroscope equipped with
an ocular micrometer. Whenever possible, cobreeding females were matched in
size to minimize any size-related differences in oviposition time between
beetles, such as have been described for cobreeding female N.
tomentosus (Scott, 1997).
The carcasses used for breeding were large laboratory mice (30 ± 0.3 g)
that had been frozen and thawed immediately before an experiment. Experiments
were initiated 2 h before the end of the photophase, when beetles are actively
searching for carcasses. We first presented females with a carcass in a small
transparent plastic box (10 x 10 x 8 cm) for 10-15 min, thereby
ensuring that females encountered each other and that they discovered the
carcass simultaneously. The carcass, along with both females, was then
transferred to a larger transparent plastic box (20 x 20 x 10 cm)
half filled with moist peat and returned to the environmental chamber. Upon
burial of the carcass, containers were transferred to a dark chamber at
20°C. We checked each container under red light at 12-h intervals to
record the location of the two females.
The beetles and their carcass were transferred to a new container with clean peat when they had been on the carcass for 40 h. At this time, we collected the eggs from the old container by searching through small amounts of peat at a time with flexible forceps. Eggs were stored on moist tissue paper in small petri dishes in the same chamber as the beetles to ensure that eggs and females were exposed to similar temperatures, because the duration of embryonic development is highly temperature-dependent (Müller JK, unpublished data). The containers were rotated among the shelves of the environmental chamber every time the beetles or eggs were checked to reduce the effect of slight differences in temperature between boxes that might result from a vertical temperature gradient inside the chamber.
We checked eggs every 4 h beginning 64 h after the start of the experiment, and placed any new larvae on the carcass. To facilitate subsequent observations, the cobreeding females and their carcass were transferred to a clean container with only 2 cm of compacted peat when their first larva had hatched. Before adding larvae, we carefully checked the carcass, turning it with forceps to check for the presence of beetles or larvae underneath, if necessary. We recorded the location of both females relative to the carcass, whether larvae were present, and, if so, which instar(s) they were. Every 48 h, the beetles were again transferred to new containers, and the old ones searched for eggs. We removed females if they had been continuously absent from the carcass for 48 h.
Experiment 1 : Observations of unmanipulated cobreeders
The first experiment was designed to determine the oviposition pattern of
cobreeding females. Pairs in which one female died (n = 4) or laid
fewer than five eggs (n = 1) were excluded ; in addition, we excluded
one pair in which less than 30% of one female's eggs hatched, leaving us with
23 of an initial 29 pairs for analysis. The experiment followed the general
experimental procedure described above. The maximum difference in pronotum
width between cobreeders was 0.2 mm, or 2.9%. We marked females individually
by piercing their elytra with insect pins (000) in number-encoding patterns,
and for speedier identification during trials, the elytra of cobreeders were
marked, one with black, the other with white paint. Two color strains are
established in our laboratory : light beetles without any black band in the
posterior half of the elytra and dark ones with a black band that covers the
posterior half of the elytra. When males from these laboratory strains are
paired with wild-type females, their offspring are easily distinguishable (see
also Müller and Eggert, 1989), which enabled us
to ascertain the maternity of surviving larvae : In each pair of cobreeders,
one of the females had been inseminated by a male from one color strain and
the other female by a male from the other strain. All surviving offspring
could be clearly attributed to one or the other of the females.
Experiment 2 : Synchrony and brood sharing in manipulated pairs
In experiment 2, we assessed the effects of experimentally manipulated
hatching synchrony on larval survival and various other features of the
surviving brood. Synchrony was manipulated by exchanging partners between
cobreeding females so as to generate pairs with high synchrony (i.e., those
whose larvae hatched more or less synchronously), and pairs with low synchrony
(i.e., those whose larvae hatched at different times). Females were paired
with their final partners when the larvae of at least one female had started
to hatch. Because we were unable to anticipate each female's oviposition
pattern, we paired females whose larvae started to hatch within 12 h of each
other to generate synchronous pairs, and females in which the onset of
hatching differed by 24 h or more to generate asynchronous pairs. In our
analyses, we used the continuous variable "oviposition synchrony"
(see explanation of variables) instead of the two treatment groups because it
is a much more accurate descriptor of the females' relative oviposition
behavior.
Re-pairing cobreeders was possible because cobreeding females typically do
not recognize their partners individually. Trumbo and Wilson
(1993) observed that
cobreeding N. defodiens violently attacked nonbreeding females but
not their own nest mates. A study in our laboratory during which female N.
vespilloides were observed for 15 min after re-pairing demonstrated
similar attacks against nonbreeding replacement females but revealed the
absence of such aggressive interactions when replacements were taken from
other cobreeding associations (Beck,
1995). The re-pairing allowed us to generate systematic variation
in hatching synchrony that was independent of the early interactions of the
original cobreeding pairs because females were exchanged before the first
larvae were placed on the carcass.
In this experiment, we used females from the two color strains that had been inseminated by males of their own color strain (see above). Whenever possible, we re-paired females with partners from the other strain, which allowed us to establish the maternity of surviving larvae. All surviving offspring could be clearly attributed to one or the other of the eventual cobreeders, except in five cases in which we had paired females from the same strain. The experiment followed the same general experimental procedure as described above. Two separate series of observations were initiated with 22 and 30 pairs of females. For our analyses, we pooled data from both series for higher statistical power because we found no significant effect of series or the interactions between series and synchrony on any of the dependent variables (two-way ANOVA, p >.3 in all instances). In the analysis of total brood size, we used data from 40 pairs (in 12 pairs, 1 of the cobreeders had died, or laid no eggs, or all her eggs were unfertilized). In the analysis of brood sharing, another 5 pairs could not be used because they consisted of females of the same color, which precluded an analysis of maternity, leaving a sample size of 35 pairs for analyses of brood sharing. Despite the re-pairing, body size differences between final cobreeders were relatively small, with only 5 out of the 40 pairs differing by more than 12% in their pronotum width (median difference = 4.9%).
Experiment 3 : Onset and synchrony of oviposition in single breeders
and cobreeders
Experiment 3 was designed to assess oviposition strategies of cobreeding
females compared to single females. As in experiment 1, we used female
F1 offspring of field-collected beetles, this time inseminated by
unrelated males from the same group. The experimental design was as described
above, except that three experimental treatments were established. Treatment 1
consisted of females paired individually with males on carcasses (single
breeders, n = 40). If synchrony between cobreeders simply results
from interindividual variation in oviposition time regardless of the presence
or absence of another breeder, oviposition synchrony should be the same for
cobreeders as for random pairs (dyads) of separately breeding females. To
assess the degree of synchrony in the separately breeding females, each single
breeder was randomly assigned another single breeder within this treatment,
yielding a total of 20 dyads. Treatment 2 consisted of pairs of females
cobreeding on the same carcass (cobreeders, n = 20 pairs). Treatment
3 consisted of cobreeding groups each made up of two cobreeding females and a
male (cobreeders with males, n = 20 groups).
Females in cobreeding pairs differed in pronotum width by a median of 1.4% (treatment 2) and 3.0% (treatment 3). As before, they had ingested different colored dyes before being placed on the carcass, and the same was true for each of the randomly assigned dyads of single breeders. Larvae were added to their mother's carcass until a brood was established. We recorded the onset of larval hatching for each individual, as well as the degree of synchrony for each pair of females, including dyads of single breeders. In some replicates, one of the females died or failed to lay any eggs, leaving a total of 18 dyads of single breeders, 17 pairs of cobreeders, and 17 groups of cobreeding females with males.
Definition of variables
Several variables used in our analyses require some elaboration. The
oviposition differential of two cobreeders is the difference in the time at
which each female's larvae begin to hatch. For this variable and for the time
to oviposition in experiment 3, we used the time at which a female's first
three larvae had hatched rather than her first larva because single eggs are
occasionally laid much earlier than the remaining clutch. We first determined
the interval (in multiples of 4 h) from the time at which we observed one
female's third larva to the time at which we recorded her cobreeder's third
larva. For any observed 4-h difference, the true difference could be anywhere
from 1 to 7 h. If actual differences occur at the same frequency, 5/8 of
these should occur between 1 and 4 h, and the remaining 3/8 should fall
between 5 and 7 h ; the latter differences would more appropriately be
assigned to the 5- to 8-h interval. We estimated the actual hatching interval
between two clutches from the distribution of observed intervals by assigning
3/8 of the original observations for each interval to the next higher 4-h
interval.
Oviposition synchrony was defined as the area of overlap between the two
females' relative hatching curves. First we determined the proportion of each
female's offspring that hatched in each 4-h interval. If
p1,i, p2,i are the proportions of
female 1's and female 2's larvae that hatch in interval i,
respectively, then synchrony (S) is defined as the sum of the smaller of the
two values over all intervals i :
 |
).
Brood sharing describes how evenly the surviving brood is shared
genetically between two cobreeding females. If p1 and
p2 are the proportions of the surviving brood that are the
offspring of female 1 and female 2, respectively, then p1
+ p2 = 1. We define brood sharing as BS =
p1*p2*4 =
p1*(1 - p1)*4 =
4p1 - 4p12. Multiplying by
four standardizes this measure to vary between one (when both females
contribute to the surviving brood equally) and zero (when only one of them
leaves offspring). This measure of reproductive partitioning in a group of two
breeders equals one minus the index of reproductive skew developed by Reeve
and Keller (Keller and Vargo,
1993 ; Reeve and Ratnieks,
1993).
Brood sharing is a variable that assesses the number of each female's larvae that survive without considering how many of each female's larvae initially arrived on the carcass. In a hypothetical brood in which all larvae survive but one of the females produces a larger number of first-instar larvae, BS < 1 despite the fact that all larvae survive. As a measure of survival that is independent of the number of larvae initially produced, we define the index of shared survival as SU = s1*s2, in which s1 and s2 are the survival rates of first-instar larvae produced by females 1 and 2, respectively, calculated as the number of a female's offspring surviving divided by the number of this female's first-instar larvae placed on the carcass. Like our brood sharing index, this index varies between zero and one.
To describe differential fecundity between females, or rather, differences in the numbers of their first-instar larvae that arrive on the carcass, we use an index similar to brood sharing. We define egg sharing as ES = e1*e2* 4, e1 and e2 being the proportions of all first-instar larvae produced by female 1 and 2, respectively ; the constant (4) again standardizes the measure to vary between zero and one.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Observations of unmanipulated cobreeders : experiment 1
Oviposition differentials and synchrony
To assess the opportunity for temporally based infanticide, we determined the oviposition differential and synchrony (for definitions, see Methods) of cobreeding pairs. The oviposition differential was frequently quite long (median = 26.5 h, first quartile = 8.5 h, third quartile = 50 h, n = 23, see Figure 1). We estimated that five pairs started ovipositing within 8 h of each other, whereas in the remaining 18 pairs the hatching differential was longer than 8 h. Based on earlier observations of single females (Müller and Eggert, 1990
), this should have been sufficient to allow for some
temporally based cannibalism. The distribution of oviposition synchrony did
not deviate significantly from normality (Shapiro-Wilk W = 0.930,
p =.11, n = 23). Synchrony values were well below 1 in all
pairs, with a mean of 0.20 (SD = 0.15, minimum = 0, maximum = 0.51, n
= 23).
|
Female presence on the carcass and disappearance of larvae
We made no systematic attempt to observe acts of infanticide directly, but
on two separate occasions, we saw females consume larvae newly arriving on a
carcass. Neither did we attempt to follow up on the fate of individual larvae,
but our visual inspections at 4-h intervals frequently revealed that larvae
placed on carcasses had disappeared 4 h later.
With respect to the cobreeders' presence on the carcass, we observed two discretely different situations, which we describe as exclusive and nonexclusive pairs. In our cobreeding associations, larvae hatched over a median period of 20 intervals (first quartile = 16.25, third quartile = 25.75 intervals, n = 23). In exclusive pairs, one of the females (the "noncaring" female) was seen on the carcass during no more than two inspections during the period of larval hatching, indicating that the other female (the "caring" female) had excluded her and monopolized the carcass. Nonexclusive pairs, on the other hand, were those in which both females were found on the carcass for a minimum of five different observations during larval hatching. Intermediate values of three or four observations on the carcass were conspicuously absent. Severe injuries, involving the loss of tibiae, femora, and entire legs were rare but did occur in four exclusive and six nonexclusive pairs.
Exclusive pairs constituted about half of our sample (12/23). In 10 of these cases, live larvae were first observed on the carcass in the interval immediately following the one in which the caring female's own first larvae had hatched and had been added to the carcass, indicating that the female started to accept larvae at exactly the time her own larvae had begun to hatch. In the two remaining cases, the caring female's own first larvae disappeared, but the larvae added 4 h later did not. In four cases larvae of the noncaring female that had hatched earlier than the caring female's disappeared from the carcass. In each of these cases, all early larvae disappeared, involving several batches of larvae that had been added at different intervals (Table 1). The disappearance of larvae after the caring female had begun to accept and feed larvae was evident only if larvae were added at a time when all the larvae present on the carcass had already molted into second instars ; this requires a gap of at least 16 h. We observed the disappearance of larvae that appeared on the carcass too late to be the caring female's offspring in four pairs (see Table 1) ; in two of these cases, the noncaring female left no surviving offspring. There were no instances in which both early and late larvae disappeared from the same carcass. Survival rates were higher for the caring female's larvae (range 19-100%, median = 70%) than for the noncaring female's larvae (range 0-89%, median = 19% ; Wilcoxon matched-pairs signed-rank test, p =.025).
|
In nonexclusive pairs, we frequently noted the disappearance of larvae as well. However, in this situation the disappearance did not simply coincide with the time at which one or both female's larvae started to hatch. In all of these pairs larvae still disappeared after the time at which both females' larvae had started to hatch (Table 1). In two instances, larvae disappeared after they had apparently been accepted (i.e., live larvae had been found in the crater on the carcass during one or several inspections). In all of the pairs we observed, more larvae disappeared after both females' larvae had started to hatch than when only one female's larvae had begun to hatch. We did not observe the disappearance of late larvae in any of these pairs. Overall survival rates of larvae in this group (24-43%, median = 33%, n = 11) were significantly lower than in exclusive pairs (39-62%, median = 53.5%, n = 12 ; Mann-Whitney U = 130, p <.0001). In some pairs, continued infanticide led to the extinction of both females' entire first clutches, forcing them to produce replacement clutches. Oviposition frequently continued (with interruptions) over 4 or 5 days ; in one pair, larvae continued to hatch over a period of 7 days.
Synchrony and partitioning of reproductive success
In unmanipulated cobreeders, the degree of oviposition synchrony was
significantly correlated with brood sharing (rs
=.571, p =.007, n = 23) but not with the index of shared
survival (rs =.310, p =.146, n
= 23).
Differential fecundity as a potential factor affecting brood
sharing
In addition to oviposition synchrony, initial differences in the number of
eggs laid, or the number of first-instar larvae produced, could also affect
the degree of brood sharing in our pairs. To assess the potential impact of
such factors on brood sharing, we determined partial rank correlations between
egg sharing, synchrony, and brood sharing. There was no significant
correlation between egg sharing and brood sharing when synchrony was held
constant (rs partial =.392, p =.064,
n = 23), no significant correlation between egg sharing and synchrony
when brood sharing was held constant (rs partial
=.121, p >.5, n = 23), but the correlation between
synchrony and brood sharing remained highly significant when egg sharing was
held constant (rs partial =.569, p
=.008, n = 23).
In exclusive pairs, access to the carcass appeared to affect differential fecundity but not the timing of oviposition : caring females laid more eggs (mean = 35, SE = 3.0, n = 12) than their noncaring partners (mean = 24, SE = 3.3, n = 12 ; paired t test : t = 2.22, p =.048) but did not oviposit earlier (time to oviposition in caring females : mean = 55.0 h, SE = 10.9 h, n = 12, in noncaring females : mean = 60.7 h, SE = 8.2 h, n = 12 ; paired t test : t = 0.582, p =.573).
Manipulations of synchrony : experiment 2
Synchrony and reproductive success of pairs
This experiment was designed to manipulate synchrony between cobreeders so
as to generate especially synchronous or especially asynchronous pairs and to
assess their subsequent reproductive success as individuals and as a group.
The majority of re-paired associations of cobreeding females were nonexclusive
(33/40), with the remaining pairs following the "exclusive"
pattern described above. As in the previous experiment, many of the larvae
added to the carcass disappeared. In several nonexclusive pairs (15/33),
entire broods vanished after they had apparently been accepted and larvae had
been present on the carcass for three or more consecutive inspections.
Due to our attempts to create pairs with high and low synchrony, synchrony values showed significant deviations from normality (Shapiro-Wilk W = 0.917, n = 40, p =.007), and thus the following correlations are nonparametric Spearman rank correlations. Synchrony of re-paired cobreeders varied between 0 and 0.82, with a median of 0.31 (n = 40). Synchrony was significantly correlated with shared survival (Figure 2a : rs = 0.763, p <.0001, n = 35), brood sharing (Figure 2b : rs =.741, p <.0001, n = 35), and the total number of offspring produced (Figure 2c : rs =.656, p <.0001, n = 40). Thus, when females were re-paired with other females whose timing of oviposition was similar to their own, both females' larvae experienced improved survival rates compared to females that were re-paired with asynchronous partners. More synchronous pairs were also more productive, and they partitioned reproduction more evenly (i.e., reproductive skew was lower).
|
Differential fecundity as a potential factor affecting brood
sharing
As in our first experiment, we determined partial rank correlations between
egg sharing, synchrony, and brood sharing. There was no significant
correlation between egg sharing and brood sharing when synchrony was held
constant (rs partial = -.028, p >.7,
n = 35), and no significant correlation between egg sharing and
synchrony when brood sharing was held constant (rs
partial =.013, p >.9, n = 35), but the correlation
between synchrony and brood sharing remained highly significant when egg
sharing was held constant (rs partial =.741,
p <.0001, n = 35). These results indicate that the
effects of the differential production of first-instar larvae on brood
sharing, or reproductive skew, were negligible.
Comparing oviposition in single breeders and cobreeders : experiment
3
The previous results indicate that conflicts between cobreeders would be
minimized, and overall reproductive success maximized, if females oviposited
synchronously. If the behavior of cobreeding females is designed to reduce
conflicts or maximize total reproductive success, they should oviposit more
synchronously than random combinations of single breeders. On the other hand,
if the behavior of cobreeders is designed to maximize individual opportunities
for selective infanticide, females should postpone oviposition because delayed
oviposition may enable them to recognize and selectively destroy their
cobreeder's offspring.
The observed times to oviposition showed significant deviations from normality (Shapiro-Wilk test, all p <.05), but synchrony values did not (all p >.10). Treatment had a significant effect on the time to oviposition (Kruskal-Wallis test, H2 = 36.64, p <.0001) : two cobreeding females initiated oviposition later than did single breeders (see Figure 3a). When only the first-laying of the two females in a pair or dyad was considered, this effect persisted (Kruskal-Wallis test, H2 = 19.93, p <.0001 ; Figure 3b). Females in cobreeding pairs clearly did not oviposit more synchronously than randomly assigned pairs of single breeders on the same carcass size, and if anything, they appeared to oviposit less synchronously (Table 2). The presence or absence of male appeared to have no effect on the timing of oviposition.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Disappearance of larvae and infanticide
Many of the first-instar larvae that arrived on large carcasses with cobreeding female N. vespilloides disappeared before larvae were finally accepted and cared for. We consider the disappearance of early larvae to be indicative of cannibalism by adults. N. vespilloides larvae that are accepted aggregate in a small crater created by the parents in the carrion ball and are thus readily visible. Larvae can survive without parental feeding (Eggert et al., 1998
),
and therefore a lack of parental care cannot explain their rapid
disappearance. Females that are not ready to accept larvae were observed to
cannibalize larvae in an earlier study
(Müller and
Eggert, 1990) as well as in the current one. A female
cannibalizing a larva picks it up from the carrion ball and lifts it up in her
mandibles, raising her head and thorax while lowering her abdomen in a
rearing-up posture we have not observed in other contexts. The larva is then
crushed between the female's mandibles and consumed. However, we are not
certain that females likewise selectively killed late larvae arriving on the
carcass after a brood had already been established. Late-arriving larvae may
be hard to see among their larger siblings, and sibling competition and the
miry condition of the carcass interior may contribute to their demise even in
the absence of adult intervention.
Parental phases in burying beetles
Parental phases, or parental windows of time, outside of which adults
cannibalize young coming to the carcass, are likely to occur in both sexes of
all burying beetle species. During the parental phase, there is no evidence of
any discrimination against nonkin ; unrelated conspecific larvae are accepted
and fed
(Müller and
Eggert, 1990 ; Trumbo and
Wilson, 1993), as are larvae of other species. For example, N.
vespilloides parents can be used to rear larvae of N. vespillo, N.
humator, N. defodiens, N. pustulatus, N. orbicollis, and even larvae of a
Japanese species in the sister genus of Nicrophorus, Ptomascopus
(Müller and
Eggert, 1990 ; Müller JK,
unpublished data). Trumbo
(1994) found that N.
orbicollis females accepted heterospecific larvae added during the
parental phase, and that they are susceptible to interspecific parasitism in
natural broods of the former. Female N. defodiens, N. sayi, and
N. marginatus all reject larvae arriving on the carcass at the wrong
time (Trumbo ST, personal communication), as do male N. vespilloides
and N. orbicollis
(Müller and
Eggert, 1990 ; Müller JK, Eggert
A-K, unpublished data ; Scott MP, personal communication).
Exclusive and nonexclusive cobreeders
Our observations revealed the existence of exclusive cobreeders even on the
large (30-g) carcasses, suggesting that noncaring females were denied access
to the carcass by caring females. Some of the noncaring females might have
chosen to abandon the carcass after oviposition if they had been given the
opportunity. However, severe injuries indicate that aggressive interactions
must have occurred even in nonexclusive associations. We do not understand all
the factors that affect the exclusive or nonexclusive nature of the
relationship between cobreeders, but carcass size clearly plays a role. On
medium-sized (15-g) carcasses, exclusive interactions between female N.
vespilloides appear to be the rule
(Müller et
al., 1990a). The smaller female is continually attacked by the
larger female and leaves the carcass after only a few days with little direct
access to the carcass, but some of her offspring are reared with the larger
female's brood ("intraspecific brood parasitism"). Associations on
large (30-g) carcasses are frequently nonexclusive : Both females frequently
feed young, sometimes side-by-side ("joint breeding" ;
Eggert and
Müller, 1992), as do females of
N. defodiens, N. orbicollis, and N. tomentosus on carcasses
that are large ("communal breeding" ;
Trumbo, 1992 ;
"cooperative associations" ;
Trumbo and Wilson, 1993). The
females' relative body sizes may also affect their relationship because a
large size discrepancy may facilitate monopolization of the carcass by the
larger female. A role for absolute body size is suggested by interspecific
differences in the tendency to exhibit communal breeding on carcasses of a
given size (Trumbo, 1992 ;
Trumbo and Wilson, 1993).
Switching from infanticidal to parental behavior in cobreeding
pairs
The timing of the switch from infanticidal to parental behavior in female
N. vespilloides is clearly affected by the social situation on the
carcass. In the absence of other females, this switch occurs well before the
female's own larvae hatch. The time at which these females accept larvae
varies from 20-17 h to 8-5 h before their own larvae hatch
(Müller and
Eggert, 1990) When another female has detected the same carcass
but can be denied access to the carcass, the switch seems to move to a
slightly later point in time. In such exclusive or brood parasitic pairs, the
caring or host female has the opportunity to selectively destroy larvae coming
to the carcass at the wrong time. In this situation, with its greater threat
of competition from another female's larvae, females start to accept larvae
closer to the exact time their own larvae hatch, occasionally killing some of
their own first larvae. When another female has continued access to the
carcass, the switch is further delayed, frequently well into the period during
which each female's own larvae hatch. In most nonexclusive pairs, larvae were
not allowed to survive until long after the time at which both females' larvae
started to hatch. Being unable to discriminate against unrelated offspring at
this time, females must frequently kill their own larvae along with those of
their cobreeder's. If the onset and continuation of parental care is regulated
by hormones such as juvenile hormone, as has been suggested by Trumbo
(1997), the hormone level
itself or its action on the respective neurons must be modified by cues
associated with the initial, or continued, presence of another female on the
carcass.
Synchrony and reproductive success
When cobreeding pairs oviposit more synchronously, larger proportions of
both females' larvae survive, reproduction is partitioned more evenly, and
overall productivity increases. We contend that this is due to reduced
opportunities for temporally based selective infanticide of unrelated larvae
under conditions of synchrony, regardless of the type of association between
females (exclusive or nonexclusive). The relative timing of oviposition, and,
concomitantly, larval hatching, plays an important role in determining the
reproductive skew and the reproductive output of the pair. Little is known
about the factors that determine the time at which a female begins to oviposit
or her specific oviposition pattern. The female's physiological condition at
the time she detects the carcass may play a role, and our data indicate that
social factors, such as the presence of a conspecific female, may have an
effect as well. We think that synchrony is likely to be a more or less
accidental and unpredictable consequence of both female's oviposition times
because the two females likely have no reliable information about each other's
oviposition pattern. However, the observed delay in oviposition relative to
single breeders suggests that the cobreeding female N. vespilloides
do not behave in ways that minimize conflict or maximize group reproductive
output. Instead, the observable behavior of cobreeding females likely results
from each female's selfish attempts to increase her own representation in the
surviving brood.
Timing of oviposition in cobreeders : an evolutionary game ?
Despite the fact that we have little knowledge about possible causes of
variation in female oviposition behavior, we suggest that oviposition in
cobreeding females might be viewed as an evolutionary game in which females
can choose the onset of oviposition and the length of the period over which
they oviposit but are limited in the number of eggs they can produce in a
certain time interval. Physiological constraints appear to limit the number of
eggs produced before hatching of the first larva in several
Nicrophorus species
(Müller et
al., 1990b ; Trumbo,
1992). Females could oviposit as soon as possible after carcass
detection, as late as the carcass remains usable as larval food, or at any
time in between these extremes. Larvae hatching early would be able to utilize
the carcass while it is in optimal condition, but they would be highly
vulnerable to infanticide by a later-laying female. Larvae from late clutches,
while relatively safe against infanticide, would have to cope with a carcass
of significantly reduced value. Spreading oviposition over a long period of
time virtually ensures that some offspring will survive, but it also entails
an exceptionally high risk of accepting the cobreeder's larvae. This tactic is
vulnerable to partners that lay their eggs in a short amount of time and are
then able to selectively kill earlier and later larvae. Shortening the
oviposition period affords the female a fairly precise means of identifying
her own larvae but entails a high risk of complete loss of this clutch. A
formal model might prove more fruitful than these considerations, which do not
reveal an optimal oviposition tactic.
Other factors affecting reproductive skew
Reproductive skew, or brood sharing, describes the final result of a joint
breeding attempt, which can be determined by different mechanisms. In burying
beetles, differential reproductive success of cobreeders can be influenced by
a whole host of factors (e.g., see Eggert
and Müller, 1997). Effects of the
breeding association (one or several females) on patterns of oviposition and
parental care as revealed in the present study were largely ignored in earlier
attempts to explain patterns of joint care in burying beetles
(Eggert and
Müller, 1992 ; Scott,
1994,
1997 ;
Trumbo and Wilson, 1993),
leading authors to believe that cobreeders oviposit more or less synchronously
and that infanticide, if it occurs at all, is restricted to brief, early
periods of a reproductive attempt (Eggert
and Müller, 1992 ;
Scott, 1994 ;
Trumbo and Wilson, 1993).
Indeed, Scott (1996,
1997) ruled out any role for
selective infanticide in the reproductive skew observed in cobreeding N.
tomentosus, attributing it instead to early behavioral interactions,
differential fecundity, and selective oophagy.
In the present study, factors affecting the number of firstinstar larvae on
a carcass, such as differential oophagy or differential fecundity, may have
had some effect on final reproductive skew, but such effects were weak
compared to the overwhelming impact of oviposition synchrony. Our experimental
design should have provided ample opportunity for selective ovicide, but
currently, we have no reason to assume that cobreeding N.
vespilloides exhibit this behavior. Even in N. tomentosus,
evidence for oophagy is indirect, and underlying mechanisms remain unclear
(Scott, 1996,
1997). Species differences are
likely because breeding associations of N. tomentosus may differ from
N. vespilloides in other respects as well : Females begin laying soon
after they have detected the carcass, both females finish laying within 48 h
after burial, and hatching is restricted to a period of about 24 h
(Scott, 1997).
We suspect that body size differences might have played a larger role had
our cobreeders been less closely matched in size. Size-related fecundity
differences have been demonstrated in N. vespilloides
(Bartlett and Ashworth, 1988),
and the outcome of aggressive interactions becomes highly predictable when
body size differences are more substantial
(Müller et
al., 1990a).
Müller et
al. (1990a) and Trumbo et al.
(1995) suggested that
dominance-related differential access to the carcass may have quantitative
effects on clutch size, an idea supported by the larger clutch sizes for
caring females in exclusive breeding associations in our first experiment. It
is likely that such differential access exacerbates size-related fecundity
differences. In N. tomentosus, reduced access to the carcass also
seems to delay the onset of oviposition
(Scott, 1997), but the present
study found no evidence of such a delay in N. vespilloides.
Consequences of infanticide : counter-adaptations and communal
breeding
Both larvae and adults may have developed counter-adaptations to larval
cannibalism, which occurs in cobreeding females (this paper), during or after
takeovers (Trumbo, 1990 a,
b), and in the context of
brood reduction (Bartlett,
1987). During time periods in which no surviving larvae were
observed on the carcass, we occasionally observed a few firstinstar larvae off
the carcass, moving around in the soil. We think that such larvae may have
escaped cannibalism by temporarily abandoning the carcass. In one instance,
our maternity data demonstrated the survival of several larvae that we had not
seen during one visual inspection. These larvae may have been concealed inside
the carcass, which could have provided some protection against infanticide.
The most important counter-adaptation for cobreeders, however, may be their
delayed oviposition. Such a delay likely entails significant costs due to the
deterioration of carcass quality and the consumption of carrion by adults.
Overall, cobreeding appears to be associated with significant costs to
females. Cobreeders still incur injuries, they may have to produce very large
numbers of eggs to compensate at least partially for the losses due to
cannibalism, and they may eventually produce fewer offspring because of low
oviposition synchrony. In view of the fact that potential benefits of communal
breeding would have to be quite large to outweight these costs, we maintain
that breeding associations are not preferable to females, but are merely a
consequence of the females' inability to monopolize large carcasses. We
suggest that the opportunity to bias the composition of the surviving brood
through larval infanticide constitutes a major incentive for a second female
to remain on the carcass.
|
ACKNOWLEDGEMENTS |
|---|
We thank Scott Sakaluk for comments on the manuscript, Jean-René Grèzes from the Institut für Chemie der Universität Freiburg for recommending and supplying the dyes, and Sabine Götter and Hilda Winterhalder for helping us search for thousands of red and blue beetle eggs. Many thanks also to Ursula Eggert, Donna, John, and Marion Sakaluk for their care of Hannah during research and manuscript preparation. The study was supported by grants from the Deutsche Forschungsgemeinschaft to J.K.M. (DFG Mu-1175/1-1 and Mu-1175/2-1).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Bartlett J, 1987. Filial cannibalism in burying beetles. Behav Ecol Sociobiol 21 : 179-183.
Bartlett J, Ashworth CM, 1988. Brood size and fitness in Nicrophorus vespilloides (Coleoptera : Silphidae). Behav Ecol Sociobiol 22 : 429-434.
Beck D, 1995. Erkennen Totengräber Nicrophorus vespilloides Herbst (Coleoptera : Silphidae) einander am Aas ? Diplomarbeit (MSc thesis). Freiburg : Universität Freiburg.
Eggert A-K, 1992. Alternative male mate-finding
tactics in burying beetles. Behav Ecol 3
: 243-256.
Eggert A-K, Müller JK, 1989. Mating success of pheromone-emitting Necrophorus males : do attracted females discriminate against resource owners ? Behaviour 110 : 248-258.
Eggert A-K, Müller JK, 1992. Joint breeding in female burying beetles. Behav Ecol Sociobiol 31: 237-242.
Eggert A-K, Müller JK, 1997. Biparental care and social evolution in burying beetles : lessons from the larder. In : The evolution of social behavior in insects and arachnids (Choe JC, Crespi BJ, eds). Cambridge, Massachusetts : Cambridge University Press; 216-236.
Eggert A-K, Reinking M, Müller JK, 1998. Parental care improves offspring survival and growth in burying beetles. Anim Behav 55: 97-107.[ISI][Medline]
Elwood RW, 1991. Parental states as mechanism for kinship recognition and deception about relatedness. In : Kin recognition (Hepper P, ed). Cambridge, Massachusetts : Cambridge University Press; 289-307.
Elwood RW, 1994. Temporal-based kinship recognition—a switch in time saves mine. Behav Proc 33: 15-24.
Emlen ST, Wrege PH, 1986. Forced copulations and intra-specific parasitism : two costs of social living in the white-fronted bee-eater. Ethology 71: 2-29.[ISI]
Hrdy SB, Hausfater G, 1984. Infanticide : comparative and evolutionary perspectives. New York : Aldine.
Keller L, Vargo EL, 1993. Reproductive structure and reproductive roles in colonies of eusocial insects. In : Queen number and sociality in insects (Keller L, ed). Oxford : Oxford University Press; 16-44.
Macedo RHF, Bianchi CA, 1997. When birds go bad : circumstantial evidence for infanticide in the communal South-American Guira Cuckoo. Ethol Ecol Evol 9: 45-54.
Møller AP, 1987. Intraspecific nest parasitism and anti-parasite behaviour in swallow, Hirundo rustica. Anim Behav 35: 247-54.
Müller JK, 1987. Replacement of a lost clutch : a strategy for optimal resource utilization in Necrophorus vespilloides. Ethology 76: 74-80.
Müller JK, Eggert A-K, 1987. Effects of carrion-independent pheromone emission by male burying beetles (Silphidae : Necrophorus). Ethology 76: 297-304.
Müller JK, Eggert A-K, 1989. Paternity assurance by "helpful" males : adaptations to sperm competition in burying beetles. Behav Ecol Sociobiol 24: 245-249.
Müller JK, Eggert A-K, 1990. Time-dependent shifts between infanticidal and parental behavior in female burying beetles : a mechanism of indirect mother-offspring recognition. Behav Ecol Sociobiol 27: 11-16.
Müller JK, Eggert A-K, Dressel J, 1990a. Intraspecific brood parasitism in the burying beetle Necrophorus vespilloides Herbst (Coleoptera : Silphidae). Anim Behav 40: 491-499.
Müller JK, Eggert A-K, Furlkröger E, 1990b. Clutch size regulation in the burying beetle Necrophorus vespilloides Herbst (Coleoptera : Silphidae). J Insect Behav 3: 265-270.
Mumme RL, Koenig WD, Pitelka FA, 1983. Reproductive competition in the communal acorn woodpecker : sisters destroy each others' eggs. Nature 306: 583-584.
Otronen M, 1988. The effect of body size on the outcome of fights in burying beetles (Nicrophorus). Ann Zool Fenn 25: 191-201.
Pukowski E, 1933. Ökologische Untersuchungen and Necrophorus F. Z Morphol Ökol Tiere 27: 518-586.
Reeve HK, Ratnieks FLW, 1993. Queen-queen conflicts in polygynous societies : mutual tolerance and reproductive skew. In : Queen number and sociality in insects (Keller L, ed). Oxford : Oxford University Press; 45-85.
Rosenblatt JS, Siegel HI, 1983. Physiological and behavioral changes during pregnancy and parturition underlying the onset of maternal behaviour in rodents. In : Parental behaviour of rodents (Elwood RW, ed). Chichester, UK : John Wiley & Sons; 23-66.
Scott MP, 1990. Brood guarding and the evolution of male parental care in burying beetles. Behav Ecol Sociobiol 26: 31-39.
Scott MP, 1994. Competition with flies promotes communal breeding in the burying beetle, Nicrophorus tomentosus. Behav Ecol Sociobiol 34: 367-373.
Scott MP, 1996. Communal breeding in burying beetles. Am Sci 84: 376-382.
Scott MP, 1997. Reproductive dominance and differential ovicide in the communally breeding burying beetle Nicrophorus tomentosus. Behav Ecol Sociobiol 40: 313-320.
Scott MP, Williams SM, 1993. Comparative reproductive
success of communally breeding burying beetles as assessed by PCR with
randomly amplified polymorphic DNA. Proc Natl Acad Sci USA
90: 2242-2245.
Stouffer PC, Kennedy ED, Power HW, 1987. Recognition and removal of intraspecific parasite eggs by starlings. Anim Behav 35: 1583-1584.
Trumbo ST, 1990a. Interference competition among burying beetles (Silphidae, Nicrophorus). Ecol Entomol 15: 347-355.
Trumbo ST, 1990b. Reproductive benefits of infanticide in a biparental burying beetle Nicrophorus orbicollis. Behav Ecol Sociobiol 27: 269-273.
Trumbo ST, 1992. Monogamy to communal breeding : exploitation of a broad resource base by burying beetles (Nicrophorus). Ecol Entomol 17: 289-298.
Trumbo ST, 1994. Interspecific competition, brood parasitism, and the evolution of biparental cooperation in burying beetles. Oikos 69: 241-249.
Trumbo ST, 1997. Juvenile hormone-mediated reproduction in burying beetles : from behavior to physiology. Arch Insect Biochem Physiol 35: 479-490.
Trumbo ST, Borst DW, Robinson GE, 1995. Rapid elevation of juvenile hormone titer during behavioral assessment of the breeding resource by the burying beetle, Nicrophorus orbicollis. J Insect Physiol 41: 535-543.
Trumbo ST, Fiore AT, 1994. Interspecific competition and the evolution of communal breeding in burying beetles. Am Midl Nat 131: 169-174.
Trumbo ST, Wilson DS, 1993. Brood discrimination, nest
mate discrimination, and determinants of social behavior in facultatively
quasisocial beetles (Nicrophorus spp.). Behav Ecol
4: 332-339.
Vehrencamp SL, 1977. Relative fecundity and parental
effort in communally nesting anis, Crotophaga sulcirostris.
Science 197:
403-405.
Wilson DS, Knollenberg WG, 1984. Food discrimination and ovarian development in burying beetles (Coleoptera : Silphidae : Nicrophorus). Ann Entomol Soc Am 77: 165-170.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. K. Muller, V. Braunisch, W. Hwang, and A.-K. Eggert Alternative tactics and individual reproductive success in natural associations of the burying beetle, Nicrophorus vespilloides Behav. Ecol., January 1, 2007; 18(1): 196 - 203. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. G. Zink and H. K. Reeve Predicting the temporal dynamics of reproductive skew and group membership in communal breeders Behav. Ecol., September 1, 2005; 16(5): 880 - 888. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. K. Muller, A.-K. Eggert, and T. Elsner Nestmate recognition in burying beetles: the "breeder's badge" as a cue used by females to distinguish their mates from male intruders Behav. Ecol., March 1, 2003; 14(2): 212 - 220. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||