A Comparative Study Of The Costs Of Alternative Mayfly Oviposition

    Drowning Fish. Most females had laid their eggs before drowning or being eaten by a fish (5090%). however, groups with oviposi- tion behaviors most exposed to the water zoology.wisc.edu.

  • Id: # 9c0f2fa
  • File Type: Pdf
  • Author: zoology.wisc.edu
  • File Size 688.2 kB
  • Read by User: 19 Times
  • Published: Monday, January 5, 2015
  • index: Drowning Fish


  • Read Online
A Comparative Study Of The Costs Of Alternative Mayfly Oviposition
A comparative study of the costs of alternative mayfly
oviposition behaviors
Andrea C. Encalada & Barbara L. Pec karsky
Received: 10 April 2006 / Revised: 5 February 2007 /Accepted: 7 February 2007 / Published online: 2 March 2007
Springer-Verlag 2007
Abstract Oviposition behavior of insects has associated
fitness costs related to the probability that females survive
to oviposit. During summer 2003, we observed the
oviposition behavior and compared the mortality rates of
females of 17 mayfly species in one western Colorado
watershed. We dissected adult females collected on terres-
trial sticky traps, in drift nets submerged in streams, and in
stomachs of brook trout to determine whether the mayflies
had oviposited before capture, drowning, or consumption.
Females oviposited by either splashing on the water surface
releasing all their eggs (splashers), dropping their eggs from
the air (bombers), dipping their abdomens multiple times
releasing a few eggs at a time (dippers), landing on rocks
and ovipositing on the undersides (landers), or floating
downstream while releasing their eggs (floaters). Almost
100% of lander and 50% of dipper females had not
oviposited when captured on sticky traps, increasing their
vulnerability to preoviposition mortality by aerial predators
compared to mayflies with other behaviors. In contrast,
most females had laid their eggs before drowning or being
eaten by a fish (50–90%). However, groups with oviposi-
tion behaviors most exposed to the water surfa ce (floaters,
then splashers, dippers, and landers) were more vulnerable
to drowning before completing oviposition. In addition,
splashers and floaters were most vulnerable to predation by
brook trout before ovipositing. These data suggest that
fitness costs associated with preoviposition mortality may
be considerable depending on mayfly oviposition behavior.
Furthermore, previously demonstrated benefits of low
predation rates on eggs of lander species may be offset in
part by costs to female survival.
Keywords Ephemeroptera
Oviposition behavior
Fish predation
Oviposition behaviors in insects are extremely diverse and
range from relatively unselective to highly specialized
(Hinton 1981). Females with alternative oviposition behav-
iors experience variable risks and consequently incur
different relative costs and benefits (Alcock 1998). Female
insects can increase their reproductive success by oviposit-
ing at locations that maximize the survival of their offspring
or minimize egg mortality (Kouki 1991;Thompson
and Pellmyr 1991; Price et al. 1998). Females can also
increase their fitness by minimizing preoviposition mor-
tality (Michiels and Dhondt 1990; McMill an 2000).
Behav Ecol Sociobiol (2007) 61:1437–1448
DOI 10.1007/s00265-007-0376-4
Communicated by G. Wilkinson
Electronic supplementary material The online version of this article
(doi:10.1007/s00265-007-0376-4) contains supplementary material,
which is available to authorized users.
A. C. Encalada
B. L. Peckarsky
Department of Entomology, Cornell University,
Ithaca, NY 14853, USA
B. L. Peckarsky
e-mail: [email protected]
A. C. Encalada
B. L. Peckarsky
Rocky Mountain Biological Laboratory,
P.O. Box 519, Crested Butte, CO, USA
B. L. Peckarsky
Department of Zoology, University of Wisconsin,
Madison, WI 53706, USA
A. C. Encalada (*)
Departamento de Ciencias Biológicas y Ambientales,
Universidad San Francisco de Quito,
Quito, Ecuador
e-mail: [email protected]
A Comparative Study Of The Costs Of Alternative Mayfly Oviposition
In general, aquatic insects have been considered to have
nonselective oviposition behaviors (Hinton 1981). Howev-
er, some Odonata (Wildermuth and Spinner 1991), Diptera
(Blaustein and Kotler 1993; Canyon et al. 1999), Hemiptera
(Smith 1976), Trichoptera (Reich and Downes 2003), and
Ephemeroptera (Elliott and Humpesch 1983; Peckarsky et
al. 2000) have highly specialized oviposition behaviors
with the potential to increase egg or offspring survival.
Nonetheless, adaptive values of different oviposition
behaviors in aquatic insects have rarely been evaluated
either from the perspective of postoviposition (Fincke
1986) or preoviposition costs (but see Michiels and Dhondt
Adult females of all species of mayflies (Ephemeroptera)
are terrestrial, but return to the water to lay their eggs and
have no parental care or egg-guarding behaviors (Needham
et al. 1905). Oviposition behaviors of mayflies are highl y
diverse (Edmunds et al. 1976; Elliott and Humpesc h 1983),
which is likely to result in variable rates of preoviposition
mortality. For example, depending on their oviposition
behavior, females may be more or less vulnerab le to
consumption by aquatic predators (principally fish ), aerial
predators (e.g., predat ory insects or birds), or drowning
before releasing their eggs. Jackson and Fisher (1986)
estimated that only 3% of the biomass of emerged aquatic
insects returned to the stream to oviposit, which suggests
high preoviposition mortality rates of adult females.
Little is known about the mortality rates of female
mayflies during oviposition, or the relative cost or benefits
of alternative oviposition behaviors. The goals of this study
were to (1) describe the oviposition behaviors of different
mayfly species in the East River drainage basin near the
Rocky Mountain Biological Laboratory (RMBL) in western
Colorado, and (2) compare the levels of female preovipo-
sition mortality among mayflies with different oviposition
behaviors. Using this comparative approach we developed
hypotheses about the costs associated with different
oviposition behavior s in mayflies and de termined that
certain types of oviposition behaviors are associated with
a significantly higher probability of egg mortality before
Materials and methods
Study area
We studied mayfly popula tions ovipositing in the East
River near the RMBL at ∼2,900 m elevation in the West
Elk Mountains, Gunnison County, CO, USA (38°57′30″N,
106°59′15″W). At RMBL the East River is a third-order
trout stream, 10–15 m wide, with mean discharge decreas-
ing from ∼4 to 0.50 m
/s throughout the flight season of
these mayflies (June to September). The riparian vegetation
of the study site is dominated by Engleman spruce (Picea
englemanii), Douglas Fir (Pseudotsuga menziesii), aspen
(Populus tremuloides), and many species of willows (Salix
spp.). However, the canopy above these streams is
generally open, providing a high sunlight environment for
observation of mayfly flight behavior during swarming,
mating, and oviposition.
Studies of the stream insects of this area have reported
the larval biology of mayflies in the East River near RMBL,
mostly in the families Baetidae, Heptageniidae, Ephemer-
ellidae, and Ameletidae (Peckarsky 1985, 1991). Previous
studies have not considered the adult biology of mayfly
species other than Baetis bicaudatus (Peckarsky et al.
Observations of mayfly oviposition and preoviposition
During the flight period in summer 2003 we recorded the
oviposition behaviors of 17 different mayfly species
(Fig. 1) and compared potential causes of female mortality
and effects of oviposition behavior on preoviposition
mortality. We obtained comparative data using three general
1. Traps to estimate relative abundance of ovipositing
females (potential risk to females of aerial predation
before ovipositing)
Because direct measures of aerial predation by birds,
bats, or insects were not feasible, we estimated the number
of females caught during flight with or without eggs as a
surrogate of preoviposition risk of aerial predation. We
estimated the relative abundance of flying females over
daily and seasonal cycles from collections of female
mayflies on two transects of sticky traps, evenly spaced
(50 cm apart) and suspen ded vertically 3–8 cm above the
East River (to catch females flying near the stream surface).
Sticky traps (11 total) consisted of transparent acetate sheets
(27.9×21.6 cm) with tree pest adhesive (Tanglefoot
applied to upstream and downstream sides. Sticky traps
were deployed on 7, 14, 21, and 28 July and 4 August, and
changed two or three times daily.
We also placed 13 horizontal “platform traps” 3–8cm
above the stream surface to catch females ovipositing in a
150-m reach of the East River. Platforms were made of
plywood (27.9×21.6 cm) and anchored to the streambed
using rebar. Platforms were distributed at random (>3 m
apart) and remained in place for 24 h on the same dates as
the vertical traps. We applied tree Tanglefoot
to three
transparent acetate sheets stacked on top of each platform.
One of the three sheets was removed at 0800, 1400 and
2000 hours.
1438 Behav Ecol Sociobiol (2007) 61:1437–1448
A Comparative Study Of The Costs Of Alternative Mayfly Oviposition
In the laboratory we removed mayflies from the sticky
traps by placing each sheet in a pan of solvent (Googone
We subsequently identified mayflies according to species,
sex, and subimago vs imago. We also measured total body
length excluding cerci and antennae and dissected females
to estimate the percentage of eggs remaining in the
To supplement the trap data, every week from 3–31 July,
we placed a ~15-m-long by 1-m-tall mesh net across a 13-
m-wide transect of the East River for 5 min at 1300 and
1600 hours. We collected all mayfly subimagoes, imagoes,
and exuvia in 70% ethanol for further analysis in the
laboratory. We also made daily collections of flying
mayflies using a Malaise trap above one location in the
East River from 15 July to 15 October. Finally, once per
week for 9 weeks (June to August) during the flight period
(0730–2130 hours) we made continuous observations of
mayfly ovipos ition behavior at three locations along the
Ameletus velox (splashers)
Acentrella turbida (floaters)
Baetis bicaudatus (landers)
Baetis sp nov. (landers)
Diphetor hageni (dippers)
Ephemerella infrequens (bombers)
Drunella doddsi, D. grandis &
D. coloradensis (bombers)
Cinygmula mimus (dippers)
Cinygmula sp. (dippers)
Rhithrogena hageni (dippers)
Rhithrogena robusta (dippers)
Epeorus deceptivus (dippers)
Epeorus longimanus (dippers)
Paraleptophlebia heteronea (dippers)
Siphlonurus occidentalis (splashers)
Fig. 1 Flight periods of 17
species of Ephemeroptera in the
East River watershed near
RMBL. M May, J June and July,
A August, S September, O Oc-
tober. Black lines show duration,
dotted lines show flight periods
when less than 20 individuals
were observed, and arrows ap-
proximate peak flight period.
Information compiled from hor-
izontal and vertical sticky traps,
malaise traps, and weekly
observations in 2003. Vertical
dashed lines indicate approxi-
mate dates of ice off and ice on.
Behav Ecol Sociobiol (2007) 61:1437–1448 1439
A Comparative Study Of The Costs Of Alternative Mayfly Oviposition
East River. After observations, individuals were collected
with sweep nets to identify the mayfly species exhibiting
each type of oviposition behavior.
2. Drift nets to estimate relative mortality of females that
drown before ovipositing
To estimate the relative proportions of females with and
without eggs that got carried away in the water column
during oviposition, we placed at random locations in the
East River six drift nets (opening 45×29 cm) on 7 July and
four drift nets on 14, 21, and 28 July and 4 August 2003.
Nets were deployed two times daily at 0900 and 1300 hours
and were left in the stream for 3 h. The nets were fastened
to the stream bottom with rebar and completely submerged
to avoid collecti ng females touching the water surfa ce
during oviposition. The discharge passing through each net
averaged overall dates was ~0.12 m
/s. Imagoes and
subimagoes that were captured in the drift nets were
preserved in 70% ethanol. In the lab we sorted mayflies
according to species and sex and dissected females to
estimate the percentage of eggs remaining in each female,
as described above.
Finally, to determine whether ovipositing females could
reattach onto rocks once entrained in the current during July
2003, we collected 20 ovipositing female B. bicaudatus and
released them into a 5-m-long wooden trough in the East
River with eight rocks (200–450 cm
) protruding from the
water surface and a drift net at the downstream end. The
current velocity in the trough was 0.43±0.12 m/s, which
was comparable to the mean velocity in the stream at that
time (0.52±0.23 m/s). Ten min utes after the release we
checked every rock and counted the females remaining and
females that drifted into the net.
3. Relative mortality of females consumed by fish before
Possible aquatic predators of female mayflies in this
system include brook trout (Salvelinus fontinalis Mitchell),
which feed on aquatic and terrestrial stages of aquatic
insects and swallow their prey whole (Allan 1978, 1981).
Although brook trout are not native to this system (Pister
2001), they were the only species of predatory fish at the
East River study site. To estimate the relative abundance of
females with and without eggs that were consumed by
brook trout, we electroshocked the East River on 7, 14, 21,
and 28 July and 4 August 2003 for 2–3 h during the
morning, afternoon, and evening (total ~8 h/day), collecting
30 brook trout in total. We flushed the stomachs of each
fish using stream water in a plastic squirt bottle attached to
Ty gon
tubing (Kamler 2001). Fish regurgitates were
preserved in 90% ethanol and analyzed later in the
laboratory. We sorted mayflies according to species, sex,
and imagoes vs subimagoes, measured body lengths, and
dissected females to estimate the percentage of eggs
remaining in their abdomens.
Data analysis
To compare the proportion of total females among different
oviposition behaviors tha t still ha d their eggs when capture d
in flight (surrogate of risk of aerial predation), in drift nets
(drowned), or fish guts (consumed), we performed a general
linear model procedure (SAS Institute 2003)foratwo-way
repeated measures analysis of variance (ANOVA). Our
response variable was the arcsine transformed proportion
of females with eggs at times T1 through T5 (Sokal and
Rohlf 1981). The fixed factors were (1) trapping method
with three levels flying, drowning, and eaten by fish, and
(2) female oviposition category (five levels). Because
“time” was not a significant factor in the within-subject
effects (Encalada 2005), we reassigned the time replicates
two-way ANOVA. We performed a posteriori Tukey’s
Studentized Range [honestly significant difference (HSD)]
tests for both f actors to interpret specific comparisons.
Finally, we use d one-way ANOVA to test for d ifferences
in mean size of all females (with and without eggs) captured
while flying, drowned, and consumed for 11 mayfly species
for which we had sufficient data. We also performed a two-
way ANOVA to test if sizes of all mayflies combined
differed between females with and without eggs (fa ctor 1)
that drowned or were eaten by fish (factor 2). We made a
posteriori pairwise comparisons using Tukey’sHSDtests
when there were significant main effects or interacti ons.
Mayfly oviposition behaviors
Mayfly flight periods (emergence, swarming, and oviposi-
tion) extended from June to October in the u pper East River
watershed (Fig. 1). However, these patterns may vary
among years depending on winter and summer precipita-
tion; high water years usually have earlier and longer
emergence periods than low water years (Peckarsky et al.
2000; Encalada 2005 ). Flight periods of some species were
more synchronous (e.g., B. bicaudatus and Cinygmula
mimus), and others extended for longer periods (e.g.,
Paraleptophlebia heteronea). Daily oviposition period of
most mayflies was from 0900 to 1700 hours with peak
flight activity around midday (1200– 1300 hours). Night-
time female catches included very few Ephemero ptera
(Encalada 2005). Nonetheless, we observed large male
swarms of Epeorus spp. from 1700 to 2000 hours (as in
Flecker et al. 1988) and also during early morning starting
around 0730 hours. Some female C. mimus were captured
at night, suggesting the possibility of nighttime oviposition
(Encalada 2005).
1440 Behav Ecol Sociobiol (2007) 61:1437–1448
A Comparative Study Of The Costs Of Alternative Mayfly Oviposition
Based on our observations and the published literature
(Needham et al. 1905; Elliott and Humpesch 1983), we
classified 17 mayfly species into five functional oviposition
categories: splashers, bombers, dippers, floaters, and land-
ers (see Fig. 2 and S1 for animations of the different flight
(1) Splashers. (e.g., Fig. 2a) This category included two
species, Ameletus velox and Siphlonurus occidentalis,
whose females flew in aggregations. Both species
bounced above the water surface while flying up-
stream. After alighting on the water surface they got
carried downstream for 1–1.5 m, discharged all their
eggs at once, and then flew away. Most mayflies from
this group oviposited in fast velocity (∼65 cms
locations with no turbulence.
2) Bombers. (e.g., Fig. 2 b) This category included all
species of Ephemerellidae. Egg masses were extruded
from the gonopore of the seventh abdominal segment of
females during relatively slow downstream or upstream
flight, very close to the water surface. Usually the entire
egg mass was released from the air; but if the extruded
egg mass touched the water surface it was relea sed
immediately. Female bombers were observ ed to drop
their eggs in fast and slow velocity places.
3) Dippers. (e.g., Fig. 2c, d) This category was very
taxonomically diverse, including most species of may-
flies in the Eas t Rive r cat chment (all sp ecies o f
Heptagenii dae, so me Baeti dae, and one species of
Leptophlebiidae). Females flew upstream or down-
stream above the water surface and dipped their abdo-
mens in the water several times, releasing a few eggs
each time. Some species flew very fast covering great
distances (e.g., Fig. 2c, Rhithrogena hageni), while
others flew slowly and dipped at short distances (e.g.,
Fig. 2d, Diphetor h ageni ).
0.8 - 1m
Siphlonurus occidentalis
0.8 - 1m
Siphlonurus occidentalis
0.8 - 1m
Siphlonurus occidentalis
0.8 - 1m
Siphlonurus occidentalis
Drunella doddsi
Drunella doddsi
Drunella doddsi
upstream flight
30 - 50 cm
1 - 2 m
Rhithrogena hageni
upstream flight
30 - 50 cm
1 - 2 m
Rhithrogena hageni
upstream flight
30 - 50 cm
1 - 2 m
Rhithrogena hageni
Downstream flight
15 - 30 cm
Diphetor hageni
Downstream flight
15 - 30 cm
Diphetor hageni
Downstream flight
15 - 30 cm
Downstream flight
15 - 30 cm
Diphetor hageni
e Landers
Baetis bicaudatus
0.5 m
Baetis bicaudatus
0.5 m
Baetis bicaudatus
0.5 m0.5 m
f Floaters
Acentrella turbinae
0.4 m
Acentrella turbinae
0.4 m
Acentrella turbinae
0.4 m0.4 m
Fig. 2 Oviposition behaviors of
different mayflies (Ephemerop-
tera) observed in the upper East
River drainage basin, near
RMBL in western Colorado. a
Splashers (e.g., S. occidentalis);
b bombers (e.g., Drunella
doddsi); c, d dippers (e.g., R.
hageni, D. hageni, and E. long-
imanus); e landers (e.g., B.
bicaudatus); and f floaters (e.g.,
Acentrella turbinae). See Fig. 1
for a complete list of species
showing each behavior.
Behav Ecol Sociobiol (2007) 61:1437–1448 1441
A Comparative Study Of The Costs Of Alternative Mayfly Oviposition
Heptageniidae had some interesting behaviors not
observed in other mayflies. Females and males flew in the
same swarms (~80 to >500 individuals). While most males
performed pendular flights 3–7 m above the water surface,
females flew very close to the water (<1 m). Most
remarkably, some males followed females while they were
ovipositing and copulated with females between dippi ng
episodes, exhibiting male-guarding behavior that may be
similar to some Odonata (Waage 1987). Chases between
males were very common, suggesting male– male competi-
tion. Females oviposited in relatively fast velocity, turbulent
places (riffles). We also observed Epeorus deceptivus and
Epeorus longimanus flying toget her with R. hageni in
mixed species swarms.
We observed two other variations in the oviposition
behavior of this group. P. heteronea females flew very fast
in small groups (6–8 individuals) near the water surface and
primarily dipped close to the stream banks in slow or fast
velocity water. We never observed males of this species
guarding females or mating with them during oviposition.
Finally, single D. hageni (Baetidae) females flew rather
slowly just above the water surface (15–30 cm), and dipped
in nonturbulent places, usually near the stream banks.
4) Landers. (e.g., Fig. 2e) Females of all observed species
of Baetis have been reported to land on a rock
protruding from the stream in high velocity locations,
crawl under water on the downstream side of the rock,
and lay one egg mass on the underside (Eaton 1888;
Peckarsky et al. 2000). Females face the current during
oviposition and lay eggs while moving their abdomens
from side to side, resulting in egg masses shaped like
5) Floaters. (e.g., Fig. 2f) Acentrella turbida (Baetidae)
females had two alternative oviposition behaviors.
Some females landed on fast or slow water surfaces,
drifted until they contacted a large rock, then immedi-
ately crawled under the rock and laid a small (mean
) irregular egg mass. We suspect that other
females released some eggs while drifting and then
flew off the water surface because m any females
caught in horizontal and vertical sticky traps had only
a small percentage of eggs remaining in their abdo-
mens. It is interesting to note that we never observed or
captured male subimagoes or imagoes of A. turbida ,
suggesting this speci es may be parthenogenetic.
Preoviposition mortality
three methods differed with respect to oviposition category
(Fig. 3; F
=9.9; P<0.001). However, the proportions of
total females that still had their eggs when captured while
flying, in drift nets, or fish guts (consumed) did not vary
through the season (no significant time effect; F
=0.59, P=
0.58 adjusted for sphericity test; SAS Institute 2003).
Consumed by fish
Consumed by fish
Oviposition Type
Proportion of Mayfly Females with Eggs
Consumed by fish
Fig. 3 Average proport ion of
total mayfly females of each
oviposition category that had
not yet oviposited their eggs
when captured during flight (on
vertical and horizontal traps),
drowned (in submerged drift
nets), and eaten by fish (in fish
regurgitates). Bars represent ±1
SE. Histograms with same let-
ters were not significantly dif-
ferent (Tukey’s HSD tests) as a
result of pairwise comparisons
within each oviposition type.
See Table 1 for pairwise com-
parisons among oviposition
1442 Behav Ecol Sociobiol (2007) 61:1437–1448
A Comparative Study Of The Costs Of Alternative Mayfly Oviposition
Therefore, we reran the ANOVA using dates as true replicates.
The proportion of females with eggs differed significantly
among trapping methods (F
=12.03, P=0.0001) and among
oviposition types (F
=10.76, P=0.0001). Most importantly,
there was a significant interaction term (F
=12.08, P=
0.0001), indicating that variation among trapping methods in
the proportion of females that were captured before oviposit-
ing depended on oviposition type.
Pairwise comparisons within each oviposition category
showed that splashers were more vulnerable to being eaten
by fish before ovipositing than by drowning or while flying
(Fig. 3). It is interesting to note that more females than
males were observed in fish guts and in some cases only
females were found, reflecting their increased vulnerability
to predation while oviposit ing. In contrast, a significantly
higher proportion of dippers and landers captured while
flying still had their eggs than those captured in drift nets or
in fish stomachs (Fig. 3).
Pairwise comparisons between oviposition categories
showed that landers and dippers had a higher proportion
of females with eggs caught flying near the stream than any
other group (Fig. 3, Table 1). It is interesting to note that
almost all (95% percent) female landers captured in sticky
traps had not yet oviposited, and most of the 5% captured
without eggs were infested with mermithid nematodes. In
addition, floaters were more vulnerable to drowning and
bombers were the least vulnerable to drowning than other
mayf ly groups (Fig. 3,Table1). Finally, brook trout
consumed a higher proportion of females with eggs of
splashers and floaters than other mayfly groups (Fig. 3,
Table 1), making those oviposition behaviors more vulner-
able to trout predation.
Variation in mortality with female size
The one-way ANOVA revealed size-dependent v ariation in
overall female mortality for 4 of 11 species of mayflies
analyzed (Table 2). While drowned and consumed females
of A. velox [mean body length (L)=12.5 mm] were larger
than their flying counterparts (L=7.5 mm), drowned
females of B. bicaudatus (L=5 mm), Ephemerella infre-
quens (L=6.4 mm), and P. heteronea (L=6 mm) were
smaller than females caught while flying (L=8.3, 10, and
11 mm, respectively). Similarly, B. bicaudatus females that
drifted off rocks before ovipositing w ere significantly
smaller than those who had successfully oviposited.
Finally, for all species combined, females eaten by fish
were significantly larger than females that drowned (F
19.81, P<0.0001). However, there was no significant
difference between sizes of females that died before or
after ovipositing (F
=0.73, P=0.39). It is interesting to
note that the interaction term was also not significant (F
0.22, P=0.64), indicating that variation in size of females
that drowned or were eaten by fish did not depend on
whether they had oviposited.
Table 1 Summary of a posteriori pairwise comparisons (Tukey’s
HSD test) among the proportion of females captured with eggs in each
“trapped ” category (flying, drowned, or eaten by fish) between
oviposition types (splashers, bombers, dippers, landers, and floaters)
(comparisons within columns)
Factor 2: oviposition type Factor 1: trapped
Flying Drowned Eaten by fish
Splashers a b b
Bombers a a a
Dippers b b a
Floaters a c b
Landers c b a
Different letters denote significantly different (alpha=0.05) propor-
tions of females captured with eggs in pairwise comparisons among
oviposition types within each trapping method. Groups with the same
letter were not significantly different and denote low (a) to high (c)
risk behaviors. See Fig. 3.
Table 2 Summary of one-way ANOVA testing for differences in the mean body length of female mayflies caught on sticky traps (F flying), in
drift nets (D drowned), or in fish regurgitates (C consumed)
SPECIES df (par,error) FPvalue Tukey’s HSD (α =0.05)
Acentrella turbida 2,80 0.096 0.9085
Ameletus velox 2,7 7.75 0.0168 F>D=C
Baetis bicaudatus 2,237 5.49 0.0047 F>D=C
Cinygmula mimus 2,76 1.06 0.3509
Diphetor hageni 2,118 0.82 0.4423
Epeorus deceptivus 2,40 2.07 0.1391
Ephemerella infrequens 2,143 6.79 0.0015 F>D>C
Paraleptophlebia heteronea 2,288 7.80 0.0005 F>D=C
Rhithrogena hageni 2,282 0.72 0.4871
Siphlonurus occidentalis 2,70 0.69 0.5072
Analyses were performed for each species individually. A posteriori HSD Tukey’s tests indicate pairwise comparisons in cases with significant
main effects.
Behav Ecol Sociobiol (2007) 61:1437–1448 1443
A Comparative Study Of The Costs Of Alternative Mayfly Oviposition
We observed impressive variation in behaviors of mayflies
ovipositing in the East River. Females oviposited by either
splashing on the water surface releasing all their eggs,
dropping their eggs from the air, dipping their abdomens
multiple times releasing a few eggs at a time, landing on
rocks and ovipositing on the undersides, or f loating
downstream while releasing their eggs. Although we
estimated ∼30 to 40% overall preoviposition mortality of
adult females, some mayf lies performed better than others,
depending on their strategies during oviposition. Bombers
had the lowest overall preoviposition mortality, and groups
with more contact with surface water (floaters and splash-
ers) were more likely to drown or be eaten before
ovipositing (Fig. 3). While landers and dippers had
intermediate rates of preovi position drowning and preda-
tion, high rates of capture of egg-bearing females flying
near the water suggested the potential for greater exposure
to aerial predators.
Effects of oviposition behavior on preoviposition mortality
Seasonal periodicity Fligh t periods of mayflies were highly
seasonal (Fig. 1), which is typical for other sites at high
altitudes and latitudes (Elliott and Humpesc h 1983), in
contrast to extended or aseasonal flight perio ds observed
for mayflies at lower altitudes (Jackson and Fisher 1986)or
latitudes (Jackson and Sweeney 1995). Such seasonality
has the potential to affect mortality risk of females during
oviposition. For example, high discharge early in the
summer might constrain the onset of oviposition b y
increasing the probability of females drowning before
ovipositing, especially for those with higher contact with
water (e.g., splashers, floaters, and landers). It is interesting
to note that the species that oviposited earliest in the season
were dippers (Fig. 1), a strategy with relatively low
probability of drowning before ovipositing (Fig. 3). On
the other hand, ovipositing later in the summer when
snowmelt waters recede and water clarity improves could
increase the vulnerability of females to trout predation. We
observed that species ovipos iting latest in the season were
dippers, bombers, or landers (Fig. 1), a behavior that was
least susceptible to preoviposition trout predation (Fig. 3).
Nonetheless, ovipositing too late in the season might affect
the probabili ty of finding reasonable oviposition sites and
increase the probability of egg desiccation for lander s
(Encalada and Peckarsky 2006).
Diel periodicity Diel periodicity of oviposition might also
be related to the abunda nce and activity patterns of
predators (brook trout, insects, and birds). Insectivorous
birds, for example, are more active at dawn and dusk
(McCarty and Winkler 1999), which could constrain the
diel periodicity of mayfly oviposition. It is interesting to
note that females rarely oviposited at dawn or dusk in these
streams, with the exception of some Heptageniidae that
swarmed and oviposited at dusk (Encalada 2005). Howev-
er, gut contents of brook trout indicated that they fed on
ovipositing females throughout the day from 0800 to 2000
hours. Thus, females could only avoid trout predation if
they oviposited at night, which was rare in this study
(Encalada 2005).
Aggregation It is interesting to note that some females of
dippers, landers, and floaters were observed flying in small
groups (∼7 to 15 individuals). Although reports of female
mayfly aggregations are rare, Ephemerella spp. have been
observed to form female swarms before oviposition
(Pleskot and Pomeisl 1952 ). Communal oviposition might
result from females being attracted to common oviposition
sites (Waage 1987; Allan and Flecker 1989), or by
association with swarming males (Savolainen et al. 1993).
Regardless of the proximate mechanisms explaining this
behavior, aggregation during oviposition could ultimately
affect preoviposition mortality by reducing individual risk
of predation via dilution (Waage 1987;Michielsand
Dhondt 1990), predator satiation (Sweeney and Vannote
1982), or predator confusion (Krakauer 1995). Heptagenii-
dae males usual ly swarm above open canopy, unshaded,
and highly reflective sites (Flecker et al. 1988; Allan and
Flecker 1989), which makes it difficult for visual salmonid
predators to distinguish ovipositing females at the air–water
interface against the bright background (Elliott and
Humpesch 1983). Alternatively, aggregation during ovipo-
sition might attract predators and increase the probability of
being eaten (Corbet 1964; McCarty and Winkler 1999).
Further data are necessary to test definitively the costs and
benefits of female aggregation during oviposition.
Mate guarding An unusual behavior for mayflies observed
in this study was contact and noncontact guarding of
females by male Heptageniidae during flight. Males were
commonly observed to harass females and force repeated
matings during female dipping episodes. Because mayfly
eggs are fertilized as they travel from the oviducts to the
genital duct (Needham et al. 1935), repeated mating in
species with dipping oviposition behavior ma y ensure
paternity in these species. Similar harassment behavior
has been observed in Odonata resulting in ~95% fertiliza-
tion precedence by the last male to mate (e.g., Michiels and
Dhondt 1990 ). In some species of Odonata male guarding
has been reported to decrease female mortality risk (Fincke
1986; Michiels and Dhondt 1988). However, male presence
might also increase female reproductive costs related to
time and energy expenditure during oviposition (Michiels
1444 Behav Ecol Sociobiol (2007) 61:1437–1448
A Comparative Study Of The Costs Of Alternative Mayfly Oviposition
and Dhondt 1990). The fitness costs or benefits of mate-
guarding behaviors are not known in mayflies.
Site sel ection The oviposition behavior of mayflies in
general has been described as nonselective (Hinton 1981),
given that females (except for landers) do not precisely
control the destination of their eggs. Studies of terrestrial
herbivorous insects suggest a strong evolutionary link
between female oviposition preference and offspring
performance (Thompson 1988; Craig et al. 1989; Ohgushi
1992; Price 1994). However, selective oviposition behavior
by mayflies may have little influence on the survival of
their offspring because larvae are highly mobile generalist
grazers capable of selecting their own foraging habitat
(Richards and Minshall 1988; Palmer 1995). Consequently,
we argue that mayfly oviposition behavior should be under
stronger selection pressure to minimize female preoviposi-
tion mortality and maximize egg survival than to maximize
offspring performance.
Nonetheless, we observed some site selection in all
oviposition strategies, potentially resulting in tradeoffs
between minimizing female mortality and maximizing egg
survival. For example, most females in all categories of
female behaviors oviposited in fast velocity locations
(Encalada 2005), which could have both costs (increase
probability of preoviposition drowning of females) and
benefits related to increasing egg dispersal. Egg dispersal
by stream current might reduce the probability of mass
predation on eggs, or maximize unpredictability of off-
spring dispersion (Thompson and Pellmyr 1991). Although
eggs of most mayflies have adhesive devices to prevent
drifting long distances (Needham et al. 1935; Hinton 1981),
eggs of Palingenia longicauda have been reported to drift
as far as 2.5 km from the oviposition site (Russev 1959).
Drifting eggs could also end up in suboptimal environments
for development, such as trapped in sediment, locations
vulnerable to egg desiccati on, or suboptimal temperatures
(Elliott and Humpesch 1980). Therefore, the lander strategy
provides the most benefits in terms of increasing the
probability of egg survival (Encalada and Peckarsky
2006), while trading off the observed costs of potentially
high preoviposition predation.
Egg release strategies vs predation risk Terrestrial stages of
aquatic insects have been shown to provide important food
subsidies for terrestrial predators (Murakami and Nakano
2001;SaboandPower2002). In fact, Nakano and
Murakami (2001) showed that aquatic prey represent
25.6% of the annual total energy demand of the entire bird
assemblage in a forested stream. One study showed that
mayflies comprised more than 50% of the diets of a
population of tree swallows during the summer months
(Wayland et al. 1998). Other studies found that densities of
emerging insects were highly correlated with densities of
flycatchers (Gray 1993), as well as with spatial distribution
and abundances of lizards (Sabo and Power 2002). We
observed tree swallows, flycatchers, and warblers common-
ly feeding on swarming male mayflies and on females
during oviposition flights. Thus, our sticky trap data and
observations suggest that some of the observed oviposition
strategies (landers and dippers) could have high potential
costs in the form of bird predat ion on egg-bearing females.
However, direct comparisons of aerial predation of mayflies
are needed to draw more definitive conclusions regarding
the relative fitness costs and benefits among the different
oviposition behaviors and how they affect variability in
preoviposition mortality.
Another aspect of fema le oviposition behavior that could
affect female survival and egg survival is whether eggs
were released all at once (e.g., splashers and bombers) or a
few at a time (e.g., dippers, floaters, and landers). Laying a
few eggs at a time may increase the probability some eggs
would be relea sed before females died, while the
“all eggs
in one basket” strategy may reduce time of exposure of
egg-bearing females to predation by aerial predators. Our
data were not consistent with this argument even though
female splashers, which lay all their eggs at once, had the
lowest probability of being caught in aerial traps before
ovipositing (Fig. 3). However, bombers, the other group
that lay all their eggs at once, had similar probabilities of
aerial capture to floaters (Fig. 3), which release a few eggs
at a time.
Females were also vulnerable to trout predation before
ovipositing. Allan (1978) reported that 14% of emerging
aquatic insects were eaten by brook trout in a nearby
stream; and adult forms of aquatic insects made up more
than 20% of invertebrate biomass in diets of coho salmon in
an Alaskan stream (Allan et al. 2003). However, previous
studies do not provide information about fish predation on
ovipositing aquatic insects. Our data showed that terrestrial
stages of aquatic insects constituted a large proportion of
the diets of brook trout (∼ 30 to 40% from June to
September), a third of which was ovipositing females.
Furthermore, preoviposition predation rates varied with the
amount of time females spent on the water surface (Fig. 3).
Bombers, which do not contact the water surface, had the
lowest vulnerability to fish predation, and splashers and
floaters, which drift on the water surface while ovipositing,
had the highest rates of brook trout predation. However, we
do not know whether mayfly eggs are viable after passing
through the digestive track of fish, as has been seen in some
aquatic invertebrates (Jarnagin et al. 2000). It is interesting
to note that most lander females consumed by brook trout
had already laid their eggs, indicating a reduced cost due to
fish predation for females that have this type of oviposition
Behav Ecol Sociobiol (2007) 61:1437–1448 1445
A Comparative Study Of The Costs Of Alternative Mayfly Oviposition
A very high proportion of female landers captured flying
close to the stream surface still had their eggs, most of the
other 5% with no eggs captured in sticky traps were
infested with mermithid nematodes. Parasitized females do
not develop eggs and, when ready to oviposit, extrude the
mermithid instead (Flecker and Allan 1988; Vance 1996).
These data suggest that lander females rarely flew again
after ovipositing under a rock, as corrob orated by experi-
mental releases of females in the in-stream wooden flumes,
which showed that 97% of drifting females did not settle on
rocks once they were in the water column.
Tradeoffs involved in female size
Size at ovipos ition may affect the probability of female
survival and determines the fecundity of survivors (Elliott
and Humpesch 1983; Peckarsky et al. 200 1). Small
individuals of three species (B. bicaudatus, P. heteronea,
and E. infrequens, a lander, dipper, and bomber, respec-
tively) were more vulnerable to drowning than large ones,
suggesting another cost to being a small female in these
groups. In contrast, large individuals of A. velox females
(splashers) were more vulnera ble to drowning than small
ones. Females of all species combined that were eaten by
fish wer e significantly larger than females that drow ned,
which could partially offset the general fecundity benefit to
insects of being a large female. This pattern may be
explained by larger prey being more profitable or more
visible to predators (cf. Li et al. 1985), or as an experimental
artifact of size-selective predation (i.e., fish could have eaten
the larger females before they were caught in the drift nets).
Large egg-bearing females may also be more conspicuous to
bird predators. In a study of diet selectivity of tree swallows,
McCarty and Winkler (1999) showed that selectivity for
large insects increased with insect abundance, a pattern that
holds among insec tivorous birds in general (Hespenheide
1971). Therefore, variation in the size of ovipositing female
mayflies may reflect tradeoffs between vulnerability to
predation, drowning and fecundity.
In summary, preoviposition mortality of female mayflies
by drowning and by fish predation was substantial and
dependent on their type of oviposition behavior. Females
that did not touch the water surface had the lowest risks of
fish predation and drowning before ovipositing. In contrast,
females with longer exposure to the water surface had the
highest preoviposition rates of mortality. Diel and seasonal
timing of oviposition, aggregation behavior, variation in the
numbers of eggs released at one time, and female size all
affected female mortality, with potential for offsetting
benefits in egg survival. Such information about factors
affecting female mortality during oviposition and the
consequences of female behavior for egg or larva l survival
is necessary to understand the adaptive values of different
oviposition behaviors.
Acknowledgments We thank Maruxa Alvarez, Wendy Brown, Matt
Harper, Bryan Horn, Mark Wallin, Brad Taylor, Esteban Suárez, and
Juan Esteban Suárez-Encalada for help with fieldwork. Boris
Kondratieff identified the adult mayfly specimens. Maruxa Alvárez
and Juan Esteban Suárez-Encalada helped unglue mayflies from sticky
traps. Esteban Suárez helped check and prepare the malaise traps and
Wendy Brown collected specimens during fall 2003. Thanks to “The
Big Benthette Boys” and “The Godfather” for electrofishing,
regurgitating fish, and avoiding the beavers. Francoise Vermeylen
provided statistical advice. Special thanks to Juan Esteban for sitting
with his mom for hours at the edge of the East River to observe the
amazing behaviors of mayflies. The final version of this paper was
greatly improved with comments by Alex Flecker, Cole Gilbert,
Nelson Hairston, and two anonymous reviewers. This research was
supported by funds from NSF Doctoral Dissertation Improvement
Grant DEB-0206095 to ACE, partial funds from NSF grant DEB-
0089863 to BLP, and Cornell University HATCH grant 139-402 to
BLP. ACE was also supported by the Ecuadorian Foundation for
Science and Technology (FUNDACYT).
Declaration The authors of this manuscript declared that the
experiments in this study comply with the current USA laws.
Electronic supp lementary material
Below is the link to the electronic supplementary material.
S1Power-point presentation showing animations of oviposition behav-
iors of different mayflies (Ephemeroptera) observed in the upper East
River drainage basin, near RMBL in western Colorado. A. Splashers
(e.g., Siphlonurus occidentalis), B. Bombers (e.g., Drunella doddsi),
C. Dippers (e.g., Rhithrogena hageni, Diphetor hageni, Epeorus
longimanus), D. Landers (e.g., Baetis bicaudatus), E. Floaters (e.g.,
Acentrella turbinae). (To activate the animations, click on “ppt
presentation” or the F5 key, and use the “return” key to move through
the different animations (PPT 232 KB).
Alcock J (1998) Animal behavior: an evolutionary approach, 6th edn.
Sinauer, Sunderland, MA
Allan DJ (1978) Diet of brook trout (Salvelinus fontinalis MITCHILL)
and brown trout (Salmo trutta L.) in an alpine stream. Verh
Internat Verein Limnol 20:2045–2050
Allan DJ (1981) Determinants of diet of brook trout (Salvelinus
fontinalis) in a mountain stream. Can J Fish Aquat Sci 38:184–192
Allan JD, Flecker AS (1989) The mating biology of a mass swarming
mayfly. Anim Behav 37:361–371
Allan JD, Wipfli MS, Caouette JP, Prussian A, Rodgers J (2003)
Influence of streamsi de vegetation on inputs of terrestrial
invertebrates to salmonid food webs. Can J Fish Aquat Sci
Blaustein L, Kotler BP (1993) Oviposition habitat selection by the
mosquito Culiseta longiareolata: effects of conspecifics, food
and green toad tadpoles. Ecol Entomol 18:104–108
Canyon DV, Hii JLK, Muller R (1999) Adaptation of Aedes aegypti
(Diptera: Culicidae) oviposition behavior in response to humidity
and diet. J Insect Physiol 45:959–964
1446 Behav Ecol Sociobiol (2007) 61:1437–1448
A Comparative Study Of The Costs Of Alternative Mayfly Oviposition
Corbet PS (1964) Temporal patterns of emergence in aquatic insects.
Can Entomol 96:264–279
Craig TP, Itami JK, Price PW (1989) A strong relationship between
oviposition preference and larval performance in a shoot galling
sawfly. Ecology 70:1691–1699
Eaton AE (1888) A revisional monograph of the recent Ephemeridae
or mayflies. Trans Linn Soc 2:III
Edmunds GF, Jensen SL, Berner L (1976) The mayflies of North and
Central America. University of Minnesota Press, Minneapolis, MN
Elliott JM, Humpesch UH (1980) Eggs of Ephemeroptera. Annu Rep
Freshw Biol Assoc 48:41–52
Elliott JM, Humpesch UH (1983) A key to the adults of the British
Ephemeroptera with notes on their ecology. Freshwater Biological
Association, scientific publication no. 47, Ambleside, Cumbria
Encalada AC (2005) Patterns, mechanisms and consequences of
selective oviposition behavior of the mayfly Baetis bicaudatus
in Colorado high-altitude streams. Ph.D. thesis, Cornell Univer-
sity, Ithaca
Encalada AC, Peckarsky BL (2006) Selective oviposition behavior of
the mayfly Baetis bicaudatus. Oecologia 148:526–537
Fincke OM (1986) Underwater oviposition in a damselfly (Odonata:
Coenagrionidae) favors male vigilance, and multiple mating by
females. Behav Ecol Sociobiol 18:405–412
Flecker A, Allan JD (1988) Flight direction in some rocky mountain
mayflies (Ephemeroptera) with observation of parasitism. Aquat
Insects 10:33–42
Flecker AS, Allan JD, McClintock NL (1988) Male body size and
mating success in swarms of the mayfly Epeorus longimanus.
Holarct Ecol 11:280–285
Gray LJ (1993) Response of insectivorous birds to emerging aquatic
insects in riparian habitats of a tallgrass prairie stream. Am Midl
Nat 129:288–300
Hespenheide HA (1971) Food preference and extent of overlap in
some insectivorous birds, with special reference to Tyrannidae.
Ibis 113:59–72
Hinton HE (1981) Biology of insect eggs, vols. I, II, III. Pergamon,
Jackson JK, Fisher SG (1986) Secondary production, emergence, and
export of aquatic insects of a Sonoran desert stream. Ecology
Jackson JK, Sweeney BW (1995) Egg and larval development times
for 35 species of tropical stream insects from Costa Rica. J North
Am Benthol Soc 14:115–130
Jarnagin ST, Swan BK, Kerfoot CW (2000) Fish as vectors of
dispersal of Bythotrephes cederstroemi diapausing eggs survive
passage through the gut. Freshw Biol 43:579–590
Kamler JF (2001) Nonlethal methods of examining fish stomach
contents. Rev Fish Sci 9:1–11
Kouki J (1991) Tracking spatially variable resources: an experimental study
on the oviposition of the water-lily beetle. Oikos 61:243–249
Krakauer DC (1995) Groups confuse predators by exploiting
perceptual bottlenecks: a connectionist model of the confusion
effect. Behav Ecol Sociobiol 36:421–429
Li KT, Wetterer JK, Hairston NG (1985) Fish size, visual resolution,
and prey selectivity. Ecology 66:1729–1735
McCarty JP, Winkler DW (1999) Foraging ecology and diet selectivity
of tree swallows feeding nestlings. Condor 101:246–254
McMillan VE (2000) Aggregation behavior during oviposition in the
dragonfly Sympetrum vicinum (Hagen) (Odonata: Libellulidae).
Am Midl Nat 144:11–18
Michiels NK, Dhondt AA (1988) Direct and indirect estimates of sperm
precedence and displacement in the dragonfly Sympetrum danae
(Odonata: Libellulidae). Behav Ecol Sociobiol 23:257–263
Michiels NK, Dhondt AA (1990) Costs and benefits associated with
oviposition site selection in the dragonfly Sympetrum danae
(Odonata, Libellulidae). Anim Behav 40:668–678
Murakami M, Nakano S (2001) Species-specific foraging behavior of
birds in a riparian forest. Ecol Res 16:913–923
Nakano S, Murakami M (2001) Reciprocal subsidies: dynamic
interdependence between terrestrial and aquatic food webs. Proc
Natl Acad Sci U S A 98:166–170
Needham JG, Morton KJ, Johannsen OA (1905) Mayflies and midges
of New York. Bull N Y St Mus 86(Entomology 23):1–352 (New
York State Education Department, Albany, NY)
Needham JG, Traver JR, Hsu Y-C (1935) The biology of mayflies.
Comstock, Ithaca
Ohgushi T (1992) Resource limitation on insect herbivore popula-
tions. In: Hunter MD, Ohgushi T, Price PW (eds) Effects of
resource distribution on animal–plant interactions. Academic,
San Diego, pp 199–241
Palmer TM (1995) The influence of spatial heterogeneity on the
behavior and growth of 2 herbivorous stream insects. Oecologia
Peckarsky BL (1985) Do predaceous stoneflies and siltation affect the
structure of stream insect communities colonizing enclosures.
Can J Zool 63:1519–1530
Peckarsky BL (1991) Habitat selection by stream dwelling predatory
stoneflies. Can J Fish Aquat Sci 48:1069–1076
Peckarsky BL, Taylor BW, Caudill CC (2000) Hydrologic and
behavioral constraints on oviposition of stream insects: implica-
tions for adult dispersal. Oecologia 125:186–200
Peckarsky BL, Taylor BW, McIntosh AR, McPeek MA, Lytle DA
(2001) Variation in mayfly size at metamorphosis as a develop-
mental response to risk of predation. Ecology 82:740–757
Peckarsky BL, McIntosh AR, Caudill CC, Dahl J (2002) Swarming
and mating behavior of a mayfly Baetis bicaudatus suggest
stabilizing selection for male body size. Behav Ecol Sociobiol
Pister EP (2001) Wilderness fish stocking: History and perspective.
Ecosystems 4:279–286
Pleskot G, Pomeisl E (1952) Bedeutung der Lichtintensitat beim
Schlupfen und bei der Eiablage von aquatischen Insekten, im
besonderen von Torleya belgica. Wett Leben Sonderheft I:41–47
Price PW (1994) Phylogenetic constraints, adaptive syndromes, and
emergent properties: from individuals to population dynamics.
Res Popul Ecol 36:3–14
Price PW, Craig TP, Hunter MD (1998) Population ecology of a gall-
inducing sawfly, Euura lasiolepis, and relatives. In: Dempster JP,
Mc Lean IFG (eds) Insect populations, 1st edn. Kluwer,
Dordrecht, pp 324–340
Reich P, Downes BJ (2003) Experimental evidence for physical cues
involved in oviposition site selection of lotic hydrobiosid
caddisflies. Oecologia 136:465–475
Richards C, Minshall GW (1988) The influence of periphyton
abundance on Baetis bicaudatus distribution and colonization in
a small stream. J North Am Benthol Soc 7:77–86
Russev B (1959) Vol de compensation pour la ponte de Palingenia
longicauda Oliv. (Ephem.) countre le courant de Danube. C R
Acad Bulg Sci 12:165–168
Sabo JL, Power ME (2002) Numerical response of lizards to aquatic
insects and short-term consequences for terrestrial prey. Ecology
SAS Institute Inc. (2003) User's Guide for SAS Software Navigator.
SAS Institute, Cary, NC
Savolainen E, Saura A, Hantula J (1993) Mode of swarming in
relation to reproductive isolation in mayflies. Evolution 47:1796–
Smith RL (1976) Male brooding behavior of the water bug Abedus
herberti (Hemiptera: Belostomatidae). Ann Entomol Soc Am
Sokal RR, Rohlf JF (1981) Biometry, 2nd edn. W.H. Freeman and
Company, San Francisco
Behav Ecol Sociobiol (2007) 61:1437–1448 1447
A Comparative Study Of The Costs Of Alternative Mayfly Oviposition
Sweeney BW, Vannote RL (1982) Population synchrony in mayflies: a
predator satiation hypothesis. Evolution 36:810–821
Thompson JN (1 988) Evolutionary ecology of the relationship
between oviposition preference and performance of offspring in
phytophagous insects. Entomol Exp Appl 47:3–14
Thompson JN, Pellmyr O (1991) Evolution of oviposition behavior
and host preference in Lepidoptera. Annu Rev Entomol 36:65–
Vance SA (1996) Morphological and behavioural sex reversal in
mermithid-infected mayflies. Proc R Soc Lond B 263:907–912
Waage JK (1987) Choice and utilization of oviposition sites by female
Calopteryx maculata (Odonata: Calopterygidae). I. Influence of
site size and the presence of other females. Behav Ecol Sociobiol
Wayland M, Trudeau S, Marchant T, Parker D, Hobson KA (1998)
The effect of pulp and paper mill effluent on an insectivorous
bird, the tree swallow. Ecotoxicology 7:237–251
Wildermuth H, Spinner W (1991) Visual cues in oviposition site
selection by Somatochlora artica (Zetterstedt) (Anisoptera:
Corduliidae). Odonatologica 20:357–367
1448 Behav Ecol Sociobiol (2007) 61:1437–1448