Insect Science (2016) 00, 112, DOI 10.1111/1744-7917.

12317

ORIGINAL ARTICLE

Does host plant quality affect the oviposition decisions

of an omnivore?

Meghan A. Vankosky and Sherah L. VanLaerhoven

Department of Biological Sciences, University of Windsor, 401 Sunset Avenue Windsor, Ontario, N9B 3P4, Canada

Abstract Optimal oviposition theory predicts a positive relationship between female pref-
erence for oviposition hosts and offspring performance. Interspecies effects on oviposition
preference have been widely investigated, especially for herbivores. However, intraspecies
variation, such as nitrogen content, might also influence female preference for oviposition
hosts and subsequent offspring performance. To evaluate this possibility, we investigated
the oviposition preference of a zoophytophagous omnivore and the development and sur-
vival of its nymphs on a single species of host plant that varied in nitrogen content. In choice
and no-choice experiments without prey, female omnivores were allowed to oviposit on
plants that had been fertilized using 4 rates of nitrogen fertilizer (39, 78, 156, and 311 mg/L
nitrogen) for 72 h. After 72 h, the most females were found on tomato plants receiving
high concentrations of nitrogen fertilizer and more eggs were laid on those plants. First
instar nymphs developed more rapidly on high-nitrogen plants and third instar nymphs
developed faster on low-nitrogen plants. Plant nitrogen did not affect nymph survival to
the adult stage, or the probability of survival over time. Although female omnivores did
discriminate between potential oviposition hosts based on plant nitrogen, their choices did
not significantly impact nymph development or survival. This is the first study to show that
intraspecies variation in nitrogen content between plants affects the oviposition preference
of female omnivores, but not offspring performance.
Key words biological control; Dicyphus hesperus; Hemiptera; nitrogen; omnivore; op-
timal oviposition

Introduction theory, or the preferenceperformance hypothesis is tested

Due to the costs associated with parental care (Krebs &
Davies, 1993), many insect species simply deposit their
eggs in presumably suitable habitats and leave the resul-
tant offspring to fend for themselves (Jaenike, 1978; Tal-
lamy, 1999). For these species, optimal oviposition theory
predicts that females should endeavor to maximize their
own fitness by carefully selecting oviposition hosts, or
locations, that will result in the greatest offspring perfor-
mance and survival (Jaenike, 1978). Optimal oviposition
Correspondence: Meghan A. Vankosky, Department of En-
tomology, University of California Riverside, 900 University
Avenue, Riverside, CA 92521, USA. Tel: +1 951 827 4360; fax:
+1 951 827 3086; email: meghan.vankosky@gmail.com
by evaluating both female preference for oviposition sites
and the performance of offspring at those oviposition sites
(Thompson, 1988).
Strong correlations between female preference and off-
spring performance have been observed for herbivores in-
cluding Polygonia c-album (L.) (Lepidoptera: Nymphali-
dae) (Nylin & Janz, 1993); predators including Coccinella
septempunctata L. (Coleoptera: Coccinellidae), which
avoids laying eggs near declining aphid (Hemiptera: Aphi-
didae) colonies (Hemptinne et al., 1993); and omnivores
including Orius insidiosus (Say) (Hemiptera: Anthocori-
dae) (Coll, 1996). However, for many species no strong
correlations between female preference and offspring per-
formance have been observed (Thompson, 1988; Jaenike,
1990; Mayhew, 1997).

1

C 2016 Institute of Zoology, Chinese Academy of Sciences
2 M. A. Vankosky & S. L. VanLaerhoven
There are a number of explanations for the absence of
preferenceperformance correlations. For example, dif-
ferent measures of offspring performance might corre-
late with female preference better than others (Thompson,
1988; Nylin et al., 1996). Nylin et al. (1996) suggest that
in order to fully test optimal oviposition theory for any
given species, the most appropriate measure of offspring
performance is their fitness, but to measure offspring fit-
ness each individual would have to be followed for its
entire life cycle. This would be extremely difficult and
time consuming, especially for long-lived species. Other
possible explanations for the lack of correlation between
preference and performance include the Time Hypothe-
sis, the Patch Dynamics Hypothesis, and the Enemy Free
Space Hypothesis (reviewed by Thompson, 1988). More
recently, it has been suggested that many species are bad
mothers that select host plants that maximize their own
performance and fecundity at the expense of their off-
spring (Scheirs et al., 2000; Mayhew, 2001; Scheirs & De
Bruyn, 2002), as has been observed for the grass miner
Chromatomyia nigra (Meigen) (Diptera: Agromyzidae)
(Scheirs et al., 2000).
Host plant quality varies in the field and may also con-
tribute to the weakness of correlations between adult pref-
erence and offspring performance, as Nylin et al. (1996)
suggest. For example, Leather (1985) observed that Pano-
lis flammea Dennis and Schiffermueller (Lepidoptera:
Noctuidae) preferred to oviposit on Pinus contorta Dou-
glas (Pinales: Pinaceae) from specific geographic regions
with different climatic conditions. Secondary metabolites
of host plants also vary in the field and this variation
is known to affect host plant preference for oviposition
(Wheeler & Ordung, 2005). For example, the terpenoid
profile of Melaleuca quinquenervia (Cav.) S.T. Blake
(Myrtales: Myrtaceae) affects plant selection by Boreio-
glycaspis melaleucae Moore (Hemiptera: Psyllidae), such
that plants with a terpenoid profile primarily consist-
ing of viridiflorol were preferred relative to those with
a terpenoid profile consisting of primarily of E-nerolidol
(Wheeler & Ordung, 2005). Plant nitrogen (N) levels also
tend to vary greatly within plant species, both tempo-
rally and spatially (Mattson, 1980). Nitrogen content af-
fects the oviposition preferences of herbivores, includ-
ing the greenhouse whitefly, Trialeurodes vaporariorum
Westwood (Hemiptera: Aleyrodidae), such that increasing
nitrogen content resulted in increasing numbers of eggs
deposited (Jauset et al., 1998). This is logical because in-
creasing plant nitrogen content is known to contribute to
increasing size, fecundity, survival, and development rate
of many herbivores (Mattson, 1980; Hunt et al., 1992).
Because nitrogen content varies between individual plants

C 2016 Institute of Zoology, Chinese Academy of Sciences, 00, 112
of the same species, it is likely that this variation affects
the oviposition behavior of a wide range of herbivorous
insects.
Omnivores obtain nutrients and energy from multiple
trophic levels (Coll & Guershon, 2002). The cues driving
oviposition choices of omnivores are not as well under-
stood as those of herbivores and predators. Their selection
of sites for oviposition may depend on host plant quality,
prey quality, or a combination of both, as well as prey
abundance. Work has been done to identify preferred host
plant species of omnivores (Coll, 1996; Coll & Guer-
shon, 2002; Sanchez et al., 2004; Lundgren & Fergen,
2006; Seagraves & Lundgren, 2006). However, few stud-
ies address the fact that plant suitability for oviposition
and juvenile development likely varies between plants of
the same species. Thus, the goal of this research was to
determine if plant nitrogen content affects the oviposi-
tion decisions of an omnivore, Dicyphus hesperus Knight
(Hemiptera: Miridae).
Dicyphus hesperus is a zoophytophagous omnivore
that consumes a diet dominated by animal foods com-
plimented by water and nutrients gained via consump-
tion of plant materials (Gillespie & McGregor, 2000;
Coll & Guershon, 2002). Therefore, this omnivore spends
the majority of its time feeding on prey (Vankosky &
VanLaerhoven, 2015). Its preferred prey species include
greenhouse whitefly nymphs, two-spotted spider mites
(Tetranychus urticae Koch; Acari: Tetranychidae), and
Ephestia kuehniella Zeller (Lepidoptera: Pyralidae) eggs
(McGregor et al., 1999; VanLaerhoven et al., 2006). In
the absence of prey foods, D. hesperus can survive on a
strictly plant-based diet, but development and reproduc-
tion requires prey feeding or that D. hesperus be reared
on mullein (Verbascum thapsus L.; Lamiales: Scorphu-
lariaceae) (Sanchez et al., 2004). Dicyphus hesperus is
used as a biological control agent of greenhouse white-
fly (McGregor et al., 1999) in environments where plant
nitrogen levels are often manipulated to maximize plant
yield. Therefore, testing its oviposition preferences has
both applied and theoretical value.
The oviposition preferences of D. hesperus were exam-
ined using choice and no-choice tests. For both tests, fe-
male omnivores were expected to prefer plants with more
nitrogen (plants receiving 156 mg/L nitrogen fertilizer)
relative to plants with less nitrogen (plants receiving <156
mg/L nitrogen fertilizer) as oviposition and feeding hosts.
Based on the results of these oviposition preference ex-
periments, we then observed the development time and
survival of D. hesperus nymphs on the most and least
preferred plants to determine if females made optimal
decisions.

Materials and methods
Insect and plant rearing
A colony of D. hesperus was maintained at the Univer-
sity of Windsor since 2005. The colony was maintained
in a facility built to the standards of the Canadian Coun-
cil on Animal Care (www.ccac.ca). The CCAC does not
require permits for experimentation using insect subjects;
however, all efforts were made to ensure that insect mor-
tality was minimized during experiments. The colony was
originally established with insects collected from Stachys
albens A. Gray (Lamiales: Lamiaceae) (white stem hedge
nettle) in California, USA (3542 15 N, 11850 00 W, ap-
proximately 500-m elevation) (McGregor & Gillespie,
2004). Conditions in the rearing laboratory at the Uni-
versity of Windsor were maintained at 20 5 C with a
light : dark photoperiod of 16 : 8 h and 50% 10% rela-
tive humidity provided by a humidifier (Nortec Humidity
Inc., Ogdensburg, NY, USA).
In the rearing colony, adult D. hesperus were held on
one of their host plants, Nicotiana tabacum L. (Solanales:
Solanaceae) (McGregor et al., 1999). Rearing cages were
modified 61 cm3 collapsible cages with mesh sides and
top (BioQuip Products, Rancho Dominguez, CA, USA).
Each cage held 2 or 3 N. tabacum plants with 48 large
leaves. After 710 d, adults were removed from the ovipo-
sition plants using an aspirator and the adult-free plants
were placed inside a second rearing cage to allow nymphs
to hatch and develop isolated from adults, as D. hespe-
rus are opportunistic cannibals (Laycock et al., 2006).
Dicyphus hesperus have 4 nymphal instars and the devel-
opment period from oviposition to adult eclosion ranges
from approximately 1530 d at temperatures ranging
from 19 to 27 C (Gillespie et al., 2004). In the rear-
ing room, 4 colony cages were maintained: 1 oviposition
cage, and 3 developmental cages, resulting in a new co-
hort of adult insects emerging approximately once per
week. Approximately 364 cohorts of D. hesperus were
produced between the initiation of the colony and when
this study was conducted. Throughout development, D.
hesperus nymphs were provided with preserved Ephestia
kuehniella Zeller (Lepidoptera: Pyralidae) (Beneficial In-
sectary Inc., Guelph, Ontario, Canada), eggs as prey ad
libitum (Gillespie et al., 2004).
Tomato plants, Solanum lycopersicum L. (Solanales:
Solanaceae), were used as the experimental host plant.
The development and reproduction of D. hesperus has
largely been evaluated on tomato plants (McGregor et al.,
1999; Gillespie & McGregor, 2000; Sanchez et al., 2004),
as tomatoes are an important greenhouse crop in Canada,
worth $337 million in exports in 2013 (Darrach, 2014).

C 2016 Institute of Zoology, Chinese Academy of Sciences, 00, 112
Omnivore oviposition behavior 3
Patio Hybrid tomato seeds (Stokes Seeds Canada,
Thorold, Ontario, Canada) were sown in BM6 potting
soil containing 78% peat moss plus perlite (Berger, Saint-
Modeste, Quebec, Canada) in 8 cm 8 cm 9 cm square
black pots, to a depth of 2 cm. In the greenhouse, seedlings
sprouted under natural light and temperature regimes. At
the 4-leaf stage seedlings were transplanted into 8-inch
( 20 cm) green plant pots, filled to 2.5 cm from the top
with the same potting medium.
Transplanted seedlings were assigned to 1 of 4 ni-
trogen fertilizer treatments: 38.87, 77.75, 155.49, and
310.99 mg/L nitrogen (herein referred to as 39, 78, 156,
and 311 mg/L). Fertilizer solutions were mixed using
6-11-31 Hydroponic Fertilizer and 15.5-0-0 calcium ni-
R
trate (Plant-Prod Canada, Brampton, ON, Canada) as
described by Vankosky and VanLaerhoven (2015) and
were based on the recommended rate of nitrogen fertilizer
inputs for greenhouse tomatoes of 200 ppm of nitrogen
(156 mg/L nitrogen) (Alberta Agriculture & Rural De-
velopment, 2009). Fertilization began the day following
transplantation. Each plant received 100 mL (after Jauset
et al., 1998) of the appropriate fertilizer solution daily for
7 d (after Stout et al., 1998), and every second day there-
after (after Jauset et al., 1998). This fertilization regime
was maintained throughout the experiments and tomato
plants received water as necessary to avoid wilting.
Plants were used in oviposition and development exper-
iments 2430 d after the first day of fertilization. Plants
used in the choice experiment were measured prior to their
placement in choice arenas. The height, stem diameter,
and BBCH growth stage (Feller et al., 1995) was recorded.
At the end of the choice experiment, the aboveground
biomass of each plant was collected, dried in brown pa-
per bags at 35 5 C in the greenhouse and weighed
using a Precisa BJ 100M balance (Precisa Ltd., Tongwell,
Milton Keynes, UK). Following the completion of the
no-choice tests, a single mid-canopy branch from 4 ran-
domly selected tomato plants from each nitrogen fertilizer
treatment was harvested for nitrogen content analysis and
frozen. Before analysis, branches were thawed and dried
inside brown paper bags in the greenhouse at 35 5 C
and crushed using a mortar and pestle. Nitrogen analyses
were conducted by the Trophic Ecology/Chemical Tracers
Laboratory at the Great Lakes Institute for Environmen-
tal Research (GLIER) using continuous flow isotope ratio
mass spectroscopy; results are presented as mean percent
nitrogen content.
No-choice and choice preference trials
In no-choice trials, individual plants were enclosed us-
ing a mesh bag tied closed around the top edge of the
4 M. A. Vankosky & S. L. VanLaerhoven
pot. Five adult female D. hesperus, 710 d old, were col-
lected from the oviposition cage in the rearing colony and
transferred to each plant pot inside a plastic vial. The vial
was placed against the main stem of the plant, with the
cap off. Female D. hesperus were allowed to oviposit for
72 h, after which adults were removed from the cages.
Plants were inspected daily for emerging nymphs be-
ginning 7 d following the removal of females from the
cages. The first nymph collection occurred on the first
day that nymphs were observed. Nymphs were collected
on 2 additional days following the first collection, result-
ing in 3 collection dates over 6 d. Eggs require 1015 d
to hatch at temperatures between 22 and 35 C (Gillespie
et al., 2004) and mean monthly temperatures in the green-
house ranged from 25 to 28 C during the summer months
when the experiment took place. Therefore, the dates of
nymph collection were broad enough that all eggs should
have hatched before plants were destroyed on the final
day of collection; plants were destroyed so that each leaf,
branch, and the main stem could be carefully inspected for
remaining nymphs. Nymphs were collected using a fine
haired paintbrush, counted, and preserved in 70% ethanol.
Twenty plants from each nitrogen fertilizer treatment were
used for this experiment (n = 20).
In choice trials, 4 tomato plants, 1 from each fertilizer
treatment, were placed inside a 55 55 60 cm mesh-
walled cage. Plants were placed in the 4 corners of the
cage so that the plants did not touch. The arrangement of
the 4 plants was pseudo-randomized, such that in a block
of 4 cages, no 2 cages had the same arrangement. Six
blocks of 4 cages were used in the experiment (n = 24).
Female D. hesperus, 710 d old, were collected into vials
and released into each cage by placing the opened vial
on the floor in the center of the cage, equidistant from all
plants. Five female D. hesperus per plant were released
into each cage (20 females per cage). Females were al-
lowed 72 h to oviposit, after which they were removed
from the cages and the 4 plants were isolated to prevent
the movement of hatched nymphs between plants. The
location of all collected females was recorded. Nymphs
were collected following the protocol described for the
no-choice experiment.
Egg mortality
As female D. hesperus insert eggs into plant tissues
during oviposition (Gillespie et al., 2004), the number
of emerging nymphs was used as a proxy measurement
of oviposition in these experiments. To ensure that this
was an accurate proxy, egg mortality on plants from the
4 fertilizer treatments was determined in the laboratory.
Tomato branches were collected from plants that had been

C 2016 Institute of Zoology, Chinese Academy of Sciences, 00, 112
fertilized for approximately 24 d. The stems were inserted

R
into a 60 mL Solo cup (Lake Forest, IL, USA) filled
with water through a hole in the lid. At 16:00 h, female
D. hesperus were collected from the oviposition cage of
the colony and placed upon the tomato branches, with 2
or 3 females per branch. Branches were under constant
surveillance by an observer between 16:00 and 21:00 h.
The observer recorded the time of the first oviposition
event, the number of oviposition events, and all attempts at
oviposition. At the end of the observation period, females
were removed from the tomato branches, the branches
were isolated and emerging nymphs were collected from
the branches using the same protocol as that used in the
choice and no-choice tests.
Statistical analysisoviposition trials
One-way analysis of variance (ANOVA) was used to de-
termine if nitrogen fertilizer had an effect on plant height,
stem diameter, dry biomass, and plant nitrogen content
using the GLM procedure (SAS Institute, 2009). Data
for plant growth stage was not normal or homoscedastic
and could not be transformed to meet those assumptions.
Therefore, the KruskalWallis nonparametric ANOVA
equivalent was used; significant KruskalWallis results
were followed up with paired Wilcoxon tests to compare
nitrogen fertilizer treatments, with adjusted using the
Bonferroni correction (SAS Institute, 2009). Dry biomass
data were transformed using the natural log (LN) trans-
formation; raw data are presented. Where significant ef-
fects were observed for plant height, stem diameter, and
dry biomass, means were compared using the LSMEANS
statement with the PDIFF and ADJUST = BON functions
specified in the model (SAS Institute, 2009). The PDIFF
statement compares all possible pairs of treatments and
the ADJUST = BON statement returns Bonferroni ad-
justed P-values that can be compared to = 0.05 (SAS
Institute, 2009). The relationship between nitrogen fer-
tilizer and dry plant biomass was tested using the REG
procedure (SAS Institute, 2009) and biomass was tested
as a potential covariate using the GLM procedure as re-
lationships between host size and offspring number have
been previously observed (Clark et al., 2011).
No-choice trials were analyzed with one-way ANOVA
using the MIXED procedure with replicate number speci-
fied as a random effect (SAS Institute, 2009). Where sig-
nificant fertilizer treatment effects were observed, means
were compared using the PDIFF function with the AD-
JUST = BON statement (SAS Institute, 2009).
Biomass did not have a significant effect on ovipo-
sition preference in the choice experiment (see Re-
sults section). Therefore, multivariate analysis of

Fig. 1 A clip cage enclosing a single Dicyphus hesperus
nymph on the leaflet of a tomato plant.
variance (MANOVA) was used to determine if plant qual-
ity affected the preference of female D. hesperus for
oviposition and feeding hosts. We used MANOVA be-
cause choice experiments involve the acceptance of a host
and the simultaneous rejection of alternative hosts, which
violates the assumption of treatment independence re-
quired for univariate testing (Roa, 1992). The MANOVA
was conducted using the repeated measures model in
SPSS to calculate within-treatment effect statistics (Lock-
wood, 1998; IBM SPSS Statistics, 21.0, 2012), where
= 0.05. Where MANOVA results were significant, treat-
ment means were separated using pairwise comparisons,
and was adjusted using the Bonferroni correction.
The assumption that egg mortality was equal on plants
from each nitrogen fertilizer treatment was tested with chi-
square analysis using the FREQ procedure (SAS Institute,
2009). For comparison, the percentage of hatched nymphs
was calculated based on the number of oviposition events
observed and the number of emerged nymphs that were
collected. For the analysis, = 0.05.
Development and survival of D. hesperus nymphs
To evaluate the performance of D. hesperus offspring,
we monitored their development and survival on tomato
plants fertilized with 156 mg/L nitrogen (high nitro-
gen) and 39 mg/L nitrogen (low nitrogen), which were
most and least preferred by ovipositing female D. hespe-
rus, respectively (see Results section). Tomato plants
receiving 156 and 311 mg/L nitrogen differed in nitro-
gen content, but not in female preference. Therefore, we
opted to use 156 mg/L nitrogen-fertilized tomato plants,
as this rate of fertilization is most likely to be encountered
in agroecosystems. Two experiments were conducted:

C 2016 Institute of Zoology, Chinese Academy of Sciences, 00, 112
Omnivore oviposition behavior 5
(i) effect of plant qualityno prey, in which nymphs were
provided with only tomato leaves as food for the entirety
of development; and (ii) effect of plant qualitywith prey,
in which nymphs were provided with 5 E. kuehniella eggs
on Days 1, 4, 7, and 10 of the experiment.
For both experiments, first instar D. hesperus nymphs
were collected from N. tabacum plants in the rearing
colony within 24 h of emergence (following McGre-
gor et al., 1999; Gillespie & McGregor, 2000; Gillespie
et al., 2004), using a fine-bristled paintbrush. All emerged
nymphs were cleaned from N. tabacum plants the day be-
fore experiments were initiated to ensure that all nymphs
were less than 24 h old. Individual nymphs were placed
inside clip-cages on whole tomato plants in the green-
house. Clip-cages were handmade using 5.5 cm 5.5 cm
cardboard squares covered with white felt and a 50-mm
diameter Petri dish. Small holes were drilled along the
edges of the Petri dishes and covered with white fabric to
allow air circulation. Leaves of whole tomato plants were
pressed between the felt and the Petri dish and the clip-
cages were secured in place with elastic bands, as pictured
in Fig. 1. Three clip-cages, holding 1 nymph each, were
placed on 12 high-nitrogen and 12 low-nitrogen tomato
plants. Nymphs were observed daily and the date upon
which each nymph molted, emerged as an adult, or died
was recorded (McGregor et al., 1999; Gillespie & Mc-
Gregor, 2000; Gillespie et al., 2004).
A Smartbutton data logger (ACR Systems Inc., Sur-
rey, British Columbia, Canada) was used to record the
temperature in the greenhouse on an hourly basis during
nymph development. Using daily temperature data, we
calculated degree-days (d) using: d = [(Tmax + Tmin )/2]
Tbase , where Tmax and Tmin are the daily maximum and
minimum temperatures and Tbase is the developmental
threshold temperature for the study species (McMaster &
Wilhelm, 1997). For D. hesperus, we used Tbase values of
8.01 C for males and 7.8 C for females (Gillespie et al.,
2004). The total (accumulated) d required for develop-
ment of a single D. hesperus nymph was calculated using
thermal summation (Gullan & Cranston, 2005). To avoid
overlap in days between instars, we defined an instar as
the first day that a given instar was observed to the day be-
fore the exoskeleton from the molt to the subsequent instar
was found inside the clip-cage. Where nymphs reached the
adult stage, gender specific Tbase values could be used, for
nymphs that died, gender average d values were used for
thermal summation.
Statistical analysisdevelopment trials
We tested 3 predictions for both development experi-
ments (Effect of plant qualityno prey and Effect of plant
6 M. A. Vankosky & S. L. VanLaerhoven

Table 1 Treatment means ( SE) for measurements of physical plant characteristics were compared with one-way ANOVA; in each
row, means with the same letters were not significantly different (P > 0.05).

Nitrogen fertilizer treatment

Characteristic
39 mg/L 78 mg/L 156 mg/L 311 mg/L

Growth stage (n = 24) 56 1a 57 1a 61 1b 61 1b

Stem diameter (mm) (n = 24) 8.96 0.16 ab 9.63 0.21 ab 10.44 0.25 b 9.91 0.22 b
Height (cm) (n = 24) 29.89 0.60 a 29.50 0.58 a 33.34 0.77 b 33.73 0.62 b
Dry biomass (g) (n = 24) 14.52 0.47 a 20.36 0.63 b 31.50 1.02 c 41.05 1.59 d
N-content (%) (n = 4) 0.77 0.05 a 0.93 0.04 a 1.51 0.06 b 2.32 0.16 c

Growth stage was recorded to the nearest whole growth stage using guidelines for tomato plants (Feller et al., 1995).

qualitywith prey). We first predicted that nymphs would a log-rank chi-square test (SAS Institute, 2009), where
develop faster on high-nitrogen tomato plants; to test this = 0.05. The analysis was repeated for both develop-
prediction, we used the GLM procedure (SAS Institute, ment time experiments.
2009) to determine the effect of plant nitrogen content
on the accumulated d required for development of each
Results
instar and from the first instar to adulthood. If necessary,
data was transformed using the square root transforma-
Tomato plant characteristics
tion; where data could not be transformed to meet the
assumptions of ANOVA, the KruskalWallis test was per-
Tomato plants that received 156 and 311 mg/L nitrogen
formed using the NPAR1WAY procedure (SAS Institute,
fertilizer were at a more advanced growth stage at the
2009). Results are summarized using raw values. For the
start of the choice experiments compared to plants that
analysis, = 0.05.
received 39 and 78 mg/L nitrogen fertilizer ( 2 = 22.16,
Our second prediction was that plant nitrogen con-
df = 3, P < 0.0001; Table 1). Similarly, nitrogen fertil-
tent would affect the survival of D. hesperus nymphs
izer treatment affected stem diameter (F3,92 = 8.50, P <
to the adult stage. To test this prediction, the proportion
0.0001), plant height (F3,92 = 11.89, P < 0.0001), dry
of nymphs that survived was calculated and compared
biomass (F3,92 = 173.09, P < 0.0001), and plant nitrogen
between high- and low-nitrogen plants using chi-square
content (F3,12 = 56.49, P < 0.0001) (Table 1). Plants that
tests. Proportions were used for comparison because some
received more nitrogen were larger and had more nitrogen
nymphs escaped or were killed during the experiment
in their leaves. Regression analysis showed that nitrogen
and those individuals had to be excluded from the anal-
fertilizer treatment had a significant linear effect on plant
ysis. Chi-square tests were conducted using the FREQ
biomass (F1,94 = 318.73, P < 0.0001, R2 = 0.7722), where
procedure (SAS Institute, 2009). For the Effect of plant
biomass increased as fertilizer nitrogen concentration in-
qualitywith prey experiment the CATMOD procedure,
creased. Biomass differed between all fertilizer treatments
which allows for 2-factor analyses of count data (SAS
(Table 1). Although not quantified, tomato plants that re-
Institute, 2009) was used to determine if plant quality
ceived 156 and 311 mg/L nitrogen fertilizer had dark
and time (accumulated d) since last prey feeding on Day
green foliage and plants that received 39 and 78 mg/L had
10 affected the survival of D. hesperus nymphs. For all
yellowgreen foliage.
chi-square analyses, P-values were compared to = 0.05.
Finally, we predicted that nymphs reared on high-
nitrogen tomato plants would have a greater probabil- Oviposition preference
ity of survival over time. To test this prediction, the
survivor functions of the nymphs were generated and In no-choice trials, nitrogen fertilizer treatment had an
the slopes of the survivor functions were compared us- effect on the number of nymphs per plant (F3, 57 = 2.86,
ing the LIFETEST procedure (SAS Institute, 2009). P = 0.0449; Fig. 2A). Tomato plants receiving 156 mg/L
Nonparametric estimates of the survivor functions were nitrogen fertilizer had more nymphs than plants treated
calculated for nymphs in the high-nitrogen and low- with 78 mg/L fertilizer. The mean numbers of nymphs on
nitrogen treatments using the KaplanMeier method and plants that received 39 and 311 mg/L nitrogen fertilizer
the resultant survivor functions were compared using were not different from the other treatments.


C 2016 Institute of Zoology, Chinese Academy of Sciences, 00, 112

Fig. 2 Mean (SE) number of Dicyphus hesperus nymphs col-
lected from tomato plants from each nitrogen fertilizer treatment
in A: no-choice arenas (n = 20), and B: choice arenas (n = 24).
Means were compared using one-way ANOVA and MANOVA,
respectively and means with the same letters were not signifi-
cantly different (Bonferroni adjusted P > 0.05).
Although dry plant biomass increased linearly as the
rate of nitrogen fertilization increased, it did not have an
effect on the mean number of D. hesperus nymphs per
plant (F1,94 = 0.00, P = 0.98) in choice tests; therefore,
biomass was not treated as a covariate in the model and
MANOVA was used determine the effect of nitrogen fer-
tilizer on the oviposition preferences of female D. hespe-
rus. When given a choice of host plant, the location of
adult D. hesperus was affected by nitrogen fertilizer treat-
ment (Hotellings Trace = 30.559, df = 3,21, P < 0.0001).
After 72 h, the most females were found on plants receiv-
ing 156 mg/L nitrogen fertilizer and the fewest were found
on plants receiving 39 mg/L nitrogen fertilizer (Fig. 3).
The number of D. hesperus nymphs collected per plant
in choice arenas was also affected by fertilizer treatment
(Hotellings Trace = 10.624, df = 3, 21, P < 0.0001),
with the most nymphs found on plants fertilized with 156
and 311 mg/L fertilizer, and the fewest on plants treated
with 39 mg/L fertilizer (Fig. 2B).

C 2016 Institute of Zoology, Chinese Academy of Sciences, 00, 112
Omnivore oviposition behavior 7
Fig. 3 Mean ( SE) number of adult Dicyphus hesperus located
on plants of each nitrogen fertilizer treatment after 72 h in choice
arenas (n = 24); means were compared with one-way MANOVA
and means with the same letters were not significantly different
(Bonferroni adjusted P > 0.05).
In the 5 h during which oviposition events were directly
observed in the laboratory, female D. hesperus given no
choice of host plant only attempted to oviposit on the
stems or large leaf veins of tomato plants. Oviposition
events were observed on 4 of 7 (57%) and on 3 of 5 (60%)
tomato branches grown with 39 and 78 mg/L nitrogen fer-
tilizer, respectively. Oviposition events were observed on
4 of 4 (100%) tomato branches grown with 156 and 311
mg/L nitrogen fertilizer. Only branches with oviposition
events were included in the analyses (n = 4, 3, 4, and 4
for 39, 78, 156, and 311 mg/L nitrogen fertilizer, respec-
tively). The mean time ( SE) to the first oviposition event
was 100 49 min, 197 41 min, 82 9 min, and 94
41 min for females on 39, 78, 156, and 311 mg/L nitrogen-
fertilized tomato branches, respectively. The number of
oviposition events on individual branches ranged from 1
5 across all nitrogen treatments in the 5 h observation
period. Egg mortality was less than 20% on plants from
all treatments, with no difference in mortality between
treatments ( 2 = 2.8092, df = 3, P = 0.4220); means of
83%, 100%, 81%, and 82% of eggs hatched on 39, 78,
156, and 311 mg/L nitrogen-fertilized tomato branches,
respectively.
Development and survival of D. hesperus nymphs
Plant quality did affect the development of first instar
nymphs that were reared on whole plants with no prey
provided (F1,21 = 8.18, P = 0.0094), such that these
nymphs developed faster when reared on high-nitrogen
tomato (Table 2). Plant quality did not affect the develop-
ment of second (F1,17 = 1.41, P = 0.2507), third (F1,10 =
3.73, P = 0.0824), or fourth instar nymphs ( 2 = 0.3333,
8 M. A. Vankosky & S. L. VanLaerhoven

Table 2 The mean ( SE) accumulated degree days (d) required for the development of Dicyphus hesperus nymphs reared on whole
high- and low-nitrogen tomato plants.

Tomato plant quality

Developmental stage
Low nitrogen High nitrogen

Effect of plant qualityno prey

First instar 129.19 15.18 (n = 11) 84.97 5.12 (n = 12)
Second instar 142.59 30.22 (n = 9) 106.26 10.24 (n = 10)
Third instar 113.48 7.22 (n = 6) 143.40 13.72 (n = 6)
Fourth instar 192.28 7.92 (n = 2) 214.91 36.68 (n = 3)
First to adult 460.53 13.17 (n = 2) 464.87 53.48 (n = 3)
Effect of plant qualitywith prey
First instar 76.69 3.65 (n = 12) 77.98 3.46 (n = 12)
Second instar 63.95 2.10 (n = 12) 62.94 2.25 (n = 12)
Third instar 71.19 1.52 (n = 12) 74.86 1.92 (n = 12)
Fourth instar 109.95 12.39 (n = 4) 177.48 34.76 (n = 8)
First to adult 306.21 16.56 (n = 4) 396.98 36.36 (n = 8)
Effect of plant qualitywith prey, survivors only
First instar 73.51 7.85 (n = 4) 73.58 3.59 (n = 8)
Second instar 58.01 2.58 (n = 4) 64.76 5.26 (n = 8)
Third instar 64.96 2.68 (n = 4) 81.39 2.71 (n = 8)
Fourth instar 109.95 12.39 (n = 4) 177.48 34.76 (n = 8)
First to adult 306.21 16.56 (n = 4) 396.98 36.36 (n = 8)
Note: Means were compared using ANOVA; significant results are shown in bold font (P < 0.05).

df = 1, P = 0.5637) when nymphs were reared in clip-

cages on whole plants with no prey provided (Table 2).
Seventy-two D. hesperus nymphs were reared on whole
tomato plants without prey. Of those nymphs, 60 died,
6 were lost or killed, and 6 survived to the adult stage;
9.1% of nymphs reared on both high- and low-nitrogen
tomato plants survived to the adult stage. Plant quality
did not affect the probability of nymph survival over time
(log-rank 2 = 0.1129, df = 1, P = 0.7369; Fig. 4).
When nymphs were provided with prey, plant quality
did not affect the development of nymphs at any stage
of development (first instar: F1,22 = 0.07, P = 0.7992;
second instar: F1,22 = 0.11, P = 0.7459; third instar: Fig. 4 The survivor functions describing the probability of sur-
F1,22 = 2.23, P = 0.1494; fourth instar: 2 = 2.3447, vival of Dicyphus hesperus reared on tomato plants fertilized
df = 1, P = 0.1257; Table 2). Because survival was with 39 (---) and 156 () mg/L nitrogen fertilizer, without prey,
greater in this experiment than in the experiment with- over time (accumulated degree days).
out prey, we repeated the analysis for only the surviving
nymphs. Of these, nymphs reared on low-nitrogen tomato
required less time for third instar development compared nymphs to develop from the first instar to the adult stage
to nymphs reared on high-nitrogen tomato (F1,10 = 14.48, (F1,10 = 2.84, P = 0.1227; Table 2). Most nymphs had
P = 0.0035; Table 2) but nitrogen treatment did not af- reached the third instar when feeding ceased.
fect the development time of any other instar (first instar: Of the 72 nymphs reared with prey until Day 10, 3
F1,10 = 0.00, P = 0.9925; second instar: F1,10 = 0.75, were lost and 17 survived to adulthood. There was no
P = 0.4079; fourth instar: F1,10 = 1.75, P = 0.2135; difference in the percent survival of nymphs reared on
Table 2). Plant nitrogen did not affect the time required for high-nitrogen (27.3%) and low-nitrogen (22.2%) tomato


C 2016 Institute of Zoology, Chinese Academy of Sciences, 00, 112

Fig. 5 The survivor functions, beginning on Day 10, describing
the probability of survival of Dicyphus hesperus nymphs reared
on 39 (---) and 156 () mg/L nitrogen-fertilized tomato plants
when nymphs were provided with prey food; for nymphs reared
on plants from both levels of nitrogen, the probability of survival
never fell to zero, meaning that the nymphs that survived for the
longest time reached the adult stage.
( 2 = 0.5102, df = 1, P = 0.4751). Of the nymphs reared
on low-nitrogen tomato, 7 of the 8 survivors emerged
as adults in the first 200 d after prey feeding ceased;
the final survivor emerged as an adult after 210 d. Of the
nymphs reared on high-nitrogen tomato, 9 adults emerged.
Of those 9, 4 emerged within the first 200 d after the
cessation of feeding and 5 emerged after accumulating
more than 200 d (ranging from 210 to 425 d). Using
the percentages of nymphs that died or survived after
Day 10, we observed that both plant nitrogen content
( 2 = 6.25, df = 1, P = 0.0124) and accumulated d since
prey feeding ceased ( 2 = 63.18, df = 1, P < 0.0001)
affected D. hesperus nymphs, such that nymphs reared
on low-nitrogen plants emerged as adults faster, but that
a greater percentage of nymphs reared on high-nitrogen
tomato reached adulthood after accumulating more than
200 d.
The probability of nymph survival over time was not
affected by nitrogen fertilizer treatment (log-rank 2 =
3.2494, df = 1, P = 0.0715). When survivor functions
were calculated beginning on Day 10, when prey feed-
ing ceased, nitrogen fertilizer treatment did not affect the
probability of nymph survival over time (log-rank 2 =
2.7214, df = 1, P = 0.0990; Fig. 5).
Discussion
Nitrogen is vital for insect development and is a limiting
resource (Mattson, 1980). Therefore, we predicted that
omnivores would prefer to lay their eggs on tomato plants
treated with 156 mg/L nitrogen fertilizer (high nitrogen)

C 2016 Institute of Zoology, Chinese Academy of Sciences, 00, 112
Omnivore oviposition behavior 9
in choice and no-choice arenas, relative to tomato plants
treated with 78 mg/L nitrogen fertilizer (low nitrogen).
Our results support this prediction, as the most nymphs
were collected from 156 mg/L nitrogen-fertilized plants in
both choice and no-choice arenas, with no difference be-
tween 156 and 311 mg/L nitrogen-fertilized tomato plants.
In addition, female D. hesperus observed in the labora-
tory only laid eggs on half of the tomato branches offered
from low nitrogen plants compared to all of the branches
from high-nitrogen tomato plants. In a study similar to
ours, Groenteman et al. (2006) observed that plant ni-
trogen content affected the behavior and oviposition site
selection of Orius albidipennis Reuter (Hemiptera: An-
thocoridae) at the within-plant level, but observed that
nitrogen content did not affect its oviposition preferences
for different host plants of the same species. To our knowl-
edge, ours is the first study to demonstrate that omnivores
may detect and respond to intraspecies variation in host
plant quality during oviposition.
One shortcoming of our tests of oviposition preference
is that neither was designed to identify the plant cues
used by D. hesperus when discriminating between plants
of different nitrogen content. In olfactometer studies and
greenhouse releases, D. hesperus were attracted to plant
odors (McGregor & Gillespie, 2004) and in Petri dish are-
nas, Hazard (2008) found that D. hesperus used a com-
bination prey odors and visual cues to detect prey. Based
on these works, we expect that D. hesperus select plant
hosts using a combination of visual and olfactory cues;
future work is needed to answer this important question
in order to fully understand the oviposition decisions of
this zoophytophagous omnivore.
The development and survival of D. hesperus nymphs
on whole tomato plants with high nitrogen (156 mg/L) and
low nitrogen (39 mg/L), in the presence and absence of
prey, was assessed following the oviposition preference
tests to determine if female D. hesperus made choices
that optimized offspring performance, as theory predicts
(Jaenike, 1978; Thompson, 1988). Plants of differing ni-
trogen content did not affect the survival of D. hesperus
nymphs over time. This was true for both experiments,
with and without prey, although the effect of nitrogen
neared significance in the experiment where prey was pro-
vided until Day 10. We suspect that this may be attributed
to the abrupt loss of prey food from the nymphs diet af-
ter Day 10, such that nymphs forced to adopt a strictly
plant-based diet did respond to differences in plant nitro-
gen content. This effect would likely have been statisti-
cally significant if a larger sample size had been used for
the experiment.
The effects of plant nitrogen content on the develop-
ment of D. hesperus nymphs differed between instars:
10 M. A. Vankosky & S. L. VanLaerhoven
first instar nymphs developed faster when reared on high-
nitrogen tomato and third instar nymphs developed faster
when reared on low-nitrogen tomato. The first instars of
some predatory insects are not able to prey feed (Coll
& Guershon, 2002); whether or not first instar D. hes-
perus nymphs can prey feed is unknown. If they cannot
prey feed, then females may exhibit a preference for high-
nitrogen oviposition hosts in order to provide first instar
nymphs with a developmental advantage. However, the
effect of plant nitrogen content on development does not
extend to the second instar, and was reversed during third
instar development. Therefore, the overall results from
our development and survival study do not support our
prediction that D. hesperus females make optimal ovipo-
sition decisions. Other aspects of offspring performance
that were not measured in this study (i.e., lifetime fitness
of offspring, fecundity, size, weight gain, etc.) might cor-
relate with female preference; these should be assessed in
the future.
In the choice test, we collected more adult female
D. hesperus from 156 and 311 mg/L nitrogen-fertilized
plants after 72 h than from 39 and 78 mg/L nitrogen-
fertilized plants and subsequently found more nymphs on
tomato plants with high-nitrogen content. Because female
D. hesperus preferred high-nitrogen tomato plants for
habitat and diet, it is possible that they only deposited eggs
on these particular plants out of convenience, suggesting
that female D. hesperus may be optimal bad mothers.
However, it is important to note that female D. hesperus
did not have access to prey in these arenas, as the current
study was designed to provide insight into the oviposition
decisions of D. hesperus in habitats where prey is absent.
The availability of prey affects D. hesperus preference for
different host plant species (Sanchez et al., 2004), and
their patch residence time (VanLaerhoven et al., 2006). In
test arenas with prey available, plant nitrogen content may
be of lesser importance when selecting oviposition hosts,
as prey are more nutrient-rich and D. hesperus spend more
time feeding on prey than on any other behavior in their
repertoire (Vankosky & VanLaerhoven, 2015). Because
the development of nymphs provisioned with prey was
not affected by plant nitrogen content, we might predict
that female D. hesperus should express no preference for
prey-infested tomato that vary in nitrogen content. There
are other factors that might also influence the oviposition
preferences of D. hesperus in the presence of prey; for
example, we know that other predators avoid declining
prey colonies (Hemptinne et al., 1993).
It is also possible that female D. hesperus do not as-
sess potential host plants in terms of suitability for their
offspring, but in terms of suitability for their prey. Green-
house whitefly is a preferred prey species of D. hespe-

C 2016 Institute of Zoology, Chinese Academy of Sciences, 00, 112
rus that prefer to oviposit on high-nitrogen tomato plants
(Jauset et al., 1998; Ortega-Arenas et al., 2006) and
whose populations grow faster on high-nitrogen tomato
plants (Jauset et al., 2000). Female D. hesperus in this
experiment preferred the same tomato plants that their
prey would prefer. Some predators are known to dis-
criminate between oviposition sites based on present
and future prey availability, such as C. septempunc-
tata (Hemptinne et al., 1993). Therefore, it is possible
that D. hesperus assessed the tomato plants based on
their potential as prey hosts, not as hosts for their off-
spring. This is an interesting possibility that deserves
investigation.
Although female D. hesperus exhibited a preference
for high-nitrogen tomato for habitat and oviposition, a
substantial number of nymphs were collected from low-
nitrogen tomato (no-choice: 17.9 2.1; choice: 7.9
1.1). Time-limited dispersers may lay eggs on low-quality
plants as a tradeoff between current opportunity and lost
opportunity if search time for a better quality host is ex-
cessive (Levins & MacArthur, 1969; Mayhew, 1997). We
do not believe that this is the case for D. hesperus, as
adults may live for upwards of 28 d (Gillespie & Quiring,
2005) and are more likely to be egg-limited. If D. hespe-
rus females are egg-limited, they should be selective and
avoid suboptimal host plants, unless time becomes short
and an egg absolutely must be deposited, as predicted
by Jaenike (1978) and Mayhew (1997). We expect that
density dependent factors led to D. hesperus oviposition
on low-nitrogen tomato. Fretwell and Lucas (1970) sug-
gested that competitors of equivalent ability should dis-
tribute themselves evenly across the landscape over time,
such that each individual has equal access to resources.
Dicyphus hesperus are reared in high-density conditions
with more than 100 adults per N. tabacum plant and no
opportunity for dispersal (M.A. Vankosky, personal ob-
servation). In choice arenas, 20 females were released
(5 per plant), and relatively low numbers of D. hespe-
rus adults were collected on low-nitrogen plants at the
end of the oviposition period, suggesting that D. hesperus
distributed themselves throughout the arena based on re-
source quality. We expect that low-nitrogen plants might
be perceived as good alternative hosts due to less com-
petition between adults and potentially less competition
between nymphs. Because D. hesperus are opportunis-
tic cannibals (Laycock et al., 2006), low-nitrogen plants
may also represent enemy free space, which can influence
oviposition decisions (Thompson, 1988).
Within a biological control framework, our results
should be encouraging for scientists and producers inter-
ested in employing omnivores as biological control agents.
When provided with plants of the same species, but of

varying nitrogen content, the oviposition behavior of D.
hesperus, and potentially other omnivores, is predictable.
Although D. hesperus females do not appear to make op-
timal decisions, they do prefer to oviposit on the plants
that are most likely to support large populations of their
preferred prey, the greenhouse whitefly. Therefore, if fe-
male D. hesperus are released into a habitat without prey,
or into a habitat with low prey density, D. hesperus pop-
ulations may begin to establish where their prey are most
likely be in the future. This is important because early es-
tablishment of natural enemies is necessary to ensure that
natural enemies have a significant impact on pest popula-
tions (Wiedenmann & Smith, 1997). Our results add to the
existing knowledge of D. hesperus behavior and provide
a framework in which to study other zoophytophagous
omnivores in order to understand how their reproductive
behaviors may affect their distribution and impact over
the landscape.
Acknowledgments
We thank V. Pacheco and K. Hans for their assistance
whenever it was needed. Nitrogen content of plants was
analyzed by A. Hussey, K. Johnson, and A. Fisk from the
Trophic Ecology/Chemical Tracers Laboratory at GLIER.
Sources of funding for this research include the Natural
Sciences and Engineering Research Council of Canada,
the Ontario Graduate Scholarship, the Government of On-
tario Early Researcher Award, the Canadian Foundation
for Innovation, the Ontario Innovation Trust, and the Uni-
versity of Windsor.
Disclosure
The authors acknowledge no potential conflicts of interest.
References
Alberta Agriculture and Rural Development (2009) Com-
mercial greenhouse tomato production: Introduction.
www1.agric.gov.ab.ca/$department/deptdocs.nsf/all/
opp7556#4 (accessed February 09, 2011).
Clark, K.E., Hartley, S.E. and Johnson, S.N. (2011) Does
mother know best? The preference-performance hypothesis
and parent-offspring conflict in aboveground-belowground
herbivore life cycles. Ecological Entomology, 36, 117124.
Coll, M. (1996) Feeding and ovipositing on plants by an omniv-
orous insect predator. Oecologia, 105, 214220.
Coll, M. and Guershon, M. (2002) Omnivory in terrestrial
arthropods: mixing plant and prey diets. Annual Review of
Entomology, 47, 267297.

C 2016 Institute of Zoology, Chinese Academy of Sciences, 00, 112
Omnivore oviposition behavior 11
Darrach, M. (2014) Statistical overview of the Canadian veg-
etable industry, 2013. Agriculture and Agri-Food Canada,
AAFC No: 12294E.
Feller, C., Bleiholder, H., Buhr, L., Hack, H., Hess, M.,
Klose, R., Meier, U., Stauss, R., Van Den Boom, T.
and Weber, E. (1995) Phanologische Entwicklungstadien
von Gemusepflanzen: II. Fruchtgemuse und Hulsenfruchte.
Nachrichtenblatt deutschen Pflanzenschutzdienstes, 47, 217
232.
Fretwell, S.D. and Lucas, H.L. (1970) On territorial behavior
and other factors influencing habitat distribution in birds. I.
Theoretical development. Acta Biotheoretica, 19, 1636.
Gillespie, D.R. and McGregor, R.R. (2000) The function of
plant feeding in the omnivorous predator Dicyphus hesperus:
water places limits on predation. Ecological Entomology, 25,
380386.
Gillespie, D.R. and Quiring, D.M.J. (2005) Diapause induction
under greenhouse conditions in two populations of Dicyphus
hesperus (Hemiptera: Miridae). Biocontrol Science and Tech-
nology, 15, 571583.
Gillespie, D.R., Sanchez, J.A.S. and McGregor, R.R. (2004) Cu-
mulative temperature requirements and development thresh-
olds in two populations of Dicyphus hesperus (Hemiptera:
Miridae). The Canadian Entomologist, 136, 675683.
Groenteman, R., Guershon, M. and Coll, M. (2006) Effects of
leaf nitrogen content on oviposition site selection, offspring
performance, and intraspecific interactions in an omnivorous
bug. Ecological Entomology, 31, 155161.
Gullan, P.J. and Cranston, P.J. (2005) The Insects: An Outline
of Entomology, 3rd edn, pp. 166169. Blackwell Publishing
Ltd., Malden.
Hazard, T.J.R. (2008) Close-range prey location, recognition
and foraging decisions by the generalist predator, Dicyphus
hesperus Knight (Heteroptera: Miridae). MSc Thesis. Simon
Fraser University, British Colombia, Canada.
Hemptinne, J.-L., Dixon, A.F.G., Doucet, J.-L. and Petersen,
J.-E. (1993) Optimal foraging by hoverflies (Diptera: Sryphi-
dae) and ladybirds (Coleoptera: Coccinellidae): mechanisms.
European Journal of Entomology, 90, 451455.
Hunt, D.W.A., Drury, C.F. and Maw, H.E.L. (1992) Influence
of nitrogen on the performance of Colorado potato beetle
(Coleoptera: Chrysomelidae) on tomato. Environmental En-
tomology, 21, 817821.
IBM SPSS Statistics 21.0 (2012) IBM Corporation Software
Group, Somers.
Jaenike, J. (1978) On optimal oviposition behavior in phy-
tophagous insects. Theoretical Population Biology, 14, 350
356.
Jaenike, J. (1990) Host specialization in phytophagous insects.
Annual Review of Ecology and Systematics, 21, 243273.
Jauset, A.M., Sarasua, M.J., Avilla, J. and Albajes, R. (1998)
The impact of nitrogen fertilization of tomato on feeding
12 M. A. Vankosky & S. L. VanLaerhoven
site selection and oviposition by Trialeurodes vaporariorum.
Entomologia Experimentalis et Applicata, 86, 175182.
Jauset, A.M., Sarasua, M.J., Avilla, J., and Albajes, R. (2000)
Effect of nitrogen fertilization level applied to tomato on the
greenhouse whitefly. Crop Protection, 19, 255261.
Krebs, J.R. and Davies, N.B. (1993) An Introduction to Be-
havioural Ecology, 3rd edn, pp. 4072. Blackwell Science
Ltd., Malden.
Laycock, A., Camm, E., VanLaerhoven, S. and Gillespie, D.
(2006) Cannibalism in a zoophytophagous omnivore is medi-
ated by prey availability and plant substrate. Journal of Insect
Behavior, 19, 219229.
Leather, S.R. (1985) Oviposition preferences in relation to larval
growth and survival in the pine beauty moth, Panolis flammea.
Ecological Entomology, 10, 213217.
Levins, R. and MacArthur, R. (1969) An hypothesis to explain
the incidence of monophagy. Ecology, 50, 910911.
Lockwood, J.R. (1998) On the statistical analysis of multiple-
choice feeding preference experiments. Oecologia, 116, 475
481.
Lundgren, J.G. and Fergen, J.K. (2006) The oviposition behavior
of the predator Orius insidiosus: acceptability and preference
for different plants. BioControl, 51, 217227.
Mattson, W.J. (1980) Herbivory in relation to plant nitrogen
content. Annual Review of Ecology and Systematics, 11, 119
161.
Mayhew, P.J. (1997) Adaptive patterns of host-plant selection
by phytophagous insects. Oikos, 79, 417428.
Mayhew, P.J. (2001) Herbivore host choice and optimal bad
motherhood. Trends in Ecology and Evolution, 16, 165167.
McGregor, R.R. and Gillespie, D.R. (2004) Olfactory responses
of the omnivorous generalist predator Dicyphus hesperus to
plant and prey odours. Entomologia Experimentalis et Appli-
cata, 112, 201205.
McGregor, R.R., Gillespie, D.R., Quiring, D.M.J. and Foisy,
M.R.J. (1999) Potential use of Dicyphus hesperus Knight
(Heteroptera: Miridae) for biological control of pests of green-
house tomatoes. Biological Control, 16, 104110.
McMaster, G.S. and Wilhelm, W.W. (1997) Growing degree-
days: one equation, two interpretations. Agricultural and For-
est Meteorology, 87, 291300.
Nylin, S. and Janz, N. (1993) Oviposition preference and larval
performance in Polygonia c-album (Lepidoptera: Nymphali-
dae): the choice between bad and worse. Ecological Entomol-
ogy, 18, 394398.
Nylin, S., Janz, N. and Wedell, N. (1996) Oviposition plant
preference and offspring performance in the comma butter-
fly: correlations and conflicts. Entomologia Experimentalis
et Applicata, 80, 141144.
Ortega-Arenas, L.D., Miranda-Aragon, D.A. and Sandoval-
Villa, M. (2006) Densidad de huevos y ninfas de mosca blanca

C 2016 Institute of Zoology, Chinese Academy of Sciences, 00, 112
Trialeurodes vaporariorum (West.) en Gerbera jamesonii
H. Bolus con diferentes regimens de fertilizacion nitroge-
nada [Whitefly Trialeurodes vaporariorum (West.) eggs and
nymphs density on Gerbera jamesonii H. Bolus under differ-
ent nitrogenous fertilizer regimes.] Agrociencia, 40, 363371.
Roa, R. (1992) Design and analysis of multiple-choice feeding-
preference experiments. Oecologia, 89, 509515.
Sanchez, J.A., Gillespie, D.R. and McGregor, R.R. (2004) Plant
preference in relation to life history traits in the zoophy-
tophagous predator Dicyphus hesperus. Entomologia Exper-
imentalis et Applicata, 112, 719.
R
SAS Institute (2009) SAS/STAT 9.2 Users Guide, 2nd edn.
SAS Institute Inc., Cary.
Scheirs, J. and De Bruyn, L. (2002) Integrating optimal foraging
and optimal oviposition theory in plant-insect research. Oikos,
96, 187191.
Scheirs, J, De Bruyn, L. and Verhagen, R. (2000) Optimization
of adult performance determines host choice in a grass miner.
Proceedings of the Royal Society of London: Biological Sci-
ences, 267, 20652069.
Seagraves, M.P. and Lundgren, J.G. (2006) Oviposition response
by Orius insidiosus (Hemiptera: Anthocoridae) to plant qual-
ity and prey availability. Biological Control, 55, 174177.
Stout, M.J., Brovont, R.A. and Duffy, S.S. (1998) Effect of ni-
trogen availability on expression of constitutive and inducible
chemical defenses in tomato. Journal of Chemical Ecology,
24, 945963.
Tallamy, D.W. (1999) Child care among the insects. Scientific
American, 280, 7277.
Thompson, J.N. (1988) Evolutionary ecology of the relationship
between oviposition preference and performance of offspring
in phytophagous insects. Entomologia Experimentalis et Ap-
plicata, 47, 314.
Vankosky, M.A. and VanLaerhoven, S.L. (2015) Plant and prey
quality interact to influence the foraging behavior of an om-
nivorous insect, Dicyphus hesperus. Animal Behaviour, 108,
109116.
VanLaerhoven, S.L., Gillespie, D.R. and Roitberg, B.D. (2006)
Patch retention time in an omnivore, Dicyphus hesperus is
dependent on both host plant and prey type. Journal of Insect
Behavior, 19, 613621.
Wheeler, G.S. and Ordung, K.M. (2005) Secondary metabolite
variation affects the oviposition preference but has little effect
on the performance of Boreioglycaspis melaleucae: a biolog-
ical control agent of Melaleuca quinquenervia. Biological
Control, 35, 115123.
Wiedenmann, R.N. and Smith, J.W. (1997) Attributes of natural
enemies in ephemeral crop habitats. Biological Control, 10,
1622.
Accepted January 11, 2016