Ecological Applications, 20(1), 2010, pp. 242–250
Ó 2010 by the Ecological Society of America
Host-adapted parasitoids in biological control: Does source matter?
LEE M. HENRY,1,5 NIGEL MAY,1,2 SUSANNA ACHEAMPONG,3,4 DAVID R. GILLESPIE,1,3
AND
BERNARD D. ROITBERG1
1
Department of Biological Sciences, Simon Fraser University, Burnaby, British Columbia V5A 1S6 Canada
School of Computing, British Columbia Institute of Technology, Burnaby, British Columbia V5G 3H2 Canada
3
Pacific Agri-Food Research Centre, Agriculture and Agri-Food Canada, POB 1000, Agassiz, British Columbia V0M 1A0 Canada
4
British Columbia Ministry of Agriculture and Lands, 200–1690 Powick Road, Kelowna, British Columbia V1X 7G5 Canada
2
Abstract. It has been hypothesized that the success of a biological control introduction is,
in part, dependent on the ability of the control agent to become established in its new
environment or to its new population of hosts through local adaptation. Despite this, few
studies have investigated the influence of the recent coevolutionary history of pest species and
natural enemies on the efficacy of biological control agents, especially for agents that are massreared for release in agriculture. We investigate the evolutionary potential of a biological
control agent Aphidius ervi to adapt to a key pest species, the foxglove aphid Aulacorthum
solani, through components essential to the evolution of parasitoid virulence. We explored (1)
the influence of genetic variation from natural source populations on the ability to parasitize
natal and non-natal host species; (2) the heritability of key traits related to parasitoid fitness;
and (3) the efficacy of parasitoid host-selection lines in a greenhouse system. Source
populations maintained genetic variation in the ability to utilize natal and non-natal host
species; however, only some of the traits sampled suggested local adaptation of parasitoid
populations. The ability to parasitize a host was found to be genetically determined and
strongly heritable, irrespective of host species. The greenhouse study demonstrated the
potential of parasitoid selection lines to substantially increase performance of parasitoids for
target pest species. This research provides insight into novel techniques that can be used to
increase the quality of biological control agents through the development of lines of natural
enemies adapted to particular pest species.
Key words: Acyrthosiphon pisum; adaptive evolution; Aphidius ervi; Aulacorthum solani; biological
control; foxglove aphid; heritability; parasitoid; pea aphid; reciprocal transplant; selection experiment.
INTRODUCTION
Research in population ecology has been an essential
component in the practical application of natural
enemies in controlling pests, but the importance of
genetics and adaptive evolution in biological control
remains a controversial topic (Force 1967, Remington
1968, Messenger and van den Bosch 1971, Roush 1990,
Hopper et al. 1993, Holt and Hochberg 1997, Jervis
1997, Hufbauer 2002, Hufbauer and Roderick 2005,
Phillips et al. 2008). Recent research has sought to better
understand the genetic structure of natural enemy
populations in an attempt to identify populations that
are locally adapted to environmental conditions or the
organisms with which they interact (Lozier et al. 2008,
Philips et al. 2008). However, few studies attempting to
increase the efficacy of biological control agents have
considered the recent coevolutionary history between
pest species and natural enemies. Furthermore, the role
of adaptive evolution in biological control has been
addressed primarily from classic introductions of
natural enemies to control exotic pests, with the
Manuscript received 10 October 2008; revised 16 March
2009; accepted 21 April 2009. Corresponding Editor: M. P.
Ayres.
5 E-mail: [email protected]
242
adaptive potential of natural enemies used in greenhouses and other areas of agriculture being virtually
ignored.
The value of natural enemies as an integrative method
of controlling pests in greenhouses is well recognized
(van Lenteren 2000). A zero-tolerance policy for pest
damage in commercial vegetable and flower markets and
an increasing desire to reduce pesticide applications has
made the continual improvement of biological control
efforts a top priority of the greenhouse industry in
general (van Lenteren 1997). Difficulty establishing
natural enemy populations, unpredictable control of
problematic pests by natural enemies, and lag periods
associated with natural enemy adjustment to novel host
species are some of the major current issues in biological
control. Recently a movement to improve biological
control efforts has emerged in the form of breeding pest
and disease resistance in plants (Cuartero et al. 1999) or
breeding plants that facilitate natural enemy foraging.
Examples include selected cucumber lines with fewer
hairs to increase searching efficiency of parasitoids (van
Lenteren et al. 1995) and the cultivation of plant lines
that produce greater amounts of natural enemy-attractive volatiles to decrease the response time to infestation
(Dicke 1999). This artificial selection approach has
demonstrated the potential of breeding programs in pest
January 2010
HOST-ADAPTED PARASITOIDS IN BIOCONTROL
control; however, surprisingly little to no effort has been
made to actually breed the natural enemies themselves in
an attempt to improve efficacy of biological control
agents. (For an exception see Croft et al. 1999.)
Insect parasitoids are one of the most common
natural enemies used in greenhouse pest management.
Generalist parasitoids used as biological control agents
are often assumed to be equally capable of effectively
utilizing all recorded host species. However, generalist
parasitoids, like most generalist insects that are largely
polyphagous at the species level, often exhibit limited
host species use at the population or community level
(Fox and Morrow 1981, Smith et al. 2007). Variation in
the ability to parasitize natal and non-natal host species
has been recorded for several aphid parasitoids,
suggesting a genetic basis for virulence and resistance
across a variety of systems (Powell and Wright 1988,
Pennacchio et al. 1994, Henter 1995, Pike et al. 1999,
Antolin et al. 2006). Recent research by Henry et al.
(2008) demonstrated the potential of Aphidius ervi to
adapt to a key greenhouse pest species, the foxglove
aphid Aulacorthum solani, under controlled laboratory
conditions. Evidence of host adaptation in generalist
parasitoids has also been revealed in several biological
control introductions in natural systems, in that
parasitoids were shown to have dramatically increased
their reproductive potential on a novel host species
following the initial introductions (van den Bosch 1964,
Lemasurier and Waage 1993). These examples provide
strong evidence for the capability of generalist parasitoids to adapt to key pest species. However, to the
author’s knowledge, no studies have initiated the first
steps toward breeding programs for pest-adapted
biological control agents or directly tested the efficacy
of host-adapted parasitoids in a commercial greenhouse
setting.
Theoretically, several possible evolutionary interactions may occur between parasitoids and their hosts,
including evolved host resistance to parasitoid attack
and the evolution of parasitoid virulence to overcome
host defenses. The interplay of these coevolving factors
is fundamental to the patterns of parasitism found in
communities of parasitoids and their hosts. The primary
objective of this study is to focus on the evolutionary
potential of parasitoids in their ability to parasitize
different host species to determine if parasitoid efficacy
can be increased. The potential for hosts to evolve
resistance to parasitism is not explicitly measured within
this study; however, it is recognized as an essential
component of host–parasitoid coevolution in natural
and agricultural systems.
The parasitoid A. ervi represents an exemplary system
to address the adaptive evolution of natural enemies
used in biological control because it is not only a
commonly used parasitoid to control aphids in agriculture, but also a model system in ecology and evolutionary biology (e.g., Sequeira and Mackauer 1992, Henter
1995). The potential of A. ervi to adapt locally to two
243
aphids, Acyrthosiphon pisum and A. solani, was investigated. These aphids represent important examples of
hosts with which A. ervi is associated in agriculture, as
A. pisum is the most common host used to commercially
mass rear colonies of A. ervi, and A. solani is an
extremely problematic aphid pest worldwide that is
historically difficult to control with biological control
introductions (Sanchez et al. 2007). In the past, Aphidius
parasitoids have been used as the primary biological
agent to control A. solani, although the results are often
unpredictable.
We used a series of experiments to investigate the
potential of A. ervi to adapt to different host species.
First, we determined if natural populations of parasitoids are locally adapted to the host species most
prevalent in their environment. This addresses the
importance of source populations in collecting parasitoids for use in mass-rearing programs. Second, we
investigated the heritability of three traits fundamental
to parasitoid performance as biological control agents.
Establishing trait heritability is an important first step in
the implementation of artificial selection and breeding
programs of pest-adapted biological control agents.
Third, using a ‘‘quasi-natural’’ selection experiment, we
established the potential of parasitoids that have been
preadapted to a pest species to outperform parasitoids
that have been reared on a standard rearing host in a
simulated commercial greenhouse setting.
MATERIALS
AND
METHODS
Parasitoid and aphid collections and rearing
Parasitoids and aphids used in the reciprocal transplant and heritability experiments were collected from
foxglove aphids Aulacorthum solani (Kaltenbach) (Hemiptera: Aphididae), feeding in large organic potato
(Solanum tuberosum L.) fields (;10 km2 field) in Delta,
British Columbia, Canada, and pea aphids Acyrthosiphon pisum (Hemiptera: Aphididae), feeding in alfalfa
(Medicago sativa L.) fields (;3 km2 field) in Agassiz,
British Columbia, Canada. Several hundred parasitoids
were collected by the removal of mummies from leaf
material as well as by rearing aphids that harbored
parasitoids. (A mummy is an aphid exoskeleton
containing a parasitoid pupa. This occurs after the
larval parasitoid has consumed the host aphid and is
metamorphosing into an adult.) All aphids were
collected using sweep nets. Upon emergence, female
parasitoids were individually given access to ;30 mixed
instars of their natal hosts; then the females were
identified, and the offspring of each identified female
were used to propagate species-specific colonies of
Aphidius ervi Haliday (Hymenoptera: Aphidiidae).
Parasitoids and aphids were identified by taxonomic
keys and by comparison to identified, exemplar specimens at the Pacific Agri-Food Research Centre in
Agassiz (Agriculture Canada). Pea aphids were the only
aphid species collected from the alfalfa field, but the
potato fields harbored a community of aphid species
244
LEE M. HENRY ET AL.
dominated by foxglove, potato (Macrosiphum euphorbiae) (Thomas), and green peach aphids (Myzus
persicae) (Sulver), and potentially other species at lower
frequencies. However, A. ervi from the potato fields that
were used in this study were all confirmed to be from
foxglove aphids by identifying and isolating the aphids
(many of which were already parasitized), then collecting mummies from the isolated foxglove aphid population.
Aphids were maintained on cuttings of their host
species of origin, with the petiole inserted into water in
clear plastic cups covered with tissue paper. A. ervi were
maintained using the same system, with the addition of
dilute honey for sustenance. Twenty replicate cups were
maintained for each aphid host species, with 10 cups per
host species maintaining the parasitoid populations.
Reciprocal transplant experiment
A. ervi collected from A. solani on potato and A.
pisum on alfalfa were experimentally assayed on the host
species from which they were originally collected (natal
host), and the novel (non-natal) host species using a
reciprocal transplant design. Natal and non-natal will be
used as terms to describe the host species from which the
parasitoids were collected; however this does not reflect
the history of host use, as individuals from both
populations may have previously utilized alternate host
species. Individual females (N ¼ 40) were given access to
a patch of 20 second-instar pea or foxglove aphids
feeding on a leaf cutting. Each female was allowed to
forage on a single aphid patch for two hours, at which
time the parasitoid was removed, and the aphids were
left on the leaf to develop. Each parasitoid was used only
once and then preserved in 75% ethanol after each trial
to confirm the parasitoid species if necessary. Total
parasitism (number of mummies formed per 20 aphids),
the proportion of offspring emerging from the mummies, and the sex ratio of the offspring (proportion
female) served as proxies for parasitoid fitness. Natal
host species of the parasitoid, the host species the
parasitoid was assayed on (exposure host), and the
interaction between natal and exposure host species were
compared for each of the proxies. The number of
mummies formed was analyzed by ANOVA. The
proportion of emergence and sex ratio were not
normally distributed and were therefore analyzed using
a generalized linear model (GLM) with a binomial
distribution and a logit function.
Heritability experiment
A mother 3 daughter regression was used to
determine the narrow-sense heritability (h2) of three
traits integral to parasitoid fitness: total parasitism,
proportion of offspring emergence from mummies, and
offspring sex ratio. Only A. ervi from the population
collected on pea aphids, feeding on alfalfa, were used for
this experiment; however, heritability estimates were
determined for performance on both pea and foxglove
Ecological Applications
Vol. 20, No. 1
aphids. Female parasitoids (mothers) were assayed on
their natal host species (N ¼ 42) using the same protocol
as in the previous experiment (i.e., 20 second-instar
aphids in a petri dish for two hours). Upon emergence,
each mother’s offspring were allowed to inbreed to
reduce genetic variability, and then four female offspring
were randomly selected and assayed on pea and foxglove
aphids (two daughters assayed on pea and two
daughters assayed on foxglove). Zero trait values of
daughters were removed from the analyses, as we could
not determine if a value of zero was due to genetic
components or other biological reasons (e.g., zero sex
ratio could be due to not mating, or zero parasitism may
be due to an unresponsive female). Heritability was
determined using mothers and daughters assayed on the
same host species (i.e., genetic component only,
controlling for environment). A comparison of inbred
sisters from the same family assayed on the two different
host species was used to determine if the expression of a
phenotype is influenced by the species of the host (i.e.,
environmental effects). This design allowed us to
determine trait heritability using a mother 3 daughter
regression while controlling for potential environmental
effects caused by developing in different host species.
Narrow-sense heritability of the aforementioned traits
was estimated using a one-parent 3 mid-offspring (mean
trait value of daughters) regression, with the coefficient
of the slope of the linear regression line approximating
heritability (Falconer and Mackay 1989). As only female
phenotypes could be measured, the estimate of heritability (h2) is double the coefficient of the slope (singleparent heritability). Each trait was analyzed using a
linear regression comparing mothers and daughters
reared on pea aphids and foxglove aphids separately,
blocking for mother (six linear regressions in total). The
relationship between sisters reared on separate host
species was assessed using a linear regression, blocking
for mother, by plotting the trait value of pea aphidreared sisters against the trait of foxglove aphid-reared
sisters for each fitness proxy. The relationship between
sister traits was analyzed separately for families that had
emerged from pea and foxglove aphids, in order to
control for potential maternal effects caused by developing in different host species.
Host-adapted parasitoid greenhouse experiment
Parasitoids used in the greenhouse experiment were
from selection lines that had been isolated and
maintained in replication on either pea or foxglove
aphids at Simon Fraser University for two years (;50
parasitoid generations), as described in Henry et al.
(2008). A. ervi selection lines were assayed for their
performance on foxglove aphids in a controlled greenhouse experiment at the Pacific Agri-Food Research
Centre in Agassiz. Mature pepper plants (;two months
old), isolated in screened individual enclosures, in a 6.4
3 15 m greenhouse, were each heavily infested with
several thousand foxglove aphids and left for a week for
January 2010
HOST-ADAPTED PARASITOIDS IN BIOCONTROL
245
the aphids to become established. Plants were infested to
the point where parasitoids would not be limited by host
abundance. Parasitoid mummies were randomly chosen
from the replicate populations of host-selection-line
parasitoids. Parasitoids were isolated by host species and
allowed to mate; then 10 females were released in each
enclosure by leaning an unsealed test tube containing the
parasitoids against the basal stem of the pepper plant (N
¼ 10 replicates for each selection line). The parasitoids
were left to forage on the infested plants within the
enclosures for a 12-d period. This time frame was long
enough that the parasitoids from the initial release had
died, and no new parasitoids had emerged to additionally parasitize the aphids. Plants were supplied with
water using a drip irrigation system but were otherwise
not tended for the course of the experiment.
After the plants were removed from the enclosures
and dissected, all mummies were removed and counted.
The dissected plants were then individually placed in
Plexiglas containers with fresh pepper plants to allow
the aphids still harboring parasitoid larvae access to
fresh plant material, where larvae developed into
mummies. This methodology resulted in a precise value
for total parasitism per plant (total mummies). Total
parasitism was compared by parasitoid selection line
(ANOVA).
All statistical analyses were performed using JMP 7.0
statistical software (SAS Institute 2007).
RESULTS
Reciprocal transplant
The total number of mummies formed differed when
comparing parasitoids from the two host species
populations (F3, 154 ¼ 39.3, P , 0.0001). Parasitoids
collected from the pea aphid population produced more
mummies (12.6 6 0.45) in general than parasitoids from
the foxglove population (10.4 6 0.49) (natal host: F1, 154
¼ 10.2, P ¼ 0.0017), and parasitoids generally produced
more mummies when exposed to foxglove aphids (14.6
6 0.46) compared to pea aphids (8.5 6 0.48) (exposure
host: F1, 154 ¼ 87.9, P , 0.0001). All values are mean 6
SE. Parasitoids produced significantly more mummies
on their natal host (natal 3 exposure: F1, 154 ¼ 25.4, P ,
0.0001), as indicated by the strong interaction between
natal and exposure host species, which suggests parasitoids are locally adapted to their natal host species (Fig.
1A). The proportion of emergence from mummies
(GLM: v23; 140 ¼ 93.65, P , 0.0001) differed between
the two parasitoid populations. The proportion of
emergence was greater from foxglove aphids at 0.90 6
0.03 compared to pea aphids at 0.81 6 0.03 (exposure
host: v21; 140 ¼ 22.90, P , 0.0001). Parasitoids from the
pea aphid population generally had higher emergence on
both host species (0.91 6 0.03) compared to parasitoids
from the foxglove population (0.80 6 0.03) (natal host:
v21; 140 ¼ 68.62, P , 0.0001; Fig. 1B). Parasitoids from the
foxglove population had a substantially lower proportion of offspring emerging from pea aphids (0.69 6 0.04)
FIG. 1. Hymenopteran parasitoids (Aphidius ervi) collected
from pea aphid (Acyrthosiphon pisum) and foxglove aphid
(Aulacorthum solani ) populations assayed on natal and nonnatal host species. Parasitoid performance is indicated by (A)
the parasitism of 20 aphids; (B) the proportion of offspring
emerging from parasitized aphids (i.e., emergence success); and
(C) the sex ratio of the emerging offspring (proportion of
females). Values are means 6 SE.
compared to emergence from their natal host species
(0.88 6 0.04) (natal 3 exposure: v21; 140 ¼ 9.15, P ¼
0.0025). No variation was detected in the sex ratio of
emerging offspring (v23; 108 ¼ 2.71, P ¼ 0.44; Fig. 1C).
246
Ecological Applications
Vol. 20, No. 1
LEE M. HENRY ET AL.
from mummies for parasitoids assayed on pea aphids did
not display heritable variation (h2 ¼ 0.18 6 0.26) (F1,46 ¼
0.49, P ¼ 0.49). The sex ratio of offspring from pea
aphid-assayed parasitoids did not show heritable variation (h2 ¼ 1.72 6 0.85) (F1,20 ¼ 4.11, P ¼ 0.07; Table 1).
The comparison of the parasitism of sisters assayed on
pea and foxglove aphids had a significant negative
relationship when the families had emerged from pea
aphids (F1,47 ¼ 8.79, r 2 ¼ 0.16, P ¼ 0.0048; Fig. 3B).
However, no relationship was detected in the parasitism
of sisters assayed on the two host species when families
had emerged from foxglove aphids (F1,57 ¼ 0.99, r 2 ¼
0.02, P ¼ 0.32; Fig. 3A). The proportion of offspring
emergence of sisters assayed on different hosts did not
show a relationship when the families had emerged from
foxglove (F1,52 ¼ 0.89, r 2 ¼ 0.02, P ¼ 0.35) or pea aphids
(F1,39 ¼ 0.79, r 2 ¼ 0.02, P ¼ 0.38). No relationship was
detected in the sex ratio of the offspring from families
reared on foxglove (F1,20 ¼ 0.22, r 2 ¼ 0.01, P ¼ 0.64) or
pea aphids (F1,15 ¼ 0.001, r 2 ¼ 9.7 3 105, P ¼ 0.97).
Host-adapted parasitoid greenhouse experiment
Parasitoids from the foxglove and pea aphid selection
lines differed in their ability to parasitize foxglove aphids
in a greenhouse setting (F1,16 ¼ 5.58, P ¼ 0.03).
Parasitoids from the foxglove aphid selection line
produced an average of 233.8 6 51.7 mummies, whereas
parasitoids from the pea aphid selection line produced
an average of 86.5 6 31.4 mummies.
DISCUSSION
FIG. 2. Heritability of parasitism using a one-parent
(mother) 3 offspring (daughter) regression for parasitoids
reared on (A) foxglove aphids and (B) pea aphids. Each point
represents one family, with the mother’s parasitism on the xaxis and the mean value of two of her daughters (mid-daughter)
on the y-axis. Axis scales (the number of mummies formed out
of 20 aphids [i.e., parasitism]) represent the deviation from the
mean value (0) of all parents (x-axis) and all offspring (y-axis).
Estimates of heritability
Significant heritable variation in the ability to
parasitize foxglove aphids was demonstrated by the
mother 3 daughter regression (h2 ¼ 0.95 6 0.16) (linear
regression: F1,59 ¼ 35.2, P , 0.0001; Fig. 2A). The
proportion of offspring emergence from mummies did
not display heritable variation for foxglove-assayed
parasitoids (h2 ¼ 0.30 6 0.46) (F1,58 ¼ 0.41, P ¼ 0.52).
Additionally, offspring sex ratio did not display heritable
variation when parasitoids were assayed on foxglove
aphids (h2 ¼ 0.64 6 0.56) (F1,42 ¼ 1.31, P ¼ 0.27; Table 1).
A similar pattern emerged when parasitoids were
assayed on pea aphids. Significant heritable variation
was detected in the parasitism of mothers and daughters
exposed to pea aphids (0.75 6 0.28) (F1,47 ¼ 7.06, P ¼
0.011; Fig. 2B). The proportion of offspring emergence
It has been suggested that the success of biological
control is primarily based on the quality of the natural
enemies produced by commercial suppliers (van Lenteren 1997). Developments in the mass production,
quality control, storage, and shipment of natural
enemies have decreased production costs and led to
better product quality (van Lenteren 2000). However,
the importance of genetic variation of natural enemies,
biotypes for mass rearing, and recent coevolutionary
history of parasitoids with host species has not been
addressed for biological control agents used in agriculture. This is surprising, given that the closed nature of
greenhouses in particular make them ideal for releasing
preadapted biological control agents, and commercial
biocontrol-rearing facilities are an ideal arrangement for
the establishment of natural enemy breeding programs.
Patterns of parasitism and evidence
of local host adaptation
The reciprocal transplant experiment provides evidence that wild host-associated parasitoid populations
maintain genetic differences that influence their ability
to utilize natal and non-natal host species. However,
only some of the traits sampled indicate patterns of local
adaptation. The patterns of parasitism from the
reciprocal transplant experiment suggest that parasitoids
are locally adapted to their natal host populations, given
January 2010
HOST-ADAPTED PARASITOIDS IN BIOCONTROL
247
TABLE 1. Heritability estimates (h2) of a one-parent 3 mid-offspring regression, using mothers and
daughters of the hymentopteran parasitoid Aphidius ervi reared on either pea or foxglove aphids.
F
P
r2
h2, mean 6 SE
35.2
0.41
1.31
,0.0001
0.52
0.27
0.38
0.01
0.06
0.95 6 0.16
0.30 6 0.46
0.64 6 0.56
7.06
0.49
4.11
0.011
0.49
0.07
0.13
0.02
0.31
0.75 6 0.28
0.18 6 0.26
1.72 6 0.85
Species/phenotypic trait
Pea aphid (Acyrthosiphon pisum)
Parasitism
Proportion of offspring emerging
Sex ratio
Foxglove aphid (Aulacorthum solani )
Parasitism
Proportion of offspring emerging
Sex ratio
Note: Mothers and daughters were always assayed on the same host species.
that parasitoids from both populations maintained a
higher level of parasitism on their natal host (Fig. 1A).
Parasitoids from the pea aphid population maintained a
relatively high level of parasitism on both natal and nonnatal hosts, whereas the foxglove population only
performed well on its natal host, with low levels of
parasitism on pea aphids. Although this pattern suggests
a genetic basis for parasitism, the large difference in the
two parasitoids’ ability to parasitize pea aphids, and the
levels of parasitism in general, may also be influenced by
nongenetic components. For instance, parasitism is most
likely influenced by the extent of phenotypic plasticity
exhibited by parasitoids developing in different host
species. This includes host fidelity or morphological
variables such as adult body size that are primarily
environmentally determined (i.e., induced by developing
in a host species), which can have a strong influence on
the willingness to utilize a host species or the ability to
overcome host defenses (Henry et al. 2008). Parasitoids
from the foxglove population are, on average, half the
size of parasitoids that have developed in pea aphids due
to growth restrictions imposed by the size of the host
species, which can influence oviposition success when
attacking large aggressive host species such as the pea
aphid (Henry et al. 2006). Although this may partially
account for the low levels of pea aphid parasitism from
foxglove aphid parasitoids, it does not account for the
foxglove parasitoids outperforming pea aphid parasitoids on their natal host species, which supports the
hypothesis of local adaptation.
Offspring emergence showed significant variation
between the two parasitoid populations, which supports
the theory of genetic differentiation in the ability to
successfully develop in a host. However, this pattern
does not support local host species adaptation, in that
parasitoids from the pea aphid population had higher
offspring emergence on both host species (Fig. 1B). The
host populations these parasitoids were collected from
may account for this deviation in that they differed in
the composition and availability of host species in their
local environment. Parasitoids from the alfalfa field
most likely did not have access to alternate host species,
as the pea aphid was the only host species found in the
field, whereas parasitoids from the potato field had
several different host species to choose from. In an
environment with the option of utilizing alternate host
species, parasitoids may be practicing host switching,
which is thought to inherently destabilize specialization
through the obliteration of locally adapted gene pools
(Kawecki and Ebert 2004). The aphids themselves may
have also experienced different degrees of selective
pressures from parasitism. Foxglove aphids may have
experienced a refuge from parasitism due to the presence
FIG. 3. The relationship between sisters in their ability to
parasitize foxglove and pea aphids. Families emerging from (A)
foxglove aphids and (B) pea aphids are presented separately.
Each point represents one family of four sisters, with the mean
value of two sisters assayed on pea aphids (y-axis) and the mean
value of two sisters assayed on foxglove aphids (x-axis) making
up each point.
248
LEE M. HENRY ET AL.
of alternate hosts, where as pea aphids are more likely to
have been exposed to greater selective pressures from
Aphidius ervi, resulting in a more highly evolved
defensive system (such as an increase in the frequency
of endosymbionts that confer resistance to parasitism
[Oliver et al. 2005]), which may partially explain the
lower levels of parasitism and emergence from pea
aphids in general. Further research is required to
determine the effects of alternate host species on
parasitoid specialization and local host adaptation.
Local host adaptation in natural populations of
parasitoids utilizing different host species is not uncommon (e.g., Vaughn and Antolin 1998, Althoff and
Thompson 2001, Morehead et al. 2001, Hayward and
Stone 2006, Stireman et al. 2006). Antolin et al. (2006)
found that Diaeretiella rapae from two adjacent fields
containing Russian wheat aphids (Diuaphis noxia) and
cabbage aphids (Brevicoryne brassicae) were genetically
differentiated and locally adapted by host species.
Similarly, changes in Microctonus hyperdoae biotype
frequencies introduced to New Zealand for control of
the weevil Listronotus bonariensis were consistent with
strong directional selection, providing one of the first
clear demonstrations of adaptive evolution improving
success rates of biological control introductions (Winder
et al. 2003, Philips et al. 2008). These studies highlight
the importance of source populations and natural
genetic variation in natural enemy populations that are
used in biological control of pests. Although sampling
methods for detecting locally adapted populations have
primarily focused on parasitoids for classical biological
control introductions (Lozier et al. 2008), the techniques
have clear applications for selecting the highest-quality
source populations for the mass rearing of parasitoids to
be released in agriculture.
Heritability of fitness-related traits
Knowing the heritability of life history traits is
necessary for a mechanistic understanding of evolutionary processes. Results of the parent 3 offspring
regression suggest there is significant heritable genetic
variation in the capacity to parasitize a patch of aphids
when assayed on both foxglove and pea aphids
independently (Fig. 2A, B), indicating that this trait
has a strong genetic component and that the genetic
basis of parasitism is universally expressed, irrespective
of host species. Heritability of parasitism on pea aphids
(h2 ¼ 0.75) was lower than on foxglove aphids (h2 ¼
0.95), which is not surprising, as the original population
of parasitoids was collected from pea aphids, and
selection should act to reduce additive genetic variation
within our sample population for performance on pea
aphids (Fisher 1930, Mousseau and Roff 1987).
However, both of these values are still considered to
be moderately high, as heritability for life history traits
generally averages around 0.25 (Roff 1992). This
suggests that genes involved in pea aphid parasitism
have not reached fixation from our sample populations,
Ecological Applications
Vol. 20, No. 1
which is interesting given that the pea aphid was the only
aphid species collected from this study site. This could
result from interspecific, frequency-dependent selection
or migrant individuals from alternate host species or
alternate host patches with differing genotypes for
resistance (Henter 1995, Henter and Via 1995). Perhaps
even more interesting is the high heritability of
parasitism on foxglove aphids from individuals collected
from the pea aphid population (Fig. 2B), which suggests
that a significant amount of standing genetic variation
exists within this population for the ability to parasitize
foxglove aphids. No trait heritability was detected in the
emergence of offspring from mummies or in the sex ratio
of offspring, indicating that these traits may be primarily
determined by environmental conditions, have reached
allele fixation, or may be solely influenced by maternal
effects.
The relationship between sisters assayed on the two
different host species allowed us to investigate the
influence of the environment (i.e., host species) on trait
expression, using closely related individuals, with little
maternal influence, as families emerged from the same
host species. Results indicate that developing in different
host species influences gene expression in that sister’s
traits were either negatively correlated or had no
detectable correlation. A strong negative relationship
was detected in parasitism between sisters assayed on the
two host species (Fig. 3B), when the families emerged
from pea aphids. This negative fitness correlation
indicates that a trade-off exists in the ability to
successfully parasitize both host species, which suggests
an antagonistic interaction between one or more genes.
However, when parasitoids were reared on foxglove
aphids, no significant relationship was detected in
parasitism between sisters (Fig. 3A). One possible
explanation for this deviation in the relationship
between sisters when reared on different hosts is that
parasitoids emerging from foxglove aphids are relatively
small; previous research has shown that foxglove-reared
parasitoids are limited in their ability to handle and
successfully ovisposit in large, well-defended hosts such
as the pea aphids (Henry et al. 2006). Thus the
morphological limitations from developing in foxglove
aphids may have masked the actual relationship between
sisters. Additionally, no relationship between sisters was
detected in the emergence or sex ratio of offspring,
supporting the earlier claims that these traits may be
determined by a combination of factors.
Our results support the research by Henry et al.
(2008), in that the ability to parasitize a host appears to
be genetically determined and strongly heritable, irrespective of the host species in which the parasitoids are
developing. Identification of heritable traits in natural
enemies used in biological control is an essential first
step in the initiation of breeding programs. Although
selectively breeding natural enemies is not a common
practice by commercial producers of biological control
agents, our results demonstrate the potential of artifi-
January 2010
HOST-ADAPTED PARASITOIDS IN BIOCONTROL
cially breeding for specific traits to increase the efficacy
of parasitoids on key pest species. Furthermore, our
results suggest a negative relationship between genotypes that indicates an inability of parasitoids to
simultaneously maximize fitness on more than one host,
which coincides with results of Henry et al. (2008).
Previous research has demonstrated that there can be a
physiological mechanism involved in overcoming host
defenses and that this process may be under directional
selection in certain host–parasitoid systems (Carton et
al. 1989, Kraaijeveld et al. 2001). Our result calls into
question the oversimplification of using generalist
parasitoids to control multiple aphid species in greenhouses, as well as the common practice of rearing
parasitoids on a host species that it is not implicitly used
to control, such as the pea aphid. If success of biological
control is partly based on the quality of the natural
enemies (van Lenteren 2000), improving the performance of parasitoids by selective breeding has great
potential to improve parasitoid efficacy, which in turn
increases economical viability of biological control
agents.
Greenhouse performance of host-adapted parasitoids
Previous research by Henry et al. (2008) demonstrated
that a foxglove selection line dramatically increased their
capacity to parasitize foxglove aphids in a controlled
laboratory experiment. Although compelling, these
results do not confirm the effectiveness of these
parasitoids as a biological control agent of foxglove
aphids in commercial greenhouses, as there are many
unpredictable factors in greenhouses that may hamper
parasitoid performance. Parasitoids from the foxglove
selection line produced an average of 63% more
mummies than the pea aphid selection line when assayed
on foxglove aphids in the greenhouse experiment. This
clearly demonstrates the potential of using host selection
lines to increase the efficacy and consistency of
parasitoids for problematic aphid pests in greenhouses.
Increased performance of selected parasitoids occurs not
only through a greater physiological virulence for
overcoming host defenses but also by an increase in
host fidelity (Henry et al. 2008). Host fidelity and host
location are highly plastic traits in many parasitoid
systems (Villagra et al. 2007). The predictive nature of
these plastic responses has great potential in biological
control by establishing fidelity for a host species prior to
release, which increases the probability that the natural
enemy population will become established through a
reduction in emigration and an increased capacity to
locate hosts (Dicke 1999).
These results provide insight into a novel technique to
increase the quality of biological control agents through
selection designs. Artificial selection designs can also be
extended to nonprey resources, such as preadapting a
natural enemy to the spatial structure or temperature of
the intended foraging environment. Ideally, reduction in
the cost of biological control to growers makes it easier
249
and more economical to apply. The importance of
increasing natural enemy efficiency is evident in the
recent movement to breed plant varieties that facilitate
the movement and response of natural enemies (van
Lenteren 2000). Thus it seems a logical progression to
develop the biological control agents themselves through
discerning collections from source populations and
careful breeding designs. However, the caveat of such
an approach is that it not only requires greater
communication and extensive collaboration of researchers with growers but also with the commercial suppliers
of natural enemies themselves.
ACKNOWLEDGMENTS
The authors thank Agriculture Canada for use of their
facilities. Thanks to the Roitberg lab for advice and support.
We also thank NSERC and the BC Greenhouse Growers
Association for funding.
LITERATURE CITED
Althoff, D. M., and J. N. Thompson. 2001. Geographic
structure on the searching behaviour of a specialist parasitoid: combining molecular and behavioural approaches.
Journal of Evolutionary Biology 14:406–417.
Antolin, M. F., T. A. Bjorkstein, and T. T. Vaughn. 2006.
Host-related fitness trade-offs in a presumed generalist
parasitoid, Diaeretiella rapae (Hymenoptera: Aphidiidae).
Ecological Entomology 31:242–254.
Carton, Y., P. Capy, and A. J. Nappi. 1989. Genetic variability
of host–parasite relationship traits: utilization of isofemale
lines in a Drosophila simulans parasitic wasp. Genetics,
Selection, Evolution 21:437–446.
Croft, P., J. Fenlon, R. J. Jacobson, and J. Dubas. 1999. Effect
of tomato conditioning on Phytoseiulus persimilis AthiasHenriot (Acari: Phytoseiidae) population growth. Bulletin
OILB/SROP 22:45–48.
Cuartero, J., H. Laterrot, and J. C. van Lenteren. 1999. Hostplant resistance to pathogens and arthropod pests. Pages
124–138 in R. Albajes, M. L. Gullino, and J. C. van
Lenteren, editors. Integrated pest and disease management in
greenhouse crops. Kluwer Publishers, Dordrecht, The
Netherlands.
Dicke, M. 1999. Direct and indirect effects of plants on
beneficial organisms. Pages 105–153 in J. R. Ruberson,
editor. Handbook of pest management. Marcel Dekker, New
York, New York, USA.
Falconer, D. S., and T. F. Mackay. 1989. Introduction to
quantitative genetics. Longman Group, Harlow, UK.
Fisher, R. A. 1930. The genetical theory of natural selection.
Clarendon Press, Oxford, UK.
Force, D. C. 1967. Genetics in the colonization of natural
enemies for biological control. Annals of the Entomological
Society of America 60:722–729.
Fox, L. R., and P. A. Morrow. 1981. Specialization: species
property or local phenomenon? Science 211:887–893.
Hayward, A., and G. N. Stone. 2006. Comparative phylogeography across two trophic levels: the oak gall wasp Andricus
kollari and its chalcid parasitoid Megastigmus stigmatizans.
Molecular Ecology 15:479–489.
Henry, L. M., B. D. Roitberg, and D. R. Gillespie. 2006.
Covariance of phenotypically plastic traits induces an
adaptive shift in host selection behaviour. Proceedings of
the Royal Society B 273:2893–2899.
Henry, L. M., B. D. Roitberg, and D. R. Gillespie. 2008. Hostrange evolution in Aphidius parasitoids: fidelity, virulence
and fitness trade-offs on an ancestral host. Evolution 62:689–
699.
250
LEE M. HENRY ET AL.
Henter, H. J. 1995. The potential for coevolution in a host–
parasitoid system. II. Genetic variation within a population
of wasps in the ability to parasitize an aphid host. Evolution
49:439–445.
Henter, H. J., and S. Via. 1995. The potential for coevolution in
a host–parasitoid system. I. Genetic variation within an
aphid population in susceptibility to a parasitic wasp.
Evolution 49:427–438.
Holt, R. D., and M. E. Hochberg. 1997. When is biological
control evolutionarily stable? Ecology 78:1673–1683.
Hopper, K. R., R. T. Roush, and W. Powell. 1993.
Management of genetics of biological-control introductions.
Annual Review of Entomology 38:27–51.
Hufbauer, R. A. 2002. Evidence for nonadaptive evolution in
parasitoid virulence following a biological control introduction. Ecological Applications 12:66–78.
Hufbauer, R. A., and G. K. Roderick. 2005. Microevolution in
biological control: mechanisms, patterns, and processes.
Biological Control 35:227–239.
Jervis, M. A. 1997. Parasitoids as limiting and selective factors:
can biological control be evolutionarily stable? Trends in
Ecology and Evolution 12:378–379.
Kawecki, T. J., and D. Ebert. 2004. Conceptual issues in local
adaptation. Ecology Letters 7:1225–1241.
Kraaijeveld, A. R., K. A. Hutcheson, E. C. Limentani, and
H. C. J. Godfray. 2001. Costs of counterdefenses to host
resistance in a parasitoid of Drosophila. Evolution 55:1815–
1821.
Lemasurier, A. D., and J. K. Waage. 1993. A comparison of
attack rates in a natal and an introduced population of the
parasitoid Cotesia glomerata. Biocontrol Science and Technology 3:467–474.
Lozier, J. D., G. K. Roderick, and N. J. Mills. 2008.
Evolutionarily significant units in natural enemies: identifying regional populations of Aphidius transcaspicus (Hymenoptera: Braconidae) for use in biological control of mealy
plum aphid. Biological Control 46:532–541.
Messenger, P. S., and R. van den Bosch. 1971. The adaptability
of introduced biological agents. Pages 68–92 in C. B.
Huffacker, editor. Biological control. Plenum, New York,
New York, USA.
Morehead, S. A., J. Seger, D. H. Feener, Jr., and B. V. Brown.
2001. Evidence for a cryptic species complex in the ant
parasitoid Apocephalus paraponerae (Diptera: Phoridae).
Evolutionary Ecology Research 3:273–284.
Mousseau, T. A., and D. A. Roff. 1987. Natural selection and
the heritability of fitness components. Heredity 59:181–197.
Oliver, K. M., N. A. Moran, and M. S. Hunter. 2005. Variation
in resistance to parasitism in aphids is due to symbionts not
host genotype. Proceedings of the National Academy of
Sciences (USA) 102:12795–12800.
Pennacchio, F., M. C. Digilio, E. Trembley, and A. Tranfaglia.
1994. Host recognition and acceptance behaviour in two
aphid parasitoid species: Aphidius ervi and Aphidius microlophi (Hymenoptera: Braconidae). Bulletin of Entomological
Research 84:57–64.
Phillips, C. B., D. B. Baird, L. L. Lline, M. R. McNeill, J. R.
Proffitt, S. L. Goldson, and J. M. Kean. 2008. East meets
west: adaptive evolution of an insect introduced for
biological control. Journal of Applied Ecology 45:948–956.
Pike, K. S., P. Stary, T. Miller, D. Alison, L. Boydston, and G.
Graf. 1999. Host range and habitats of the aphid parasitoid
Diaeretiella rapae (Hymenoptera: Braconidae) in Washington State. Environmental Entomology 28:61–71.
Ecological Applications
Vol. 20, No. 1
Powell, W., and A. F. Wright. 1988. The abilities of the aphid
parasitoids Aphidius ervi Haliday and A. rhopalosiphi De
Stephani Perez (Hymenoptera: Braconidae) to transfer
between different known host species and the implications
for the use of alternate hosts in pest control strategies.
Bulletin of Entomological Research 78:683–693.
Remington, C. L. 1968. The population genetics of insect
introductions. Annual Review of Entomology 13:425–426.
Roff, D. A. 1992. The evolution of life histories: theory and
analysis. Chapman and Hall, New York, New York, USA.
Roush, R. T. 1990. Genetic considerations in the propagation
of entomophagous species. Pages 373–387 in R. R. Baker and
P. E. Dunn, editors. New directions in biological control.
Liss, New York, New York, USA.
Sanchez, J. A., F. Canovas, and A. Lacasa. 2007. Thresholds
and management strategies for Aulacorthum solani (Hemiptera: Aphididae) in greenhouse pepper. Journal of Economic
Entomology 100:123–130.
SAS Institute. 2007. JMP. Version 7.0. SAS Institute, Cary,
North Carolina, USA.
Sequeira, R., and M. Mackauer. 1992. Nutritional ecology of
an insect parasitoid association: the pea aphid Aphidius ervi
system. Ecology 73:183–189.
Smith, M. A., D. M. Wood, D. H. Janzen, W. Hallwachs, and
P. D. N. Herbert. 2007. DNA barcodes affirm that 16 species
of apparently generalist tropical parasitoid flies (Diptera,
Tachinidae) are not all generalists. Proceedings of the
National Academy of Sciences (USA) 104:4967–4972.
Stireman, J. O., J. D. Nason, S. Heard, and J. M. Seehawer.
2006. Cascading host-associated genetic differentiation in
parasitoids of phytophagous insects. Proceedings of the
Royal Society B 273:523–530.
van den Bosch, R. 1964. Encapsulation of the eggs of
Bathyplectes curculionis (Thompson) (Hymenoptera: Ichneumonidae) in larvae of Hypera brunneipennis (Boheman) and
Hypera positica (Glyllenhal) (Coleoptera: Curculionidae).
Journal of Invertebrate Pathology 6:343–367.
van Lenteren, J. C. 1997. Commercial mass production and
pricing of organisms for biological control of pests in Europe.
Biological Control 10:143–149.
van Lenteren, J. C. 2000. A greenhouse without pesticides: fact
or fantasy? Crop Protection 19:375–384.
van Lenteren, J. C., L. Z. Hua, and J. W. Kamerman. 1995. The
parasite–host relationship between Encarsia formosa (Hymenoptera: Aphelinidae) and Trialeurodes vaporariorum (Homoptera: Aleyrodidae). XXVI. Leaf hairs reduce the capacity
of Encarsia to control greenhouse whitefly on cucumber.
Journal of Applied Entomology 119:553–559.
Vaughn, T. T., and M. F. Antolin. 1998. Population genetics of
an opportunistic parasitoid in an agricultural landscape.
Heredity 80:152–162.
Villagra, C. A., F. Pennacchio, and H. M. Niemeyer. 2007. The
effect of larval and early adult experience on behavioural
plasticity of the aphid parasitoid Aphidius ervi (Hymenoptera, Braconidae, Aphidiinae). Naturwissenschaften 94:903–
910.
Winder, L. M., C. B. Phillips, C. L. Williams, R. C. Cane, K.
Paterson, C. L. Vink, and S. L. Goldson. 2003. PCR-based
DNA analysis corroborates phenotypic evidence of transAndean variation in the South American parasitoid Microctonus hyperodae. Bulletin of Entomological Research 95:
289–298.