Entomopathogenic Nematodes: Host-Finding

Ability in the Presence of Plant Roots

HO YUL CHOO,' HARRY K. KAYA,2·3T. M. BURLANDO,2

AND RANDY GAUGLER'

Environ. Entomol. 18(6): 1136-1140 (1989)

ABSTRACT The host-finding ability of Heterorhabditis heliothidis (Khan, Brooks, and
Hirschmann) and Steinernema feltiae Filipjev in the presence of plant roots in sandy soil
was investigated. During a 7-d period, host-finding ability of H. heliothidis was not impaired
when the dried root weights of corn, tomato, or marigold plants were <1.6 g or of S. feltiae
when dried root weights were <3.6 g. For example, host mortality by H. heliothidis was

Downloaded from http://ee.oxfordjournals.org/ at University of Nebraska-Lincoln Libraries on October 16, 2015

85% for controls (no roots) and 85% for corn roots (1.6 g); host mortality by S. feltiae was
21% for controls and 16% for corn roots (3.6 g). However, host mortalities between controls
and corn roots for H. heliothidis during a 2-d period showed significant differences when
root weights were> 1.5 g. Thus, treatments with root weights> 1.5 g had significantly lower
host mortalities (range, 27-50%) compared with control treatments (range, 73-96%); treat-
ments with root weights <1.2 g showed no significant difference (range, 79-80%) compared
with control treatments (range, 82-85%). Linear regression suggested a relationship between
root weight and host mortality. Increasing root weight (x) significantly lowered percentage
(y) host mortality (y = 81.8 -7.5(x), r2 = 0.57, P = 0.0001). Although significant differences
between controls and roots> 1.5 g were observed for H. heliothidis, this nematode species
showed greater motility, hence greater host-finding ability, than S. feltiae in the presence
of roots because host mortality occurred over a shorter period of time with higher mortality
rates and with fewer nematodes.

KEY WORDS Nematoda, biological control, host finding, plant roots

ENTOMOPATHOGENIC NEMATODES have shown to the nematode or host 'resistance to nematode

promise for biological control since the early 1930s infection.
(Poinar 1979, Gaugler 1981, Kaya 1985), but only An attribute of steinernematid and heterorhab-
recently has the technology to produce steinernem- ditid nematodes shared with few microbial and no
atid and heterorhabditid nematodes commercially chemical insecticides is the ability to find and infect
been achieved (Bedding 1984, Georgis 1987). Be- insect hosts. Accordingly, dispersal behavior of these
cause the natural habitat of these nematodes is the nematodes in the soil has been studied extensively.
soil, their primary application has been directed as Although both nematode genera tend to remain
biological insecticides against a number of soil-in- where they are placed, some will disperse vertically
habiting pests, especially coleopterous and lepi- and horizontally (Yloyle & Kaya 19131; Georgis &
dopterous insects (Gaugler 1981, Kaya 1985). How- Poinar 1983a,b,c; Sandner 1986; Schroeder & Bea-
ever, results of field trials have been inconsistent. vers 1987). The presence of an insect host increases
The basis underlying unsuccessful trials is poorly dispersal (Moyle & Kaya 1981; Georgis & Poinar
understood, although extremes of the abiotic soil 1983a,b); Heterorhabditis spp. show greater mo-
environment (e.g., temperature, moisture, and tex- tility in soil than Steinernema spp. (Georgis &
ture) can adversely affect survival or behavior and Poinar 1983c). Several factors increase dispersal of
consequently, their efficacy (Kaya 1985). Biotic entomopathogenic nematodes. Steinernema jel-
factors also are capable of negatively affecting tiae Filipjev, the species on which most studies have
nematode populations (Epsky et al. 1988, Timper been conducted, is attracted to various chemical
& Kaya 1988). Alternative reasons for the incon- stimuli from the host (Pye & Burman 1978; Schmidt
sistencies include nonattractiveness of the insects & All 1978, 1979), chemical ions and bacteria (Pye
& Burman 1981), heat emitted by the host (Byers
I Visiting scholar, Department of Nematology, University of & Poinar 1982), and carbon dioxide (Gaugler et a!.
California, Davis;sponsoredby the Korea Scienceand Engineer- 1980). Moreover, Bird & Bird (1986) demonstrated
ing Foundation. Current address: Department of Agricultural Bi-
ology,GyeongsangNational University,Chinju 660-701,Republic that S. glaseri (Steiner), a close relative of S. jeltiae,
of Korea. was attracted to the meristematic region of plant
• Department of Nematology, Universityof California, Davis, roots, perhaps in response to the release of carbon
Calif. 95616.
dioxide. Because many root-feeding insects are
3 To whom reprint requests should be addressed.

, Department of Entomology, Rutgers University,New Bruns- considered primary targets for these nematodes,
wick, N.J. 08903. we investigated the influence of roots on the host-

0046-225x/89/1136-1140$02.00/0 © 1989 EntomologicalSocietyof America

December 1989 CHOO ET AL.: HOST-FINDING ABILITY OF NEMATODES 1137

finding behavior of S. feltiae and Heterorhabditis moist filter paper (Whatman No.1), and the roots
hcliothidis (Khan, Brooks, and Hirschmann). were washed, placed on paper towels, oven-dried
at 105°C for 24 h, and weighed. Soil moisture was
measured gravimetrically in some experiments at
Materials and Methods
the time of insect removal. The insects were held
Nematode and Insect Culture. S. feltiae (All for 7 d, and all cadavers were dissected to confirm
strain) and H. he/iothidis (NC strain) were pro- infection by the nematode.
duced in Galleria mel/onel/a L. (Dutky et al. 1964) Statistical Analyses. Gal/eria mortalities are pre-
stored at 100e at 20,000 infective nematodes per sented as f ± 1 SD, and comparisons were made
ml and were used within 6 wk of production. The between controls (no roots) and roots at P < 0.05
test insect was last-instar G. mellonel/a reared on in the nematode treatments by using the equality
a bran diet or purchased from a commercial source of proportions statistic (Freund 1979). Galleria
(Northern Bait Co., Chetek, Wis.). mortality (dependent variable) for H. heliothidis

Downloaded from http://ee.oxfordjournals.org/ at University of Nebraska-Lincoln Libraries on October 16, 2015

Plants. Marigold, Tagetes patu/a L. 'Cinnabar' and corn root weight data (independent variable)
(Chas. H. Lily Co., Portland, Oreg.); tomato, Ly- were subjected to linear regression. Arcsine trans-
copersicon escu/entum L. 'Murietta' (Moran Seed, formation was made before the analysis. The soil
Davis, Calif.); and corn, Zea mays L. 'White Cloud' moisture data are expressed as f ± SE and t tests
(Northrup King Seeds, Minneapolis, Minn.) were were conducted to detect significant differences (P
planted as seeds in 455-ml (14-oz) styrofoam cups < 0.05).
with four drainage holes (3-4 mm) in the bottom
ami maintained in a pesticide-free greenhouse until
Results
brought to the laboratory for testing. A marigold
or tomato seed was sown about 1.5 cm deep in sand No significant differences were observed in Gal-
(particle size 0.21-0.85 mm). Three corn seeds at leria mortality between 50 and 100 S. fe/tiae or
about 2.0 cm in depth and spaced about 2.5 cm between 5 and 10 H. heliothidis infective nema-
from each other were planted in each cup. A cup todes per cup, and the data were combined. Gal-
with sand but without seed served as a control. leria mortality in the no-nematode treatments was
E"achcup contained about 440 cc of sand, providing less than 5%, and none of the dead insects were
a depth of 11 cm from the top to the bottom of infected by nematodes.
the cup. All cups were watered daily either with At 7 d after inoculation, Galleria mortality by
a half-strength Hoagland's solution or with deion- S. fe/tiae in a comparison between controls and
ized water. Watering was withheld from cups with- roots of corn (dried root weight 1.52 ± 0.21 g),
out plants 3 d before a test because previous ex- marigold (0.19 ± 0.03 g), or tomato (0.30 ± 0.19
perience showed that high moisture levels in these g) showed no significant difference (Fig. lA). In a
cups adversely affected the survival of test insects. comparison between sparse corn roots (0.34 ± 0.05
The plants were watered only with deionized water g) and controls, Galleria mortality by S. feltiae was
for the last 3 d before a test. 11.1 ± 7.7% and 11.1 ± 3.8%, respectively; be-
Bioassay. A hole (3 cm deep) was made with a tween dense corn roots (3.57 ± 0.39 g) and controls,
No.5 cork borer in the bottom center of each cup. Galleria mortality was 20.8 ± 6.2% and 15.5 ±
A last-instar Galleria was placed in the hole and 3.9%, respectively. In both instances, the differ-
covered with sand. The hole was plugged with ences were not significant. In the sparse-dense root
colton, and the cup was returned to the upright experiments, soil moisture was not significantly dif-
position. The treatments were cups with plants and ferent at 16.3 ± 0.3% for sparse roots and 15.7 ±
without plants (control); each cup received 50 or 0.3% for controls but was significantly different (t
100 infective juveniles of S. fe/tiae, 5 or 10 infective = 8.42, df = 88) at 5.6 ± 0.6% for dense roots and
juveniles of H. heliothidis, or no nematodes. S. 11.2 ± 0.2% for controls.
fe/tia<' in 0.5 ml of water were pipetted by using Although Galleria mortality by H. heliothidis
a Pipetman, whereas H. he/iothidis in about 0.05 was generally higher than by S. feltiae at 7 dafter
ml of water were pi petted by mouth to the sand inoculation, the results were similar in that no sig-
surface at the center of the cup. Different numbers nificant differences in mortality were obtained with
of S. fe/tiae and H. heliothidis were needed to corn (dried root weight 1.63 ± 0.75 g), marigold
bracket Gal/eria mortality between 50 and 90% (0.19 ± 0.08 g), or tomato (0.59 ± 0.44 g) roots
bl'cause S. fe/tiae disperses poorly and H. helio- and controls (Fig. IB). When the insects were re-
thidis dispPTses readily. Each treatment was rep- moved at 2 d after nematode inoculation, Galleria
licated at least three times with 8-15 cups per mortality in tomato roots versus controls also showed
replicate. plants were not watered but were cov- no significant difference. In two experiments, Gal-
ered with a plastic sheet to reduce water loss and leria mortality was 66.7 ± 20.0% in the root treat-
maintained in the laboratory at 24 ± 2°e. ments (0.22 ± 0.07 g) and 75.6 ± 10.1 % in the
At either 2 or 7 d after addition of the nematodes, controls, and 71.1 ± 21.4% with roots (1.20 ± 0.08
the plant stems were cut at the soil level, the insects g) and 80.0 ± 17.6% in the controls. In the presence
were removed from the cups and placed individ- of sparse ( < 1.20 g) and dense (> 1.55 g) corn roots,
ually in Petri dishes (60 by 15 mm) containing a significantly lower Galleria mortality occurred
1138 ENVIRONMENTAL ENTOMOLOGY Vol. 18, no. 6

A Heterorhabditis

a
100
Steinernema a
100 a a a
a a a
80
80
60 ~
iii
40 t:0 60
b
~ E

~
0
20
-••
0
J:
40
b

E
;; B tft
Heterorhabditis 20
0
a

Downloaded from http://ee.oxfordjournals.org/ at University of Nebraska-Lincoln Libraries on October 16, 2015

J: 100

:R.
0
80

60
e ,., Ol
e Ol
~c:
Ol
~c:
Ol
g Ol

C ,., C 0
"!~ '"
Ii? ,.,'" c:
a
::;:
40 0
<..l a 0
<..l
a
<..l ~ a
<..l Iii <..l «i

20 Fig. 2. Mortality of G. mel/onella by H. heliothidis

in control (no roots) and corn roots 2 d after nematode
application. Numbers on the abscissa are the dried root
'"
0 ~ '"
0 ~ o'"o weights. Different lower case letters above each pair of
~c: 0
a:
0
a: a: bars indicate significant differences (P < 0.05) using the
0
C c:
o
() c: 0
()
"0 () o equality of proportion statistics.
0 0 iii
() Ol E
.~ o
I-
~ species, the presence of an insect host may have
Fig. 1. Mortality of G. mellonella by S. feltiae (A) enhanced nematode motility as shown by previous
and H. heliothidis (B) in control (no roots) and corn, research (Moyle & Kaya 1981; Georgis & Poinar
marigold, or tomato roots 7 d after nematode application. 1983a,c). Although steinernematids are attracted
Comparisons were made using the equality of proportion
statistics between control and roots. The same lower case to roots (Bird & Bird 1986) and carbon dioxide
letter above each pair of bars indicates no significant (Gaugler et al. 1980), the presence of an insect host
differences (P > 0.05). may have been a stronger attraction for S. feltiae
than the presence of roots. No studies have been
conducted to indicate that H. heliothidis is attract-
when root weights were 1.55 ± 0.28 (Z = 4.87), ed to roots, but this nematode has many charac-
5.32 ± 0.57 (Z = 4.25), and 6.41 ± 1.59 g (Z = teristics of steinernematids, including directed ori-
3.95) compared with the controls (Fig. 2). There entation to insect hosts and carbon dioxide (R.G.,
was a correlation between increasing root weights unpublished data) and is presumed to respond to
(x) and lower percentage (y) Galleria mortality (y roots in a similar fashion.
= 81.8 - 7.5(x), r2 = 0.57, P = 0.0001). In these The host-finding ability of H. heliothidis was
root density tests, no significant difference was ob- reduced by the presence of dense roots, while that
served in Galleria mortality when moisture levels of S. feltiae was not, suggesting that the former
were significantly different (t = 3.41, df = 58) was more sensitive to root density. The question of
between controls (4.6 ± 0.3%) and roots (8.2 ± why dense roots affected the host-finding ability of
0.6%) with a dry weight averaging 1.20 g. On the H. heliothidis but not S. feltiae remains to be re-
other hand, significant differences in Galleria mor- solved. Because H. heliothidis has greater dispersal
tality were observed when the moisture levels were ability than S. feltiae (Georgis & Poinar 1983c),
not significantly different between controls (10.9 dense roots may have inhibited motility of H. he-
± 0.2%) and dense roots (9.7 ± 0.8%) with a dry liothidis to a greater extent than the less motile S.
weight of 6.41 g. feltiae. It also could be argued that carbon dioxide
and root exudate levels were sufficiently high to
"confuse" the behavioral pattern of H. heliothidis
but not of S. feltiae. Perhaps our presumption is
Discussion
incorrect, and H. heliothidis is more sensitive to
No significant difference in Galleria mortality root carbon dioxide or exudates than S. feltiae.
by S. feltiae was observed in the presence or ab- Other factors accounting for these results may be
sence of corn, marigold, or tomato roots, indicating that dense roots changed the abiotic or biotic char-
that roots did not influence the nematode's host- acteristics of the soil such as porosity, chemical
finding ability. Similar results were obtained for composition, and microorganism population which
H. heliothidis when the roots were classified as affected H. heliothidis but not S. feltiae. Dense
sparse « 1.2 g dry weight). For these two nematode roots may have physically obstructed H. heliothidis
December 1989 CHOO ET AL.: HOST-FINDING ABILITY OF NEMATODES 1139

but not S. feltiae, but because both species are References Cited
similar in size, this possibility appears unlikely. In-
teraction among factors also may account for the Bedding, R. A. 1984. Large scale production, storage
results. and transport of the insect-pathogenic nematodes
Soil moisture, another factor which could have Neoaplectana spp. and Heterorhabditis spp. Ann.
accounted for the results, was a variable which was App!. Bio!. 101: 117-120.
difficult to control in our experiments. In some tests, Bird, A. F. & J. Bird. 1986. Observations on the use
it was similar between absence and presence of of insect parasitic nematodes as a means of biological
roots, but in other tests there were significant dif- control of root-knot nematodes. J. Parasito!. 16: 511-
516.
ferences. Soil moisture levels, however, did not seem
Byers, J. A. & G. O. Poinar, Jr. 1982. Location of
to affect the host-finding ability of either S. feltiae insect hosts by a nematode, Neoaplectana carpocap-
or H. heliothidis in our tests. Although they ranged sae, in response to temperature. Behaviour 79: 1-10.
from 4.6 to 16.3%, no correlation in Galleria mor- Dutky, S. R., J. V. Thompson & G. E. Cantwell. 1964.

Downloaded from http://ee.oxfordjournals.org/ at University of Nebraska-Lincoln Libraries on October 16, 2015

tality because of nematode infection was observed A technique for the mass propagation of the 00-136
when soil moisture was high or low. nematode. J. Insect Patho!. 6: 417-422.
There were two major differences between the Epsky, N. 0., D. E. Walter & J. L. Capinera. 1988.
experiments conducted for H. heliothidis and S. Potential role of nematophagous microarthropods as
feltiae: number of nematodes per cup and number biotic mortality factors of entomogenous nematodes
(Rhabditida: Steinernematidae, Heterorhabditidae).
of days of insect exposure to the nematodes. Few
J. Econ. Entomo!. 81: 821-825.
H. heliothidis were required for host infection Freund, J. E. 1979. Modern elementary statistics.
compared with S. feltiae. Our aim was to use the Prentice-Hall, Englewood Cliffs, N.J.
minimum number of nematodes to obtain between Gaugler, R. 1981. Biological control potential of neo-
50 and 90% Galleria mortality to test our hypoth- aplectanid nematodes. J. Nematol. 13: 241-249.
esis that roots may influence the nematodes' ability Gaugler, R., L. LeBeck, B. Nakagaki & G. M. Boush.
to find hosts. Initially, the 7-d period appeared to 1980. Orientation of the entomogenous nematode
be satisfactory for both nematode species, but as Neoaplectana carpocapsae to carbon dioxide. Envi-
more tests were conducted, 7 d were required for ron. Entomo!. 9: 649-652.
S. feltiae (which disperses poorly), whereas 2 d Gaugler, R., T. McGuire & J. Campbell. 1989. Ge-
netic variability among strains of the entomopatho-
were better for showing differences for H. helioth-
genic nematode Steinernema feltiae. J. Nematol. 21:
idis (which disperses readily). If the 2-d test period 247-253.
with H. heliothidis was extended to 7 d, no dif- Georgis, R. 1987. Nematodes for biological control of
ferences in host mortality in dense roots would urban insects, pp. 816-821. In 194th Meeting, Amer-
probably have been observed. ican Chemical Society, New Orleans, Vol. 27.
Heterorhabditis heliothidis is clearly superior to Georgis, R. & G. O. Poinar, Jr. 1983a. Effect of soil
S. feltiae in host-finding ability and in our esti- texture on the distribution and infectivity of Neo-
mation has greater potential as a biological control aplectana carpocapsae (Nematoda: Steinernemati-
agent for insect pests in soil because fewer H. he- dae). J. Nemato!. 15: 308-311.
1983b. Effect of soil texture on the distribution and
liothidis caused equal or greater Galleria mortality infectivity of Neoaplectana glaseri (Nematoda: Stei-
in 2 d than did S. feltiae in 7 d. Although H. nernematidae). J. Nemato!. 15: 329-332.
heliothidis was affected by dense roots confined in 1983c. Vertical migration of Heterorhabditis bacte-
a cup and S. feltiae was not, H. heliothidis in the riophora and H. heliothidis (Nematoda: Heterorhab-
field will not have to contend with a bound root ditidae) in sandy loam soil. J. Nematol. 15: 652-654.
system, and its host-finding ability should not be Kaya, H. K. 1985. Entomogenous nematodes for in-
impaired by dense roots. Because fewer H. helio- sect control in IPM systems, pp. 283-302. In M. A
thidis are needed for efficacious results, this fact Hoy & O. C. Herzog [eds.], Biological control in ag-
should translate into better economics for insect ricultural systems. Academic, New York.
Moyle, P. L. & H. K. Kaya. 1981. Dispersal and
control, assuming mass production, formulation,
infectivity of the entomogenous nematode, Neo-
and other costs are comparable with S. feltiae. aplectana carpocapsae Weiser (Rhabditida: Steiner-
Recent studies initiated by Gaugler et al. (1989) nematidae), in sand. J. Nemato!. 13: 295-300.
have increased the host-finding ability of S. feltiae; Poinar, G. 0., Jr. 1979. Nematodes for biological
this nematode eventually may produce strains of control of insects. CRC, Boca Raton, Fla.
nematodes as effective as those of H. heliothidis. Pye, A. E. & M. Burman. 1978. Neoaplectana car-
pocapsae: infection and reproduction in large pine
weevil larvae, Hylobius abietis. Exp. Parasito!. 46: I-
ll.
Acknowledgment 1981. Neoaplectana carpocapsae: nematode accu-
mulations on chemical and bacterial gradients. Exp.
We thank Sherita Normington for technical assistance; Parasito!. 51: 13-20.
and Mary Barbercheck, Bruce Jaffee, and Patricia Tim- Sandner, H. 1986. Movement of invasive larvae of
per for their comments and suggestions on the manu- Heterorhabditis bacteriophora Poinar in soil. Ann.
script. This research was supported in part by USDA Wars. Agric. Univ. SGGW-AR Hort. 20: 45-47.
Competitive Grants, Biological Stress on Plants Program Schmidt, J. & J. N. All. 1978. Chemical attraction of
No. 84 CRCR-I-1462. Neoaplectana carpocapsae (Nematoda: Steinerne-
1140 ENVIRONMENTAL ENTOMOLOGY Vol. 18, no. 6

matidae) to insect larvae. Environ. Entomol. 7: 605-- Timper, P. & H. K. Kaya. 1988. Role of the second-
607. stage cuticle in determining the susceptibility of en-
1979. Attraction of Neoaplectana carpocapsae (Nem- tomopathogenic nematodes to the endoparasitic fun-
atoda: Steinernematidae) to common excretory prod- gus Hirsutella rhossiliensis. J. Nematol. 20: 661.
ucts of insects. Environ. Entomol. 8: 55-61.
Schroeder, W. J. & J. B. Beavers. 1987. Movement Received for publication 9 December 1988; accepted
of the entomogenous nematodes of the families Het- 15 March 1989.
erorhabditidae and Steinernematidae in soil. J. Nem-
atol. 19: 257-259.

Downloaded from http://ee.oxfordjournals.org/ at University of Nebraska-Lincoln Libraries on October 16, 2015