Role of Oviposition Preference in an Invasive Crambid Impacting Two
Graminaceous Host Crops
F.P.F. REAY-JONES,1,2L. T. WILSON,3A. T. SHOWLER,4T. E. REAGAN,1AND M. O. WAY3
Environ. Entomol. 36(4): 938Ð951 (2007)
Oviposition preference studies of the Mexican rice borer, Eoreuma loftini (Dyar),
on sugarcane, Saccharum spp., and rice, Oryza sativa L., showed that drought stressed sugarcane
susceptible sugarcane cultivar LCP 85Ð384 was 1.6-fold more attractive than HoCP 85Ð845 based
on numbers of eggs per egg mass. Egg masses were 9.2-fold more abundant and 2.3-fold larger on
sugarcane than on rice. Rice, however, was preferred to sugarcane on a plant biomass basis.
Oviposition on sugarcane occurred exclusively on dry leaf material, which increased under
drought stress. Egg masses per plant increased on drought stressed sugarcane and were correlated
with several foliar free amino acids essential for insect growth and development. The more
levels of several free amino acids than the susceptible cultivar Cocodrie. The association of host
to sugarcane and rice production areas were estimated for Texas and Louisiana based on the
availability of each host in different regions of each state. These results suggest that, where
because of this cropÕs substantially greater biomass compared with rice. Abundance later in the
season would also favor sugarcane; however, the abundance on rice would be greater than
expected solely based on host availability, largely because of the greater preference per gram of
rice plant dry weight.
Mexican rice borer, Eoreuma loftini, sugarcane, rice, oviposition
Oviposition of many lepidopterans is a critical step in
instars (Feeny et al. 1983, Showler 2002, Showler and
Moran 2003). Visual, olfactory, gustatory, and me-
plant selection (Ramaswamy 1988). Plant phenotypic
characters that inßuence acceptability for insect ovi-
position include leaf pubescence (Sosa 1988), color
(Levinsonetal.2003),phenologicalstage(More ´ etal.
2003), and leaf shape (Mackay and Jones 1989). In
status (Myers 1985, Showler and Moran 2003), and
secondary metabolites (Feeny et al. 1983) inßuence
host selection by insect herbivores. Determining
quantifying oviposition preference for host crops can
assist in the development of pest management strat-
egies (Renwick and Chew 1994, Showler 2004a).
The availability of host plant free amino acids
(FAAs) is a critical factor in population growth of
many insect herbivores (McNeil and Southwood
1978), and insects can respond to changes in the nu-
tritional quality of a plant (Rhoades 1983, Showler
host plant FAAs have been associated in many plants
with numerous stresses (Rabe 1994, Showler 2004b),
including drought (Gzik 1996, Showler 2002). Accu-
mulated FAAs lower the water potential of cells
and may reduce water loss through osmoregulation
The Mexican rice borer, Eoreuma loftini (Dyar), is
mays L., sorghum, Sorghum bicolor L. Moench (Youm
et al. 1988), rice, Oryza sativa L. (Reay-Jones et al.
1994). Eoreuma loftini originated in Mexico and be-
came the dominant insect pest of sugarcane in the
Lower Rio Grande Valley of Texas since it became
established in 1980 (Johnson 1984), now representing
?95% of the sugarcane stalk borer population, which
1Department of Entomology, Louisiana Agricultural Experiment
2Corresponding author: Clemson University, Department of
Entomology, Soils and Plant Sciences, Pee Dee Research and
Education Center, 2200 Pocket Rd., Florence, SC 29506-9727
3Texas A&M University System Agricultural Research and Exten-
sion Center, Beaumont, TX 77713.
la Garza Subtropical Agricultural Research Center, USDAÐARS,
Weslaco, TX 78596.
0046-225X/07/0938Ð0951$04.00/0 ? 2007 Entomological Society of America
saccharalis (F.) (Legaspi et al. 1999b). By 1989, it
moved into the rice production area of east Texas
(Browning et al. 1989), where it is responsible for
major yield loss in rice (Reay-Jones et al. 2005b).
Invasion of Western Louisiana, where sugarcane and
rice are grown in close proximity, is likely imminent
(Reay-Jones et al. 2007). The objectives of this study
and rice cultivars at different phenological stages, (2)
behind these relationships, and (3) to estimate ovi-
position patterns on sugarcane and rice in Texas and
projected patterns in Louisiana.
Materials and Methods
This study was conducted at the Texas A&M Agri-
cultural Experiment Station in Weslaco, TX, during
the summers of 2003 and 2004. Sugarcane plants (cul-
tivars LCP 85Ð384 and HoCP 85Ð845) were grown in
a greenhouse in 3.8-liter pots containing nursery pot-
Seba Beach, Canada). Sugarcane nodes collected in
Þelds in the Lower Rio Grande Valley of Texas were
planted individually in pots and fertilized with 200 ml
of Peters Professional water-soluble general purpose
20-20-20 N-P-K (Scotts-Sierra Horticultural Products,
Maryville, OH) at 18.3 g/liter of water approximately
once every 4 wk. Plants were watered with 1.5 liters
three times per week. The two phenological stages of
sugarcane used in this study were plants with Þve
elongated nodes (?89 cm from soil surface to plant
apex of the stalk) and elongated 10 nodes (?158 cm
from soil to apex of the stalk). In the drought-stressed
treatment, sugarcane plants were watered once a
week with 1.5 liters for 2 wk before starting the ex-
the normal irrigation regimen. The watering treat-
ments were initiated before the plants reached the 5-
and 10-node stages.
greenhouse in 1.1 liter of potted soil (3 plants/pot),
and received two applications of 0.79 g/pot of urea
(46% N) corresponding to 207 kg/ha of N at 1 and 5
wk after emergence. Rice was ßooded 6 wk after
emergence. The four phenological stages used in this
study were the 5Ð6 node tillering (1 wk after emer-
gence), 9Ð11 node tillering (3 wk after emergence),
(10Ð11 wk after emergence) (Vergara 1991).
Oviposition Choice Tests. Eoreuma loftini adults
were obtained from a laboratory colony at the Texas
A&M System Agricultural Research and Extension
Center in Weslaco that was initiated from larvae col-
lected in sugacrane Þelds in the Lower Rio Grande
Valley of Texas. Every year, Þeld-collected larvae
artiÞcial diet in an environmental chamber (Martinez
(L:D). Pupae were separated by sex and placed in
3.8-liter plastic containers for emergence under the
same conditions. Adults used in these experiments
were 48 h old.
Seven oviposition experiments were conducted,
with four treatments in each, covering the 16 treat-
ments described in Table 1. Each experiment had
either sugarcane (1 and 7), rice (3Ð5), or sugarcane
and rice (2 and 6). Each test was a randomized com-
samples) of each of the four treatments within each
block. A greenhouse cage (2 by 2 by 2 m) was used as
initiated with the release of 30 male and 30 female
moths in each cage and ended 6 d later. Numbers of
eggs and egg masses and location of the eggs on the
host plant were recorded. For the remainder of this
paper, data from an experiment are referred to as
datasets 1Ð7 (as depicted in Table 1).
Plant Measurements. At the end of each experi-
counted on each sugarcane plant. Rice tillers and
green and dry leaves were counted. Dry weight was
determined for plants in three pots of each treatment
after 5 d in an oven at 75?C. Weights on a per plant
Table 1.Design of E. loftini oviposition studies, Weslaco, TX, 2003–2004
Species Cultivar Stage
SugarcaneLCP 85Ð3845 nodes NonÐdrought stressed
HoCP 85Ð845 5 nodesX
Rice Cocodrie Tillering 5Ð6 nodes
Tillering 9Ð11 nodes
Tillering 5Ð6 nodes
Tillering 9Ð11 nodes
August 2007REAY-JONES ET AL.: OVIPOSITION PREFERENCE IN E. loftini939
basis are presented in Table 2. In experiments 1 and 3
(Table 1), the third leaf from the apex of each sug-
arcane plant (n ? 8 per treatment) was excised, and
water potential was measured immediately with a
model 610 (PMS Instrument Co., Corvalis, OR) pres-
sure bomb at 0900 hours. The second leaf from the
rice treatments in experiments 2 and 6 (n ? 8 per
treatment). Each 1-g leaf tissue sample was homoge-
nized with 10 ml of 0.1 N HCl using a Virtishear
homogenizer (Virtis, Gardiner, NY). Five grams of
homogenate from each sample was placed in separate
10-ml tubes and centrifuged at 10,000 rpm for 30 min.
Samples were stored at ?80?C.
One milliliter of supernatant from each sample was
Þltered through a 0.5-?l Þlter (EconoÞlter; Agilent,
Santa Clara, CA; pore size ? 0.45 ?m, diameter ? 25
mm) Þtted to a 5-ml plastic syringe. Samples were
placed in the autosampler of an Agilent 1100 Series
(Agilent Technologies, Atlanta, GA) reversed-phase
high-performance liquid chromatograph (HPLC)
with a binary pump delivering solvent A (1.36 g so-
acid to bring the pH to 7.2 ? 0.05 [95% CI]) and
solvent B (1.36 g sodium acetate trihydrate ? 100 ml
puriÞed HPLC grade water [acetic acid added to this
mixture to bring the pH to 7.2 ? 0.05; 95% CI] ? 200
ml acetonitrile ? 200 ml methanol) at 100 and 1.0
ml/min on a Zorbax Eclipse AAA 4.6 by a 150-mm
262 and 338 nm were monitored on a variable wave-
length detector for 48 min/sample. The autosampler
pH 10.2 in water), 1 ?l 9-ßuorenylmethylchlorofor-
mate (FMOC), and 1 ?l ophthalaldehyde (OPA) de-
2 ?l for chromatographic separation of free amino
acids (FAAs). IdentiÞcation and quantiÞcation of 17
teine, glutamic acid, glycine, histidine, isoleucine,
leucine, lysine, methionine, phenylalanine, proline,
by calibrating with a standard mixture of amino acids.
Peak integration accuracy was enhanced by manual
establishment of peak baselines using Agilent soft-
Data Analyses. Oviposition choice, based on egg
masses, eggs per egg mass and total eggs per plant,
ness by performing ?2analyses of contingency tables
(Zar 1999) for each dataset and overall for the entire
experiment. Expected frequencies of egg laying for
each of the four treatments in each dataset were one-
would occur if oviposition choice was random.
Preference can be quantiÞed as departure from the
probability of randomly selecting a host based on
availability and has been used to predict patterns of
herbivore oviposition, parasitism, and predation on
different hosts (Murdoch 1969, Manly et al. 1972,
Chesson 1978, Wilson and Gutierrez 1980, Pickett et
al. 1989, Murphy et al. 1991). When multiple hosts are
simultaneously made available to an herbivore, pref-
host dry mass basis can be derived using equation 1.
where ? ˆij? estimated preference shown for the ith
per plant or per gram of plant dry mass, respectively,
on the ith treatment for the jth dataset; and max nj?
mean maximum number of eggs laid per plant or per
for the jth dataset.
Table 2. Sugarcane and rice measurements from greenhouse oviposition test, Weslaco, TX, 2003–2004
SugarcaneLCP 85Ð3845 nodes No
HoCP 85Ð8455 nodes
RiceCocodrie Tillering 5Ð6 nodes
Tillering 9Ð11 nodes
Tillering 5Ð6 nodes
Tillering 9Ð11 nodes
P ? F
Means within the same column followed by the same letter are not signiÞcantly different (P ? 0.05; TukeyÕs ?1953? HSD).
adf ? 15,68;bdf ? 15,96;
cdf ? 7,24;ddf ? 7,56.
Vol. 36, no. 4
When combined with estimates of the relative den-
sity (i.e., plants per pot: one for sugarcane, three for
rice) or mass of each host for a particular dataset,
preference coefÞcients can in turn be used to provide
estimates of relative host selection for each host type
n ˆij? nj? ˆijAi/?
where ? ˆij? the estimated relative preference shown
for the ith host for the jth dataset; n ˆij? the estimated
number of eggs, egg masses, or eggs/egg mass ovipos-
ited on the ith host; and Ai? relative density or mass
of the ith host.
Although equation 1 is extremely easy to use, pref-
erence coefÞcients derived from separate experi-
ments or datasets can only be compared if they share
two or more common hosts; otherwise, the values of
different sets of coefÞcients are not relative to each
other. In our experiment, two treatments overlap be-
tween each successive dataset, thereby providing
common hosts. Experiment-wide estimates of each
preference coefÞcient can be derived from the data
using iterative nonlinear least squares regression
based on the modiÞed Gauss-Newton method (JMP;
SAS Institute 2002) (equation 3).
predicted number of eggs, egg masses, or eggs per egg
Once the experiment-wide estimates are derived,
least squares estimates of preference coefÞcients for
each of the individual datasets can in turn be derived
wide estimates (?i) and the iteratively scaled prelim-
inary estimates ([carot]?ij) obtained from equation 1
(see equations 4 and 5 using equation 6).
?? ˆij?j? ?i?2
?ij? ? ˆij?j
?? ˆij?j? ?i?2
datasets j; ?ij? the individual experiment based pref-
erence estimates shown for the ith host for the jth
dataset; and D?,j? the minimized deviation of ob-
served from predicted preference estimates for each
dataset j across all hosts i.
coefÞcients (?ij) derived for each dataset was esti-
mated using multiple linear regression analysis
(PROC REG; SAS Institute 1999). The number of dry
leaves per plant was included in the model for sugar-
cane based on previous research (Van Leerdam et al.
1986) showing the importance of this variable for E.
measurements (PROC CORR; SAS Institute 1999).
Plant measurements were pooled across experiments
and analyzed with a one-way analysis of variance
(ANOVA; PROC MIXED; SAS Institute 1999), and
(Tukey 1953) was used for mean separation. Means
wise error rates were corrected using the stepdown
method (PROC MULTTEST; SAS Institute 1999).
Simulated oviposition patterns on sugarcane and
rice were predicted in four geographical regions: (1)
Texas rice belt west of Houston (Austin, Brazoria,
Calhoun, CO, Fort Bend, Galveston, Harris, Jackson,
Matagorda, Victoria, Waller, and Wharton counties),
(2) Texas rice belt east of Houston (Chambers, Jef-
ferson, Liberty, and Orange counties), (3) southwest
Louisiana (Acadia, Allen, Avoyelles, Beauregard, Cal-
casieu, Cameron, Evangeline, Jefferson Davis, Lafay-
ette, Pointe Coupee, Rapides, and Vermilion par-
ishes), and (4) southcentral Louisiana (Assumption,
Ascension, East Baton Rouge, Iberia, Iberville,
Lafourche, St. Charles, St. James, St. John the Baptist,
St. Martin, St. Mary, Terrebonne, and West Baton
corresponding to the four rice growth stages in our
preference experiments. Projected oviposition pat-
terns were estimated using equation 1, with the ex-
periment-wide preference coefÞcients based on E.
loftini eggs per gram of dry weight, and host availabil-
ity based on estimated sugarcane and rice plant dry
weight in each region on each date. Preference esti-
mates for the nondrought stressed sugarcane cultivar
LCP 85Ð384, the dominant sugarcane cultivar in Lou-
isiana (Legendre and Gravois 2005), and the rice cul-
Louisiana, were used in the analysis. Because prefer-
ence was only measured for two sugarcane growth
stages, the youngest of which corresponded to the
oldest rice growth stage, we approximated the pref-
erence for sugarcane at three of the four dates by
linear extrapolation. The 5- and 10-node sugarcane
stages correspondED to ?6 July and 10 August, re-
spectively. In the four regions, the area producing
sugarcane (Legendre and Gravois 2005) and rice
and Extension Center in Beaumont) were estimated.
Plant weight in each region for each of the four dates
was calculated by multiplying production area by es-
timated biomass per hectare for sugarcane (using a
et al. 1993) and 2002 Texas Þeld data for rice (Wilson
et al., unpublished data). To correct for differences in
estimated biomass given by the Florida model was
multiplied by the ratio of the average sugarcane yield
August 2007REAY-JONES ET AL.: OVIPOSITION PREFERENCE IN E. loftini 941
Florida from 2000 to 2006 (81.2 tonnes/ha in Florida,
60.0 tonnes/ha in Louisiana; USDAÐNASS). The pre-
dicted relative proportion of eggs laid on each of the
two crops is presented for each region. This method
assumes (1) the distribution of E. loftini moths in a
E. loftini moths respond similarly to all cultivars, and
(3) the growth of Louisiana and East Texas sugarcane
is similar to the growth of Florida sugarcane.
and eggs per egg mass (?2? 1688.0; df ? 21; P ?
0.0001) per plant were signiÞcantly affected by host
type. A total of 1,130 egg masses (mean ? 7.5 ? 0.66
[SE] egg masses per plant) and 29,337 eggs (mean ?
194 ? 16.3 eggs per plant) were laid in this study,
(r2? 0.965), and eggs per plant (r2? 0.967). The
preference coefÞcients (Fig. 1) showed values rang-
85Ð384 at the 5-node stage) to 0.0 (both rice cultivars
at the 5Ð6 node tillering stage). Sugarcane was more
egg masses per plant, 2.3-fold based on eggs per egg
mass, and 12.9-fold based on eggs per plant. Drought
masses and 1.6-fold based on eggs per plant. Egg
masses per plant (1.2-fold), eggs per egg mass (1.6-
fold), and eggs per plant (1.5-fold) were greater on
cultivar LCP 85Ð384 than on cultivar HoCP 85Ð845.
The young sugarcane (5 nodes) was more attractive
oviposition) to 1 (maximum preference).
Oviposition preference estimates (?SD) per plant from nonlinear regression models ranging from 0 (no
Vol. 36, no. 4
than the old sugarcane (10 nodes) based on eggs per
egg mass (1.2-fold) and eggs per plant (1.3-fold). On
rice, cultivar XL8 was more attractive than Cocodrie
by 1.4-fold based on egg masses per plant, 1.7-fold
per plant. Preference estimates for egg masses per
plant, eggs per egg mass, and eggs per plant increased
with rice phenology on Cocodrie and for egg masses
per plant on XL8. The boot stage was the most attrac-
tive for cultivar XL8 based on eggs per egg mass and
eggs per plant. On sugarcane, 100% of the eggs were
On rice, 46% of the eggs were laid on dry leaves and
54% on green leaves, leaf sheaths, and on the stem.
more attractive than sugarcane using egg masses and
1.7-fold using total eggs (Fig. 2). Averaging over both
cultivars, drought stressed sugarcane at the Þve-node
stage was 6.0-fold more attractive than all other sug-
arcane treatments based on egg masses per gram of
plant dry weight and 6.2-fold based on total eggs.
Differences among the treatments were detected
were associated with oviposition estimates based on
egg masses per plant, eggs per egg mass, and eggs per
plant (Table 6). On sugarcane, correlation analyses
showed positive associations (P ? 0.05) between egg
masses per plant and both essential FAAs (arginine,
from 0 (no oviposition) to 1 (maximum preference).
Oviposition preference estimates (?SD) per gram of plant dry weight from nonlinear regression models ranging
Table 3. Multiple contrasts of plant measurements on rice and sugarcane from greenhouse oviposition test, Weslaco, TX, 2003–2004
Rice versus sugarcane
LCP 85Ð384 versus HoCP 85Ð845
Stressed sugarcane versus nonstressed
5- versus 10-node stage (sugarcane)
Cocodrie versus XL8
Tillering versus boot and heading
adf ? 1,68;bdf ? 1,96;cdf ? 1,24;ddf ? 1,56.
eP ? 0.01;fP ? 0.05;gP ? 0.05.
August 2007REAY-JONES ET AL.: OVIPOSITION PREFERENCE IN E. loftini943
Free amino acid accumulations (nmol/10 ?l juice) in rice and sugarcane leaves from greenhouse oviposition test, Weslaco, TX, 2003–2004
Means within the same row followed by the same letter are not signiÞcantly different (P ? 0.05; TukeyÕs ?1953? HSD).
adf ? 11,36.
bSum of concentrations of arginine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, threonine, and valine.
Vol. 36, no. 4
phenylalanine, and threonine) and dry leaves and
between eggs per plant and both essential FAAs (me-
thionine and threonine) and dry leaves (Table 6). On
(P ? 0.05) between egg masses per plant and both
essential FAAs (threonine and valine) and dry leaves,
between eggs per egg mass and both dry leaves and
tillers, and between eggs per plant and the essential
linear regression analyses with the sugarcane data
showed that egg masses per plant were positively as-
sociated with dry leaves (parameter estimate ?
0.0818) and methionine (0.00519) and negatively as-
df ? 4,3; P ? 0.0005; r2? 0.97). Eggs per egg mass
and alanine (0.00112) and a negative association with
P ? 0.064; r2? 0.85). Eggs per plant were positively
associated with dry leaves (0.0199), aspartic acid
P ? 0.0239; r2? 0.92). On rice, the preference coef-
Þcients were not signiÞcantly associated with any of
the plant-based estimates using regression analyses
(P ? 0.05); however, strong trends were observed,
with egg masses per plant positively associated with
oviposition test, Weslaco, TX, 2003–2004
Multiple contrasts of free amino acid accumulations (nmol/10 ?l juice) in rice and sugarcane leaves from greenhouse
LCP 85Ð384 versus
5- versus 10-node
adf ? 36.
bdf ? 1,36.
cP ? 0.05;dP ? 0.001;eP ? 0.05.
Table 6. Correlation coefficients (P < 0.1) of E. loftini oviposition estimates with plant phenology and physiochemical measurements
Plant variablenrP Plant variablenrP
Egg masses/plantDry leaves
Eggs/egg mass Dry leaves
August 2007REAY-JONES ET AL.: OVIPOSITION PREFERENCE IN E. loftini945
threonine (0.000783; F ? 79.0; df ? 2,1; P ? 0.079;
During 2004, sugarcane and rice production was 0
and 73,936 ha west of Houston, 405 and 18,836 ha east
2,187 ha in southcentral Louisiana, respectively (Fig.
3). Biomass estimates for sugarcane and rice were 2.0
and 0.12 t/ha on 1 May, 2.7 and 0.81 t/ha on 15 May,
4.9 and 7.3 t/ha on 15 June, and 7.1 and 9.3 t/ha on 6
July, respectively. The predicted relative oviposition
pattern by E. loftini on sugarcane and rice in each of
the four regions is presented in Fig. 4. These results
not consider oviposition on other hosts. On 1 May, in
areas where sugarcane is available, this crop is pro-
jected to receive 100% of the eggs, because rice is not
On 15 May, rice is more attractive than sugarcane on
a biomass basis; however, sugarcane is predicted to
receive a disproportionate amount of eggs, as a result
of its greater mass in all areas where sugarcane is
grown. On 15 June and 6 July, rice is more attractive
will approach an average of 4.7% in southeast Louisi-
ana, where rice represents only 2% of the combined
area of these two crops.
Oviposition on Sugarcane. Eggs on sugarcane were
laid exclusively on dry leaves, dry tips of leaves, and
dry leaf sheaths. Eggs have been observed on sugar-
cane in the Þeld between the leaf sheath and the stalk
leaves (Van Leerdam et al. 1984). Van Leerdam et al.
(1986) conducted a greenhouse bioassay and found
sites on dried sugarcane leaves located on the lower
part of the plant (i.e., between soil surface and 80 cm
relation between oviposition and dry leaves on sug-
arcane, with all eggs laid on dry leaves or dry tips of
leaves. The numbers of eggs laid and the number of
dry leaves per sugarcane plant increased under
showed that both E. loftini injury and production of
moths on sugarcane can be reduced by irrigation.
Preference for drought stressed sugarcane provides a
mechanism that partially explains the breakdown of
plant resistance observed in the Þeld.
Insecticide studies (Meagher et al. 1994, Legaspi et
al. 1999a, b) and extensive attempts at classical bio-
in effective E. loftini control programs. Oviposition of
E. loftini in concealed sites on dried sugarcane leaves
on the lower portion of the plant might be a mecha-
nism to protect eggs from predation, parasitism (Van
Leerdam et al. 1986), and insecticides.
Greenhouse and laboratory studies have previously
shown only slight differences in E. loftini oviposition
val establishment indicated antibiosis as a more im-
portant resistance mechanism (Meagher et al. 1996).
A Þeld study has shown that sugarcane cultivar LCP
85Ð384 was more susceptible to E. loftini than HoCP
and moth emergence per hectare (Reay-Jones et al.
2003). Cultivar LCP 85Ð384 had more dry leaves than
HoCP 85Ð845 in our study, which seemed to affect
oviposition preference. The decreased oviposition on
HoCP 85Ð845 therefore is an antixenosis mechanism
conferring resistance to E. loftini.
Relative availability of rice and sugarcane for oviposition by E. loftini across Texas and Louisiana in 2004.
Vol. 36, no. 4
Oviposition on Rice. Eoreuma loftini eggs were dis-
tributed on rice green leaves, leaf sheaths, stems, and
as on sugarcane, indicating potential increased expo-
predators, and insecticides. The relative concealment
of eggs on sugarcane might explain the preference
ber of oviposition sites (i.e., green and dry leaves) on
young rice plants. The pest status of E. loftini on rice
in the Texas Rice Belt is increasing as the insect
spreads. Field insecticide trials on rice have shown
yield losses as much as 50% or greater attributable to
stem borers [E. loftini and Diatraea saccharalis (F.)]
(Reay-Jones et al. 2005b). Insecticidal control is more
effective on rice than on sugarcane, likely because of
increased egg and larval exposure.
Drought Stress Effects on Sugarcane Physiology.
Drought stress signiÞcantly increased water potential
and levels of several FAAs (arginine, aspartic acid,
glycine, leucine, phenylalanine) in sugarcane; how-
deÞcit stress (Reay-Jones et al. 2005a, Showler 2002).
Discontinuance of daily watering of sugarcane in
2.5-fold (Muqing and Ru-Kai 1998). Other types of
stress have also increased levels of free proline in
sugarcane leaves 1.6-fold (salt stress) (Joshi and Naik
1980), 6.2-fold (Colletotrichum falcatum Went infec-
(Jain and Shrivastava 1998). When plants are subject
to dehydration, osmoregulation is achieved by accu-
mulation of free proline (Heuer 1994). Free proline
stress in sugar beets, Beta vulgaris L., by 12-fold (Gzik
(Showler and Moran 2003). Free proline seems to be
the most widespread and consistent amino acid re-
lated to drought stress (Aspinall and Paleg 1981). In
our study, reducing irrigation 2 wk before the begin-
stress symptoms, such as increased frequency of dry
leaves, were visible.
ity of nitrogen is acquired by insects through absorp-
levels of FAAs under plant-stressed conditions can
increase insect herbivore populations (White 1984).
Three potential physiological mechanisms may ex-
plain the enhanced nutritional quality of plants under
and (D) southeast Louisiana (2% rice, 98% sugarcane).
August 2007REAY-JONES ET AL.: OVIPOSITION PREFERENCE IN E. loftini947
stress: (1) FAAs are nutritionally superior to proteins,
(2) FAAs are more readily available than proteins
because of the absence of any proteinase inhibitors,
and (3) FAAs are physically more accessible because
of increased solubility (CockÞeld 1988). Certain
amino acids are known to be essential for insect de-
velopment (Vanderzant 1958, Nation 2002). ArtiÞcial
diets with amino acid distributions simulating anthers
were adequate for survival and development of the
tobacco budworm, Heliothis virescens (F.) (Hedin et
al. 1991). Moths possess contact chemoreceptors on
in accepting or rejecting a host plant based on pres-
ence or absence of secondary or primary compounds
(Sta ¨dler 1984). FAAs can elicit electrophysiological
responses of the sensilla of the adult tobacco bud-
worm, the corn earworm, Heliothis armigera (Hu ¨b-
ner), and Spodoptera littoralis (Boisduval) (Blaney
and Simmonds 1988). Oviposition of the beet army-
worm was increased on cotton under drought stress,
which was correlated with greater levels of essential
FAAs (Showler and Moran 2003). Assuming that E.
loftini can detect host plant FAA levels and that such
levels inßuence oviposition preference, levels of es-
in egg laying.
Insects often oviposit on plants that maximize their
survival and development (Showler 2001). Nonpub-
lished greenhouse studies by M. Se ´tamou and A. T.
Showler, mentioned in Reay-Jones et al. (2003), have
sugarcane is enhanced within a certain range of
drought stress. Our study showed increased attrac-
A positive correlation may exist between preference
and performance on sugarcane. However, perfor-
of E. loftini on rice has not been studied. Reay-Jones
et al. (2005b) has shown that cultivar XL8, despite
more resistant to stem borers than Cocodrie. Poor
relationships between ovipositional preference and
performance can be explained by several hypotheses
(Thompson 1988). Further studies are necessary to
determine which hypothesis best explains the rela-
tionship between performance and preference of E.
viewed as a sequence of behavioral events consisting
ing and contact evaluation, and (3) acceptance or
rejection (Renwick and Chew 1994). Alighting on a
potential host plant is the result of integrating infor-
mation perceived by the moth, which includes visual,
olfactory, gustatory, and mechanical cues (Ra-
maswamy 1988). Contact chemoreception is the most
predominant sensory modality involved in host ac-
ceptance (Ramaswamy 1988). Host location and ac-
ceptance in oviposition preference studies are re-
each egg mass might reßect the mothÕs perception of
plants that are perceived as having low suitability.
Moths may assess host acceptability and host suitabil-
ity using different mechanisms, which likely involve
different host cues. Our analyses yielded associations
oviposition parameter estimates (Table 6), which
might reßect such behavioral steps.
On a plant selection basis, drought-stressed sugar-
cane cultivar LCP 85Ð384 (Þve nodes) was the most
attractive for oviposition based on egg masses laid.
Once a female began egg laying on sugarcane, its
2.3-fold greater numbers of eggs per egg mass when
contrasting with the number of eggs per egg mass
placed on rice. From a behavioral perspective, these
results suggest this species is able to regulate its egg
deployment strategy to account for the size of the
plant host and therefore the available sites for larval
feeding. Rice plants are much smaller than sugarcane
plants and large egg masses on rice would require
greater dispersal of larvae, thus exposing them to a
greater degree of mortality. The larger number of egg
masses and egg mass size on sugarcane indicates that
this plant is not only preferred for host location and
acceptance, but is also perceived as the most suitable
plant by E. loftini. In our study, oviposition on sugar-
cane was associated with arginine (egg masses per
plant) and aspartic acid (eggs laid per plant), which
both increased under stress. Reducing plant stress
with irrigation might assist in decreasing E. loftini
oviposition in sugarcane by decreasing both the nu-
tritional value of the crop for this insect and the num-
ber of ovipositional sites (i.e., dry leaves). Young sug-
old sugarcane (10 nodes), was more attractive for
levels of several FAAs essential for insect develop-
ment (alanine and valine). On rice, associations were
established between egg masses per plant and essen-
tial FAAs (threonine and valine) and dry leaves. Rice
cultivar XL8, which was more attractive for oviposi-
essential FAA histidine. The greater resistance of cul-
inßuence the resistance of rice to stem borers
also had more tillers, a plant trait that was positively
associated with egg masses per plant. E. loftini laid
more eggs on rice plants of large biomass, a common
response in oviposition behavior among other insects
(Asman 2002, Vasconcellosneto and Monteiro 1993).
Our study suggests that host plant foliar FFAs may
affect the oviposition preference of E. loftini. Plant
volatiles can have a major role in lepidopteran ovipo-
sition (Renwick and Chew 1994), but have not yet
been identiÞed for E. loftini. Foliage weight has been
correlated with the emission of volatiles from potato
plants, Solanum tuberosum L. (Agelopoulos et al.
1999). If the oviposition of E. loftini is greatly affected
Vol. 36, no. 4
might better quantify host plant selection than on a
per plant basis. In addition to the quantity of volatiles
by the quality of volatiles emitted, such as the ratio of
several compounds (Thompson and Pellmyr 1991).
Determining preference on a plant basis might there-
fore better quantify host choice in E. loftini if quality
of plant volatile emission is a more important factor
Simulated Oviposition Patterns. Our study indi-
cates that rice is more attractive than sugarcane on a
biomass basis. However, the projected potential ovi-
position patterns of E. loftini might vary greatly with
the biomass of the available host plants (Fig. 4). Sug-
arcane plants develop biomass more quickly in the
spring than rice and are expected to receive a greater
rice is the dominant crop (Fig. 4B), rice might over-
whelmingly receive the greatest proportion of eggs as
its biomass increases. As the proportion of sugarcane
increases (Fig. 4B and C), oviposition is expected to
increase on sugarcane. When sugarcane is the domi-
nant crop (98% of the production area; Fig. 4D), ovi-
(3.5% on 6/15) because of increased attractiveness of
the biomass of rice compared with sugarcane.
Our initial assumptions made to determine these
oviposition patterns imply a somewhat simpliÞed
agroecosystem. E. loftini can develop on numerous
plant species (Reay-Jones 2005), and the role of al-
ternate hosts in the population dynamics of the insect
E. loftini between crop hosts have also not been stud-
ied. Also, the growth model for sugarcane used in our
be different from Louisiana (i.e., a shorter growing
season in Louisiana than in Florida). To correct for
this, we assumed a similar relationship between the
two states for both sugarcane yield and dry weight.
Our predictions provide insight into how E. loftini
might distribute its populations between sugarcane
data from Þeld studies on the oviposition behavior of
E. loftini are also needed to verify the validity of the
results reported here from greenhouse studies.
Early-instar E. loftini larvae have limited mobility
and must feed on or very near the plant on which the
eggs are laid. Levels of antixenosis can help control
pests of crops in some integrated pest management
(IPM) systems (Smith 1989) and might assist in de-
cultivar, stress, and phenology and were associated
with oviposition preference estimates using both cor-
relation and regression analyses. Reducing drought
stress decreases both host plant suitability and attrac-
tiveness for oviposition. Because sugarcane is more
expected to contribute to enhancing infestations on
proximate sugarcane in some areas in Louisiana and
Texas. On rice, cultivar XL8 has been shown to be
more resistant to stem borers than Cocodrie, despite
being more attractive for oviposition (Reay-Jones et
al. 2005b). The use of this resistant rice cultivar as a
trap crop within the rice agroecosystem might be
effective in reducing infestations on proximate host
crops if the resistance mechanisms are antibiotic. Our
study has shown substantial differences in E. loftini
oviposition among host plants and the preference was
associated with several host plant characteristics. Un-
derstanding the population dynamics on both sugar-
cane and rice is necessary to conceptualize areawide
This work was supported in part by grants to Drs. Reagan
and Way from the USDA (CSREES) Southern Region IPM,
Crops-at-Risk IPM and IPM Enhancement Grants Programs,
the American Sugar Cane League, and the Texas Rice Re-
M. Garcia, J. Amador, J. Huerta, J. Rivera, J. daSilva, E.
Hernandez, and T. X. Liu at the Texas A&M University
System, Agricultural Research and Extension Center at
Weslaco, and M. Nun ˜ez, G. Wallace, and B. Pearson at the
Texas A&M University System, Agricultural Research and
Extension Center at Beaumont, and J. Cavazos (USDA-ARS
publication by the Director of the Louisiana Agricultural
Experiment Station as manuscript 06-26-0054.
Agelopoulos, N. G., K. Chamberlain, and J. A. Pickett. 1999.
of ratios. J. Chem. Ecol. 26: 497Ð511.
Asman,K. 2002. Trapcroppingeffectonovipositionbehav-
ior of the leek moth Acrolepiopsis assectella and the dia-
mondback moth Plutella xylostella. Entomol. Exp. Appl.
Aspinall, D., and L. G. Paleg. 1981. Proline accumulation:
physiological aspects, pp. 205Ð241. In L. G. Paleg and D.
Aspinall (eds.), The physiology and biochemistry of
drought resistance in plants. Academic, Australia.
Blaney, W. M., and M.S.J. Simmonds. 1988. Food selection
in adult and larvae of three species of Lepidoptera: a
Appl. 49: 111Ð121.
Brodbeck, B., and D. Strong. 1987. Amino acid nutrition of
In P. Barbosa and J. C. Schultz (eds.), Insect outbreaks.
Academic, London, UK.
Browning, H. W., M. O. Way and B. M. Drees. 1989. Man-
aging the Mexican rice borer in Texas.
Chaudhary, R. C., G. S. Khush, and E. A. Heinrichs. 1984.
Varietal resistance to rice stem-borers in Asia. Insect Sci.
Appl. 5: 447Ð463.
Chesson, M. 1978. Measuring preference in selective pre-
dation. Ecology 59: 211Ð215.
Coale, F. J., C. A. Sanchez, F. T. Izuno, and A. B. Bottcher.
1993. Nutrient accumulation and removal by sugarcane
grown on Everglades Histosols. Agron. J. 85: 310Ð315.
Cockfield, S. D. 1988. Relative availability of nitrogen in
host plants of invertebrate herbivores: three possible nu-
August 2007REAY-JONES ET AL.: OVIPOSITION PREFERENCE IN E. loftini 949
tritional and physiological deÞnitions. Oecologia (Berl.)
Feeny, P., L. Rosenberry, and M. Carter. 1983. Chemical
S. Ahmad (eds.), Herbivorous insects: host-seeking be-
havior and mechanisms. Academic, New York.
Flanders,S.E. 1930. Mexicansugarcaneborersandthepar-
asite Trichogramma. J. Econ. Entomol. 23: 603Ð606.
Gzik, A. 1996. Accumulation of proline and pattern of
alpha-amino acids in sugar beet plants in response to
osmotic, water, and salt stress. Environ. Exp. Bot. 36:
Hedin, P. A., W. L. Parrott, and J. W. Jenkins. 1991. Effects
Heinrichs, E. A. 1994. Host plant resistance, pp. 517Ð548. In
E. A. Heinrichs (ed.), Biology and management of rice
insects, Wiley, New York.
Heuer, B. 1994. Osmoregulatory role of proline in water-
and salt-stressed plants, pp. 363Ð381. In M. Pessarakli
(ed.), Plant and crop stress. Marcel Dekker, New York.
Jain, R., and A. K. Shrivastava. 1998. Biochemical studies
on healthy and chlorotic sugarcane leaves. Sugar Cane
Johnson, K.J.R. 1984. IdentiÞcation of Eoreuma loftini
ner for other sugarcane boring pest immigrants from
Mexico? Bull Entomol. Soc. Am. 30: 47Ð52.
Joshi, G. V., and G. R. Naik. 1980. Response of sugarcane to
different types of salt stress. Plant Soil 56: 255Ð263.
Legaspi, J. C., B. C. Legaspi, Jr., J. E. Irvine, J. Johnson, R. L.
Meagher, Jr., and N. Rozeff. 1999a. Stalkborer damage
on yield and quality of sugarcane in Lower Rio Grande
Valley of Texas. J. Econ. Entomol. 92: 228Ð234.
Legaspi, J. C., B. C. Legaspi, Jr., and R. R. Saldan ˜a. 1999b.
Laboratory and Þeld evaluations of biorational insecti-
dae) and a parasitoid (Hymenoptera: Braconidae). J.
Econ. Entomol. 92: 804Ð810.
Legendre,B.L.,andK.A.Gravois. 2005. The2004Louisiana
variety survey. Sugar Bull 83: 15Ð21.
Levinson, H., A. Levinson, and E. Osterried. 2003. Orange-
derived stimuli regulating oviposition in the Mediterra-
nean fruit ßy. J. Appl. Entomol. 127: 269Ð275.
Mackay, D. A., and R. E. Jones. 1989. Leaf shape and the
host-Þnding behaviour of two ovipositing monophagous
butterßy species. Ecol. Entomol. 14: 423Ð431.
Manly, B. F., P. Miller, and L. M. Cook. 1972. Analysis of a
selective predation experiment. Am. Nat. 106: 719Ð736.
Martinez, A. J., J. Baird, and T. Holler. 1988. Mass rearing
of parasites Allorhogas pyralophagus and Rhacanotus
McNeil, S., and T.R.E. Southwood. 1978. The role of nitro-
gen in the development of insect plant relationships, pp.
77Ð98. In Harborne, J. B. (eds.), Biochemical aspects of
plant and animal coevolution. Academic, London, UK.
Meagher, R. L., Jr., J. W. Smith, and K.J.R. Johnson. 1994.
Insecticidal management of Eoreuma loftini (Lepidop-
tera: Pyralidae) on Texas sugarcane: a critical review. J.
Econ. Entomol. 87: 1332Ð1344.
Meagher, R. L., Jr., J. E. Irvine, R. G. Breene, R. S. Pfannen-
stiel, and M. Gallo-Meagher. 1996. Resistance mecha-
nisms of sugarcane to Mexican rice borer (Lepidoptera:
Pyralidae). J. Econ. Entomol. 89: 536Ð543.
Saldan ˜a. 1998. Sugarcane stemborers and their parasites
in southern Texas. Environ. Entomol. 27: 759Ð766.
More ´,M.,E.V.Trumper,andM.J.Prola. 2003. Inßuenceof
corn, Zea mays, phenological stages on Diatraea saccha-
ralis F. (Lep.: Crambidae) oviposition. J. Appl. Entomol.
Muqing, R., and C. Ru-Kai. 1998. Osmotic adjustment in
leaves of sugarcane in response to water stress. Sugar
Cane 5: 3Ð7.
Murdoch, W. W. 1969. Switching in general predators: ex-
periments on predator speciÞcity and stability of prey
populations. Ecol. Monogr. 39: 335Ð354.
Murphy, B. C., R. V. Dowell, and L. T. Wilson. 1991. Quan-
tifying apple maggot (Diptera: Tephritidae) preference
for apples to optimize the distribution of traps among
trees. Environ. Entomol. 20: 981Ð987.
Myers, J. H. 1985. Effect of physiological condition of host
plant on the ovipositional choice of the cabbage white
butterßy, Pieris rapae. J. Anim. Ecol. 54: 193:204.
Nation, J. L. 2002. Insect physiology and biochemistry.
CRC, Boca Raton, FL.
1989. Measuring the host preference of parasites: an aid
in evaluating biotypes of Anagrus epos (Hym.: Mymari-
dae). Entomophaga 34: 551Ð558.
Rabe, E. 1994. Altered nitrogen metabolism under environ-
mental stress conditions, pp. 261Ð276. In M. Pessarakli
(ed.), Plant and crop stress. Marcel Dekker, New York.
Ramaswamy,S.B. 1988. HostÞndingbymoths:sensorymo-
dalities and behaviours. J. Insect Physiol. 34: 235Ð249.
Reay-Jones,F.P.F. 2005. Integratedpestmanagementofthe
rice. PhD dissertation, Louisiana State University, Baton
Reay-Jones, F.P.F., M. O. Way, M. Se ´tamou, B. L. Legendre,
and T. E. Reagan. 2003. Resistance to the Mexican rice
borer (Lepidoptera: Crambidae) among Louisiana and
Texas Sugarcane Cultivars. J. Econ. Entomol. 96: 1929Ð
Reay-Jones, F.P.F., T. E. Reagan, A. T. Showler, B. L.
Legendre, M. O. Way, and E. B. Moser. 2005a. Inte-
grated tactics for managing the Mexican rice borer
(Lepidoptera: Crambidae) in sugarcane. Environ. En-
tomol. 34: 1558Ð1565.
Reay-Jones, F.P.F., T. E. Reagan, M. O. Way, and B. L.
Legendre. 2005b. Conceptsofareawidemanagementof
Cane Int. 23: 20Ð24.
Reay-Jones, F.P.F., L. T. Wilson, M. O. Way, C. E. Carlton,
and T. E. Reagan. 2007. Movement of the Mexican rice
borer (Lepidoptera: Crambidae) through the Texas rice
belt. J. Econ. Entomol. 100: 54Ð60.
Renwick,J.A.A.,andF.S.Chew. 1994. Ovipositionbehavior
in Lepidoptera. Annu. Rev. Entomol. 39: 377Ð400.
Rhoades, D. F. 1983. Herbivore population dynamics and
plant chemistry, pp. 155Ð220. In R. F. Denno and M. S.
and managed systems. Academic, New York.
SAS Institute. 1999. UserÕs manual, version 8.0. SAS Institute,
SASInstitute. 2002. JMPuserÕsguide.Version5.SASInstitute,
Showler, A. T. 2001. Spodoptera exigua oviposition and
larval feeding preferences for pigweed, Amaranthus
hybridus, over squaring cotton, Gossypium hirsutum,
and a comparison of free amino acids in each host plant.
J. Chem. Ecol. 27: 2013Ð2028.
Vol. 36, no. 4
Showler, A. T. 2002. Effects of water deÞcit stress, shade, Download full-text
weed competition, and kaolin in particle Þlm on selected
hirsutum (L). J. Chem. Ecol. 28: 615Ð635.
Showler, A. T., and P. J. Moran. 2003. Effects of drought
stressed cotton, Gossypium hirsutum L., on beet army-
Showler,A.T. 2004a. Inßuenceofcottonfruitstagesasfood
sources on boll weevil (Coleoptera: Curculionidae) fe-
Showler, A. T. 2004b. Stress-induced host plant free amino
acids and insects, pp. 2139Ð2142. In J. L. Capinera (eds.),
Encyclopedia of entomology, vol. 3. Kluwer Academic,
Dordrecht, The Netherlands.
Showler, A. T., and J.R.C. Robinson. 2005. Proactive spray-
ing against boll weevils (Coleoptera: Curculionidae) re-
duces insecticide applications and increases cotton yield
and economic return. J. Econ. Entomol. 98: 1977Ð1983.
Singh, J., M. Singh, P. Chandra, G. P. Rao, and H. N. Singh.
1993. Biochemical studies on resistance to red rot in
sugarcane. Sugar Cane 1993: 16Ð19.
Smith,C.M. 1989. Plantresistancetoinsects,afundamental
approach. Willey Interscience, New York.
Sosa, O., Jr. 1988. Pubescence in sugarcane as a plant resis-
tance character affecting oviposition and mobility by the
sugarcane borer (Lepidoptera: Pyralidae). J. Econ. En-
tomol. 81: 663Ð667.
Sta ¨dler,E. 1984. Contactchemoreception,pp.3Ð35.InW.J.
Bell and R. T. Carde (eds.), Chemical ecology of insects.
Sinauer, Sunderland, MA.
Thompson,J.N. 1988. Evolutionaryecologyoftherelation-
phytophagous insects. Entomol. Exp. Appl. 47: 3Ð14.
Thompson, J. N., and O. Pellmyr. 1991. Evolution of ovipo-
sition behavior and host preference in Lepidoptera.
Annu. Rev. Entomol. 36: 65Ð89.
Tukey, J. W. 1953. The problem of multiple comparisons.
Department of Statistics, Princeton University, Prince-
Vanderzant, E. S. 1958. The amino acids requirements of
the pink bollworm. J. Econ. Entomol. 51: 309Ð311.
Van Leerdam, M. B., K.J.R. Johnson, and J. W. Smith, Jr.
1984. Effects of physical characteristics and orientation
on oviposition by Eoreuma loftini (Lepidoptera: Pyrali-
dae). Environ. Entomol. 13: 800Ð802.
Ovopositional sites of Eoreuma loftini (Lepidoptera: Pyrali-
dae) in sugarcane. Environ. Entomol. 15: 75Ð78.
Van Zwaluwenburg, R. H. 1926. Insects enemies of sugar-
cane in western Mexico. J. Econ. Entomol. 19: 664Ð669.
Vasconcellosneto, J., and R. F. Monteiro. 1993. Inspection
and evaluation of host plant by the butterßy Mechanitis-
lysimnia (nymph, Ithomiinae) before laying eggsÑa
mechanism to reduce intraspeciÞc competition. Oecolo-
gia (Berl.) 95: 431Ð438.
Vergara, B. S. 1991. Rice plant growth and development,
pp. 13Ð22. In B. S. Luh (ed.), Rice production, vol. I.
Van Nostrand Reinhold, New York.
White, T.C.R. 1984. The abundance of invertebrate herbi-
vores in relation to the availability of nitrogen in stressed
food plants. Oecologia (Berl.) 63: 90Ð103.
Wilson, L. T., and A. P. Gutierrez. 1980. Fruit predation
submodel: Heliothis larvae feeding upon cotton fruiting
structures. Hilgardia 48: 24Ð36.
Youm, O., H. W. Browning, and F. E. Gilstrap. 1988. Pop-
ulation dynamics of stalk borers attacking corn and sor-
ghum in the Texas Rio Grande Valley. Southwest. Ento-
mol. 13: 199Ð204.
Zar, J. H. 1999. Biostatistical analysis. Prentice Hall, Upper
Saddle River, NJ.
August 2007REAY-JONES ET AL.: OVIPOSITION PREFERENCE IN E. loftini 951