Infestation of Wheat and Downy Brome Grass by Wheat Stem Sawfly ...

Download PDF
0 downloads 0 Views 486KB Size Report
Comment
The level of infestation in downy brome grass was almost two-fold higher than for wheat throughout the growing season. Larval mortality was not detected in ...
COMMUNITY AND ECOSYSTEM ECOLOGY

Infestation of Wheat and Downy Brome Grass by Wheat Stem Sawfly and Subsequent Larval Performance JOEL PEREZ-MENDOZA, DAVID K. WEAVER,1

AND

WENDELL L. MORRILL

Montana State University, Department of Land Resources and Environmental Sciences, Bozeman, MT 59717

Environ. Entomol. 35(5): 1279Ð1285 (2006)

ABSTRACT Wheat stem sawßy, Cephus cinctus Norton (Hymenoptera: Cephidae), oviposition patterns and larval development in downy brome grass, Bromus tectorum L., and wheat, Triticum aestivum L., were compared in a commercially grown wheat Þeld infested with downy brome grass in Montana. Seven weekly randomly selected samples of stems for each plant species were collected at points where both plants were growing together. The level of infestation in downy brome grass was almost two-fold higher than for wheat throughout the growing season. Larval mortality was not detected in either plant species early in the season, but mortality was more frequent in downy brome grass as the host plants matured. Mortality of late-instar larvae late in the season was signiÞcantly higher in mature downy brome grass than in mature wheat stems. The weight of these late-instar larvae from wheat was almost four-fold heavier than larvae from downy brome grass. Stem height and seed weight in wheat were signiÞcantly reduced by larval sawßies feeding. In contrast, stem height, stem diameter, seed weight, and seed number in grass stems were signiÞcantly greater in infested downy brome grass stems compared with uninfested plants. Our results suggest that downy brome grass, a serious weed in wheat cropping, may play an important role in the survival and dynamics of wheat stem sawßy populations in the northern Great Plains. KEY WORDS wheat stem sawßy, downy brome grass, wheat, oviposition, larval development

The wheat stem sawßy, Cephus cinctus Norton (Hymenoptera: Cephidae), is the major pest of wheat, Triticum aestivum L., in the northern Great Plains (Montana, Wyoming, and North and South Dakota) of the United States and the Canadian Prairies (Manitoba, Saskatchewan, and Alberta) (Weiss and Morrill 1992). Annual economic losses caused by this insect have been estimated to exceed $25 million in Montana (Montana State University Extension Service 1996) and more than $ 100 million in the entire northern Great Plains (USDA-ARS 2005). Although larval feeding reduces the yield (10 Ð22% yield loss), test weight, and quality of wheat, main losses are caused by insectinduced stem lodging (Morrill et al. 1992). Mature larvae cut a ring around the inside of the stem near the soil, and plants frequently break and heads may be lost during harvest (Morrill et al. 1992). The study of the dynamic relationship between the populations of insects in uncultivated hosts and those in cultivated crops is an important aspect of sustainable integrated pest management (IPM), and a more complete understanding of the role of those uncultivated hosts in insect outbreaks could suggest new management strategies. C. cinctus is an insect that was 1 Corresponding author: Department of Land Resources and Environmental Sciences, Montana State University, 334 Leon Johnson Hall, Bozeman, MT 59717Ð3120 (e-mail: [email protected]).

originally found in large hollow-stemmed feral grasses in the western states of the United States and in Manitoba, Canada (Criddle 1917, Ainslie 1920). Various grass species in the genus Agropyron and Elymus seem to have been the original hosts (Ainslie 1920, Criddle 1922). Despite these early reports, the role that alternate grasses hosts have in the population dynamics of sawßies in wheat Þelds has not been addressed. These studies are particularly important for understanding population ßuctuations, especially because early records suggest that wheat stem sawßy may be dependent on wild grasses for its perpetuation (Criddle 1917). Recently, wheat stem sawßy larvae have been found feeding in large stems of downy brome grass, also known as “cheatgrass,” Bromus tectorum L., in several counties in Montana (W.L.M and D.K.W., unpublished data). B. tectorum is a winter annual grass species that was introduced into North America from Mediterranean Europe during the 19th century (Mack 1981). This grass species has developed into a severe weed in several agricultural production systems throughout North America, particularly on spring and winter wheat, and in rangeland. The ecology of downy brome grass is similar to wheat (Morrow and Stahlman 1984). Because downy brome grass occurs in most wheat Þelds in Montana (Trainor and Bussan 2002), it is important to evaluate the role of this grass as an

0046-225X/06/1279Ð1285$04.00/0 䉷 2006 Entomological Society of America

1280

ENVIRONMENTAL ENTOMOLOGY

alternate host in the life cycle and population dynamics of the wheat stem sawßy. Therefore, the objectives of this study were (1) to assess the oviposition and larval development of wheat stem sawßy in downy brome grass and spring wheat and (2) to examine the impact of sawßy larval feeding in both plant species. Materials and Methods Selection of Infested Field. This study was performed in a commercially grown wheat Þeld infested with downy brome grass that had a history of sawßy infestations. Traditional wheat production in the semiarid regions of Montana is alternate-year summer fallow (creating adjacent crop and idle Þelds in which the previous yearÕs crop was located). The monitored Þeld portion for this study was 3.5 ha (250 by 140 m) of a 40-ha Þeld (250 by 1,600 m). The Þeld was located near Amsterdam (Gallatin County), MT. ÔMcNealÕ spring wheat was planted on 7 April 2004 at the rate of 67.3 kg/ha (60 lb/ac) and was seeded directly into the fallow without spring tillage. The row spacing was 17.8 cm (7 in). This variety is the most common spring wheat variety currently grown by wheat producers in Montana and is a hollow stem variety that is very susceptible to sawßy infestation (Kushnak and Talbert 2004). Wheat was grown according to local agronomic practices, including the use of broadleaf and grassy weed herbicides, and fertilizer was applied at seeding at the rate of 60-40-7 actual units of nitrogen, phosphorus, and potassium, respectively. The downy brome grass was quite sparse and patchily distributed within the wheat Þeld, with a population density that was always ⬍10% of the available stems within the dense wheat crop, even in the areas of highest downy brome grass infestation. Assessment of Infestation and Early Larval Establishment. The rate of infestation of sawßies in wheat and downy brome grass was determined by estimating the percentage of infested stems with eggs and/or larvae. Sampling of stems began on 30 June 2004 at the beginning of the sawßy ßight period. Wheat was suitable for oviposition at Zadoks (Zadoks et al. 1974) stages ranging from 32 to 47 (⬃5Ð 6 wk after planting date). At these early growth stages, the wheat plants are very susceptible to oviposition by female sawßies. At this time, the majority of the downy brome grass stems had already ßowered. Stem sampling was repeated weekly for 6 wk, until the Þrst week of August, 1 wk before the wheat crop was harvested. Four unidirectional transects were established for sampling at right angles to the long axis of the Þeld crop. The Þrst transect was located at the northeast edge of the Þeld, and the remainder was located at intervals of 35 m. Wheat and downy brome grass stem samples were collected from an area of 3 m2 at Þve random points within each transect. These points contained both plant species. Sampling points were spaced at 50-m intervals along each transect, resulting in a total of 20 points within the wheat Þeld. Five plants for each plant species were randomly collected in each sampling point.

Vol. 35, no. 5

Plant samples were taken to a laboratory for examination. Because detection of C. cinctus eggs or larvae requires stem dissection, and individual plants consist of several stems, it is limiting to collect large number of samples each week. To overcome this problem, we decided to select two stems for dissection in each sampled plant in both plant species. Therefore, on each sampling date, 10 downy brome grass stems and 10 adjacent wheat stems were randomly collected in each of the 20 sampling points, to provide a 200-stem sample per plant species per sampling date. Stems were split lengthwise with a X-ACTO Knife (Hunt, Statesville, NC) and examined under magniÞcation for sawßy infestation; the number of infested stems and the number of eggs and larvae present in each stem was recorded. Sawßy eggs and larvae are easily distinguished inside the stems. Larval cadavers or cadaver fragments were also determined, where possible, at each sampling date. The amount of larval mortality in individual stems caused by cannibalism is impossible to measure in C. cinctus, so, in this study, we assessed larval survival by counting live larvae within the stem. Infestation was determined by the presence of live larvae within the stem or by the presence of characteristic frass, resulting from the feeding of the sawßy larvae, when no larval remains were found inside the stem. To characterize the phenological differences between wheat and downy brome grass during the sawßy oviposition period, we measured the height of the stem and the stem diameter at each internode, midway along each node from the bottom of the plant (Morrill et al. 1992), in both plant species at the second and third sampling date. Stem diameter to the nearest 0.01 mm was measured with a digital caliper (model MCD-Lite-S; Mitotoyo, Hiroshima, Japan). Fifty Þnal-instar larvae were collected from each plant species on 5 August (before sawßy cutting) and weighed using a microbalance (model AHAUS-Explorer; OHAUS, Florham Park, NJ). Larval Performance and Damage. To assess larval performance, we measured the proportion of larval establishment and survivorship for stems collected from each plant species at the end of the crop season. Because evidence of larval feeding is provided by characteristic frass accumulated within the stem, we assessed survival and establishment of the colonizing population in stems from both plant species. Larval establishment was deÞned by larval feeding within the stem, even if the larva died before reaching maturity; therefore, if a stem contained frass but no cadaver, the immature was recorded as having died before reaching the last larval instar. Larval survivorship was deÞned as the presence of live last-instar larvae within the stem at the time of sampling. To comparatively assess these parameters, mature downy brome grass and mature wheat plant stems were randomly selected and collected from the same transects in the crop a few days before harvest (18 August) by uprooting entire plants. Five downy brome grass and Þve adjacent wheat plants were collected at each of the 20 sampling points to provide 100 plants for each species. These plants provide a total of 542 and 384 stem samples of

October 2006

PEREZ-MENDOZA ET AL.: WHEAT STEM SAWFLY INFESTATION

1281

Table 1. Proportion of field-collected wheat and downy brome grass stems containing with wheat stem sawfly eggs and larvae (n ⴝ 200 stems for each plant species in each sampling date) Sampling date 6/30/04 7:7:04 7/14/04 7/22/04 7/27/04 8:5:04

a

Percentage of infested stems containing

Plant species

Proportion infested (mean ⫾ SEM)

Eggs

Larvae

Eggs ⫹ larvae

Downy brome Wheat Downy brome Wheat Downy brome Wheat Downy brome Wheat Downy brome Wheat Downy brome Wheat

0.625 ⫾ 0.072 0.220 ⫾ 0.061 0.575 ⫾ 0.074 0.250 ⫾ 0.067 0.675 ⫾ 0.074 0.390 ⫾ 0.087 0.665 ⫾ 0.078 0.340 ⫾ 0.094 0.755 ⫾ 0.050 0.355 ⫾ 0.091 0.765 ⫾ 0.060 0.380 ⫾ 0.080

87.2 79.6 35.7 30.0 5.9 10.3 0.0 0.0 0.0 0.0 0.0 0.0

2.4 13.6 34.8 51.7 69.1 59.0 100.0 100.0 100.0 100.0 100.0 100.0

10.4 6.8 29.5 18.3 25.0 30.7 0.0 0.0 0.0 0.0 0.0 0.0

Eggs/stema (mean ⫾ SEM)

Larvae/stema (mean ⫾ SEM)

2.14 ⫾ 0.15 1.69 ⫾ 0.20 2.13 ⫾ 0.19 1.69 ⫾ 0.23 1.21 ⫾ 0.08 1.75 ⫾ 0.18 0.0 ⫾ 0.0 0.0 ⫾ 0.0 0.0 ⫾ 0.0 0.0 ⫾ 0.0 0.0 ⫾ 0.0 0.0 ⫾ 0.0

1.00 ⫾ 0.00 1.00 ⫾ 0.00 1.30 ⫾ 0.07 1.12 ⫾ 0.05 1.46 ⫾ 0.06 1.10 ⫾ 0.04 1.30 ⫾ 0.05 1.04 ⫾ 0.03 1.16 ⫾ 0.03 1.10 ⫾ 0.04 1.15 ⫾ 0.03 1.04 ⫾ 0.02

In infested stems.

downy brome grass and wheat, respectively. The total number of stems that were cut by sawßy larvae was recorded for both plant species. The lower portions of the sawßyÐ cut stems were dissected lengthwise to conÞrm the presence of an overwintering larva. Uncut stems were also dissected lengthwise and examined for evidence of infestation. Infestation of uncut stems was determined by the presence of either a live larva or by the presence of characteristic frass if no larva was found. The total number of infested stems included both sawßy-cut and any infested stems that were not cut. To determine the effect of insect feeding in both plant species, heads were removed from the collected stems and individually threshed. The kernels were weighed to the nearest 0.01 g on an electronic balance (model SC2020; OHAUS), and the number of seeds was also recorded. We measured height and stem diameter of infested and uninfested mature wheat and downy brome grass stems. As for the growing plants, stem diameter was measured at midway between the two top nodes (Morrill et al. 1992) to the nearest 0.01 mm with a digital caliper. Loss estimates were made by comparing heights, head weights, and number of seeds produced by infested and uninfested stems within each plant species. Statistical Analysis. Variation in infestation rates, number of eggs and larvae, and larval survivorship in wheat and downy brome grass over the sampling dates was analyzed by using a two-way analysis of variance (ANOVA) with sampling date as a random factor (PROC GLM [RANDOM Statement]; SAS Institute 1998). ANOVA (PROC GLM; SAS Institute 1998) also was used to analyze the variation of weight in larvae and variation in the stem height and diameter for infested and uninfested young plants. Data for infestation rates and number of eggs and larvae were square root (x ⫹ 0.5) transformed to homogenize variances before analysis. Data for weight of larvae was square root transformed before analysis. Untransformed data are reported to simplify interpretation, but analysis results are based on the transformed data when a transformation was used.

Variation in the proportion of larval establishment, larval survivorship, and cut stems for mature wheat and downy brome grass, as well as variation in stem height and diameter, seed weight, and numbers of seeds produced for infested and uninfested plants of each plant species, were analyzed using ANOVA (PROC GLM; SAS Institute 1998). Relationships between the proportion of infested stems, cut stems, live larvae, dead larvae, stem height, stem diameter, threshed weight of seeds, and number of seeds within each plant species collected on the last sampling date were determined by linear correlation (PROC CORR; SAS Institute 1998) and by stepwise regression (PROC REG; SAS Institute 1998). We used the sequential Bonferroni technique (Rice 1989) to control the error rate for the 20 correlations at the 0.05 level. Results Infestation and Early Larval Establishment. Mean rate of infestation of downy brome grass and wheat by wheat stem sawßy varied with sampling date (F ⫽ 8.1; df ⫽ 5,5; P ⬍ 0.01) and plant species (F ⫽ 308.1; df ⫽ 1,5; P ⬍ 0.01). However, the interaction between sampling date and plant species was not signiÞcant (F ⫽ 1.1; df ⫽ 5,2388; P ⫽ 0.34). Downy brome grass was more heavily infested than wheat. The proportion of infested stems varied from 0.575 to 0.765 in downy brome grass and from 0.22 to 0.395 in wheat (Table 1). The mean number of eggs collected during the Þrst three sampling dates did not differ among sampling dates (F ⫽ 1.2; df ⫽ 2,2; P ⫽ 0.46) or with plant species (F ⫽ 3.1; df ⫽ 1,2; P ⫽ 0.22), but the sampling date ⫻ plant species interaction was signiÞcant (F ⫽ 25.8; df ⫽ 2,1194; P ⬍ 0.01). We looked more closely at the sampling date ⫻ plant species interaction, and there was an obvious decrease in the mean number of eggs collected for both plant species as the crop season advances, and this explains the interaction (Table 1). During the Þrst three sampling dates, ⬎40% of the infested stems of both plants species contained eggs only (Table 1), and of those, ⬎60% were infested with a single egg (Fig. 1). The number of eggs per individ-

1282

ENVIRONMENTAL ENTOMOLOGY

Fig. 1. Frequency distribution of C. cinctus eggs in infested downy brome grass stems and infested wheat stems from a wheat Þeld near Amsterdam, MT, for the initial three sampling dates (30 June to 14 July 2004).

ual stem ranged from one to eight in downy brome grass and from one to six in wheat. The mean number of larvae collected from all stems over the growing season differed with sampling date (F ⫽ 6.2; df ⫽ 5,5; P ⫽ 0.03) and with plant species (F ⫽ 22.2; df ⫽ 1,5; P ⬍ 0.01). The interaction of sampling date ⫻ plant species was also signiÞcant. Infestation by sawßy larvae increased in both plant species as the growing season advanced, and for the infested stems, ⬎80% contained only one larva (Fig. 2). During the Þrst three sampling dates, no dead larvae were found in the infested stems of either plant species. After the fourth sampling date, we started to Þnd infested stems with a clear indication of sawßy infestation (presence of frass), but without live larvae. The proportion of infested stems without live larvae increased as the plants matured (Fig. 3). Weight of mature larvae varied with plant species (F ⫽ 243.3; df ⫽ 1,99; P ⬍ 0.01). Larvae developing on wheat were almost four-fold heavier than larvae developing in downy brome grass (14.9 ⫾ 0.99 versus 4.2 ⫾ 0.30 mg, respectively).

Fig. 2. Frequency distribution of C. cinctus larvae in infested downy brome grass stems and infested wheat stems from a wheat Þeld near Amsterdam, MT, for all sampling dates (30 June to 5 August 2004).

Vol. 35, no. 5

Fig. 3. Proportion survivorship for C. cinctus larvae in infested downy brome grass stems and infested wheat stems from a wheat Þeld near Amsterdam, MT, over the crop season.

Diameter of young stems (during sawßy female oviposition period) varied with plant species (F ⫽ 832.8; df ⫽ 1,185; P ⬍ 0.01) and with internode (F ⫽ 10.1; df ⫽ 4,185; P ⬍ 0.01). The interaction of plant species ⫻ internode was also signiÞcant (F ⫽ 9.5; df ⫽ 4,185; P ⬍ 0.01). Both plants species had Þve internodes between the root and the inßorescence in the stems collected, and the mean diameter of wheat internodes was almost two-fold wider than that of downy brome grass internodes (3.53 ⫾ 0.06 and 1.87 ⫾ 0.04 mm, respectively; Fig. 4). The mean height of the stems did not vary with plant species (F ⫽ 0.28; df ⫽ 1,40; P ⫽ 0.60). Larval Performance and Damage. The mean proportion of larval establishment varied between plant species in preharvest samples collected on 18 August (F ⫽ 29.5; df ⫽ 1,39; P ⬍ 0.01). Larval establishment was greater in downy brome grass stems (0.669 ⫾ 0.05) than in wheat stems (0.269 ⫾ 0.05). Only one live larva was found per infested stem in both plant species at this time. The mean proportion of larval survivorship in infested stems also varied signiÞcantly with plant species (F ⫽ 27.6; df ⫽ 1,39, P ⬍ 0.01). Larval survivorship was signiÞcantly higher in wheat stems

Fig. 4. Differences in internode diameters between young downy brome grass stems and wheat stems from a wheat Þeld near Amsterdam, MT, in 2004.

October 2006

PEREZ-MENDOZA ET AL.: WHEAT STEM SAWFLY INFESTATION

1283

Table 2. Comparison of height, diameter, weight of seeds, and no. of seeds produced by infested and uninfested wheat and downy brome grass stems (mean ⴞ SEM) Plant species Agronomic parameters

Wheat

Height (cm) Diameter (mm) Weight (g) No. of seeds

Downy brome grass

Infested

Uninfested

Infested

Uninfested

703.9 ⫾ 28.4 3.35 ⫾ 0.09 0.938 ⫾ 0.063 28.52 ⫾ 1.33

782.0 ⫾ 7.9 3.43 ⫾ 0.03 1.132 ⫾ 0.021 30.61 ⫾ 0.50

715.3 ⫾ 11.4 2.43 ⫾ 0.03 0.208 ⫾ 0.009 66.76 ⫾ 2.83

579.4 ⫾ 21.2 2.00 ⫾ 0.05 0.148 ⫾ 0.011 48.90 ⫾ 2.98

(0.836 ⫾ 0.06) than in downy brome grass stems (0.391 ⫾ 0.06). The mean proportion of infested plant stems cut by last-instar larvae also varied signiÞcantly with plant species (F ⫽ 23.3; df ⫽ 1,39; P ⬍ 0.01). Proportion of wheat stems cut by larvae (0.724 ⫾ 0.06) was almost 2.5-fold higher than the proportion of downy brome grass stems cut (0.296 ⫾ 0.06). Mean height of mature wheat stems (F ⫽ 11.1; df ⫽ 1,213; P ⬍ 0.01) and the mean weight of the seeds produced (F ⫽ 10.4; df ⫽ 1,213; P ⬍ 0.01) differed between infested and uninfested plants, with the values for infested plants being lower (Table 2). However, the mean diameter of internode (F ⫽ 0.62; df ⫽ 1,213; P ⫽ 0.43) and the mean number of seeds produced by wheat plants was not different for infested and uninfested plants (F ⫽ 2.2; df ⫽ 1,213; P ⫽ 0.14; Table 2). Infestation of downy brome grass by sawßy had a signiÞcant effect on mean mature stem height (F ⫽ 37.7; df ⫽ 1,334; P ⬍ 0.01), mean internode diameter (F ⫽ 55.5; df ⫽ 1,334; P ⬍ 0.01), mean seed weight (F ⫽ 14.3; df ⫽ 1,334; P ⬍ 0.01), and mean seed number (F ⫽ 14.7; df ⫽ 1,334; P ⬍ 0.01). However, contrary to the data for wheat, those parameters tended to be higher on infested downy brome stems compared with uninfested stems (Table 2). Stepwise regression showed that only one factor at each time was signiÞcant for predicting the relationships between stem heights, diameter of internode, weight of seeds, and number of seeds produced versus the proportion of sawßy infested stems, presence of live larvae, cut plants, and dead larvae in both plant species. Thus, we used correlation analysis to explore these relationships. Correlation analyses gave different results within each plant species. In wheat, only the proportion of sawßy infested stems was negatively correlated with height of the stems and with weight of seeds produced (Table 3), whereas in downy brome grass, the proportion of infested stems was positively correlated with stem height, diameter, weight, and number of seeds (Table 3).

Discussion Over time, the mean percentage of infestation of downy brome grass was more than two-fold higher than the mean percentage of infestation for wheat. These results indicate a signiÞcant use of downy brome grass by C. cinctus as an alternate host. It is interesting to speculate that this may be an established preference of C. cinctus for downy brome grass, but the lack of any historical data conÞrming this theory does not preclude that the use of downy brome grass may have begun only within the past few years. However, downy brome grass may play an important role on the perpetuation and population dynamics of this insect. This is feasible if we consider that, in the northern Great Plains, downy brome grass is abundant, particularly in wheat Þelds (Morrow and Stahlman 1984). However, the abundance of this weed does not even begin to approach that of cultivated cereals. The importance of uncultivated grasses on sawßy populations was Þrst noted by Criddle (1917), who suggested that sawßy perpetuation is dependent on certain grass species. Three years later, Criddle (1922) reported that the Canada wildrye, Elymus canadensis (C. L. Hitchcock) Jensen, and smooth brome grass, Bromus inermis Leys, were severely attacked by wheat stem sawßy, but only E. canadensis produced large numbers of adults. More recently, Youtie and Johnson (1988) reported that plants of the Great basin wildrye, Elymus cinereus Scribner and Merrill, collected at the Idaho National Engineering Laboratory and Craters of the Moon National Monument showed high levels of wheat stem sawßy infestation. Sawßy females oviposited more in the comparatively rare downy brome grass stems, even though they could oviposit in adjacent wheat plants that were at the physiological stage in which they are very susceptible and attractive to sawßy females. For the Þrst three sampling dates, ⬇62% of the infested stems collected were infested with sawßy eggs across plant

Table 3. Relationship between the proportions of sawfly-infested stems and certain agronomic parameters (height, diameter, weight, and no. of seeds) for wheat and downy brome grass Wheat

Downy brome grass

Agronomic parameters

n

CoefÞcient

P

n

CoefÞcient

P

Height of stems Diameter of stems Weight of seeds Number of seeds

213 213 213 213

⫺0.224 ⫺0.054 ⫺0.217 ⫺0.102

0.001 0.433 0.001 0.139

334 334 334 334

0.319 0.379 0.203 0.256

⬍0.001 ⬍0.001 0.001 0.001

1284

ENVIRONMENTAL ENTOMOLOGY

species. As expected, the presence of eggs decreased and the presence of larvae increased as the plants grew, particularly after oviposition ceased at the end of the ßight period of the adults. There is an important comparison to be made between plant species, however. Approximately 40% of the infested downy brome grass stems contained multiple eggs, and ⬇25% contained multiple larvae early in the growing season compared with only 36% for eggs and 8% for larvae in the infested wheat stems. However, at the end of the growing season, only one larva survived in mature stems in both plant species. This shows that the sawßy population in downy brome grass experiences much higher mortality compared with the sawßy population in wheat because of cannibalism, perhaps as a direct result of the selective distribution of eggs. Holmes (1978) reported that the larvae normally move upward through the stem, destroying other larva and eggs as they go, until only one individual survives in each stem. Furthermore, in our study, the number of infested stems containing live larvae was very similar for both plant species over the Þrst Þve sampling dates. However, at the end of the growing season, the numbers of live larvae in mature wheat stems was more than two-fold higher than the numbers of live larvae infesting mature downy brome grass stems. In addition, the weight of mature larvae collected in wheat was also more than four-fold higher than the weight of mature larvae collecting in downy brome grass. These results suggest that downy brome grass is not a highly suitable host and may not provide adequate conditions for larvae to complete development very often. However, because downy brome grass is commonly found in most wheat areas in the northern Great Plans (Morrow and Stahlman 1984), it is possible that this species may be a signiÞcant alternate host for sawßies in these areas or that it may negatively impact population dynamics because of lower immature survival. More detailed studies are required to determine the role of this plant species on wheat stem sawßy population maintenance. In comparison to wheat, downy brome grass stems infested with wheat stem sawßy larvae tended to be larger, to have a greater internode diameter of internode, and to produce more and heavier seeds compared with uninfested stems. However, ⬇75% of the uninfested downy brome grass stems had smaller diameters compared with infested stems (Fig. 5), suggesting that sawßy females oviposited more frequently in the more robust stems of the downy brome grass. Morrill et al. (2000) also noted that stem diameters of most grass species are smaller than those of wheat in the area where wheat stem sawßy causes economic damage, especially during periods of drought stress when the stems of the uncultivated grasses may be too small to support larval development. Holmes and Peterson (1960) also reported that sawßy females tended to select wheat stems with larger diameter for oviposition, and Morrill et al. (2000) found that female wheat stem sawßies that emerged from larger wheat stems were heavier, lived longer, and produced more

Vol. 35, no. 5

Fig. 5. Proportion infestation in plants as a function of stem diameter for both wheat stems and downy brome grass stems from a wheat Þeld near Amsterdam, MT, in 2004.

eggs compared with females that emerged from smaller wheat stems. In summary, the greater infestation of downy brome grass by wheat stem sawßy in comparison with wheat suggests that female sawßies prefer to oviposit in downy brome grass. However, more detailed studies on sawßy preferences under controlled environments are required to fully assess this. This research will be undertaken to provide a better understanding of this phenomenon. Additional Þeld studies, over multiple years, will help to determine the role of downy brome grass in wheat stem sawßy population dynamics and maintenance. Acknowledgments We thank the many members of the stem dissecting team for technical support during the Þeld collection of stems and J. E. Baker, J. E. Throne, and T. Macedo for comments and suggestions on an earlier version of the manuscript. This research was supported by funds from a USDA CSREES Special Research grant entitled, “Novel semiochemical- and pathogen-based management strategies for wheat stem sawßy.” Additional support was provided by the Montana Wheat and Barley Committee and by the Montana Ag Experiment Station.

October 2006

PEREZ-MENDOZA ET AL.: WHEAT STEM SAWFLY INFESTATION References Cited

Ainslie, C. N. 1920. The western grass-stem sawßy. U.S. Dep. Agric. Bull. 841. Criddle, N. 1917. Further observations upon the habits of the western wheat stem sawßy in Manitoba and Saskatchewan. Agric. Gaz. Can. 3: 176Ð177. Criddle, N. 1922. The western wheat-stem sawßy and its control. Canadian Department of Agriculture, Ottawa, Canada. Holmes, N. D. 1979. The wheat stem sawßy, p. 13. In Proceedings of the Entomological Society of Alberta. 26: 2Ð13. Holmes, N. D., and L. K. Peterson. 1960. The inßuence of the host on oviposition by the wheat stem sawßy, Cephus cinctus Nort. (Hymenoptera: Cephidae). Can. J. Plant Sci. 40: 29Ð46. Kushnak, G. D., and L. Talbert. 2004. Spring wheat variety notes & comments. Western Triangle Agricultural Research Center, Montana Agricultural Experiment Station, Conrad, MT. Mack, R. N. 1981. Invasions of Bromus tectorum L. into western North America: an ecological chronicle. Agro-Ecosystems 7: 145Ð165. Montana State University Extension Service. 1996. Montana crop health report. Special edition: wheat stem sawßy. Crop. Health Rep. 10: 1Ð7. Morrill, W. L., J. W. Gabor, and G. D. Kushnak. 1992. Wheat stem sawßy (Hymenoptera: Cephidae): damage and detection. J. Econ. Entomol. 85: 2413Ð2417.

1285

Morrill, W. L., J. A. Gabor, D. K. Weaver, G. D. Kushnak, and N. J. Irish. 2000. Effect of host plant quality on the sex ratio and Þtness of female wheat stem sawßies (Hymenoptera: Cephidae). Environ. Entomol. 29: 195Ð199. Morrow, L. A., and P. W. Stahlman. 1984. The history and distribution of downy brome (Bromus tectorum) in North America. Weed Sci. 32(suppl 1): 7Ð12. Rice, W. R. 1989. Analyzing tables of statistical tests. Evolution 43: 223Ð225. SAS Institute. 1998. The SAS System version 7 for Windows. SAS Institute, Cary, NC. Trainor, M., and A. J. Bussan. 2002. Downy brome. http:// weeds.montana.edu/crop/brome.htm. USDA–ARS NPARL. 2005. Wheat stem sawßy biological control. Northern Plains Agricultural Research Laboratory, Sidney, MT. Weiss, M. J., and W. L. Morrill. 1992. Wheat stem sawßy (Hymenoptera: Cephidae) revisited. Am. Entomol. 38: 241Ð245. Youtie, B. A., and J. B. Johnson. 1988. Association of the wheat stem sawßy with basin wildrye. J. Range. Manag. 41: 328 Ð331. Zadoks, J. C., T. T. Chang, and C. F. Konsak. 1974. A decimal code for growth stages of cereals. Weed Res. 14: 15Ð21.

Received for publication 18 May 2005; accepted 26 May 2006.