Final Report on the Ecological Impact of Bt Corn Pollen on the Monarch Butterfly in Ontario

October 3, 2001

---

M.K. Sears, D.E. Stanley-Horn, H.R. Mattila
Department of Environmental Biology
University of Guelph
Ontario N1G 2W1

Tel: 519-824-4120 ext. 3921
msears@evbhort.uoguelph.ca

---

Table of Contents

• Executive Summary
• Section 1: Background
  • Introduction
  • Objectives of Ontario Studies
• Section 2: Likelihood of Exposure of monarch larvae to Bt corn pollen
  • Spatial and temporal overlap between corn anthesis and monarch larvae: Phenology experiment
    • Introduction
    • Methods
    • Results
    • Conclusions
  • Milkweed population survey
    • Introduction
    • Methods
    • Results
    • Conclusions
  • Pollen densities in and dispersal from the field
    • Introduction
    • Methods
    • Results
    • Conclusions
• Section 3: Toxicity of Bt corn pollen to monarch larvae
  • Laboratory bioassays
    • Introduction
    • Methods
    • Results
    • Conclusions
  • Field bioassays
    • Introduction
    • Methods
    • Results
    • Conclusions
• Section 4: Risk Assessment
  • Introduction: Hazard identification
  • Methods
  • Results
• References Cited

---

Executive Summary

A significant concern regarding non-target effects of transgenic corn crops that contain genes transferred from the organism Bacillus thuringiensis (Bt) was raised following the publication of a note to the editor of the journal Nature in 1999. The Canadian Food Inspection Agency and the U.S. Environmental Protection Agency had reviewed data submitted by the registrants, which included effects of the Bt proteins on a series of non-target organisms chosen to demonstrate potential deleterious effects on non-target invertebrates and vertebrates with no discernible impacts noted. Their review concluded that the potential impact of corn pollen, which contains variable amounts of Bt protein, on sensitive larvae of lepidoptera was considered negligible due to the short distances that corn pollen dispersed from fields and the relatively low expression of Bt protein in pollen grains. Considerable media attention was focused on this issue and pronouncements of severe debilitation of monarch butterfly populations were common in the press.

In response, industry representatives and the U.S. Department of Agriculture, Agricultural Research Service initiated a collaborative research effort to specifically address the question of risk associated with Bt corn pollen on populations of the Monarch butterfly. A consortium of industry and researchers was established in June 1999 to identify and investigate the potential exposure and ecological effects of Bt corn pollen on larvae of the monarch. Our investigations in collaboration with this group began at this time with financial support form Environment Canada and the Canadian Food Inspection Agency. After the growing season, a meeting was held in November 1999 in Chicago of the participants in these and related studies to present their findings to interested parties. Following this initial review a planning meeting was held in February 2000 in Kansas City to establish specific goals and objectives for research by this collaborative group during the coming growing season in 2000. It was at this meeting that the elements of risk assessment as outlined by the U.S. Environmental Protection Agency were considered with respect to the issue of impact of Bt corn pollen on populations of the monarch butterfly. Our research objectives were integrated with those of the other research groups operating in the U.S.

Objectives of Ontario Studies

• To determine the spatial and temporal overlap of susceptible stages of monarchs with corn anthesis.
• To determine the spatial arrangement of milkweed in relation to cornfields.
• To estimate the relative importance of corn and non-agricultural habitats for monarch production.
• To determine the extent and degree to which pollen deposits on milkweed plants in and near cornfields.
• To determine the extent and degree to which pollen disperses from cornfields.
• To determine lethal and sublethal doses of Bt 176 and Bt-11corn pollen to monarch larvae in laboratory bioassays.
• To examine the impact of naturally deposited Bt and non-Bt pollen on monarchs exposed on milkweed plants in the field.

Likelihood of Exposure of Monarch larvae to Bt corn pollen

Monarchs use cornfields as a breeding habitat to the same extent as other crop lands and non-agricultural landscapes and are present as caterpillars in cornfields to some degree during corn anthesis. Larvae of second-generation monarchs in Ontario are found on milkweed leaves at approximately the same time that pollen shed occurs in cornfields. In Ontario we estimate this overlap of larval feeding with the period of pollen shed to include about 40-60% of the population. Across the Corn Belt, this estimate is about 25% of the population. Because milkweed densities are much lower in cornfields, the relative contribution of cornfields to monarch populations is much lower than that of non-agricultural areas. However, cornfields represent about 21% of the land cover in SW Ontario, whereas the percentage of land cover that is non-agricultural monarch habitat is estimated to be 16%. Across the entire Corn Belt, we estimate that about 20% of the landscape consists of cornfields. Finally, Bt corn comprises 19% of the corn grown in North America in 2001 and 2000, although this figure is higher for Ontario.

The likelihood that monarch larvae are exposed to Bt corn pollen is a function of the area in their breeding habitat that is composed of corn and that fraction of corn that is Bt corn. Also, that portion of the monarch population exposed to corn pollen as larvae during pollen shed is critical to this estimate. For example, if 1) the breeding range of monarchs contains 20% corn, 2) 20% of that corn is Bt corn and 3) 25% of monarch larvae are exposed, the likelihood of exposure to Bt corn pollen is 1% of the population.

Toxicity of Bt corn pollen to monarch larvae

Although pollen does not disperse very far from the field (eg. 90% of the pollen falls within about 5m of the field edge), monarchs utilise milkweed as food plants within the cornfield where they may be exposed to pollen densities of several hundred grains/cm² during the period of pollen shed. For example, in Ontario we found pollen densities of 70 and 98 pollen grains/cm² on milkweed leaves 6 days and 11 days after the commencement of anthesis, respectively. Data from our collaborative research across the Corn Belt indicates that the average deposition of corn pollen is about 170 pollen grains/cm².

Toxicity of Bt pollen to monarch larvae depends on the amount of Bt toxin expressed in the pollen and the amount of pollen to which larvae are exposed. Pollen from hybrids containing the event 176 transformation poses a risk to monarch larvae. Exposure to pollen levels of as little as 10 pollen grains/cm², which are commonly observed in cornfields, are likely to cause sublethal effects such as reduced weight gain and delayed development, while exposure to 50-100 pollen grains/cm² may cause lethal effects. However, event 176 derived hybrids have never comprised more than 2% of the Bt corn plantings in North America. Bt11 and Mon810 derived hybrids, which comprise 99% of the corn grown in North America, do not pose a significant risk to monarchs in the field because of their relatively low level of protein expression in pollen compared with hybrids from event 176. In our laboratory studies doses as high as 4000 grains/cm² on milkweed leaves did not cause lethal or sublethal effects. In our combined studies, a threshold of exposure was established at 1000 pollen grains/cm². The likelihood that a larva would encounter this density of pollen on milkweed leaves is less than 1%, therefore the likelihood of intoxication is also less than 1%.

Risk Analysis

In addressing risk posed by Bt corn pollen, two factors must be considered: 1) the likelihood of exposure of larval stages to Bt corn pollen and 2) the dose of pollen at which a measurable effect can be determined to larvae or subsequent stages of the monarch. Our data indicate that the likelihood of exposure is 1% while the likelihood of encountering a toxic dose is less than 1%. Therefore, risk to monarchs from Bt corn pollen is less than 1/100 of 1%. Event 176 Bt corn will not be re-submitted for registration and supplies will be phased out by 2003. Other events, such as Bt 11 and MON 810, and recently registered events express such small amounts of Cry protein in their pollen that toxicity cannot be recorded at nominal doses. Effects of chronic exposure of larvae to pollen were not considered here and may require further study to determine whether any impact to the population would likely result. Our studies of the deposition and effects of Bt pollen suggest that risk of deleterious effects from consumption by monarch larvae on milkweed leaves is very unlikely, if not negligible.

Resulting from these studies is a series of peer-reviewed publications in the Proceedings of the National Academy of Sciences (US), in which details of the collaborative research is presented. These publications can be found at the journal's website at http://www.pnas.org under the banner Special Edition for September 14, 2001. The data are organized and presented in four papers which include, milkweed distribution and utilization by monarchs, pollen deposition and distribution, laboratory assessment of toxicity of pollen and fields trials to determine impact of pollen on monarch larvae. The fifth paper describes a detailed risk assessment utilizing data presented in the preceding papers.

---

Section 1: Background

Introduction

A significant concern regarding non-target effects of transgenic Bt corn crops containing transgenes from the organism Bacillus thuringiensis, was raised following the publication of a note to the editor of the journal Nature by Losey, et al. (1999). The Canadian Food Inspection Agency and the U.S. Environmental Protection Agency had reviewed data submitted by the registrants that included effects of the Bt proteins on a series of non-target organisms chosen to demonstrate potential deleterious effects on non-target invertebrates and vertebrates with no discernible impacts noted (U.S. EPA 1999). Their review concluded that the potential impact of corn pollen, which contains variable amounts of Bt protein, on sensitive larvae of lepidoptera was considered negligible due to the short distances that corn pollen dispersed from fields and the relatively low expression of Bacillus thuringiensisU.S. EPA 1995).

Bt corn has been transformed with an insecticidal gene from a common soil bacterium Bacillus thuringiensis (Bt) that encodes the Cry1A(b) endotoxin (Koziel et al. 1993). Toxicity of the Cry1A(b) toxin is limited to certain lepidopteran species, including pests that feed on corn tissue, thus impact on non-target organisms has been considered negligible (Ostlie Bt 1997, Orr and Landis 1997, Pilcher Bt 1997, Schuler Bt 1999, but see Hilbeck Bt, 1998). Most commercial Bt corn hybrids express toxin in their pollen (Fearing et al. 1997). Corn pollen is wind-dispersed at least 60 meters (Raynor Bt 1972) where it may be deposited and consumed on host plants of non-target species that feed in and near cornfields.

In particular, Bt corn pollen may present a hazard to monarch butterflies, Danaus plexippus (Losey Bt 1999, Hansen and Obrycki 2000) because monarch larvae feed exclusively on milkweed species (Malcolm Bt 1993), primarily the common milkweed, Asclepias syriaca (Malcolm et al. 1989). Milkweed is present in and near cornfields in north eastern United States and south eastern Canada. The note published in Nature (Losey et al. 1999), indicated that exposure to pollen from transgenic corn plants expressing a Bacillus thuringiensis endotoxin from a Bt-11 hybrid (N4640-Bt corn) resulted in increased mortality and delayed development compared with ingestion of non-Bt pollen. Decreased consumption was observed following ingestion of either pollen type compared with control larvae that were fed only leaves. Because amounts of pollen used in the aforementioned study were not reported, and because exposure of non-target lepidopteran species to toxic doses of Bt pollen was not addressed, the risk of Bt-corn pollen to monarch populations could not be determined.

The breeding range of the monarch butterfly coincides to a large degree with that of the major corn-growing regions of North America. About half of the monarch population that migrates to Mexico during winter originates from the Corn Belt in the Midwestern United States (Wassenaar and Hobson, 1998) and cornfields represent approximately 18.9% of the suitable monarch habitat in the central U.S. (Taylor and Shields, 2000). Bt corn represents approximately 25% of the corn acreage in Canada and the United States, thus a portion of the migratory monarch population is exposed to Bt pollen.

Although a hazard had been identified, an assessment of the risk to the monarch butterfly requires a detailed analysis of the doses of pollen that are toxic to monarch larvae and the likelihood that monarchs are exposed to toxic doses of pollen. The research reported here is part of a larger, collaborative effort with researchers from several States to better understand the risks of Bt corn pollen to the monarch butterfly, and addresses several aspects of the toxicity of pollen and the potential for exposure to toxic doses of pollen.

Objectives of Ontario Studies

• To determine the spatial and temporal overlap of susceptible stages of monarchs with corn anthesis.
• To determine the spatial arrangement of milkweed in relation to cornfields.
• To estimate the relative importance of corn and non-agricultural habitats for monarch production.
• To determine the extent and degree to which pollen deposits on milkweed plants in and near cornfields.
• To determine the extent and degree to which pollen disperses from cornfields.
• To determine lethal and sublethal doses of Bt 176 and Bt-11 corn pollen to monarch larvae in laboratory bioassays.
• To examine the impact of Bt and non-Bt pollen on monarch larvae exposed to natural deposits on milkweed plants in the field.

---

Section 2: Likelihood of Exposure of monarch larvae to Bt corn pollen

Spatial and Temporal Overlap Between Corn Anthesis and Monarch Larvae: Phenology Experiment

Introduction

The degree of exposure of monarch larvae to toxic doses of pollen depends, in part, on the likelihood that monarch larvae are present in and near cornfields (i.e. a spatial overlap between monarch larvae and corn). At the time of anthesis, corn plants are substantially taller than milkweed plants, and some research suggests that females are less likely to lay eggs on milkweed plants surrounded by higher vegetation (Prysby and Oberhauser personal communication). In addition, in a greenhouse study by Losey (2000), females laid fewer eggs on milkweed plants surrounded with corn. Thus, the degree of spatial overlap may be limited by female behavior. Also, the likelihood that monarch larvae are present in and near cornfields when corn pollen is abundant must be determined (i.e. a temporal overlap between the presence of monarch larvae and corn anthesis). Recent modeling efforts (Calvin and Taylor 2000 personal communication) and monarch larval monitoring data (Prysby and Oberhauser www.monarchlab.umn.edu) suggest that there is temporal overlap between corn anthesis and susceptible monarchs in some parts of the US, however, the overlap in Ontario is unknown. Finally, the significance of an overlap between susceptible monarch stages and corn anthesis will depend on the proportion of the monarch population that is affected, which can be estimated to some extent by the relative numbers in different habitats and the proportion of land covered by those habitats.

In the summer of 2000, a phenological survey was conducted to determine the relative usage and preference of monarchs in non-Bt cornfields versus non-agricultural areas and the temporal overlap of monarchs with corn anthesis.

Methods

The study was conducted over a six-week period from 10 July to 18 August, in and around commercial cornfields within Waterloo and Wellington counties, ON. The intent was to bracket the period of pollen shed by two weeks before and after. However, because the period of pollen shed was both late and variable this year, anthesis occurred during the last two weeks of the study in most fields. Five locations were monitored, and each location included two habitats: a cornfield, and a non-crop habitat (milkweed growing outside of a crop field and more than 100 meters from a cornfield). The two habitats were monitored weekly throughout the six-week period at each location. Because our goal was to measure importance of cornfields to monarchs, and not effects of Bt pollen (which are better measured with manipulative experiments), corn habitat at each location consisted of non-transgenic corn. The most important site selection criterion was naturally-growing milkweed at densities of approximately 10 or more plants (ramets) per hectare.

A general location description was made for each habitat and surrounding area. This description included area of the monitored habitat (measured directly) and a map and description of adjacent habitats (gathered from direct observation and county plat book land descriptions). We measured milkweed density (in ramets/ m²) at each habitat by sampling m² quadrats located along randomly positioned transects (Elzinga Bt 1998). Density measurements were conducted at the beginning and end of the monitoring season to track any changes in milkweed density over time.

Every week, we surveyed milkweed plants in each habitat to estimate monarch density (expressed as monarchs/milkweed ramet). For each habitat, we recorded numbers of milkweeds examined and numbers of monarchs found of each stage (eggs, larval instars 1-5, pupae). Monarch presence was determined by samples of 400 ramets located along randomly placed belt transects. We measured characteristics of a random sample of both milkweed and corn, including height, reproductive status, and signs of senescence. In addition, for each milkweed plant with a monarch on it, we measured the distance to closest corn plant and the number of other milkweed plants within a square meter that has the focal plant at its center. During our weekly visit to each location, we recorded rainfall over the past week, the presence of flowering plants in or near each habitat, and weather conditions on the monitoring day (temperature in sun and shade and under the corn canopy, wind speed and direction (both ambient and under corn canopy) and humidity (ambient and under the corn canopy).

Each week, we collected all of the 4^th and 5^th instar monarch larvae we found, recording the habitat from which they are taken. These were reared indoors to determine the presence of parasitoids and parasites. Common parasites include Tachinid flies (Prysby and Oberhauser personal communication) and the protozoan Ophryocystis elektroscirrha (Altizer and Oberhauser 1999). Wasp parasitoids and viral and bacterial infections have also been observed, though the latter two are primarily found in laboratory populations (Prysby and Oberhauser personal communication). The final status of each monarch collected (healthy, parasitized by flies, etc.) was recorded, and the sex, wing length, and mass determined for every resulting healthy monarch.

Results

A) Spatial overlap of monarchs and cornfields

Monarch larvae were observed in both cornfields and non-agricultural areas throughout the sampling period (Fig. 1). This was true even after the corn canopy had closed and corn plants were much taller than milkweed plants in cornfields (Figure 2). The numbers of monarchs/milkweed plant in a 400-plant sample did not differ between the two habitats. However, survivorship was higher in cornfields compared with non-agricultural areas (Weibull frequency distribution, P < 0.05)(Fig. 3).

Figure 1. Presence of monarchs (eggs and larvae) in cornfields and non-agricultural areas (non-ag) throughout July and August in Wellington County, ON. (The data represent all five sites, each of which was visited once per week.)

Click on image for larger view

Figure 2. Monarch presence on milkweed plants in cornfields in relation to corn height.

Figure 3. Survivorship of monarch larvae in cornfields and non-agricultural areas. Survivorship was higher in cornfields than in non-agricultural areas (Weibull frequency distribution (P < 0.05)).

B) Temporal overlap of corn anthesis and monarch larvae

Corn anthesis coincided with the second generation of monarchs; the overlap of the egg stage and the commencement of the period of pollen shed was substantial (Fig. 4). Temporal overlap between larvae of the second generation (peak migratory population) and corn anthesis ranged from 27% to 75%.

Figure 4. Spatial and temporal overlap of monarch larvae with corn anthesis.

C) Densities of milkweed plants in cornfields and non-agricultural habitats.

The relative importance of various habitats for monarch production depends on the number of milkweed plants in that habitat and the area of land covered by those habitats. Densities were typically 2-4 times greater in the non-agricultural areas than in the cornfields. In one location, there was 12-14 times more milkweed in surrounding natural areas than in cornfields (Figure 5).

Figure 5. Milkweed densities in corn and non-agricultural areas at the five monarch phenology study sites.

Conclusions

Cornfields in North America exist within the breeding range of monarch butterflies and serve as breeding habitats. In Ontario, numbers of monarch eggs in cornfields and non-agricultural areas were approximately equal in a sample of 400 plants per field suggesting that butterflies do not prefer to lay eggs on milkweed plants found in non-agricultural areas compared with those in cornfields. In other locations where similar studies were carried out, a considerable range of milkweed densities was observed. Overlap of monarch populations, specifically smaller and more sensitive larval stages, was 27 to 75% in Ontario. In other locations it was similar (MN) or quite a bit less (IA and MD). From prior data and those of other research collaborators, it is clear that a relatively small portion of monarch populations are exposed to pollen from Bt corn. In order to establish the relative importance of corn for monarch production, we must take into consideration the milkweed densities observed in our phenology fields , the milkweed densities observed in typical fields in Ontario and the area planted to corn versus other monarch habitats.

Milkweed Population Survey:

Introduction

Because cornfields and non-agricultural fields in the phenology study were chosen for the large numbers of milkweeds they contained, they do not necessarily represent typical milkweed densities in these habitats. Thus, the relative importance of each habitat for monarch production might be mis-represented. A study of milkweed density was conducted in randomly chosen fields within three representative counties within SW Ontario: Kent, Huron and Wellington counties.

Methods

Typical milkweed density in various habitats was determined by surveying agricultural, non-agricultural and field margin habitats located in Huron, Kent and Wellington counties. These habitat types were surveyed within each county, with 8 replicates of agricultural lands, each of corn, soybean and wheat fields, 8 replicates of field margin areas (corn field on one side, roadside on the other) and 10 replicates of non-agricultural areas (open grassy/weedy meadows where vegetation grew freely; not mowed, residential or grazing land). The dimensions of all agricultural and non-agricultural habitats were determined with a laser range finder, and the longest side of each site was divided into five equal sections. Four 2 x 100 m transects were located between sections, each of which projected into the field or non-agricultural site, perpendicular to its edge. Field margins were surveyed along 4 transects (2 x 100 m) which were parallel to the edge of the surveyed corn fields (and the nearest corn field if the first corn field did not have a long enough field/road margin). The number of milkweed ramets was recorded along each transect, and this count was converted to a milkweed/m² measurement. The mean field sizes for corn, soybean and wheat fields were measured (37.2 ± 5.1 acres, 48.6 ± 8.0 acres and 30.7 ± 4.5 acres, respectively). Because non-agricultural habitats were highly fragmented, care was taken to ensure that the dimensions of each of site were 100m x 100m at minimum (mean dimensions were 197.9 ± 28.5 m in length and 196.8 ± 18.5 m in width). Fields were selected for study by calling growers in each county prior to the survey, thus agricultural fields and margin habitats were blindly selected, which avoided biasing field choice towards fields that did or did not contain milkweed. Non-agricultural habitats were surveyed as soon as they were located, as long as they met the criteria for dimension and composition (freely growing open meadow).

Results

There were no differences in milkweed densities within a given habitat for the three counties (P>0.05) thus, the data were pooled (Table 1). However, non-agricultural areas contained significantly more milkweed than did other habitats ( F= 8.33 df = 4,489 P < 0.0001, Tukey's studentised range test, P < 0.05) (Fig. 6).

Table 1. Milkweed densities in various habitats within three Ontario counties, 2000.

County	Habitat	Mean ± s.e.	Range
Huron	Non-ag	0.97 ± 0.36	0 - 9.54
Huron	Margin	0.08 ± 0.02	0 - 0.65
Huron	Corn	0.02 ± 0.01	0 - 0.175
Huron	Soybean	0.02 ± 0.00	0 - 0.08
Huron	Wheat	0.02 ± 0.01	0 - 0.246
Kent	Non-ag	0.25 ± 0.06	0 - 1.125
Kent	Margin	0.21 ± 0.05	0 - 1.205
Kent	Corn	0.00 ± 0.00	0 - 0.025
Kent	Soybean	0.01 ± 0.00	0 - 0.085
Kent	Wheat	0.00 ± 0.00	0 - 0.025
Wellington	Non-ag	1.55 ± 0.72	0 - 15.98
Wellington	Margin	0.45 ± 0.11	0 - 2.14
Wellington	Corn	0.01 ± 0.00	0 - 0.075
Wellington	Soybean	0.03 ± 0.02	0 - 0.705
Wellington	Wheat	0.02 ± 0.01	0 - 0.22
Pooled	Non-ag	0.92 ± 0.27	0 - 15.98
Pooled	Margin	0.25 ± 0.04	0 - 2.14
Pooled	Corn	0.01 ± 0.00	0 - 0.175
Pooled	Soybean	0.02 ± 0.02	0 - 0.705
Pooled	Wheat	0.01 ± 0.01	0 - 0.246

Figure 6. Comparison on milkweed densities within different habitats. ( F= 8.33 df = 4,489 P < 0.0001, Tukey's studentised range test, P < 0.05).

Click on image for larger view

Although milkweed densities within the two experiments cannot be compared directly, it is evident that the fields in the phenology study contained milkweed densities that were higher than those observed in "typical" fields. For example, all of the cornfields monitored weekly had higher densities of milkweed than the cornfield average (Fig. 6). The lowest density in a phenology field was 0.01 ramets/m², whereas the average of randomly surveyed sites was 0.008 ramets/m². Using the observed monarch egg and milkweed densities to obtain the relative monarch productivity in each habitat, these results suggest that non-agricultural areas produce more monarchs than do cornfields in Ontario (Table 2).

Table 2. Estimated relative monarch productivity in cornfields compared with non-agricultural areas.

Habitat	Eggs/mw observed1	mw density2	egg density in habitat3	Monarch productivity Relative to non-ag4
Non-ag	0.0133	0.924	0.0123	1
Corn	0.0125	0.008	0.0001	0.008137

¹ From the phenology study (mw = milkweed)
² From the milkweed survey of randomly selected fields
³ Values are obtained by multiplying columns 1 and 2
⁴ Relative monarch productivity on a per area basis calculated from values in column 3.

Total contribution of cornfields to monarch production will depend on the acreage of corn relative to other habitats, or the proportion of breeding habitat that cornfields comprise. The relative acreage of various agricultural habitats within the three counties sampled in the milkweed survey is presented in Table 3.

Table 3. Acreage and percent land cover of crops and non-agricultural land in three counties within Southwestern Ontario, 1999.

Habitat	Huron acreage	Huron %	Kent acreage	Kent %	Wellington acreage	Wellington %	Average acreage	Average %
Grainsa	44,111	12.9	26,912	7.9	32,699	9.6	34,574	12.1
Corn	74,058	21.7	59,894	17.6	34,398	10.1	56,117	19.7
Soybeans	67,583	19.8	97,125	28.5	29,542	8.7	64,750	22.7
Non-agricultural	11,685	3.4	4016	1.6	23,063	8.6	23,063	4.5
Total	341,027	-	246,410	-	268,791	-	285,409	-

^a includes winter wheat, oats, barley, mixed grains and hay.

Table 4. Estimated relative contribution of corn and non-agricultural habitats for monarch production.

Habitat	Egg density In habitat1	Proportion of breeding habitat2	Total contribution relative To non-ag3	Proportion
non-ag	0.0123	0.045	0.000559	1.0
corn	0.0001	0.300	0.0000325	0.054

¹ From Table 2.
² From Table 3.
³ Relative productivity on a landscape basis calculated by multiplying egg density by proportion of the breeding habitat

Conclusions

The values presented above are rough estimates based on numbers of monarchs observed during a six week sampling period in one year, milkweed density estimates from a one-year, three county survey and general land use data for three counties in SW Ontario. Our estimate of the relative contribution of corn and non-agricultural habitats is only as good as the accuracy of available information and may change as more information becomes available. Further, an understanding of the total contribution of corn to monarch production will require an examination of the relative importance of other milkweed habitats such as soybean fields and field margins. However, the above analysis provides an important model that outlines the process required to determine the relative importance of cornfields for monarch production. Given the available information, it appears that the total contribution of cornfields relative to non-agricultural areas for monarch production is low, with about 20 times more production form non-agricultural areas than from corn fields. This is considerably different from corn growing areas in the mid-western U.S. where estimates for production of monarchs indicate 50-100 times more production is contributed from corn fields compared with non-agricultural areas (Oberhauser Bt 2001). This may not be surprising considering the extent of corn production in IA and MN compared with the amount of non-agricultural land in those areas. Clearly, across the corn belt the habitats from which monarch populations are produced vary considerably.

Pollen Density and Dispersal in the Field:

Introduction

The results of the studies presented thus far indicate that monarchs utilise cornfields and that monarch larvae are found in cornfields during anthesis, although the relative importance of cornfields compared with non-agricultural fields in Ontario is low. Given that some monarch larvae are exposed to corn pollen in their diet, the amount of pollen they may be exposed to must be determined. Dispersal and deposition of pollen from corn fields and on milkweed leaves is important to establish the degree of potential exposure.

In the summer of 1999, the distance, direction and density of pollen dispersal was examined at several field sites in Southwestern Ontario. Milkweed leaves at given distances from the field edge were collected and the extent of pollen deposition on them was determined at the end of the period of pollen shed. In 2000, pollen density on milkweed leaves was compared with densities observed on sticky plates within and near cornfields to predict more realistically the exposure of monarch larvae to corn pollen in the field.

Methods

Summer, 1999. Nine Bt-corn fields were chosen for study which were located within the Wellington, Oxford and Hamilton-Wentworth counties of Southern Ontario. An attempt was made to select small corn fields (< 20 ha) with access to all sides and little to no adjacent corn. When this was not possible, fields were selected that met these criteria on the southern and eastern sides, as these were considered the most important for study due to prevailing northwesterly winds. The study was conducted during the pollen shed period from 16 July to 31 July, although specific timing of the shed varied from field to field.

The fields were primarily rectangular and, ideally, were to have 16 transects radiating from the field's center and passing through either the four corners, or ¼, ½ or ¾ of the way down each side. Along each transect, a wooden stake (approximately 1 m high) was placed at 0, 1, 5, 10, 25, 50 and 100 m from the field's edge. Adjustments were made to these distances along angled transects in order to maintain the correct distance perpendicular to the edge of the field. In many cases, the property around the study field was inaccessible (i.e., residential, dense forest, harvesting field crop) and only the stakes closest to the edge could be erected.

Because pollen shed occurs over approximately 2 weeks, with a peak period of shed during the first 5 days, pollen dispersal was sampled over 24 hour periods for the first five days and then over 48 hour periods for the remainder of the shed. Pollen was collected on Petri plates painted with sticky material (Sticky Stuff™ or Tanglefoot™) that were attached to the top of each stake with Velcro™. After exposure, plates were labeled, replaced with a new plate and held at -17°C to prevent mould growth. Pollen was counted on each plate at 5 randomly selected 1 cm² areas and a mean pollen density per square centimetre was determined.

Summer, 2000. Pollen densities on leaves and sticky plates were compared in 3 Bt corn fields (Novartis N2555 and N27M3) and 3 non-Bt corn fields (Limagrain Pride 177, Hyland 2240 and Novartis 4064NK) in Wellington County in central southern Ontario, Canada. The fields were generally small (between 5 and 17 ha), and were planted in the first week of May (with the exception of one field planted in the last week of May) with corn plant populations from 28 000 to 30 000/ha. All fields were greater than 150 m from adjacent corn to avoid pollen contamination from neighbouring fields. Fields were equipped with a temperature logger (HOBO Pro Series, Onset Computer Corp., Pocasset, MA) and a rain gauge which was monitored daily. A drop of oil was placed in the rain gauge to prevent evaporation.

Potted milkweed plants (Asclepias syriaca) and Petri dishes coated with sticky insect trapping material (Sticky Stuff™, Olson Products Inc., Medina, OH) were used to estimate pollen densities on leaves and sticky plates, respectively. The milkweed plants were grown in a phytotron (16:8 photoperiod, 24°C/20°C day/night temperature and 65% R.H.) from seed collected the prior summer. Plants were watered and fertilized (20:20:20 NPK) every 2 days. Individual plants were transferred to 6" pots after two months and moved to a greenhouse one week before relocation to the field. Prior to corn pollen shed, wooden stakes were erected along 4 transects perpendicular to the edge of the field. The transects were evenly spaced along one or two sides of the field (one side only if it was longer than 500 m), and the stakes were placed at 4 distances along each transect at 1.5 m (inside the corn from the field's edge), 0 m (at the field's edge) and 1 m and 5 m (outside the field from its edge). A potted plant was secured to the ground beside each stake with 6 inch spikes. The mean height of the plants and plates was 76.1°C 0.9 m and 101.0°C 0.4 m, respectively. Sticky plates were fastened to the top of each stake with Velcro™ and changed every 24 hrs. This began before the onset of corn pollen shed, and these plates were examined daily to determine the start date of the shed.

Plates were changed daily for 16 days after the start of pollen shed and leaves were collected after 6 and 11 days of pollen shed. A leaf was randomly selected and carefully cut from the top, middle and bottom third of each plant. The height of the leaf from the ground, its angle relative to the stem and its aspect relative to the corn was recorded, then the leaf was sandwiched between 2 strips of contact paper (ConTact7™ Brand, Decora Manufacturing, North Ridgeville, OH) to avoid the loss of pollen from its surface. Plates and leaves were frozen at -20°C and -5°C until pollen could be counted.

To analyze pollen density on the sticky plates, the plates were stained with acid fuchsin (Sigma-Aldrich, Oakville, ON) and computer images of the stained pollen in five 1 cm² areas were created with a Panasonic WV-D5100 system digital camera mounted on a dissecting scope and AIMS Lab GrabIT II™ (v. 1.10). The image was analyzed with Scion Image Beta 4.0.2 and the pollen density in grains/cm² determined. Pollen density on milkweed leaves was evaluated by pulling the contact paper strips away from the leaf, staining them with acid fuchsin and counting the pollen in five 1 cm² areas on the top and bottom of the strips. Any pollen remaining on the leaves was also counted in five 1 cm² areas and added to the counts made on the contact paper.

Results

Pollen Dispersal From the Field Edge

Summer, 1999. The majority of the pollen counted (90%) fell within 5m of the field (Table 5). The direction of pollen dispersal from the field edge reflected the prevailing northwesterly winds in Southern Ontario. Thus, outside the south and east sides of a field more pollen collected than outside the north and west sides. For example, the plates from east side of one field collected more than 5x the amount of pollen than the west side.

Table 5. Pollen densities and proportional deposition on sticky plates placed at increasing distances from the edges of 9 cornfields.

Distance from Field edge (m)	Pollen density (grains/cm2/day) Avg.+ S.E.	Pollen density (grains/cm2/day) Range	Proportion of total Pollen deposition
0m	163.8 ± 9.1	0 - 853.8	0.44
1m	116.9 ± 7.8	0.9 - 672.0	0.75
5m	52.2 ± 4.7	0 - 686.2	0.90
10m	24.1 ± 1.9	0 - 240.8	0.95
25m	9.4 ± 1.6	0 - 261.1	0.98
50m	4.2 ± 0.4	0 - 30.2	0.99
100m	3.2 ± 0.4	0 - 38.4	1

Table has average daily pollen deposition for the 3 days when pollen shed was greatest.

Pollen Densities on Plates and Leaves in and near Cornfields (Summer, 2000)

Pollen deposition reached a peak by day 6 (the first sampling day for leaves) and occurred for a period of at least 16 days (Fig. 7). Deposition on leaves was much lower than that on plates; leaves contained approximately 6% of the amount of pollen found on plates on both sampling days (Table 6). Average accumulated rainfall for the six cornfields and three control areas during the first six and eleven days of pollen shed (prior to placing cages on plants) was 10.1 ± 1.9 mm and 13.4 ± 2.3 mm, respectively, thus some pollen may have been washed off of the plants (Pleasants Bt, 2001). Leaves taken from plants within the field had more than 7 times the amount of pollen than leaves taken from plants 5m from the field edge, where densities were about 10 grains/cm2 . Finally, leaves sampled from the middle of the plant consistently contained higher levels of pollen than those from either the top or bottom of the plant. The highest density of pollen observed on any leaf was 456 grains/cm2. In Iowa, where hot, dry conditions prevailed in the summer, 2000, densities as high as 1000 grains/cm2 were observed on a very small proportion of leaves and average densities were about 4-5 times higher than those observed in Ontario (Pleasants Bt 2001).

Figure 7. Pollen density on milkweed leaves and sticky plates inside the corn field (1.5m) during the pollen shed. Plates densities reflect accumulation over the previous 24 hours and are not cumulative over the pollen shed period.

Click on image for larger view

Table 6. Comparison of pollen deposition on plates and milkweed leaves from plants placed in and near cornfields.

Day of Shed	Distance (m)	Pollen Density (grains/cm2) Plates Avg. ± S.E. Range	en Density (grains/cm2) Plants Avg. ± S.E. Range	en Density (grains/cm2) Midleaf Avg. ± S.E. Range	% pollen on plant	% pollen on midleaf
Day 6	-1.5	777.6 ± 117.7	69.8 ± 10.0	106.1 ± 15.6	0.09	0.14
Day 6	-1.5	567 - 1301	4 - 189	4 - 264
Day 6	0	602.7 ± 94.1	38.1 ± 6.1	55.4 ± 11.0	0.06	0.09
Day 6	0	349 - 874	7 - 116	3 - 220
Day 6	1	543.7 ± 100.5	31.1 ± 5.7	46.3 ± 9.3	0.06	0.09
Day 6	1	313 - 842	3 - 133	4 - 195
Day 6	5	243.5 ± 37.8	9.4 ± 1.1	9.2 ± 1.2	0.04	0.04
Day 6	5	99 - 349	1 - 20	0 - 28
Day 11	-1.5	1333.6 ± 145.7	98.2 ± 12.0	150.5 ± 23.3	0.07	0.11
Day 11	-1.5	020 - 1880	32 - 265	24 - 454
Day 11	0	992.5 ± 109.9	59.5 ± 9.2	89.8 ± 21.8	0.06	0.09
Day 11	0	727 - 1404	6 - 188	15 - 456
Day 11	1	873.5 ± 123.8	49.1 ± 9.3	61.9 ± 10.9	0.06	0.07
Day 11	1	556 - 1310	3 - 234	1 - 231
Day 11	5	360.9 ± 42.9	13.6 ± 2.1	19.4 ± 4.6	0.04	>0.05
Day 11	5	180 - 457	2 - 41	1 - 78

Factors affecting pollen deposition on milkweed leaves

Reported levels of pollen deposition on milkweed leaves can vary widely from one study to another depending on when milkweed leaves were sampled during anthesis and where the milkweed plants were located within the cornfield. Pleasants Bt (2001) describe other factors that affect pollen deposition on milkweed leaves that include:

• Rainfall: a single rain event can wash up to 90% of the pollen off of a milkweed leaf
• Plant position in the corn canopy: plants within rows of corn have been found to accumulate more pollen than plants between corn rows
• Leaf position: leaves taken from the middle of a milkweed plant tend to accumulate more pollen than leaves from the top or bottom of the plant
• Position on the leaf: Pollen tends to collect along the midvein of the leaf compared with the laminar portions on either side of the midvein

Conclusions

Pollen densities on milkweed leaves in Ontario were no more than 100 grains/cm² leaf area within cornfields during peak pollen shed. Although densities may be 4-5 times higher under conditions of little or no rainfall, densities observed in Ontario are typical of locations where some rainfall occurs (Pleasants Bt 2001). Leaves contained only a fraction of the pollen that was shed, indicating that rainfall, wind or some other factor was removing pollen from the leaf surface. If we assume that the proportion of pollen found on leaves compared with plates is constant as one moves away from the field edge, we would predict that pollen levels would be about 1 grain/cm² or less at a distance of 10m from the field edge. Thus, the area of potential risk of exposure to Bt pollen is likely to include the cornfields themselves and a 5-10m area around a cornfield.

---

Section 3: Toxicity of Bt corn pollen to monarch larvae

Laboratory Bioassays

Introduction

The impact on monarch populations of exposure to Cry1A(b) proteins may vary for different Bt corn events due to differing amounts of toxin expressed in Bt pollen. Pollen from event 176 Bt hybrids expresses the highest level of Bt protein with toxin levels as high as 7.1 mg/gm pollen (EPA 2000a). Hybrids derived from event 176 currently represent <2% of the corn acreage in North America, and the registration of these hybrids will not be renewed for 2001. However, studies utilizing hybrids of this event are valuable because they allow assessment of the potential for Bt toxin expressed in pollen to negatively impact monarch survival and development. Studies have demonstrated that exposure of monarch larvae to event 176 hybrids can potentially reduce survivorship and growth of first instars (Hansen and Obrycki, 2000, Hellmich Bt, 2001). In comparison, Bt11 and MON810 hybrids express less than 0.09 mg Cry1A(b)/gm of pollen (EPA 2000b), thus the potential negative impacts of these hybrids on monarch populations are expected to be lower than that of event 176 hybrids.

We conducted laboratory bioassays in which first instar monarch larvae were exposed to milkweed leaves containing Bt corn pollen from events Bt176 and Bt11, non-Bt pollen and pollen-free milkweed leaves.

Methods

Bt-176 Bioassay

Pollen collection and storage: During the peak pollen shed period (a five day period between 15 July and 22 July, 1999 depending on the field location), pollen was collected by stapling paper bags over the shedding tassels of 100 corn plants at each of the nine field sites. After 24 hours, the bag and tassels were removed from the plant and taken back to the laboratory where tassels were shaken out and removed and the pollen and anthers were collected and sealed in clear plastic containers and stored at -80° C. Prior to use, pollen was sifted to remove tassel and anther material and collected in 4ml glass vials. Samples of milkweed leaves collected during the field study portion of this project were examined for pollen and other whole or fragmented corn tissues. Only pollen and intact anthers have been observed on milkweed leaves from cornfields, no other corn tissue fragments or broken anthers were observed. We have attempted top feed corn anthers to young larvae without success because they are too small to consume whole anthers, while larger larvae tend to brush the anthers aside while consuming the milkweed leaf tissue. It is critical to the laboratory bioassay that it reflects the conditions presented to monarch larvae in the field. It has been clearly demonstrated that corn plant tissues, if left to contaminate the pollen sample, can severely affect the results of the bioassay by causing mortality that is not equivalent to that which would be expected in the natural situation (Hellmich et al. 2001).

Bioassay set-up: Adult monarchs were purchased from a butterfly farm (Swallowtail Farms, Carmichael, CA) and placed in mating cages with potted milkweed plants (Asclepias curassavica) from which first instar larvae that were < 24 hours old were harvested. A cohort of 10 larvae was weighed and then moved with a camel hair brush onto the top of an excised A. syriaca leaf to which a known dose of pollen had been applied. Pollen application was achieved using a modified Potter tower. The tower was modified by replacing the nozzle (through which a liquid insecticide is normally administered) with a mesh basket. A weighed amount of pollen was forced through the mesh by allowing a stream of air to flow through a glass funnel cupped over the basket. This method resulted in an even distribution of pollen grains over a leaf placed at the bottom of the tower. Larvae received one of three treatments: Bt pollen (Bt 176 from Max357 hybrids, Novartis Seeds, Inc.), non-Bt pollen (EnerFeast 1, Novartis Seeds Inc.) or no pollen (control), at doses of 0, 133, 541, 2379, 5486 ans 8525 grains/cm². Each leaf was placed in a ventilated plastic arena, and the arenas were kept in growth cabinets set at 20° C, 60% R.H. and 16:8 photoperiod for the duration of the bioassay. The larvae were exposed to the dose of pollen for 48 hours, after which the milkweed with pollen was removed from each arena and replaced with clean milkweed.

Larval survival, development, weight gain, and consumption were evaluated over the course of the bioassay. Leaf consumption and number of feeding sites were measured by creating a digital image of the leaf with a XC-75CE black and white video camera module with a Cosmicar/Pentax 16 mm TV lens and processing it with image analysis software (Northern Exposure 2.93, Empix Imaging Inc.).

Bt-11 Bioassay

Pollen Collection and Storage: Laboratory bioassays were conducted with field-collected Bt 11 (Novartis N27M3) and non-Bt isoline (Novartis N26L6) corn pollen. Tassel bags (Lawson 404 Showerproof=d, Northfield, IL) were placed on 300 shedding Bt and non-Bt corn plants starting at 5 p.m. left overnight, and removed the following morning at 10 a.m. The contents of the tassel bags were emptied onto a separate tray for each pollen type, spread out and left to dry for 48 hours in a cool, dark and dry location (10° C and 48% R.H.). After 48 hours, the pollen was sieved first through a 150 mesh (90 microns), then a 250 mesh (63 microns, both from Canadian Forestry Equipment, Mississauga, ON) to remove all tassel material other than pollen. The sieved pollen was placed in glass vials and stored at -80° C until use.

Bioassay Set-up: The bioassay was conducted as described above. Six doses of Bt and non-Bt pollen at 150, 300, 750, 1100, 1500 and 4000 grains/cm² were applied to whole milkweed leaves (Asclepias syriaca). Each bioassay was replicated 4 times and there were 12 control replicates that consisted of whole leaves without pollen.

In addition, a second bioassay was conducted comparing only the 1500 grains/cm² dose using the same pollen types and methodology as that described above with the exception that the exposure period was extended to 5 days.

For the Bt176 bioassay the following responses were analysed: survival, development and consumption at 48 and 96 and proportional weight gain (final/initial) at 96 hours. For the Bt11 bioassay, the following responses were analysed: survivorship and development at 48 hours, 96 hours and 10 days, proportional weight gain at 96 hours and 10 days, and consumption and number of feeding sites at 48 and 96 hours. In the second Bt11 bioassay (5 day exposure) survivorship, development and proportional weight gain were recorded at 5 and 10 days and consumption and number of feeding sites at 5 days. Data were analysed with ANOVA (PROC GLM, SAS, 1991) and were tested for normality and homogeneity of variances using Spearman's Rank Correlation and Shapiro-Wilk's W test. Pairwise orthogonal contrasts were performed for the lowest pollen dose or the highest pollen dose where interactions were significant.

Results

Bt-176 Bioassay

Survivorship was affected by both the type and dose of pollen [48 hours: pollen, F = 39.39, df = 1,30, P < 0.0001, dose, F = 32.3, df = 1,30, P < 0.0001; 96 hours: pollen, F = 22.64, df = 1,27, P < 0.0001, dose, F = 25.75, df = 1,27, P < 0.0001]. In addition, there was a dose*dose*pollen effect at 96 hours [F = 9.32, df = 1,27, P < 0.005] indicating that the dose response curves differed between the pollen types; the quadratic response was significant for Bt pollen [P < 0.01] but only approached significance for the non-Bt pollen treatment group [P = 0.067]. A pairwise contrast for the lowest dose revealed that no differences in survival were observed at 133 grains/cm² (P = 0.3739). Development, weight gain and consumption were also affected by the type and dose of pollen, though no interactions were observed [(development:48 hours: pollen, F = 5.5, df = 1,25, P < 0.05, dose, F = 5.83, df = 1,25, P < 0.05; 96 hours: pollen, F = 9.12, df = 1,25, P < 0.01, dose, F = 8.49, df = 1,25, P < 0.01) (weight gain: 96 hours: pollen, F = 8.01, df = 1,26, P < 0.01, dose*dose, F = 4.32, df = 1,26, P < 0.05) (consumption: 48 hours: pollen, F = 34.30, df = 1,30, P < 0.0001, dose*dose, F = 12.07, df = 1,30, P < 0.005; 96 hours: pollen, F = 4.45, df = 1,29, P < 0.05, dose, F = 3.6, df = 1,29, P = 0.068)]. Overall, exposure to Bt176 pollen resulted in slowed development, decreased weight gain and reduced consumption compared with non-Bt pollen. The same effects were observed with increasing dose of pollen regardless of the pollen type. For survivorship, Bt pollen caused reduced survivorship at all but the lowest dose of pollen. The results for 96 hours are shown in Fig. 8.

Figure 8. Responses of first instar monarch larvae to increasing doses of Bt176 and non-Bt pollen on milkweed leaves following a 48 hour exposure period on day 4.

Bt11 Bioassay

No differences in survivorship were observed throughout the exposure period. Development was affected by pollen dose at 48 hours [dose*dose, F = 6.45, df = 1,34, P < 0.05], however, the effect was not curvilinear and was not significant at 96 hours. By day ten, there was a significant pollen*dose interaction [F=4.16, df = 1,54, P < 0.05]. However, no difference was detected between larvae fed Bt and those fed non-Bt at the highest dose [Contrast: F = 1.11, df = 1,6, P = 0.3332]. Proportional weight gain was affected by dose at both 4 and 10 days [96 hours: F = 6.26, df = 1,54, P < 0.05; 10 days: F = 9.07, df = 1,54, P < 0.005]. In addition, the pollen*dose interaction approached significance on both days [96 hours: F = 3.44, df = 1,54, P = 0.0692; 10 days: F = 3.23, df = 1,54, P < 0.0778]. However, orthogonal contrasts revealed that weight gain in response to exposure to Bt and non-Bt pollen did not differ, though on day 4 the effect approached significance [(day 4: F = 4.41, df = 1,6, P < 0.0804), (day 10: F = 3.02, df = 1,6, P < 0.1330)]. Consumption was affected by dose during the exposure period [F = 7.21, df = 1,54, P < 0.001] but not at 96 hours and there were no differences in the number of feeding sites at either sampling time. Results at 96 hours are presented in Fig. 9. Results for development and weight gain after 10 days are presented in Fig. 10.

Finally, first instar larvae exposed to either 1500 BT11 pollen grains/cm², 1500 non-Bt pollen grains/cm² or pollen-free leaves for a five day exposure period did not differ in survivorship, development, weight gain, consumption or feeding behaviour (number of sites) after either 5 days or 10 days from the commencement of the exposure period. In fact, survivorship was 100% under all conditions.

Figure 9. Responses of first instar monarch larvae to increasing doses of Bt11 and non-Bt pollen on milkweed leaves following a 48 hour exposure period on day 4.

Figure 10. Responses of first instar monarch larvae to increasing doses of Bt11 and non-Bt pollen on milkweed leaves following a 48 hour exposure period on day 10.

Conclusions

One would predict that Bt176 poses a certain degree of risk to the monarch butterfly at doses observed in the field based on the results of our bioassay. Although survivorship was not affected at the lowest dose observed (133 grains/cm²) lower survivorship was observed in response to Bt pollen compared with non-Bt pollen at higher doses. Weight gain, developmental rate and consumption of first instar larvae were reduced by exposure to Bt pollen as compared with non-Bt pollen at all doses measured although high doses of pollen, regardless of type, also reduced these responses compared with low doses of pollen or no pollen. Hellmich Bt (2001) found that first instar monarchs showed reduced weight gain in response to Bt pollen at doses as low as 8 grains/cm².

In contrast, our results for Bt11 pollen revealed that, at exposure doses considerably higher than doses observed in the field, no detrimental effects were observed. At the highest dose evaluated (4000 grains/cm²) survivorship, development and weight gain may demonstrate a downward trend, but were not significant. At 1500 grains/cm² and at an exposure period of 5 days, no detrimental effects were observed in monarch larvae exposed to Bt pollen compared with non-Bt pollen or pollen-free leaves. Our results are corroborated by those of Hellmich Bt (2001) who reported no effects of Bt11 pollen in first instar monarch larvae following a four day exposure to 1200 grains/cm². Although MON810 pollen was not included in our studies, it has about the same level of expression as Bt11 pollen and had similar effects (Hellmich Bt 2001).

Field Bioassays

Introduction

As important as laboratory bioassays are for establishing toxic and sublethal doses of pollen, they do not estimate the effect of naturally occurring levels of pollen deposition on monarch larvae. The densities of pollen required to elicit a negative response in lab bioassays are often much greater than levels found in and around the corn field. They also do not address the more subtle aspects of the experience of monarch larvae in the field, for example, the choice to feed on plant material with less pollen deposition or the rapid degradation of the protein toxin in UV light. Conversely, there is potential for survival and growth to be more limited on whole plants in the field due to the relative decrease in latex flow and cardenolide content in severed leaves, such as those used for lab bioassays (Zalucki and Malcolm 1999). The black swallowtail (Papilio polyxenes) is another non-target Lepidopteran whose host plant is found in close proximity to corn fields. A field bioassay of first instar black swallowtail larvae found that there was no relationship between larval mortality and either proximity to corn or pollen deposition on the host plant and that furthermore, increased mortality could not be induced in larvae at 50 times the highest pollen density found in the field (Wraight et al. 2000). Laboratory studies might provide results that indicate there is a risk of Bt corn pollen to non-targets, but a field bioassay would provide the opportunity to put this risk into the context of natural conditions.

We conducted a field bioassay to determine the effect of Bt 11 pollen deposition on naturally growing milkweed plants inside and outside the field on survival, development, larval weight gain, consumption, pupation time, pupal weight, time to adult emergence, adult weight and adult wingspan for both first and third instar monarch larvae.

Methods

Field bioassays were conducted during pollen shed in 6 corn fields in Wellington County in central southern Ontario, Canada. Three non-Bt fields (Limagrain Pride 177, Hyland 2240, Novartis 4064NK) and 3 Bt 11 fields (Novartis N2555 and N27M3) were selected based on the following criteria: significant amounts of milkweed were located inside the edge of the cornfield and at the edge and in the margin, and fields were more than 150 m from other cornfields. The fields were generally small (between 12 and 43 acres), and were planted in the first week of May (with the exception of one field planted in the last week of May) with planting populations from 28 000 to 30 000 seeds/ha.

Prior to pollen shed, naturally growing milkweed plants (Asclepias syriaca) were flagged along 8 transects at 3 distances along the edge of cornfields: 1-2 m inside the field (mean 1.15 m ? 0.06), 0-1 m outside the field from the edge (mean 0.58 m ? 0.04) and 4.5-6.5 m from the edge (mean 5.18 m ? 0.14). A control plot consisting of 8 milkweed plants was identified in a non-agricultural area for each of the 3 pairs of Bt/non-Bt fields. To prevent contamination with pollen, control plots were located 150 m or more from adjacent cornfields. In general, milkweed plants that were selected were in good health, had minimal herbivore damage, and were approximately 70 cm tall.

Each of the 8 transects of milkweed plants in the cornfields and 8 plants in the control plots were alternately assigned to the first or second set of bioassays. The first bioassay was conducted after 6 days of pollen shed with first instar monarch larvae and the second bioassay was conducted after 11 days of pollen shed with third instar larvae. Preliminary trials with larvae on caged and uncaged milkweed indicated that cages were required to avoid high larval mortality due to predation over the course of the bioassay. Before pollen shed, stakes (1 m tall) were placed beside each plant, each with an arm that extended over the top of the plant. Cages consisted of a tube of a sheer mesh fabric that was gathered and tied tightly to the stem at the bottom of the plant and the arm that extended over the top of the plant. Circles of wire were fastened inside the sheer tube at both ends to keep the material taut and from touching the plant. To avoid disturbing any pollen that had deposited on the leaves during pollen shed, cages were placed at the bottom of the plant prior to pollen shed so they could be drawn up around it at the start of the bioassay.

Fields were monitored daily to determine the first day of pollen shed. All fields and control areas were equipped with temperature loggers (HOBO Pro Series, Onset Computer Corp., Pocasset, MA) and rain gauges, which were checked daily. After pollen had shed for 6 or 11 days for the first and second bioassay, respectively, a cohort of 5 larvae were placed on the centre of the fifth leaf from the top of the plant with a camel hair brush, and the mesh was drawn up around the plant and secured to the overhanging arm. A sheet of acetate was placed on top of the cage to keep rain and additional pollen off the plant for the duration of the bioassay. Larvae for the first bioassay were < 24 hours old and larvae for the second bioassay had been third instars for < 24 hours. Within this timeframe, larvae were categorised as early or late based on relative size, and then evenly distributed among the cohorts. Weights of the cohort of 5 larvae (in mg) were taken prior to the bioassay with a Mettler AT250 scale, and individual larval weight was determined. Cohorts were randomly assigned to plants.

After larvae were on the caged plants for 5 days (11 and 16 days after the start of pollen shed for the first and second bioassay, respectively), cages were removed and the number and instar of remaining larvae were recorded. In order to estimate consumption of leaf area and pollen dose during the bioassay, any leaf with monarch feeding damage was collected from plants of the first bioassay. Feeding damage from third instars was so extensive in the second bioassay that total consumption could not be estimated, therefore a leaf from the top, middle and bottom third of plant was collected to estimate pollen dose. To minimize loss of pollen from leaf surfaces, all sampled leaves were sandwiched between strips of contact paper (ConTact7 Brand, Decora Manufacturing, North Ridgeville, OH). Remaining larvae from each cohort were collected, placed on clean milkweed in a 17 x 12 x 8 cm ventilated plexiglass container and returned to the laboratory where they were reared to adulthood at 24° C and 16:8 photoperiod. Weight of each cohort was recorded 24 hours after larvae were removed from the field and weight per larva was determined. Larvae were supplied with fresh milkweed leaves every day until they pupated. Pupation date was recorded for each individual and the pupa was weighed 2 days later. Date of adult emergence was recorded and 1 day later adult wingspan, sex and weight were recorded and adults were checked for the microsporidian parasite Ophryocystis elektroscirrha.

Consumption was measured by holding leaves in front of a light source and creating a digital image of the open areas of leaf consumption with a XC-75CE black and white video camera module and a Cosmicar/Pentax 16 mm TV lens. Area of consumption was determined with image analysis software (Northern Exposure 2.9e, Empix Imaging, Inc.). Pollen densities in grains/cm² were determined for leaves collected from the top, middle and bottom of plants for the first and second bioassay. The majority of pollen adhered to the contact paper strips after they were removed from the leaves, and they were stained with acid fuchsin (Sigma-Aldrich, Oakville, ON) to facilitate pollen counting. Pollen was counted within five 1 cm² areas on the top and bottom strips and then on the top and bottom of the leaf itself. Pollen counts for leaves and strips were added to estimate total pollen density on the top and bottom of each leaf.

Results

Average distances of plants within each condition were as follows: -1.15 ± 0.06 m within the field, 0.58 ± 0.04 m outside of the field, and 5.18 ± 0.14 m outside of the field. Pollen densities on leaves did not differ between corn types (Bt or Non-Bt) but decreased significantly with increasing distance from the field on day 6 (F = 14.0; df = 2,56; P < 0.0001) and day 11 (F = 70.51; df = 1,62; P < 0.0001) of anthesis (Fig. 11). On day 6, the range of pollen densities at each distance was 1.5-309, 0-176 and 0-75 grains/cm² for plants found within the field (-1 m), less than 1 m outside the field (<1 m) and 5 m outside the field, respectively. On day 11, the range of densities calculated on leaf samples taken from -1 m, <1 m and 5 m were 3.3-429, 0.2-320 and 0.8-50 respectively. Results of a concurrent study in which leaf samples from the same fields were collected on day 6 and day 11, suggest that the cage and acetate top prevented further accumulation of pollen (unpublished data). Control plants contained an average of 1.43 ± 0.3 grains per cm² and 1.33 ± 1.0 grains per cm² on day 6 and day 11 respectively, which is likely to have resulted from contamination during leaf sampling.

Neither corn type (Bt or Non-Bt) nor plant position affected survivorship or fitness of first or third instar monarchs during the exposure period in the field or in later developmental stages. There was a tendency, but non-significant, for greater weight gain and development in and near Bt cornfields compared with Non-Bt cornfields that is explained by different temperatures in the fields; average temperatures for Bt and Non-Bt fields during the period of exposure were 19.6 ± 0.85°C and 17.2 ± 3.1°C, respectively, on day 6 and 18.3 ± 1.6°C and 16.7 ± 1.7°C, respectively, on day 11. Comparison of responses to Bt and Non-Bt pollen is presented for the first bioassay (Table 7) and the second bioassay (Table 8). No significant differences in survivorship or development through to adulthood were observed between larvae in cages within the fields and those in control cages more than 150 m from any cornfield in the first bioassay (Table 9) or in the second bioassay (Table 10).

Figure 11. Pollen deposition on milkweed leaves from plants used in the first bioassay (day 6) and the second bioassay (day 11). Within = leaves from plants approximately 1m inside of the field, margin = leaves from plants taken <1m from the field edge.

Click on image for larger view

Table 7. Comparison of survivorship and growth for monarch larvae exposed to Bt11 and Non-Bt pollen as first instars.

Response variable	Mean (± SEM) Bt11	Mean (± SEM) Non-Bt	Fa	P
% Survival in field	84.4 ± 3.8	80.6 ± 3.4	F = 0.64	0.4695
Leaf consumption/larva (cm2)	1.7 ± 0.3	0.97 ± 0.1	F = 1.28	0.3216
Developmental changeb	2.4 ± 0.1	2.1 ± 0.1	F = 0.87	0.4030
Weight gain/larvac	20.4 ± 1.1	15.6 ± 1.8	F = 1.03	0.3682
% Survival to pupation	59.4 ± 5.1	57.4 ± 6.2	F =0.18	0.7025
Days to pupationd	16.1 ± 0.3	16.3 ± 0.3	F = 0.97	0.3980
Pupal weight (mg)	1280 ± 24.5	1246 ± 37.1	F = 0.25	0.6498
% Emergence from pupae	80.0 ± 6.0	95.2 ± 2.3	F = 5.83	0.0731
Adult weight (mg)	534.1 ± 12.2	522.2 ± 13.2	F = 1.38	0.3249
Adult wing length (mm)	52.7 ± 0.44	52.0 ± 0.6	F = 0.22	0.6719

^aANOVA; df = 1.4 for field measurements, df = 1.3 for laboratory measurements
^bDevelopmental change = average instar on day 5/ average instar on day 0
^cWeight gain/larva = proportional weight gain in the field (day 6 wt/day 0 wt)
^dDays to pupation from the beginning of the exposure period

Table 8. Comparison of survivorship and growth for monarchs exposed to Bt-11 and Non-Bt pollen as third instars.

Response variable	Mean (± SEM) Bt11	Mean (± SEM) Non-Bt	Fa	P
% Survival in field	91.7 ± 2.0	87.8 ± 2.8	F = 0.66	0.4635
Developmental changeb	4.2 ± 0.1	4.0 ± 0.04	F = 1.77	0.2544
Weight gain/larvac	15.1 ± 0.9	10.8 ± 1.5	F = 1.17	0.3398
% Survival to pupation	73.3 ± 3.4	63.3 ± 6.2	F = 0.45	0.5503
Days to pupationd	11.6 ± 0.2	12.1 ± 0.4	F = 0.22	0.6705
Pupal weight (mg)	1201.4 ± 24.7	1147.1 ± 34.6	F = 0.12	0.7484
% Emergence from pupae	80.2 ± 3.7	76.2 ± 6.1	F = 0.00	0.9717
Adult weight (mg)	474.9 ± 11.1	467.8 ± 15.4	F = 0.04	0.8628
Adult wing length (mm)	50.9 ± 0.4	49.8 ± 0.5	F = 1.32	0.3341

^aANOVA; df = 1.4 for field measurements, df = 1.3 for laboratory measurements
^bDevelopmental change = average instar on day 5/ average instar on day 0
^cWeight gain/larva = proportional weight gain in the field (day 6 wt/day 0 wt)
^dDays to pupation from the beginning of the exposure period

Table 9. Comparison of survivorship and growth for monarchs exposed to within field doses of Bt-11 and Non-Bt pollen or pollen-free leaves as first instars.

Response variable	Mean (± SEM) Bt11	Mean (± SEM) Non-Bt	Mean (± SEM) Control	Fa	P
% Survival in field	81.7 ± 8.7	80.0 ± 5.4	71.7 ± 8.3	F = 1.35	0.3288
Leaf consumption /larva (cm2)	1.4 ± 0.3	1.0 ± 0.2	0.8 ± 0.2	F = 2.26	0.1860
Developmental changeb	1.3 ± 0.1	0.9 ± 0.2	0.8 ± 0.1	F = 2.27	0.1850
Weight gain/larvac	19.8 ± 2.0	15.4 ± 2.6	10.5 ± 2.5	F = 1.06	0.4167
% Survival to pupation	70.9 ± 8.7	55.0 ± 11.8	40.0 ± 11.3	F = 2.0	0.2489
Days to pupationd	15.6 ± 0.3	16.7 ± 0.4	16.5 ± 0.3	F = 2.89	0.1464
Pupal weight (mg)	1294.7 ± 34.3	1213.3 ± 75.9	1304.4 ± 63.5	F = 0.75	0.5181
% Emergence from pupae	86.2 ± 6.5	97.1 ± 2.9	97.1 ± 2.9	F = 1.55	0.2998
Adult weight (mg)	533.6 ± 15.6	544.8 ± 26.9	535.2 ± 23.8	F = 0.08	0.9265
Adult wing length (mm)	52.4 ± 0.8	51.6 ± 1.2	53.0 ± 1.1	F = 0.27	0.7711

^aANOVA; df = 2.6 for field measurements, df = 2.4 for laboratory measurements
^bDevelopmental change = average instar on day 5/ average instar on day 0
^cWeight gain = proportional weight gain in the field (day 6 wt/day 0 wt)
^dDays to pupation from the beginning of the exposure period

Table 10. Comparison of survivorship and growth for monarchs exposed to within field doses of Bt-11 and Non-Bt pollen or pollen-free leaves as third instars.

Response variable	Mean (± SEM) Bt11	Mean (± SEM) Non-Bt	Mean (± SEM) Control	Fa	P
% Survival in field	90.0 ± 3.0	90.0 ± 3.9	91.7 ± 3.0	F = 0.04	0.9613
Developmental changeb	0.9 ± 0.1	0.5 ± 0.1	0.7 ± 0.1	F = 1.58	0.2816
Weight gain/larvac	14.2 ± 1.8	11.2 ± 2.8	8.4 ± 2.1	F = 0.71	0.5269
% Survival to pupation	68.3 ± 3.9	52.5 ± 9.2	60.0 ± 14.1	F = 0.28	0.7677
Days to pupationd	11.7 ± 0.5	12.7 ± 0.9	12.1 ± 0.3	F = 0.16	0.8554
Pupal weight (mg)	1182.3 ± 51.2	1110.7 ± 73.8	1065.1 ± 43.1	F = 0.25	0.7931
% Emergence from pupae	79.7 ± 7.5	95.2 ± 4.8	90.0 ± 4.8	F = 1.41	0.3443
Adult weight (mg)	465.2 ± 27.5	442.5 ± 35.9	451.7 ± 19.1	F = 0.10	0.9111
Adult wing length (mm)	51.2 ± 0.9	49.5 ± 1.0	51.2 ± 0.9	F = 1.04	0.4322

^aANOVA; df = 2.6 for field measurements, df = 2.4 for laboratory measurements
^bDevelopmental change = average instar on day 5/ average instar on day 0
^cWeight gain = proportional weight gain in the field (day 6 wt/day 0 wt)
^dDays to pupation from the beginning of the exposure period

Conclusions

Results of field bioassays corroborate those of laboratory assays. Levels of pollen observed in field cages were not high enough to cause a significant effect on survival or growth of first or third instars during the exposure period or in later development stages. Differences in susceptibility of different instars of monarchs (Oberhauser et al. unpublished data) and other species (Peacock Bt 1998) have been observed in response to exposure to Bt toxin. However, in the Ontario study, neither first nor third instars were detrimentally affected by pollen from event Bt11 hybrids following exposure to pollen densities of 59 ± 10.9 and 94.1 ± 23.0 grains/cm², respectively. Fitness of adult monarchs was also unaffected by Bt11 pollen at these doses; larvae caged in Bt11 cornfields, non-Bt cornfields or control areas developed into adults of similar weight and size within similar developmental times. This result is consistent with other studies that have shown that even where sublethal effects are observed in larvae following a period of exposure to Bt toxin, later stages are often unaffected (Hanson and Obrycki 2000, Peacock et al. 1998). Results are also consistent with those of an experiment using the Bt11 sweet corn (Stanley-Horn Bt 2001) that showed no effects on monarch survival or development after 4 days at any level of pollen deposition, though pollen levels reached 586 grains/cm². That Hanson and Obryicki (2000) observed detrimental effects in first instar larvae following exposure to 135 grains/cm² Bt11 pollen in the laboratory may have resulted from sample contamination with tassel material since this was not filtered out (see Hellmich et al. 2001). Bioassays reported in this study, however, predict monarch responses for only a 5-day window of exposure. In natural settings, larvae that hatch at the onset of anthesis may be exposed to CryIAb protein in pollen for a longer period, thus continuous exposure may result in sublethal or lethal impacts. That exposure time is critical to toxicity of Bt has been previously demonstrated (Fast and Régnière, 1984).

---

Section 4: Risk Assessment

Introduction: Hazard identification

Lepidopeteran-active Bt-protein expressed in corn may pose a risk to sensitive species in or near cornfields. Monarch butterflies have demonstrated sensitivity to certain Bt proteins that occur in certain corn pollen that is distributed in and about cornfields during anthesis. Milkweeds, Asclepias spp., the sole larval food source for Monarch butterfly larvae, are abundant throughout the corn-growing regions of North America. Hazard due to Bt corn pollen deposited on milkweed leaves in and around cornfields should be considered with regard to ecological risk to migratory monarch populations.

Conceptual Model

Risk assessment requires knowledge of four essential components: i) hazard identification, ii) nature of dose-response to a toxin, iii) probability of exposure to an effective dose, and iv) characterization of risk. Components of a risk assessment approach as applied to the case of Bt corn pollen and monarch butterfly are depicted in Fig. 12.

Figure 12. Conceptual model of components of risk assessment on the impact of Bt corn pollen on populations of the monarch butterfly.

Consideration of risk as a function of exposure and effect requires that lines of evidence be established in four areas of inquiry: 1) Is there some density of Bt pollen on milkweed leaves that represents a lethal or sublethal threat to monarch larvae or later stages of development? 2) What proportion of Bt pollen deposited on milkweed leaves in and around cornfields exceeds the toxicity threshold for larvae of monarchs? 3) What proportion of monarch populations utilize milkweed in and near cornfields? 4) What is the degree of overlap between the phenological stages of monarch larvae and corn anthesis over the shared range of these species? Portions of each of these questions were answered in the study reported here and provide the basis for risk assessment of the impact of Bt pollen to monarch butterflies.

Methods

Because pollen from corn plants is shed over a relatively short period of time during the season, potential exposure of susceptible stages of monarch larvae to pollen depends on synchrony of their development with that of corn plants. A phenological model has been developed for monarch butterflies and field corn across North America (Calvin and Taylor 2000). It involves three elements that integrate 1) the first presence of spring migratory monarch butterflies in an area, 2) heat unit requirements for egg laying and larval development of monarchs and 3) corn heat unit requirements for development of corn hybrids. We sampled these elements at four locations across the Corn Belt to establish the first occurrence and phenological development of monarch populations and corn growth and corroborate the model.

In those regions of significant overlap, acreage of corn, especially Bt-corn, as a proportion of the total habitat were obtained from crop statistics and from direct ground observations. Surveys of various habitats across the Corn Belt, including other crop fields, fallow or conservation land and roadway margins, have provided data of this nature. Density and extent of milkweed stands in these habitats indicate the potential for monarch utilization of this food resource relative to milkweed populations in Bt cornfields.

Laboratory bioassays were utilized to establish LC₅₀'s for first and third instars for each transformation event pollen type. A no effect level was estimated for mortality and for growth effects such as weight gain and rate of development of the larvae. A field bioassay was undertaken to determine the outcome of exposure of first and third instars to pollen under field conditions on milkweed plants placed in the field.

Results

Characterization of Effects of Bt Corn Pollen

Laboratory bioassays of pollen fed to 1^st instar monarchs for 2-4 days on leaf disks or whole, detached leaves of common milkweed, A. syriaca, indicate that pollen from event 176 Bt corn causes mortality and sublethal effects, such as growth inhibition, at concentrations as low as 10 pollen grains/cm². Pollen from all other events, including Mon810 and Bt11 corn hybrids as well as events not presently grown, such as Dbt418, Cbh351 and Tc1507 (expressing Cry1Ac, Cry9C, and Cry1F proteins, respectively), did not demonstrate any lethal or sublethal effects, even at densities above 1000 pollen grains/cm² (Hellmich et al. 2001). These data were used to establish a no-observable-effect-level (NOEL) for growth inhibition of larvae for event 176 pollen and for Bt11 and Mon810 pollen.

A field bioassay was undertaken to determine the outcome of exposure of larvae under field conditions on milkweed plants growing or placed in the field. Field studies incorporated natural levels of pollen from Bt corn plants and demonstrated no acute effects of Bt11 and Mon810 corn pollen on survival or growth of monarch larvae.

Figure 13. Percent growth inhibition for monarch larvae exposed to pollen from event 176 corn hybrids and for larvae exposed to pollen from Bt11/Mon810 corn hybrids. Average cumulative proportion of pollen on milkweed leaves sampled in Iowa and Ontario during 1999-2000 is also depicted.

Click on image for larger view

Dose response curves for weight gain of neonates exposed to pollen from event 176 and Bt11/Mon810 hybrids clearly demonstrate the different level of expression of protein in the two pollen types (Fig. 13). For event 176, the estimated LC₅₀ for relative weight loss after exposure is about 10-12 pollen grains/cm² (Hellmich et al. 2001). For Bt11, Mon810 and other events, no measurable effects were evident, except when response data for concentrations above 1000 grains/cm² were pooled for analysis. The response curve for Bt11/Mon810 was based on larval reaction to event 176 pollen and a hypothetical curve with a slope similar to that for event 176 was constructed (Fig. 13).

Characterization of Exposure to Bt Corn Pollen

Exposure depends on 1) the phenological overlap between monarch populations and corn anthesis, 2) the spatial overlap between milkweeds used by monarchs and cornfields, and 3) the pollen densities encountered on leaves of milkweed plants in and near cornfields.

Phenological overlap

Pollen from corn plants within a particular field is shed over a period of 7-15 days between mid-July and mid-August during the season, while larvae develop over a more prolonged period. Potential for exposure of susceptible stages of monarch larvae to corn pollen depends on synchrony of their development with pollen shed of corn plants. Locations in four corn-growing regions were monitored for phenological development of monarch populations and anthesis (Oberhauser Bt 2001). These locations were established in Iowa, Maryland, Minnesota/Wisconsin and Ontario. Overlap of the more susceptible stages of monarchs, primarily 1^st and 2^nd instars, with pollen shed was considered for purposes of risk assessment.

The presence of susceptible larvae at the time of corn anthesis varied considerably across the regions studied (Oberhauser et al. 2001). In the more northern locations (MN/WI and ON), about 40 and 62% of the larvae overlapped with pollen shed, respectively, while in areas further south (IA and MD), about 15 and 20% of the larval stages overlapped, respectively. Data from a computer simulation of monarch phenology and corn development support the general observation that overlap increases at higher latitudes across the Corn Belt. Projected overlap from these simulations are likely over-estimated because in the model, 30-year average temperature data were used, while our in-field measurements were made during 2000 in 3-5 specific fields at each of four locations.

Spatial overlap

Density of milkweed stands in cornfields compared with non-agricultural lands and data on the proportion of the landscape in corn and non-agricultural lands provided a basis on which to determine the proportion of the milkweed population that was in cornfields (Oberhauser et al. 2001). In all locations, densities were higher in non-agricultural lands than in cornfields, but the range of difference was considerable. In Minnesota/Wisconsin and in Iowa, the density of milkweed was approximately 4-7 times greater in non-agricultural fields than in cornfields, while in Ontario the density was up to 115 times greater. In areas where corn is more intensively cultivated, as in Iowa and southern Minnesota/Wisconsin, less non-agricultural land exists and the overall proportion of milkweed on a landscape basis is higher in cornfields and other crop lands than in non-agricultural land. In regions of the corn growing area where mixed habitats are more common, such as in Maryland and Ontario, milkweeds are more abundant in the non-agricultural landscape and provide proportionately greater habitat than those in cornfields (Oberhauser Bt 2001).

The likelihood of monarch larvae feeding on milkweed plants in cornfields depends not only on what proportion of milkweeds are in cornfields, but also on the relative usage by monarchs of milkweeds in cornfields relative to milkweeds in other habitats. Observations from the four regions studied indicated that monarch butterflies locate and lay eggs on milkweeds in corn despite the canopy of the crop obscuring the milkweeds. Sampling done in cornfields and in non-agricultural land in these areas, suggests that egg densities per plant are higher in corn in Iowa and Minnesota/Wisconsin, but are the same in cornfields and non-agricultural lands in Maryland and Ontario (Oberhauser Bt 2001). In Iowa, egg densities were higher also in soybean fields than in non-agricultural areas.

Pollen densities encountered

Dispersal of corn pollen was described by Raynor et al. (1972), who demonstrated deposition of pollen as much as 60 m from field edges. Because of the rapid decline in concentration of pollen from the field margin outward, risk assessment concerns are focused on the concentration of pollen on milkweed leaves within the corn field and those leaves found outside the field up to 5m from the field edge. During the period of pollen shed, samples of pollen were collected on sticky trap surfaces and on milkweed leaves (Pleasants Bt 2001). Samples were taken at various distances within and beyond the margins of cornfields to estimate the concentration of pollen that could be encountered by monarch larvae. Data from three locations, Iowa, Maryland and Ontario (Pleasants Bt 2001) demonstrated a 5-fold reduction in concentration of pollen from just within the edge of the cornfield to about 2-3m distant. Within-field densities across the different studies averaged between 65-425 pollen grains/cm² on milkweed leaves at the peak of corn anthesis with an average of 171 grains/cm². From these data, frequency histograms were derived to determine the likelihood of encounter by 1^st or 2^nd instars of different concentrations of pollen within and outside cornfields. These frequency distributions can be used to determine the probability of a larva encountering a toxic pollen dose (Fig. 14).

Figure 14. Cumulative deposition of pollen grains on milkweed leaves within cornfields as a proportion of total samples recorded over 1999-2000 in Ontario and Iowa. Average grains/cm² for various distances from the edge of the fields are also depicted.

Bt corn represented about 25% of the total acreage of corn in Ontario in 2000. Of this, less than 1% was of hybrids containing event 176 transformations of the Cry 1A(b) protein. The remainder consisted primarily of Bt 11 and MON 810 hybrids. Monarch butterflies utilize milkweed plants growing in and around cornfields in Ontario as well as host plants in other agricultural and non-agricultural lands. Estimates of the production of monarchs produced on milkweed from non-agricultural lands exceed that from cornfields by about 50 fold. The presence of monarchs on milkweed as small larvae (1^st and 2^nd instars) during the development of the migratory generation of monarchs corresponds to a considerable degree (estimated conservatively at 48%) to that of the pollen shed period of corn. Pollen is distributed primarily in and near corn fields. Approximately 90% of the pollen falls within 5m of the edge of fields, and at doses well below 100 grains/cm². Average densities of pollen observed on milkweed leaves ranged from 50 to 150 grains/cm². In Ontario, the highest single value recorded was 819 grains/cm². Obviously, some monarch larvae will be exposed to corn pollen during their development.

Risk Characterization

Exposure of monarch larvae to event 176 hybrids represents a risk that has evidently been quite small due to the relatively small acreage of corn planted to this variety and the large proportion of the population that develops away from corn fields. Event 176 hybrids have not been submitted for re-registration, so their use will be phased out through 2003. Risk to monarchs from Bt11 and Mon810 hybrids cannot be easily determined because of the low expression of Cry 1A(b) protein in their pollen. Even assuming sublethal effects would occur at exposures above 1000 grains/cm², average densities of pollen on milkweed leaves do not exceeded 150-200 grains/cm². Less than 1% of monarch larvae present during pollen shed would be exposed to doses above 1000 grains/cm² (Sears et al. 2001).

In addition, corn represents, at most, 20% of the total landscape in North America over which monarch butterflies breed, of which less than 20% consists of Bt corn hybrids. Only 25% of the larvae within this breeding range are in contact with pollen during the period of anthesis (Sears et al. 2001). Clearly, the risk to monarchs form current hybrids of Bt corn is exceedingly small, if not negligible.

References Cited

Altizer, S..M and K.S. Oberhauser. 1999. Effects of the protozoan parasite Ophryocystis elektroscirrha on the fitness of monarch butterflies (Danaus plexippus). J. Inv. Pathol. 74:76-88.

Elzinga, C.L., D.W. Salzer, J.W. Willoughby. 1998. Measuring and monitoring plant populations. BLM Technical Reference 1730-1.

Fast, P. G., and J. Regniere. 1984. Effect of exposure time to bacillus thuringiensis on mortality and recovery of the spruce budworm (lepidoptera: Tortricidae). Can. Entomol. 116: 123-130.

Fearing, P. L., D. Brown, D. Vlachos, M. Meghji, and L. Privalle. 1997. Quantitative analysis of CryIA(b) expression in Bt maize plants, tissues, and silage and stability of expression over successive generations. Molec. Breed. 3: 169-176.

Hanson, L. C., and J. J. Obrycki. 2000. Field deposition of Bt transgenic corn pollen: lethal effects on the monarch butterfly. Oecologia 125(2): 241-248.

Hellmich, R. 2000. Presentation at USDA Monarch Bt Workshop. Kansas City, Feb. 24-25, 2000.

Hellmich, R.L., B.D. Siegfried, M.K. Sears, D.E. Stanley-Horn, M.J. Daniles, H.R. Mattila, T. Spencer, K.G. Bidne, and L.C. Lewis. 2001. Monarch larvae sensitivity to Bacillus thuringiensis-purified proteins and pollen. Proc. Natl. Acad. Sci. USA 98:(In press).

Hilbeck, A., M. Baumgartner, P. M. Fried, and F. Bigler. 1998. Effects of transgenic Bacillus thuringiensis corn-fed prey on mortality and development time of immature Chysoperla carnea (Neuroptera: Chrysopidae). Environ. Entomol. 27(2): 480-487.

Koziel, M.G., G.L. Beland, C. Bowman, N.B. Carozzi, R. Crenshaw, L. Crossland, J. Dawson, N. Desai, M. Hill, S. Kadwell, K. Launis, K. Lewis, D. Maddox, K. McPherson, M.R. Meghji, E. Merlin, R. Rhodes, G.W. Warren, M. Wright, and S.V. Evola. 1993. Field performance of elite transgenic maize plants expressing an insecticidal protein derived from Bacillus thuringiensis. Biotechnology 11: 194-200.

Losey, J.E., Rayor, L.S. & Carter, M.E. 1999. Transgenic pollen harms monarch larvae. Nature 399, 214.

Losey, J.E. 2000. Presentation at USDA Monarch Bt Workshop. Kansas City, Feb. 24-25, 2000.

Malcolm, S.B. and L.P. Brower. 1989. Evolutionary and ecological implications of cardenolide sequestration in the monarch butterfly. Experientia 45(3): 284-294.

Malcolm, S.B., B.J. Cockrell, and L.P. Brower. 1993. In: Biology and Conservation of the Monarch Butterlfy Eds. S.B. Malcolm and L.P. Brower. Natural History Museum of Los Angeles County, Los Angeles. Pp, 253-267.

Oberhauser, K.S., M.D. Prysby, H.R. Mattila, D.E. Stanley-Horn, M.K. Sears, G.Dively, E. Olson, J.M. Pleasants, W.F. Lam, and R.L. Hellmich 2001. Temporal and spatial overlap between monarch larvae and corn pollen. Proc. Natl. Acad. Sci. USA 98:(In press).

Orr, D.B., and D.A. Landis. 1997. Oviposition of European corn borer (Lepidoptera: Pyralidae) and the impact of natural enemy populations in transgenic versus isogenic corn. J. of Econ. Entomol. 90: 905-909.

Ostlie, K.R., W.D. Hutchinson, and R.L. Hellmich. 1997. Bt corn and European corn borer. North Central Regional publication 602. University of Minnesota, St. Paul, MN.

Peacock J.W., D.F. Schweitzer, J.L. Carter and N.R. Dubois. 1998. Laboratory assessment of the effects of Bacillus thuringiensis on native Lepidoptera. Environ. Entomol. 27(2): 450-457.

Pilcher, C.D., J.J. Obrycki, M.E. Rice, and L.C. Lewis, 1997. Preimaginal development, survival, and field abundance of insect predators on transgenic Bacillus thuringiensis corn. Environ. Entomol. 20(2): 446-454.

Pleasants, J. 2000. Presentation at USDA Monarch Bt Workshop. Kansas City, Feb. 24-25, 2000.

Pleasants, J.M., R.L. Hellmich, G.P. Dively, M.K. Sears, D.E. Stanley-Horn, H.R. Mattila, J.E. Foster, T.L. Clarke, and G.D. Jones. 2001. Corn pollen deposition on milkweeds in and near cornfields. Proc. Natl. Acad. Sci. USA USA 98:(In press).

Prysby, M. and K. Oberhauser. 1999. Large scale monitoring of monarch populations. Proceedings of the North American conference on the Monarch Butterfly, pp. 379-384. I. Pisanty, K. Oberhauser, L. Merino and S. Price, editors.

Prysby, M. and K. Oberhauser. 1999. Monarch Larval Monitoring Project. www.monarchlab.umn.edu.

Raynor, G.S., E.C. Ogden, and J.V. Hayes. 1972. Dispersion and deposition of corn pollen from experimental sources. Agronomy Journal 64: 420-427.

SAS 1990. SAS/STAT User's Guide Fourth edition. SAS Institute Inc., Cary NC, USA.

Sears, M.K., R.L. Hellmich, D.E. Stanley-Horn, K.S. Oberhauser, J.M. Pleasants, H.R. Mattila, B.D. Siegfried, and G.P. Dively. 2001. Impact of Bt corn pollen on monarch butterfly populations: A risk assessment. Proc. Natl. Acad. Sci. USA 98:(In press).

Stanley-Horn, D.E., G.P. Dively, R.L. Hellmich, H.R. Mattila, M.K. Sears, R. Rose, L.C.H. Jesse, J.E. Losey, J.J. Obrycki, and L.C. Lewis. 2001. Assessing the impact of Cry1Ab-expressing corn pollen on monarch butterfly larvae in field studies. Proc. Natl. Acad. Sci. USA 98:(In press).

Taylor, O.R. and J. Shields. 2000. The summer breeding habitat of monarch butterflies in eastern North America. Unpublished report submitted to EPA. 52pp.

U.S. EPA. 1995. Bacillus thuringiensis Cry1A(b) k-endotoxin and the genetic material necessary for its production (plasmid vector pCIB4431) in corn. Pesticide Fact Sheet, Publ. No. EPA 731-F-95-004. U.S. Govt. Printing Office, Washington, DC.

U.S. EPA. 1999. Oral Statement Janet L. Anderson Ph.D. before the committee on science subcommittee on basic research U.S. House of Representatives. Office of pesticide Programs, U.S. Environmental Protection Agency. (http://www.epa.gov/pesticides/biopesticides/otherdocs/testimony-ohouse.htm)

U.S. EPA. 2000a. Biopesticide Fact Sheet: Bacillus thuringiensis Cry1Ab delta-endotoxin and the genetic material necessary for its production (plasmid vector pCIB4431) in corn [Event 176]. Office of Pesticide Programs, US Environmental Protection Agency

(http://www.epa.gov/pesticides/biopesticides/factsheets/fs006458t.htm)

U.S. EPA. 2000b. Biopesticides registration action document: preliminary risks and benefits sections, Bacillus thuringiensis plant-pesticides. P. IIC52. Office of Pesticide Programs, Biopesticides and Pollution Prevention Division, US Environmental Protection Agency http://www.epa.gov/scipoly/sap/2000/october/brad3_enviroassessment.pdf

Wassenaar, L.I. and K.A. Hobson. 1998. Natal origins of migratory monarch butterflies at wintering colonies in Mexico: New isotopic evidence. Proc. Natl. Acad. Sci. USA. 95:15436-15439.

Wraight, C.L., A.R. Zangerl, M.J. Carroll, and M.R. Berenbaum. 2000. Absence of toxicity of Bacillus thuringiensis pollen to black swallowtails under field conditions. Proc. Natl. Acad. Sci. USA 97:7700-7703.

Zalucki, M.P. and S.B. Malcolm. 1999. Plant latex and first-instar monarch larval growth and survival on three North American milkweed species. Journal of Chemical Ecology 25: 1827-1842.