Additional common milkweed would help Canada meet its share of the trinational eastern migratory monarch butterfly recovery target
Abstract
The eastern migratory monarch butterfly (Danaus plexippus) population has declined by ∼84% between 1993 and 2024. Population recovery in the Midwestern United States is limited by the availability of the monarch's main host plant for egg laying—common milkweed (Asclepias syriaca). The extent to which common milkweed availability is limiting in other breeding regions is unknown. Our objective was to determine whether Canada has enough common milkweed to support its share of the trinational eastern migratory monarch population recovery target, given ∼29 stems of common milkweed are needed to contribute one adult monarch into the fall migratory population. To meet this objective, we estimated the number of common milkweed stems in Canada using published common milkweed availability estimates by land cover type. We also estimated the size of the Canadian monarch population if the recovery target was achieved using published estimates of wintering monarch density in Mexico, fall migration survival rates, and the relative proportion of monarchs entering fall migration from Canada. We estimate that Canada currently has 484 million common milkweed stems (range: 111 million–1 billion stems) and increasing this amount by 1.61 times (i.e., by ∼295 million stems), or equivalently, by 61%, would support the recovery target.
Introduction
Each year millions of eastern monarch butterflies (Danaus plexippus) capture people's imaginations across North America as they migrate thousands of kilometers in a single generation from their breeding grounds in the United States and Canada to their wintering grounds in central Mexico, and then back again via a multi-generational migration (Diffendorfer et al. 2014; Gustafsson et al. 2015). However, long-term monitoring on the wintering grounds indicates that this population has declined by 84% from 1993/1994 to 2024/2025 (32-year period; Rendón-Salinas et al. 2025; Fig. 1). Many different stressors acting concurrently or at different times of their annual cycle may be contributing to the observed decline, for example, decreased nectar flower availability—particularly during fall migration, toxic effects of pesticides, climate change, parasites, and forest loss on the wintering grounds (Brower et al. 2012; Pleasants and Oberhauser 2013; Flockhart et al. 2015; Inamine et al. 2016; Thogmartin et al. 2017c; Saunders et al. 2019; Wilcox et al. 2019; Van Deynze et al. 2024). One often cited hypothesis is that loss of common milkweed (Asclepias syriaca), particularly from agro-ecosystems, is largely responsible (Pleasants and Oberhauser 2013; Flockhart et al. 2015; Oberhauser et al. 2017; Pleasants 2017; Thogmartin et al. 2017c; Pleasants et al. 2023). Milkweed species are the primary host plants used by monarchs for egg laying, although they can successfully breed on other plant genera (Greenstein et al. 2022). Importantly, latex sap (containing triterpenes and cardenolides) from milkweed ingested by larvae provides monarch caterpillars and adults with a chemical defense against vertebrate predators and parasites (Emon and Seiber 1985; Malcolm and Brower 1989; Malcolm 1994; Gowler et al. 2015; Agrawal 2017; Züst et al. 2019).
Fig. 1.
Approximately 130 species of native milkweed occur in North America, with 76 species occurring in the United States and Canada (Woodson 1954; Luna and Dumroese 2013). Twenty-three and 12 species are used by monarchs for breeding in the eastern United States and Canada, respectively (Woodson 1954; White 1996; Luna and Dumroese 2013; for common native species of milkweed used by monarchs in different geographies, refer to The Xerces Society 2021). Currently, the most broadly distributed and abundant milkweed species east of the Rockies in both countries is common milkweed (Woodson 1954; Malcolm et al. 1989; White 1996). For example, ∼72% of milkweed stems found along an urban to rural gradient in Chicago, Illinois, United States of America, were common milkweed (Johnston et al. 2019). However, the broad distribution and abundance of common milkweed is a relatively recent phenomenon. Genetic evidence indicates common milkweed rapidly expanded in range and abundance with clearing of forests in North America for agriculture in the 19th century (Boyle et al. 2023). Regardless, today, common milkweed attracts and supports some of the highest rates of monarch egg deposition and larval survival, respectively, relative to other milkweed species (Pocius et al. 2017, 2018a, 2018b), highlighting its current importance in the eastern migratory monarch butterfly's life-cycle.
Within the eastern monarch's current core breeding range in the Midwestern United States (i.e., Kansas, Illinois, Indiana, Iowa, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, Wisconsin), common milkweed availability is estimated to be below the amount that would support monarch population recovery (Flockhart et al. 2017; Pleasants 2017; Thogmartin et al. 2017b). However, the degree to which other breeding regions can support monarch population recovery is unknown. In the Midwestern United States, common milkweed availability is estimated to have declined by 40% between 1999 and 2014, with 98% of this loss occurring in soy and cornfields, which represent 67% of agricultural land covers in the region (Pleasants 2017; Crossley et al. 2021). This loss coincides with the introduction of the herbicide glyphosate and genetically modified glyphosate-resistant soy and corn in 1996 and 1998, with 94% and 89% adoption rates by 2015, respectively (Pleasants 2017; United States Department of Agriculture (USDA) 2023). Glyphosate-resistant soy and corn were introduced in Canada in 1997 and 1998, respectively (Beckie et al. 2014; Soltani et al. 2014a, 2014b). However, information on its use is not publicly available (Bacon et al. 2023) and long-term monitoring of common milkweed before and after the introduction of glyphosate did not occur. Moreover, different from the Midwestern United States, soy and corn comprise only 33% of all agricultural land covers in the breeding region of the monarch in eastern Canada (Supplementary Material 2: Milkweed Calculator). This difference in agricultural land cover composition and potential differences in glyphosate adoption rates suggests current and historical common milkweed availability may differ between the Midwestern United States and Canada.
In 2016, the leaders of Canada, the United States of America, and Mexico adopted a trinational eastern migratory monarch population recovery target of 6 ha (approximately six soccer fields in extent) of occupied area on the monarch wintering grounds in Mexico to limit extinction risk (Semmens et al. 2016; Trudeau et al. 2016). Population size is measured in terms of area occupied (ha) on the wintering grounds because of the challenges of counting individual butterflies given they number in the millions (Wassenaar and Hobson 1998; Stenoien et al. 2015; Flockhart et al. 2017; Thogmartin et al. 2017a; refer to CEC 2017 for methods). An important step in understanding whether current common milkweed availability in Canada can support Canada's share of the trinational recovery target is translating the proportion of monarchs in Mexico from Canada, into the number of monarchs that should be originating from Canada prior to fall migration if recovery were achieved. This in turn can be translated into the number of common milkweed stems that would support Canada's share of the monarch recovery target, given ∼29 stems of common milkweed are needed to contribute one adult monarch to the fall migratory population (Nail et al. 2015).
Our objectives were three-fold. Our first objective was to provide the first estimate of contemporary common milkweed availability east of the Rockies in Canada based on the best and most current land cover-specific stem density estimates available and current herbicide adoption rates by corn and soy farmers in Canada. We build upon the general approach taken by Thogmartin et al. (2017b) but also estimate variability in stem density estimates using the methods provided by Pleasants (2017). Our second objective was to estimate the fall migratory population size of monarchs from Canada that would support Canada's portion of the 6 ha target, based on published estimates of monarchs/ha on the wintering grounds, estimates of the proportion of monarchs entering the fall migratory population from Canada, and estimates of fall migration survival rates. Our final objective was to assess how many stems of common milkweed would support Canada's share of the eastern migratory monarch population recovery target and contrast this with our estimate of common milkweed availability.
Methods
Overview
We used common milkweed stem density measurements for different land covers specific to Canada when available (otherwise, measurements were from the Midwestern United States; Table 1), information on herbicide adoption rates across Canada, land cover maps, and an estimate of common milkweed range in Canada, to estimate the total number of common milkweed stems available in Canada. Next, we estimated the size of the fall migratory monarch population originating from Canada if the 6 ha winter recovery target was achieved. This estimate was calculated by adjusting the proportions of monarchs originating from each province and state by region-specific survival rates for fall migration and normalizing these proportions to sum to one. This resulted in an estimate of the proportion of monarchs on the wintering grounds from each province and state. We then multiplied these normalized proportions by an estimate of the number of monarchs on the wintering grounds if the 6 ha recovery target was achieved and used the same estimated migration survival rates mentioned above to back-calculate the number of monarchs entering fall migration from each province. Last, using an estimate of number of common milkweed stems needed to contribute one monarch to the fall migratory population, we estimated the number of common milkweed stems within the common milkweed range in Canada that would support the monarch recovery target and compared this estimate with our estimate of common milkweed stem availability.
Natural history of common milkweed
Common milkweed is a native sun-tolerant broad-leafed perennial plant from the family Asclepiadaceae currently found in open habitats throughout the north Central and Northeastern United States and in southern Canada from Manitoba to the Maritimes (Bhowmik and Bandeen 1976; Bhowmik 1978, 1994). Genetic evidence indicates common milkweed expanded in range following the retreat of the glaciers 10–20 thousand years ago and more recently, owing to clearing of forests in northeast North America for agriculture in the 19th century (Boyle et al. 2023). Common milkweed grows to ∼1 m in height and generally has seven or fewer inflorescences, each of which can support more than 100 purplish-pink flowers (Bhowmik and Bandeen 1976; Bowhmik 1994). Common milkweed is self-incompatible and is predominantly pollinated by wasps and bees; it spreads through wind dispersal of its seeds and vegetatively through root buds on lateral roots (Bhowmik and Bandeen 1976; Bhowmik 1994). It is tolerant to a variety of soil and climatic conditions but grows best in sites with well-drained loamy soils, 30% to full sunlight, July growing season temperatures of 18 ⁰C in its northern range, and at least 50 cm of precipitation from June–August (Groh and Dore 1945; Berkman 1949; Bhowmik and Bandeen 1976; Bhowmik 1978, 1994).
Objective 1
Estimating common milkweed stem availability in Canada
Base land cover map
All mapping and processing were done using ArcGIS Desktop 10.4.0 (ESRI 2021). A detailed description of the processing and assembly of our base land cover map can be found in Supplementary Materials (Supplementary Material 1: S2. Geo-spatial processing), including hyperlinks to all original mapping products used. Briefly, base land cover data for crop fields and forages (including grasslands, hay, and pasture) were from the Agriculture and Agri-food Canada (AAFC) Annual Crop Inventory Map (AAFC 2015) and all other base land covers, including developed land cover, were from the Commission for Environmental Cooperation's (CEC) North American Land Cover Map (CEC 2015). Both raster maps have a 30 m spatial resolution. Some larger roads in both maps were captured as “developed” land cover; however, smaller roads were readily missed given pixel resolution. Therefore, for both maps, roads were erased prior to merging using a shrink and expand operation. Overlay masks were then created and added for roads and their margins, as well as transmission corridors, rail rights-of-way and their margins, watercourses, the edges of water bodies and wetlands, and field margins (Supplementary Material 1: S2. Geo-spatial processing: Overlay preparation). Overlay masks allowed us to partition land cover within a single raster cell into multiple land cover types and estimate common milkweed stem availability for each land cover type accordingly. For example, two-lane roads were assumed to have a width of 12 m with 3 m of road margin on either side. Thus, if a road grid cell occurred over a land cover grid cell defined as crop, then the cell would be portioned into an area of road ditch of 3 × 2 × 30 = 180 m2, an area of road surface of 12 × 30 = 360 m2, and a remaining area of crop of 900 – 180 – 360 = 360 m2. Each of these areas would be multiplied by the appropriate common milkweed stem density (described below) to estimate the density for the full cell. Overlays for rivers, lakes, and watercourses were created from Natural Resources Canada's (NRCan) Lakes, Rivers, and Glaciers in Canada–CanVec Series–Hydrographic Features map (NRCan 2017a). Overlays for transmission line rights-of-way were created from NRCan's Mines, Energy and Communication Networks in Canada–CanVec Series–Resources Management Features map (NRCan 2017b). Overlays for roads were created from Statistics Canada's (StatCan) National Road Network–NRN–GeoBase Series Map (StatCan 2018a). Railway rights-of-way overlays were created from NRCan's National Railway Network–NRWN–GeoBase Series map (NRCan 2016). Wetland edge overlays were developed using the outer pixels of wetlands mapped in the CEC Land Cover of North America Map (CEC 2015). Last, to create an overlay for field margins, we applied an edge detector to the AAFC crop inventory data and accepted an edge as a margin if it occurred between two different agricultural land cover types, i.e., our measurement of edge does not account for field edges between two fields containing the same land cover type. However, this underestimate is partly balanced by the fact that our measure of edge also assumes a field margin exists where two different crops are planted side by side even if they are directly adjacent with no field margin between them. We posit that any overestimation or underestimation of field margin amount is unlikely to influence the general conclusions of our study because even if we underestimated common milkweed availability in field margins by a factor of two, field margins would still only support 2% of our total estimated common milkweed stem availability (Table 2). All mapping products were re-projected to the Albers projection and resampled to a 30 m spatial resolution to match the resolution of the AAFC and CEC land cover maps described above.
Common milkweed stem availability by land cover type
We used a species distribution model from Drapeau Picard et al. (2024) to define the geographic extent over which common milkweed can be found in Canada. We could not use this model to estimate common milkweed availability because it is based on presence-only data and does not account for abundance. Instead, we used common milkweed stem densities per land cover type from the literature (described below; Table 1), building off Thogmartin et al. (2017b). We refer to both the summation of stem densities per land cover type and across land cover types as stem availability.
Table 1.
| Land cover type | Mean stems/ha | SE | Lower 95% CI | Upper 95% CI |
|---|---|---|---|---|
| Corn: (≤70% herbicide adoption rate)1,2,3 | 16.88 | 3.21* | 10.58 | 23.18 |
| Corn: (>70% herbicide adoption rate)4,5,6 | 0.12 | 0.08 | 0 | 0.29 |
| Soy: (≤70% herbicide adoption rate)1,2,3 | 12.47 | 3.71* | 5.19 | 19.75 |
| Soy: (>70% herbicide adoption rate)4,5,6 | 0.12 | 0.08 | 0 | 0.29 |
| Other crops (≤70% herbicide adoption rate)1,2,3 | 72.59 | 30.14* | 13.51 | 131.67 |
| Other crops: (>70% herbicide adoption rate)4,5,7 | 7.64 | 7.27 | 0 | 21.89 |
| Forages1,2,3 | 12.53 | 2.17* | 8.28 | 16.78 |
| Field margins1,2,3 | 63.59 | 4.25* | 55.27 | 71.92 |
| Transmission line (outside developed)5,6,7,8 | 7.64 | 4.32* | 0 | 16.11 |
| Transmission line (inside developed)9 | 1.71 | 0.94 | 0 | 3.55 |
| Road margins (paved; outside developed)3,6,10 | 105.79 | 12.28* | 81.72 | 129.86 |
| Road margins (unpaved)4,5,7 | 48.77 | 27.61 | 0 | 102.89 |
| Railway margins (outside developed)5,6,7,8 | 7.64 | 4.32* | 0 | 16.11 |
| Railway margins (inside developed)9 | 1.71 | 0.94 | 0 | 3.55 |
| Watercourses4,5,7 | 151.59 | 207.39 | 0 | 558.07 |
| Waterbody edge4,5,7 | 151.59 | 207.39 | 0 | 558.07 |
| Wetland edge4,5,7 | 151.59 | 207.39 | 0 | 558.07 |
| Shrubland1,2,3,8 | 72.59 | 30.14* | 13.51 | 131.67 |
| Developed9 | 2.05 | 1.13 | 0 | 4.27 |
Note: Standard errors (SEs) used in CI estimates were either extracted directly from the literature or were calculated using the methods provided in Pleasants (2017); the latter are indicated with * in the SE column (for data used in calculations, refer to Supplementary Material 1: S1. Milkweed stem availability estimates). Numerical superscripts indicate sources of variables used for calculations.
1
Fahrig et al. (2015) for number of sites occupied and area occupied.
2
Martin et al. (2020) for number of sites occupied and area occupied.
3
Pitman et al. (2018) for stem densities for occupied sites.
4
Pleasants (2017) for stem availabilities for land cover types and variability or derivation of SEs.
5
Hartzler and Buhler (2000) for number of sites occupied and areas occupied.
6
Hartzler (2010) for number of sites occupied and areas occupied.
7
Flockhart et al. (2015) for stem densities for occupied sites.
8
Thogmartin et al. (2017b) for stem densities of land cover types and/or methods.
9
Johnston et al. (2019) for stem availabilities and variability.
10
Kasten et al. (2016) for number of sites occupied.
For agricultural land covers, we estimated stem availability for soy, corn, other crops, and forages individually following Thogmartin et al. (2017b) using data from both Canada and the Midwestern United States (Table 1). Common milkweed stem density estimates used by Thogmartin et al. (2017b) were originally measured in the Midwestern United States, where adoption rates for glyphosate-tolerant soy and corn were 94% and 89% by 2015, respectively (USDA 2023; refer to Pleasants 2017). However, fine-scale spatial information on glyphosate use in Canada is not publicly available (Bacon et al. 2023). Therefore, we used herbicide adoption rates at the county level as a proxy for glyphosate adoption rates for corn, soy, and other crops within the range of common milkweed in Canada. We felt this was a reasonable assumption given that glyphosate represents the primary active ingredient sold across all pesticide types in Canada within the agricultural sector by an order of magnitude (∼tens of millions of kg of active ingredients compared with millions for each of the other top five herbicides; Pest Management Regulatory Agency (PMRA) 2021). In eastern Ontario, where the only empirical measurements of proportion of sites occupied and area occupied by common milkweed in agricultural fields are readily available for Canada (described below), herbicide adoption rates range from 41% to 70% (StatCan 2016). In contrast, for much of southwestern Ontario and parts of southern Manitoba, herbicide adoption rates range from 71% to 90% (StatCan 2016). Therefore, for counties (refer to StatCan 2018b) in Canada where herbicide adoption rates are >70%, we used the Midwestern United States stem availability estimates for soy, corn, and other crops land cover categories from Pleasants (2017). For all other census divisions, we used stem availability estimates derived from Fahrig et al. (2015) and Martin et al. (2020) from eastern Ontario (refer to Table 1 for sources of land cover-specific common milkweed stem availability estimates). Given stem availability estimates are not available for field margins in the Midwestern United States where glyphosate use is high or in Canada where herbicide adoption rates are high, we used the same field margin estimates from eastern Ontario regardless of herbicide adoption rates or glyphosate use.
Stem availability estimates for agricultural land covers from eastern Ontario, including field margins, were derived using data on the proportion of sites occupied and area occupied by common milkweed from surveys conducted in 93 agricultural landscapes across 2011 and 2012 (Fahrig et al. 2015; Martin et al. 2020). Briefly, surveys were conducted at four sites per landscape; two surveys occurred within agricultural fields and two occurred at the field/field margin interface. Surveys were also conducted twice per site, once between 24 May and 9 July, and once between 17 July and 30 August. For this study, we only used surveys from the second sampling period because we were interested in understanding how much common milkweed is available to produce the final monarch generation that migrates to Mexico. We assumed that common milkweed availability during the second survey would be a more accurate representation of this because it integrates what is left over from the first survey period with new growth that occurred following the initial survey. Prior to the calculation of any summary statistics based on these data, we converted m2 occupied by common milkweed per 100 m2 to m2 occupied per ha by dividing by 0.01. To derive stem densities (and standard errors (SEs) for land covers from these data (Supplementary Material 1: S1. Milkweed stem availability estimates: Table S1) we first calculated median area infested by common milkweed using a random selection of half the samples. We estimated the median rather than the mean area because measurements of area infested were highly right-skewed across sites. We then used a bootstrap procedure by repeating our median calculation a total 10 000 times with replacement and used the resulting distribution to calculate the mean (±standard deviation (SD)) of the median area occupied across sampling sites.
For all agricultural land covers in counties where herbicide adoption rates are ≤70%, we converted area occupied by common milkweed (m2) per ha to stem availability per hectare using two different measurements of stem densities (stems/m2) measured in Southwestern Ontario from Pitman et al. (2018) because there are no stem density estimates available for eastern Ontario. Specifically, for soy, corn, and other crops, we used a mean estimate of 3.60 stems/m2 (SD = 7.30) measured directly in soy and cornfields (Supplementary Material 1: S1. Milkweed stem availability estimates: Table S1). For field margins and forages, we used a mean estimate of 1.60 stems/m2 (SD = 1.10) measured in restored grasslands/private properties.
For all non-agricultural land covers, we used Canadian-specific common milkweed data when available; otherwise, we used the most current published estimates from the Midwestern United States (Table 1). We assumed that common milkweed availability was zero for water, exposed/barren, forest, wetland interior, road surface, and railway track land covers. To our knowledge, no estimates of common milkweed availability within forests have been published; however, densities are predicted to be very low (Groh and Dore 1945; Bhowmik and Bandeen 1976; Bhowmik 1994).
Developed land cover in our basemap came from the CEC's North American Land Cover Map (CEC 2015) and was defined by pixels containing ≥30% human constructed materials. No common milkweed availability estimates are available for developed land cover in Canada; therefore, we used stem availability estimates from Chicago, Illinois, United States of America (Johnston et al. 2019), where common milkweed availability was measured across 14 sub-classes of developed land cover, including small-scale industrial, large-scale industrial, single family residential, common space/multi-family residential, corporate/medical, commercial, community/cultural, restricted use rights-of-way (margins), vacant lots, transitional/restricted use, minor road (margins), major rights-of-way (margins), conservation open space, and non-conservation open space land cover categories. Given that map products for developed land cover are not available at the same resolution as Johnston et al. (2019) for most developed areas in Canada, we produced a single common milkweed availability estimate for developed land cover by averaging the mean stem availability estimates and SEs for 11 of the 14 land cover types detailed above. We did not include mean stem availability estimates and SEs from conservation and non-conservation open space land cover categories form Johnston et al. (2019) in our mean because these lands in Canada were captured by other land cover categories in our basemap, i.e., watercourses, waterbody edges, wetlands and wetland edges, shrublands, forest, grassland, or open barren. We also did not use stem availability estimates for major rights-of-way (margins) from Johnston et al. (2019) in our mean because we assumed no common milkweed stems were adjacent major roads within developed areas and because we estimated common milkweed stem availability independently for railway margins and transmission corridors running through developed areas (described below). We acknowledge averaging across developed land cover types may overestimate or underestimate stem availability depending on the relative proportions of land covers present in developed areas in Canada. To assess this possibility, we compared the mean stems per acre from Johnston et al. (2019) for Chicago across the 14 land cover types listed above with an estimate of mean stem availability adjusted for land cover availability for Chicago. We found these estimates were very similar (0.94 and 0.98 stems/acre, respectively), indicating the averaging of stem availability across developed land cover types is unlikely to have introduced a strong bias into our results. We converted stems per acre to stems per hectare by dividing by 0.405.
We estimated common milkweed stem availability separately for paved and unpaved road margins outside of developed areas (Table 1). For paved road margins outside developed areas, number of sites occupied was based on data from roadside sites in Minnesota, Iowa, Wisconsin, and South Dakota (Kasten et al. 2016), area occupied was from sites in Iowa (Hartzler 2010), and stem densities within common milkweed patches were based on measurements from roadside sites in Southwestern Ontario (Pitman et al. 2018). We note that the National Road Network Database we used (StatCan 2018; Supplementary Materials: S2. Geo-spatial processing) classified roads as primary, secondary, or unpaved. Following Thogmartin et al. (2017b), we combined primary and secondary road types into a single land cover type–paved roads. For unpaved road margins, we extracted mean stem availabilities and SEs from Pleasants (2017), which were based on occupancy rates and area occupied from the “other” category in Hartzler and Buhler (2000) and an estimate of stems per common milkweed patch from Flockhart et al. (2015).
We estimated stem availability for rail rights-of-way and transmission rights-of-way separately depending on whether the corridors were found inside or outside of developed areas (Table 1). For rail and transmission rights-of-way outside developed areas, we followed Thogmartin et al. (2017b) and used their “low other crop” category for number of sites occupied and area occupied, which was from the “other” category in Hartzler and Buhler (2000) that contained surveys in rights-of-way outside urban areas. For number of stems per common milkweed patch for rail and transmission rights-of-way outside developed areas, we again used the value reported by Flockhart et al. (2015), as per Thogmartin et al. (2017b). For corridors within developed areas, we used the mean estimate and SE reported for the major right-of-way category from Johnston et al. (2019). We estimated stem availability for waterbody edges, wetland edges, and watercourses using the Conservation Reserve Program category from Pleasants (2017), which was based on Hartzler and Buhler (2000) values for waterways and terraces. For shrublands, we followed Thogmartin et al. (2017b) and used the same value estimated for our “other crops” category.
We derived SEs and confidence intervals (CIs) for all land cover-specific stem density estimates following the methods of Pleasants (2017) (Supplementary Material 1: S1. Milkweed stem availability estimates: Table S1), except for land cover types from Johnston et al. (2019) and Pleasants (2017) who provided SEs for common milkweed density estimates. If the lower CI we derived for a stem density estimate was negative, we set the lower 95% interval to zero (Table 2).
Table 2.
| Land cover type | % of total land cover with milkweed | Stem availability based on mean density estimates (106) | % of total | Stem availability based on lower 95% CIs for density estimates (106) | % of total | Stem availability based on upper 95% CIs for density estimates (106) | % of total |
|---|---|---|---|---|---|---|---|
| Other crops | 25% | 188 | 39% | 33 | 30% | 350 | 33% |
| Shrubland | 8% | 83 | 17% | 16 | 14% | 151 | 14% |
| Wetland edge | 4% | 76 | 16% | 0 | 0% | 279 | 27% |
| Forages | 28% | 51 | 11% | 34 | 31% | 69 | 7% |
| Waterbody edge | 1% | 24 | 5% | 0 | 0% | 88 | 8% |
| Corn | 11% | 14 | 3% | 9 | 8% | 20 | 2% |
| Soy | 15% | 14 | 3% | 6 | 5% | 22 | 2% |
| Road margins (paved; outside developed) | 1% | 12 | 2% | 9 | 8% | 14 | 1% |
| Watercourses | <1% | 8 | 2% | 0 | 0% | 31 | 3% |
| Road margins (unpaved) | 1% | 6 | 1% | 0 | 0% | 12 | 1% |
| Field margins | 1% | 5 | 1% | 5 | 4% | 6 | 1% |
| Developed | 5% | 2 | <1% | 0 | 0% | 3 | <1% |
| Transmission lines (outside developed) | <1% | <1 | <1% | 0 | 0% | 1 | <1% |
| Railway margins (outside developed) | <1% | <1 | <1% | 0 | 0% | <1 | <1% |
| Transmission lines (inside developed) | <1% | <1 | <1% | 0 | 0% | <1 | <1% |
| Railway margins (inside developed) | <1% | <1 | <1% | 0 | 0% | <1 | <1% |
Table 3.
| Stem availability | Number of stems to add | ||||
|---|---|---|---|---|---|
| Sum of mean estimates | Lower bound | Upper bound | to sum of mean estimates | to lower bound | to upper bound |
| 483 746 628 | 111 078 321 | 1 046 355 433 | 294 773 568 (61%) | 667 441 875 (601%) | – |
Note: Sum of mean estimates, lower bound, and upper bound estimates are sums of stem availabilities and their 95% confidence intervals (CIs) across land cover types within the common milkweed range in Canada (Supplementary Material 2: Milkweed Calculator). Number of stems to be added is the difference between our estimate of current stem availability and the estimated number of common milkweed stems that would be needed in Canada to support the eastern migratory monarch population recovery target. Numbers in parentheses represent the proportional increase in stems needed. A proportional increase of 1.61 times is equivalent to an increase of 61%. A proportional increase of 7.01 times is equivalent to an increase of 601%. The dash indicates no additional stems would be needed.
Extrapolation of common milkweed stem density estimates by land cover type across the common milkweed range in Canada to derive national-level stem availability estimates
We estimated common milkweed stem availability in Canada by summing, across land cover types, the product of (i) area (in hectares) of a given land cover type within the estimated extent of common milkweed occurrence, and (ii) the mean estimate of stems per hectare for that land cover type (Table 2; Supplementary Material 2: Milkweed Calculator). We also repeated this procedure for our lower and upper 95% CI estimates per land cover type and report these as minimum and maximum range values for our mean common milkweed availability estimate.
Objective 2
Estimating pre-migratory population size of monarchs in Canada that would support the eastern migratory monarch population recovery target
Proportion of monarchs in Mexico from each province and state
Estimates of the relative proportion of monarchs entering fall migration from each province and state were derived from Momeni-Dehaghi et al. (2021). Momeni-Dehaghi et al. (2021) used community science observations of monarchs in the fall, accounting for sampling effort (i.e., number of observers) to estimate relative monarch abundances. Note that results from Momeni-Dehaghi et al. (2021) are qualitatively similar to those estimated independently using stable isotopes (Flockhart et al. 2017). Estimates of fall migration survival rate were from Oberhauser et al. (2017), where experts (n = 5) estimated long-term mean rates of migration survival and associated uncertainty and buffered these estimates by tuning a full annual cycle matrix population model to a time series of winter population sizes in Mexico and the estimated total amount of egg production in the Midwestern United States. Oberhauser et al. (2017) delineated the monarch breeding range into three regions: south, north central, and northeast (refer to fig. 1 from Oberhauser et al. 2017) to account for expected regional differences in vital rates. We assigned each province and state to one of these three breeding regions (Supplementary Material 1: S3. Assignment of states and provinces to monarch breeding regions: Table S5). Estimates of region-specific fall migration survival rates (mean ± 95% CI) were 76% ± 65%–85% for the south region, 70% ± 60%–78% for the north central region, and 5% ± 4%–6% for the northeast region. The survival rates for the north central region represent the survival rates for the 4th generation migrating to Mexico, while survival rates for the south represent survival rates for the 3rd and 4th generations migrating to Mexico. The survival rate for the northeast region represents the survival rate for the 3rd generation migrating from the northeast to Mexico, because it was assumed by Oberhauser et al. (2017) that there is no fourth generation in the northeast (refer also to Miller et al. 2011). A small proportion of monarchs from Canada may stop to produce a fifth generation in the fall in the Southern United States; however, this proportion is considered negligible and immaterial to the overwintering population (Oberhauser et al. 2017).
To estimate the proportion of monarchs from each breeding region present on the wintering grounds in Mexico, we divided the proportion of monarchs reaching Mexico from each province in the north central, and northeast breeding regions as well as a single regional estimate for the south region by the sum of the proportions reaching Mexico (Fig. 2). In other words, we normalized mortality-corrected proportions of monarchs that reach Mexico to sum to one.
Fig. 2.
Number of monarchs on the breeding grounds in Canada prior to fall migration that would achieve the eastern migratory monarch population recovery target
Once we estimated the proportion of monarchs on the wintering grounds that originated from Canada and the United States using the pre-migration monarch distribution estimates detailed above, we back calculated the population size of monarchs entering fall migration from Canada (Fig. 2). To do this, we again used the mean estimated survival rates provided in Oberhauser et al. (2017), as well as an estimate of the number of monarchs on the wintering grounds: 21.1 million monarchs per hectare (Thogmartin et al. 2017a). If the recovery target of 6 ha is achieved, the estimated total population size in Mexico would be 126.6 million eastern migratory monarchs. We then multiplied this estimate by our estimate of proportion of monarchs on the wintering grounds originating from the north central and northeast regions in Canada to estimate the number of monarchs on the wintering grounds from both regions in Canada. Last, we divided these estimates for each region by the region-specific survival rates to estimate the number of monarchs departing Canada in the fall, in an average year, if the recovery target was achieved and assuming future monarch survival rates remain similar to contemporary estimates.
Example calculation for number of monarchs entering fall migration that would support Ontario, Canada's share of the eastern migratory monarch population recovery target
Momeni-Dehaghi et al. (2021) estimated that 5.23% of the eastern monarch butterfly fall migratory population is from Ontario, Canada. Oberhauser et al. (2017) estimated the fall migration survival rate for monarchs from the north central region of the breeding range, in which Ontario is situated, to be 0.697. The product of these two values, 3.64%, represents the percentage of the total original fall migratory population reaching Mexico from Ontario after accounting for fall migration mortality (Fig. 2; Supplementary Material 2: Milkweed Calculator). Applying the same logic to all states and provinces, we estimate that 62.20% of the total original fall migratory population east of the Rockies in the United States and Canada reaches Mexico. Dividing the percentage of the fall migratory population reaching Mexico from Ontario (3.64%) by the total percentage of the fall migratory population reaching Mexico (62.20%) results in an estimate of the total percentage of monarchs in Mexico from Ontario (5.86%). Thogmartin et al. (2017a) estimated that there are 21.1 million monarchs per ha on the wintering grounds. Multiplying this value by the recovery target of 6 ha results in a recovered population estimate of 126.6 million monarchs. We then estimate that 7 413 602 monarchs on the wintering grounds would be from Ontario if the recovery target were achieved, by multiplying the percentage of monarchs on the wintering grounds from Ontario (5.86%) by the estimate of the population size under recovery (126.6 million monarchs). Last, we estimate that 10 636 445 fall migratory monarchs would support Ontario's share of the trinational monarch population recovery target, by dividing our estimate of the number of monarchs on the wintering grounds in Mexico from Ontario (7 413 602 monarchs) by the fall migration survival rate (0.697).
Objective 3
Estimating how many stems of common milkweed in Canada would support Canada's share of the eastern migratory monarch population recovery target
To assess how many stems of common milkweed would be needed to produce the final generation of migratory monarchs in Canada that would support the 6 ha recovery target, we multiplied our estimate of the pre-migratory population size of monarchs in Canada that would achieve the recovery target by 29, which is the estimated number of common milkweed stems that would contribute one monarch to the fall migratory population in the North Central portion of the monarch's breeding range from 7 July onwards (Nail et al. 2015). We then compared these estimates with our mean stem estimates. Last, we divided the number of stems needed by our estimate of the number of stems available to determine the proportional increase in common milkweed stems that would support Canada's portion of the recovery target. We note that monarchs, on average, begin to arrive in Canada in early June (Howard and Davis 2004). We assumed that monarchs breeding prior to the production of the final migratory generation have minimal effects on common milkweed availability given high rates of egg and caterpillar mortality (Nail et al. 2015) and that a single mature common milkweed plant can sustain more than one monarch (Mitchell, personal observation). Lower and upper range limits of common milkweed stem availability reported below represent sums of lower and upper 95% stem availability CIs across landcover types, respectively.
Results
Land cover composition within the range of common milkweed in Canada
The species distribution model we used for common milkweed in Canada from Drapeau Picard et al. (2024) encompasses a terrestrial area of 43 million hectares, spanning from Manitoba in the west to Nova Scotia in the east (Fig. 3). In terms of land covers supporting common milkweed relative to all land covers, agriculture (i.e., forages/grasslands, corn, soy, other crops, and field margins) comprises 22%, shrubland comprises 2%, developed land comprises 2%, riparian areas (including wetland edges) comprise 1%, and rights-of-ways (i.e., paved road margins outside developed areas and unpaved road margins, transmission line corridors inside and outside of developed areas, and railway margins inside and outside of developed areas) comprise 1% (Supplementary Material 2: Milkweed Calculator). In terms of land covers that do not support common milkweed, forest comprises 53%, water comprises 17%, and barren lands and wetland interiors each comprise 1%.
Fig. 3.
Common milkweed stem availability
Based on mean stem density estimates per land cover, we estimate Canada currently supports 483 746 628 (range: 111 078 321–1 046 355 433; Table 3) common milkweed stems. Fifty-six percent of these stems are from agricultural land covers (i.e., forages/grasslands, corn, soy, other crops, and field margins; Table 2), of which 69% is from the “other crops” category (e.g., barley, millet, rye, oats, wheat, sunflower, potato), 19% from forages/grasslands (including natural grasslands), 5% from corn, 5% from soy, and 2% from field margins. Forty percent of common milkweed stems are from more natural land covers, of which 44% and 40% are from shrubland and wetland edges, respectively. Our estimates indicate rights-of-way and developed land cover currently support 4% and <1% of common milkweed stems in Canada, respectively.
Estimate of proportion of monarchs on the wintering grounds originating from Canada
Momeni-Dehaghi et al. (2021) estimated that 17.23% of the North American eastern migratory fall monarch population originates from Canada; 13.19% of monarchs originate in provinces that have common milkweed and the remaining 4.04% originate from Saskatchewan and Alberta, which do not have common milkweed, but which support other species of milkweed. After accounting for estimates of fall migration survival rates from Oberhauser et al. (2017) for monarchs migrating from different breeding regions, we estimate 9.88% of monarchs on the wintering grounds, on average, originate from provinces in Canada where common milkweed occurs (Manitoba to Nova Scotia; Supplementary Material 2: Milkweed Calculator).
Estimate of monarch population size originating from Canada if the eastern migratory monarch population recovery target were achieved and number of common milkweed stems that would support this population
Given our estimate that 9.88% of the wintering population in Mexico is from provinces in Canada where common milkweed occurs (Manitoba to Nova Scotia), we estimate 26 845 524 monarchs would originate from Canada if the 6 ha recovery target were achieved. Given that approximately 29 stems of common milkweed are needed to contribute one adult into the fall migratory population (Nail et al. 2015), we further estimate that 778 520 196 common milkweed stems would support a fall migratory monarch population of this size (Table 3). Our current mean estimate of common milkweed availability in Canada is 483 746 628 (range: 111 078 321–1 046 355 433) stems. This indicates that increasing common milkweed stem availability in Canada by 1.61 times (range: 0.74–7.01 times), or equivalently by 61%, would support Canada's share of the 6 ha recovery target.
Discussion
We estimate that increasing common milkweed stem availability in Canada by 1.61 times (i.e., by 61%) would support Canada's share of the 6 ha trinational recovery target. We acknowledge substantial variability surrounding this estimate; however, we posit it is highly unlikely current common milkweed availability is sufficient given Canada's adoption of glyphosate tolerant corn and soy, and multiple studies indicating glyphosate-driven common milkweed declines are a major driver of eastern migratory monarch population decline (Pleasants and Oberhauser 2013; Flockhart et al. 2015; Oberhauser et al. 2017; Pleasants 2017; Thogmartin et al. 2017c; Pleasants et al. 2023). We also note our suggested increase in stem availability is very similar to an increase of 1.64 times needed to replace common milkweed stems lost following the introduction of glyphosate in the Midwestern United States (Pleasants 2017). Together, evidence therefore strongly indicates increasing common milkweed abundance in Canada would help meet Canada's share of the current trinational eastern migratory monarch population recovery target.
Agricultural lands currently play a major role in supporting common milkweed availability in Canada. Specifically, our results indicate that agricultural land covers support 56% of common milkweed stems (Table 2). This percentage is perhaps not surprising given agricultural land covers, including grasslands/forages, corn, soy, other crops, and field margins comprise 22% of all landcovers within the range of common milkweed (Supplementary Material 2: Milkweed Calculator), and 80% of land covers with common milkweed. However, we expect this contribution to decrease in the future given increasing row crop amounts, associated herbicide use, increases in farm size at the expense of field margins, and annual decreases in amounts of agricultural land cover in Canada (Stanton et al. 2018; Liu et al. 2020; Malaj et al. 2020; Martin et al. 2021). Therefore, diversifying crop production, maintaining smaller field sizes, and minimizing herbicide use where possible would help increase common milkweed availability for monarchs in agroecosystems (refer to Martin et al. 2021). Our results also highlight the importance of more natural land covers, including shrublands, wetland edges, water body edges, and watercourses, which comprise 13% of total land cover with common milkweed, but support an estimate of 40% of Canada's common milkweed stems. Of note, wetland edges, which despite representing 4% of land covers with common milkweed in Canada, are estimated to support 16% of common milkweed stems. Given wetland land cover is generally decreasing in Canada (Fluet-Chouinard et al. 2023), protecting remaining wetlands in southern Canada would help maintain breeding habitat for the monarch butterfly.
Conservation planning to restore common milkweed warrants not only consideration of common milkweed availability across different land cover types, but also factors that may influence the likelihood of adoption (e.g., Thogmartin et al. 2017b; Johnston et al. 2019). Knowledge of adoption rates allows more realistic restoration scenarios for achieving conservation goals. While beyond the scope of this study, research on potential adoption rates in Canada by land cover type would help identify stakeholders, rightsholders, and conservation practitioners with which to engage for restoration and to identify factors that promote or hinder adoption. For example, McPherson and Buckingham (2023) conducted interviews with representatives from the agricultural sector in Ontario, Canada. Their interviews indicated that providing financial incentives would be key to establishing or restoring monarch habitat on agricultural lands. One potential is for conservation easements to be established that provide tax benefits for agricultural landowners (Thogmartin et al. 2017b). Given the extent of agricultural land that could support common milkweed in Canada, even minimal uptake could substantially increase common milkweed availability.
Our results indicate monarchs leaving Canada and arriving on the wintering grounds are predominantly from the north central region (Supplementary Material 2: Milkweed Calculator). This is because estimates of fall migration survival rates from Oberhauser et al. (2017) for the northeast are very low. An important research avenue would be to obtain empirical migration survival rate estimates for monarchs originating from different regions to validate the estimates based on expert opinion from Oberhauser et al. (2017). This is particularly important for our analytical approach, because survival rate estimates can change the restoration target. For example, if it were found that the migration survival rates for monarchs from the north central and northeast regions of Canada were half that estimated by Oberhauser et al. (2017), then the estimated increase in common milkweed stems in Canada needed to support the trinational eastern migratory monarch population recovery target would increase from 1.61 times to 1.73 times. This hypothetical example also indicates that our restoration target may be conservative. More accurate migration survival rate estimates might be achieved using existing mark–recapture data from tagging programs such as Monarch Watch (e.g., Taylor et al. 2020). Alternatively, tracking technology, such as radio transmitters, is becoming smaller and could be used in combination with automated detection arrays in a mark–recapture framework to estimate survival rates (Evans et al. 2020; Green II 2023). For example, Fisher et al. (2020) recently used 220 mg VHF transmitters to track monarch movement. Importantly, to assess potential bias in their data, they compared the behavior and flight ability of sham tagged and untagged monarchs and found no differences. Last, differences between pre-migratory distribution estimates (Momeni-Dehaghi et al. 2021) and natal origin estimates via isotopes (Wassenaar and Hobson 1998; Flockhart et al. 2017) might provide region-specific fall migration survival estimates. Ultimately, a better understanding of region-specific fall migration survival rates would not only help managers refine common milkweed recovery target estimates, but also help prioritize geographies for restoration activities.
Inconsistency in data reporting from common milkweed studies in the literature makes interpreting and comparing results challenging. Our estimation of land cover-specific common milkweed stem availabilities required three pieces of information: (i) proportion of surveyed sites containing common milkweed, (ii) measurements of area occupied by common milkweed at each survey site or milkweed patch area, and (iii) common milkweed stem density within a patch. We found that many studies do not report all three values and so we had to combine estimates from different studies to derive our stem densities (Table 1). To improve accuracy and precision of stem density estimates, recording and measuring all three values as well as reporting sample sizes for each variable would be beneficial in future research and monitoring efforts.
The milkweed calculator we developed for our analysis can easily be updated to accommodate new data (Supplementary Material 2: Milkweed Calculator). For example, Canada has developed Mission Monarch Expert, a “sister” milkweed monitoring program to the Integrated Monarch Monitoring Program developed in the United States (refer to Cariveau et al. 2019), which is designed to provide detailed milkweed availability information across relevant land cover types. Data from this monitoring program can be used to update stem availability estimates and adjust milkweed restoration targets. As a hypothetical example, if it were found common milkweed availability is 50% lower than our current estimates for corn, soy, and other crop fields in counties where the herbicide adoption rate is below 70%, our calculator indicates increasing the number of common milkweed stems by 2.05 times (i.e., 27% more than our current estimate) would be needed in Canada. Our milkweed calculator can also be used to evaluate milkweed stem increases needed to meet alternative eastern migratory monarch population recovery targets. For example, if the trinational eastern migratory monarch population recovery target was 4 ha instead of 6 ha, our data indicate an increase in common milkweed stem availability of 1.07 times or 7% would support Canada's share of this alternative target. Given an increase would still be needed at this lower hypothetical monarch population recovery target, we suggest this is further evidence that increasing common milkweed stem availability in Canada would help promote monarch recovery. Last, our milkweed calculator can easily be updated to include new migration survival rate estimates. As illustrated above, if it were found that migration survival rate was lower for monarchs originating from Canada than that estimated by Oberhauser et al. (2017), then the number of milkweed stems supporting Canada's share of the trinational eastern migratory monarch population recovery target would increase. Therefore, we strongly recommend continually updating the recovery target for common milkweed availability in Canada as new data become available to best guide conservation efforts.
Climate change may increase Canada's proportional contribution of monarchs to the wintering grounds. We estimated that 9.88% of monarchs in Mexico originate from provinces where common milkweed occurs in Canada. We acknowledge that on a short-term annual basis, given variability in suitable climate conditions during spring and fall, this proportion can vary (Flockhart et al. 2017; Crewe et al. 2019). However, over the longer term, climate change is predicted to result in a northward shift in common milkweed and monarch distributions (Lemoine 2015), indicating Canada would increasingly contribute a larger proportion of monarchs to the wintering grounds relative to the United States. Continued monitoring of monarchs via community science efforts (e.g., Cariveau et al. 2019; Drapeau Picard et al. 2024) and isotopic assignment of breeding ranges of monarchs from the wintering grounds (e.g., Flockhart et al. 2017) can help track potential range shifts and allow each country to adjust national recovery targets through time.
In conclusion, our results indicate that increasing common milkweed availability in Canada would help support Canada's share of the trinational eastern migratory monarch population recovery target. Given the precariousness of the eastern migratory monarch population (Thogmartin 2024), expeditious actions for its recovery would be beneficial. Further research can occur simultaneously with recovery actions, to iteratively improve conservation decision-making through time for this iconic species.
Acknowledgements
We thank the Trinational Monarch Conservation Science Partnership for their feedback and input on this study. We also thank Ken Tuininga, Holly Bickerton, Renee Turza, and Krista Holmes for valuable input. We thank Janet Carter and Peter Ibsen for helpful comments on earlier versions of this manuscript as well helpful comments from two anonymous reviewers. This manuscript was financially supported by Environment and Climate Change Canada. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the United States of America or Canadian Governments.
References
Agrawal A. 2017. Monarchs and milkweed: a migrating butterfly, a poisonous plant, and their remarkable story of coevolution. Princeton University Press, Princeton, New Jersey. 296p.
Agriculture and Agri-Food Canada (AAFC). 2015. Annual crop inventory 2015 [online]: Available from Annual Crop Inventory–Open Government Portal (canada.ca) [accessed 6 March 2018].
Bacon M.H., Vandelac L., Gagnon M.A., Parent L. 2023. Poisoning regulation, research, health, and the environment: the glyphosate-based herbicides case in Canada. Toxics, 11(2): 121.
Beckie H.J., Sikkema P.H., Soltani N., Blackshaw R.E., Johnson E.N. 2014. Environmental impact of glyphosate-resistant weeds in Canada. Weed Science, 62(2): 385–392.
Berkman B. 1949. Milkweed—a war strategic material and a potential industrial crop for sub-marginal lands in the United States. Economic Botany, 3: 223–239.
Bhowmik P.C. 1978. Germination, growth and development of common milkweed. Canadian Journal of Plant Science, 58(2): 493–498.
Bhowmik P.C. 1994. Biology and control of common milkweed (Asclepias syriaca). Reviews of Weed Science, 6: 227–250.
Bhowmik P.C., Bandeen J.D. 1976. The biology of Canadian weeds: 19. Asclepias syriaca L. Canadian Journal of Plant Science, 56(3): 579–589.
Boyle J.H., Strickler S., Twyford A.D., Ricono A., Powell A., Zhang J., et al. 2023. Temporal matches between monarch butterfly and milkweed population changes over the past 25,000 years. Current Biology, 33(17): 3702–3710.
Brower L.P., Taylor O.R., Williams E.H., Slayback D.A., Zubieta R.R., Ramirez MI. 2012. Decline of monarch butterflies overwintering in Mexico: is the migratory phenomenon at risk? Insect Conservation and Diversity, 5(2): 95–100.
Cariveau A.B., Holt H.L., Ward J.P., Lukens L., Kasten K., Thieme J., et al. 2019. The integrated monarch monitoring program: from design to implementation. Frontiers in Ecology and Evolution, 7: 167.
Commission for Environmental Cooperation (CEC). 2015. 2015 Land Cover of North America at 30 meters [online]: Canada Centre for Remote Sensing/Canada Centre for Mapping and Earth Observation, Natural Resources Canada, Comisión Nacional para el Conocimiento y Uso de la Biodiversidad, Comisión Nacional Forestal, Instituto Nacional de EstadÍstica y GeografÍa, U.S. Geological Survey. Available from Land Cover 30 m, 2015 (Landsat and RapidEye)–Commission for Environmental Cooperation (cec.org) [accessed 25 February 2021].
Commission for Environmental Cooperation (CEC). 2017. Monitoring monarch butteries and their habitat across North America: inventory and monitoring protocols and data standards for monarch conservation. Commission for Environmental Cooperation. 48p.
Crewe T.L., Mitchell G.W., Larrivée M. 2019. Size of the Canadian breeding population of monarch butterflies is driven by factors acting during spring migration and recolonization. Frontiers in Ecology and Evolution, 7: 308.
Crossley M.S., Burke K.D., Schoville S.D., Radeloff V.C. 2021. Recent collapse of crop belts and declining diversity of US agriculture since 1840. Global Change Biology, 27(1): 151–164.
Diffendorfer J.E., Loomis J.B., Ries L., Oberhauser K., Lopez-Hoffman L., Semmens D., et al. 2014. National valuation of monarch butterflies indicates an untapped potential for incentive-based conservation. Conservation Letters, 7(3): 253–262.
Drapeau Picard A.P., Dieni A., Moreau A., Mitchell G.W., MacNair M.L., Casajus N., et al. 2024. Mission Monarch: engaging the Canadian public for the conservation of a species at risk. Journal of Insect Conservation, 28: 225–231.
Emon J.V., Seiber J.N. 1985. Chemical constituents and energy content of two milkweeds, Asclepias speciosa and A. curassavica. Economic Botany, 39(1): 47–55.
ESRI. 2021. ArcGIS Desktop: Release 10.4.0 Redlands, CA: Environmental Systems Research Institute.
Evans D.R., Hobson K.A., Kusack J.W., Cadman M.D., Falconer C.M., Mitchell G.W. 2020. Individual condition, but not fledging phenology, carries over to affect post-fledging survival in a neotropical migratory songbird. Ibis, 162(2): 331–344.
Fahrig L., Girard J., Duro D., Pasher J., Smith A., Javorek S., et al. 2015. Farmlands with smaller crop fields have higher within-field biodiversity. Agriculture, Ecosystems & Environment, 200: 219–234.
Fisher K.E., Adelman J.S., Bradbury S.P. 2020. Employing very high frequency (VHF) radio telemetry to recreate monarch butterfly flight paths. Environmental Entomology, 49(2): 312–323.
Flockhart D.T., Brower L.P., Ramirez M.I., Hobson K.A., Wassenaar L.I., Altizer S., Norris D.R. 2017. Regional climate on the breeding grounds predicts variation in the natal origin of monarch butterflies overwintering in Mexico over 38 years. Global Change Biology, 23(7): 2565–2576.
Flockhart D.T., Pichancourt J.B., Norris D.R., Martin T.G. 2015. Unravelling the annual cycle in a migratory animal: breeding-season habitat loss drives population declines of monarch butterflies. Journal of Animal Ecology, 84(1): 155–165.
Fluet-Chouinard E., Stocker B.D., Zhang Z., Malhotra A., Melton J.R., Poulter B., et al. 2023. Extensive global wetland loss over the past three centuries. Nature, 614(7947): 281–286.
Gowler C.D., Leon K.E., Hunter M.D., de Roode J.C. 2015. Secondary defense chemicals in milkweed reduce parasite infection in monarch butterflies, Danaus plexippus. Journal of Chemical Ecology, 41(6): 520–523.
Green D.A. II. 2023. Tracking technologies: advances driving new insights into monarch migration. Current Opinion in Insect Science, 60: 101111.
Greenstein L., Steele C., Taylor C.M. 2022. Host plant specificity of the monarch butterfly Danaus plexippus: a systematic review and meta-analysis. PLoS ONE, 17(6): e0269701.
Groh H., Dore W.G. 1945. A milkweed survey in Ontario and adjacent Quebec. Scientific Agriculture, 25(8): 463–481.
Gustafsson K.M., Agrawal A.A., Lewenstein B.V., Wolf S.A. 2015. The monarch butterfly through time and space: the social construction of an icon. Bioscience, 65(6): 612–622.
Hartzler R.G. 2010. Reduction in common milkweed (Asclepias syriaca) occurrence in Iowa cropland from 1999 to 2009. Crop Protection, 29(12): 1542–1544.
Hartzler R.G., Buhler D.D. 2000. Occurrence of common milkweed (Asclepias syriaca) in cropland and adjacent areas. Crop Protection, 19(5): 363–366.
Howard E., Davis A.K. 2004. Documenting the spring movements of monarch butterflies with Journey North, a citizen science program. The monarch butterfly: biology and conservation. Cornell University Press, Ithaca, NY. 114p.
Inamine H., Ellner S.P., Springer J.P., Agrawal A.A. 2016. Linking the continental migratory cycle of the monarch butterfly to understand its population decline. Oikos, 125(8): 1081–1091.
Johnston M.K., Hasle E.M., Klinger K.R., Lambruschi M.P., Derby Lewis A., Stotz D.F., et al. 2019. Estimating milkweed abundance in metropolitan areas under existing and user-defined scenarios. Frontiers in Ecology and Evolution, 7: 210.
Kasten K., Stenoien C., Caldwell W., Oberhauser K.S. 2016. Can roadside habitat lead monarchs on a route to recovery? Journal of Insect Conservation, 20(6): 1047–1057.
Lemoine N.P. 2015. Climate change may alter breeding ground distributions of eastern migratory monarchs (Danaus plexippus) via range expansion of Asclepias host plants. PLoS ONE, 10(2): e0118614.
Liu J., Huffman T., Green M., Joosse P., Martin T. 2020. Changes in land use and management by farm type and the impact on soil cover in Canada, 1991–2011. Ecological Indicators, 116: 106531.
Luna T., Dumroese R.K. 2013. Monarchs (Danaus plexippus) and milkweeds (Asciepias species). Native Plants Journal, 14(1): 5–15.
Malaj E., Freistadt L., Morrissey C.A. 2020. Spatio-temporal patterns of crops and agrochemicals in Canada over 35 years. Frontiers in Environmental Science, 8: 556452.
Malcolm S.B. 1994. Milkweeds, monarch butterflies and the ecological significance of cardenolides. Chemoecology, 5(3–4): 101–117.
Malcolm S.B., Brower L.P. 1989. Evolutionary and ecological implications of cardenolide sequestration in the monarch butterfly. Experientia, 45(3): 284–295.
Malcolm S.B., Cockrell B.J., Brower L.P. 1989. Cardenolide fingerprint of monarch butterflies reared on common milkweed, Asclepias syriaca L. Journal of Chemical Ecology, 15(3): 819–853.
Martin A.E., Collins S.J., Crowe S., Girard J., Naujokaitis-Lewis I., Smith A.C., et al. 2020. Effects of farmland heterogeneity on biodiversity are similar to–Or even larger than—the effects of farming practices. Agriculture, Ecosystems & Environment, 288:106698.
Martin A.E., Mitchell G.W., Girard J.M., Fahrig L. 2021. More milkweed in farmlands containing small, annual crop fields and many hedgerows. Agriculture, Ecosystems & Environment, 319: 107567.
McPherson M., Buckingham D. 2023. Opportunities for establishing pollinator habitat on Ontario agricultural lands. Unpublished report prepared for: Canadian Wildlife Service–Ontario region. Environment and Climate Change Canada. 122p. Available from finalreport-2023-Opportunities-for-Establishing-Pollinator-Habitat-Ontario-1.pdf [accessed April 2025].
Miller N.G., Wassenaar L.I., Hobson K.A., Norris D.R. 2011. Monarch butterflies cross the Appalachians from the west to recolonize the east coast of North America. Biology Letters, 7(1): 43–46.
Momeni-Dehaghi I., Bennett J.R., Mitchell G.W., Rytwinski T., Fahrig L. 2021. Mapping the pre-migration distribution of eastern Monarch butterflies using community science data. Ecology and Evolution, 11(16): 11275–11281.
Nail K.R., Stenoien C., Oberhauser K.S. 2015. Immature monarch survival: effects of site characteristics, density, and time. Annals of the Entomological Society of America, 108(5): 680–690.
Natural Resources Canada (NRCan). 2016. National Railway Network–NRWN–GeoBase Series[online]. Available from National Railway Network–NRWN–GeoBase Series–Open Government Portal (canada.ca) [accessed 7 March 2018].
Natural Resources Canada (NRCan). 2017a. Lakes, rivers and glaciers in Canada–CanVec series–hydrographic features[online]. Available from Lakes, Rivers and Glaciers in Canada–CanVec Series–Hydrographic Features–Open Government Portal [accessed 6 March 2018].
Natural Resources Canada (NRCan). 2017b. Mines, Energy and Communication Networks in Canada–CanVec series–resources Management features[online]. Available from Mines, Energy and Communication Networks in Canada–CanVec Series–Resources Management Features–Open Government Portal [accessed 7 March 2018].
Oberhauser K., Wiederholt R., Diffendorfer J.E., Semmens D., Ries L., Thogmartin W.E., et al. 2017. A trans-national monarch butterfly population model and implications for regional conservation priorities. Ecological Entomology, 42(1): 51–60.
Pest Management Regulatory Agency (PMRA). 2021. Pest control products sales report for 2019. Health Canada. 49p.
Pitman G.M., Flockhart D.T., Norris D.R. 2018. Patterns and causes of oviposition in monarch butterflies: implications for milkweed restoration. Biological Conservation, 217: 54–65.
Pleasants J. 2017. Milkweed restoration in the US Midwest for monarch butterfly recovery: estimates of milkweeds lost, milkweeds remaining and milkweeds that must be added to increase the monarch population. Insect Conservation and Diversity, 10(1): 42–53.
Pleasants J., Thogmartin W.E., Oberhauser K.S., Taylor O.R., Stenoien C. 2023. A comparison of summer, fall and winter estimates of monarch population size before and after milkweed eradication from crop fields in North America. Insect Conservation and Diversity, 17(1): 51–64.
Pleasants J.M., Oberhauser K.S. 2013. Milkweed loss in agricultural fields because of herbicide use: effect on the monarch butterfly population. Insect Conservation and Diversity, 6(2): 135–144.
Pocius V.M., Debinski D.M., Pleasants J.M., Bidne K.G., Hellmich R.L. 2018a. Monarch butterflies do not place all of their eggs in one basket: oviposition on nine Midwestern milkweed species. Ecosphere, 9(1): e02064.
Pocius V.M., Debinski D.M., Pleasants J.M., Bidne K.G., Hellmich R.L., Brower L.P. 2017. Milkweed matters: monarch butterfly (Lepidoptera: Nymphalidae) survival and development on nine Midwestern milkweed species. Environmental Entomology, 46(5): 1098–1105.
Pocius V.M., Pleasants J.M., Debinski D.M., Bidne K.G., Hellmich R.L., Bradbury S.P., Blodgett S.L. 2018b. Monarch butterflies show differential utilization of nine Midwestern milkweed species. Frontiers in Ecology and Evolution, 6: 169.
Rendón-Salinas E., Fernánez-Islas A., Cruz-Piña M., Mondragón-Contreras G., Martínez-Pacheco A. 2025. Superficie forestal ocupada por las colonias de mariposas monarca en México durante la hibernación de 2024–2025[online]. Available from 2024-Monitoreo-Mariposa-Monarca-en-México-2024-2025.pdf [accessed 7 March 2025].
Saunders S.P., Ries L., Neupane N., Ramírez M.I., García-Serrano E., Rendón-Salinas E., Zipkin E.F. 2019. Multiscale seasonal factors drive the size of winter monarch colonies. Proceedings of the National Academy of Sciences, 116(17): 8609–8614.
Semmens B.X., Semmens D.J., Thogmartin W.E., Wiederholt R., López-Hoffman L., Diffendorfer J.E., et al. 2016. Quasi-extinction risk and population targets for the Eastern, migratory population of monarch butterflies (Danaus plexippus). Scientific Reports, 6(1): 1–7.
Soltani N., Nurse R.E., Sikkema P.H. 2014a. Two-pass weed management with preemergence and postemergence herbicides in glyphosate-resistant soybean. Agricultural Sciences, 5(6): 504–512.
Soltani N., Shropshire C., Sikkema P.H. 2014b. Volunteer glyphosate and glufosinate resistant corn competitiveness and control in glyphosate and glufosinate resistant corn. Agricultural Sciences, 5(5): 402–409.
Stanton R.L., Morrissey C.A., Clark R.G. 2018. Analysis of trends and agricultural drivers of farmland bird declines in North America: a review. Agriculture, Ecosystems & Environment, 254:244–254.
Statistics Canada (StatCan). 2016. Census of Agriculture[online]. Available from 2016 Census of Agriculture (statcan.gc.ca) [accessed 7 September 2021].
Statistics Canada (StatCan). 2018a. National Road Network–NRN–GeoBase Series[online]. Available from National Road Network–NRN–GeoBase Series–Open Government Portal (canada.ca) [accessed 7 March 2018].
Statistics Canada (StatCan). 2018b. Census Divisions 2016[online]. Available from Census division 2016 - Open Government Portal [accessed 7 September 2021].
Stenoien C., Nail K.R., Oberhauser K.S. 2015. Habitat productivity and temporal patterns of monarch butterfly egg densities in the eastern United States. Annals of the Entomological Society of America, 108(5): 670–679.
Taylor O.R. Jr., Pleasants J.M., Grundel R., Pecoraro S.D., Lovett J.P., Ryan A. 2020. Evaluating the migration mortality hypothesis using monarch tagging data. Frontiers in Ecology and Evolution, 8: 264.
The Xerces Society. 2021. 100 plants to feed the monarch: Create a healthy habitat to sustain North America's most beloved butterfly. Storey Publishing, North Adams, MA. 49p.
Thogmartin W.E. 2024. Non-negligible near-term risk of extinction to the eastern migratory population of monarch butterflies–An updated assessment (2006–22) (No. 2023-1097). 10 p. U.S. Geological Survey Open-File Report 2023-1097.
Thogmartin W.E., Diffendorfer J.E., López-Hoffman L., Oberhauser K., Pleasants J., Semmens B.X., et al. 2017a. Density estimates of monarch butterflies overwintering in central Mexico. PeerJ, 5: e3221.
Thogmartin W.E., López-Hoffman L., Rohweder J., Diffendorfer J., Drum R., Semmens D., et al. 2017b. Restoring monarch butterfly habitat in the midwestern US:‘all hands on deck’. Environmental Research Letters, 12(7): 074005.
Thogmartin W.E., Wiederholt R., Oberhauser K., Drum R.G., Diffendorfer J.E., Altizer S., et al. 2017c. Monarch butterfly population decline in North America: identifying the threatening processes. Royal Society Open Science, 4(9): 170760.
Trudeau J., Obama B., Peña Nieto E. 2016. North American Climate, Clean Energy, and Environment Partnership Action Plan[online]. Available from https://obamawhitehouse.archives.gov/the-press-office/2016/06/29/leaders-statement-north-american-climate-clean-energy-and-environment [accessed April 2025].
United Sates Department of Agriculture (USDA). 2023. Adoption of genetically engineered crops in the US[online]. Available from USDA ERS–Recent Trends in GE Adoption.
Van Deynze B., Swinton S.M., Hennessy D.A., Haddad N.M., Ries L. 2024. Insecticides, more than herbicides, land use, and climate, are associated with declines in butterfly species richness and abundance in the American Midwest. PLoS ONE, 19(6): e0304319.
Wassenaar L.I., Hobson K.A. 1998. Natal origins of migratory monarch butterflies at wintering colonies in Mexico: new isotopic evidence. Proceedings of the National Academy of Sciences, 95(26): 15436–15439.
White D.J. 1996. Status, distribution, and potential impact from noxious weed legislation [online]. Available from Monarch Watch.
Wilcox A.A., Flockhart D.T., Newman A.E., Norris D.R. 2019. An evaluation of studies on the potential threats contributing to the decline of eastern migratory North American monarch butterflies (Danaus plexippus). Frontiers in Ecology and Evolution, 7: 99.
Woodson R.E. 1954. The North American species of Asclepias L. Annals of the Missouri Botanical Garden, 41(1): 1–211.
Züst T., Petschenka G., Hastings A.P., Agrawal A.A. 2019. Toxicity of milkweed leaves and latex: chromatographic quantification versus biological activity of cardenolides in 16 Asclepias species. Journal of Chemical Ecology, 45: 50–60.
Supplementary material
Supplementary Material 1 (DOCX / 73 KB).
- Download
- 72.41 KB
Supplementary Material 2 (XLSX / 5.73 MB).
- Download
- 5.74 MB
Information & Authors
Information
Published In
FACETS
Volume 10 • 2025
Pages: 1 - 14
Editors: Andrea Olive and Allyson Kate Menzies
History
Received: 29 March 2024
Accepted: 6 February 2025
Version of record online: 23 April 2025
Copyright
© 2025 Fahrig, Johnston, Larrivee, Momeni-Dehaghi; and The Crown. This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.
Data Availability Statement
Data generated or analyzed during this study are provided in full within the published article and its supplementary materials. Data are also available in the Dryad repository, https://doi.org/10.5061/dryad.3bk3j9kv9.
Key Words
Sections
Subjects
Authors
Author Contributions
Conceptualization: GWM, JG, ER, EDS, WET, DP
Data curation: GWM, PK, LF, JG, MKJ, AEM, IM, WET, DP
Formal analysis: GWM, PK, JD, DP
Investigation: GWM, PK, JD, LF, JG, MKJ, ML, AEM, IM, JP, WET, DP
Methodology: GWM, PK, JD, LF, JG, MKJ, ML, AEM, IM, JP, WET, DP
Project administration: ER, EDS
Visualization: PK, GWM
Writing – original draft: GWM
Writing – review & editing: GWM, PK, JD, LF, JG, MKJ, ML, AEM, IM, JP, ER, EDS, WET, DP
Competing Interests
There are no competing interests to declare.
Metrics & Citations
Metrics
Other Metrics
Citations
Cite As
Greg W. Mitchell, Patrick Kirby, Jason Duffe, Lenore Fahrig, Judith Girard, Mark K. Johnston, Maxim Larrivée, Amanda E. Martin, Iman Momeni-Dehaghi, Jon Pasher, Elizabeth Rezek, Elisabeth D. Shapiro, Wayne E. Thogmartin, and Darren Pouliot. 2025. Additional common milkweed would help Canada meet its share of the trinational eastern migratory monarch butterfly recovery target. FACETS.
10: 1-14.
https://doi.org/10.1139/facets-2024-0063
Export Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.
There are no citations for this item