Accessing monarch butterflies, outside their colonies in México, using community-based monitoring.
ABSTRACT
Monarch butterfly (Danaus plexipus) streaming - a behavior in which adults fly outside their overwintering colonies in Mexico - has been noted but no comprehensive study has looked to explain the temporal and spatial patterns of this phenomenon. We trained local women living at 10 sites within 5 km of the El Rosario colony to count monarchs streaming by their houses 3 times daily from 4 March to 5 April. This preliminary study suggests that the prevalence of streaming monarchs is correlated with distance, temperature, and cloud cover. Data also suggests that peak streaming occurs in early March, and that the trailing edge of the migration in 2020 was between 15 March and 27 March. Further study of monarchs outside their overwintering colonies is needed to understand streaming behavior, better estimate the monarch migration seasonal timeline, and create baseline data to track climate change effects on overwintering monarchs. Continuation and expansion of community-based monitoring near the monarch's overwintering sites will create a small but valuable local, conservation-based economy, and give women not traditionally involved in scientific studies the opportunity to contribute to research.
SPECIFIC AIMS
Collect information on overwintering monarch butterfly behavior, location, and population trends outside of the main colonies in order to:
Better understand the streaming patterns of overwintering monarchs outside of the more well-studied colonies,
Better track when the monarchs begin to migrate in the spring, and detect patterns associated with the start of the migration.
Gain baseline data to see how these patterns change in response to climate change and continued human development.
Create a network of local, trained women living around the colonies to:
Collect data as citizen scientists,
Empower the community to make scientific observations and participate in monarch conservation,
Establish a conservation economy.
BACKGROUND
Each winter the eastern population of monarch butterflies (Danaus plexippus) overwinter in dense clusters in the high-elevation Oyamel fir forest of central Mexico (Urquhart and Urquhart 1976). Typically, the monarchs remain in their overwintering grounds from November to March, with the majority of their time spent inactive as a strategy to preserve lipids for their remigration north in the spring (Alonso-Mejía et al. 1997).
Overwintering activity can be broken into four phases. The first, from October to November, is marked by the arrival of monarchs to the sanctuaries. This is a period of intense flight and the formations of clusters. The second phase, from December to mid-February, is more stable, and monarch activity and flight is much reduced. The third phase, from the end of February to March, sees a dramatic increase in activity as monarchs fly to places with higher humidity. The final phase, typically beginning in late February and finishing by April, is the start of the northbound migration (Garcia-Serrano et al. 2004). During all phases, on warm, sunny days monarchs become more active, and will often stream down the mountains to drink water (Calvert and Masters 1986).
Overwintering activity, including streaming, tracks with the monarch's reproductive timeline. Monarchs arrive to the sanctuaries in sexual diapause, which is regulated by suppressed production of juvenile hormone (JH) (Herman 1975). Monarchs see little to no reproductive development during much of the overwintering season , then, triggered by longer days and warming conditions, they begin to develop their reproductive organs and enter a post-diapause phase (Herman et al. 1989). Reproductive development, mating, and flight activity all build during the third phase of overwintering, but the exact timeline varies between individuals ().
During all phases the need to conserve lipids for the remigration north in the spring is of utmost importance (Alonso-Mejía et al. 1997). Without sufficient lipids, successful remigration north becomes impossible (Masters et al. 1988). Thus, streaming can appear contradictory, as flight can be energy intensive. Despite the energy cost of streaming, it appears to be a necessary behavior for monarchs to acquire sufficient amounts of water (Calvert and Brower 1986, Brower 1996).
The need for water is a necessity for overwintering survival, as it is required to metabolize lipid reserves. Water can be generated during metabolism (e.g., metabolic water) or collected from condensation for much of the winter season (). Warming, drying conditions and reproductive development require monarchs to stream and pool at open sites with moist, ground vegetation and at live water sources (Brower 1996). Brower (1997) hypothesized that as the dry season progresses, desiccation becomes more severe and this increases the frequency of streaming.
Water is often cited as the driver for streaming (Brower 1997, Calvert and Brower 1986). Yet monarchs can be seen passing pooling sites in favor of further travel (personal observation). Since insects are capable of evaluating energy stores and breaking diapause prematurely if energy reserves are low (Hahn and Denlinger 2011), streaming could be used by low fitness individuals to increase temperature, JH production, and break diapause earlier than ambient temperatures would allow. Small male monarchs, in poorest condition, have been determined to mate sooner, a likely result of low lipid levels and a high risk of starvation before the migration begins (Van Hook 1993). These premature matings are inline with the hypothesis of Brower (1997) in which premature streaming would lead to early production of JH, which would lead to early termination of diapause. While poor conditioned males are likely to die before migrating, if enough monarchs broke diapause early, this could lead to an early spring remigration.
Whether streaming is a direct or indirect response to reproductive schedules or water requirements, the phenomenon is influenced by weather conditions and, as a result, anthropomorphic climate change. Spatial climate models of Mexico suggest a warmer, dryer future (Sáenz-Romero et al. 2010), broken by intense winter storms (Oberhauser and Peterson 2003)). An analysis of 61 studies involving 694 species found that over the last 50 years spring events have begun to occur earlier (Root et al. 2003). In Cape May, New Jersey, the fall migration timeline has shifted 16 to 19 day during a 29 year period, an increased delay of 6 days per decade (Culbertson 2022). Similar trends were predicted by Brower (1997) and there is anecdotal evidence of spring monarchs migrating sooner (personal observation). These changes will be compounded by forest degradation, including an increase in diseased trees presumably affected by drought-induced stress from climate change (Vidal. et al. 2013).
Monarch research at the overwintering colonies has been limited to scientists and a small handful of local participants (google scholar search). In other parts of the monarch’s range, most notably in the United States and Canada, citizen science - the gathering of data by members of the community - has played a crucial role in our current understanding of the monarchs. Of the 503 monarch-focused research publications cataloged from 1940 to 2014, 88 (17%) included data contributed by citizen science programs (Ries and Oberhauser 2015). Beyond scientific pursuits, citizen science can foster environmental stewardship, spread knowledge, bring local knowledge to decision making, and improve science literacy (McKinley et al. 2015).
Citizen Science has been done mostly in developed countries, where researchers tend to utilize skilled amateur volunteers and have better access to funding for such research (Ticheler et al. 1998). Citizen science participants tend to be white, highly educated, and affluent (Pandya 2012). However, for citizen science to be most effective and for science to be more democratic, researchers must try and expand the range of participants contributing to research (Bonny et al. 2015).
In communities that struggle socioeconomically, examples of modified citizen science projects have found success. By creating simple collection schemes, paying locals for their data, and developing systems to affirm data quality, researchers can expand citizen science methods to developing countries (Ticheler et al. 1998). In Zambia local fishermen were employed to collect data on their catches. Comparing the locals’ data to that of the fisheries scientists showed that local-based data collection and monitoring (also referred to as participatory monitoring and community based monitoring) could produce large quantities of reliable and relatively cheap data (Ticheler et al. 1998). Danielson et al. (2005) reviewed 15 local based monitoring projects in 13 developing countries and found that they provide relevant, reliable results that can lead to changes in attitudes and decision making.
In 2010, 27,000 people were living within the Biosphere Reserve’s buffer zone and one million were living around the Reserve. The communities are agrarian, and depend on the forest for building materials and fuels, as well as cleared land for grazing and cultivation. Logging was the basis of the local economy, but restrictions have furthered job scarcity (Vital et al. 2013). As monarch-associated tourism continues to expand it does not benefit all people equally. People not associated with tourism suffer from new logging restrictions (Monterrubio and Muñoz 2013). The creation of more sustainable job opportunities is fundamental to the monarchs’ future (Vital. et al. 2013).
RESEARCH DESIGN AND METHODS
Location
Ten households participated in our preliminary study. Surveys were conducted at participant’s houses, located 1.7 to 4.9 km from the 2019-2020 El Rosario colony. Sites were situated around El Rosario’s developed western slopes. To the extent possible, houses were paired in order to buffer missed surveys, better understand streaming dispersal, and corroborate collected data. Discrepancies in data between pairs flagged possible surveyor error and ensured data was being collected systematically.
Each household surveyed a predetermined, 25 m transect. Transects were located within 10 m of the participants’ houses for convenience. The participant stood at a predetermined point and used a predetermined marker (typically a tree or pole) to delineate the 25 m transect. Since monarchs typically fly up and down slopes rather than contour the land, transects were perpendicular to slope.
Participants
We chose half the households based on preexisting relationships. These households were either acquaintances or related to acquaintances of the lead surveyor. The other half had no prior relationship or known connection to the lead surveyor or their acquaintances. We chose these households based on location. We went to locations where gaps in our study existed and found women at their homes. These women were invited to participate in the survey as long as they planned to be home 6 days a week during study.
Each household had between 1 and 4 people participating. Having multiple people trained at each household ensured a higher probability that someone would be available to survey as much as possible.
Methods
Monarchs were counted three times daily. Monitoring began at the 1100, 1300, and 1600 hour. Each monitoring session included 4 independent one minute surveys. For each minute, participants counted and recorded all the monarchs that passed within their predetermined 25 m transect. The subsequent 3 surveys were each separated by at least 4 min. Each household was provided a digital watch with a one minute, pre-programmed timer. Timers chimed to alert participants of survey’s end, so participants could focus on counting. Watches also had 3 alarms preprogrammed to go off 5 minutes prior to each survey session.
Participants recorded the number of monarchs seen within the transect during each one minute survey. Participants also noted local weather conditions. Cloud cover was recorded as either sunny (0% - 25% cloud cover), partially cloudy (25% - 75% cloud cover), or cloudy (75% - 100% cloud cover). Wind was recorded as none, light, or very windy using a simplified Beaufort wind scale. Precipitation was recorded as either none, rain, snow/hail. We provided binders with paper data sheets and pencils. Kestrel D3 Data Loggers were placed in a shaded location (typically hanging from a shrub), one m off the ground at every site to record temperature, humidity, dew point, heat stress, station pressure, and density altitude at 10 minute intervals.
Payments
The lead surveyor and a local community member visited each house weekly to review monitoring data, upload Kestrel weather data, answer questions, and provide payment. Participants that surveyed 17 to 21 times in a week were paid $200 pesos ($10 US). Allowing for 4 missed surveys a week minimized the incentive to falsify data when a survey couldn’t be done. Participants that surveyed between 14 to 16 times in a week were paid $150 pesos ($7.35 US). A bonus of $400 pesos ($20 US) was paid to each participant at the end of the season if they were able to return their watch, data binder, and Kestrel thermometer. In our preliminary study, all participants returned their equipment.
Data Quality
We paired households to help compare counts and identify suspect data. On multiple occasions the lead surveyor conducted surveys independently, but alongside participants, to corroborate data. We also varied our payment schedule to more safely travel with payments and arrive unannounced during survey times to confirm participation.
We ranked data as either high trust, med trust, or low trust. We considered one site’s data as low trust after our quality control methods flagged their data for inconsistencies. We made multiple visits during survey times to offer support and confirm participation. They were often not surveying when we arrived at survey hour, and our counts and those of the paired house varied enough that this household’s data was considered low trust and was not included in overall analyses. Two other households raised questions of data quality, but scrutiny alleviated doubts and data was included.
Data Analysis
Five households began surveys on 23 February 2020 and 5 began 4 March 2020. All surveys finished on 5 April 2020. Monarch presence dropped to below 25 daily monarch sightings across all sites after 26 March 2020. We concentrated our analysis on data collected when all household sites were collecting data and monarchs were present (4 Mar to 26 Mar).
Weather loggers did not collect data for short periods because of technical error; neighbor kids discovering one and moving it; and a participant moving one inside when it was raining to protect it. In these cases the surveys were excluded from relevant analysis.
The closest house was analyzed alone, as this house saw a dramatic difference in monarch abundance compared to the others, and because the participant was able to complete 100% of surveys and data was ranked as high trust. Other houses were paired, and counts averaged between them, to offset the effects of missing surveys, lower trust, and lower counts. Thus site one includes one house (1.66 km from colony). Sites 2 is a house pair (1.74 km and 1.87 km from colony), as is Sites 3 (2.35 and 2.93 km from colony), Sites 4 (3.72 to 3.77 km from colony), and Sites 5 (4.83 and 4.88 km from colony).
RESULTS
Excluding one household with untrusted data, we observed 34,028 monarchs during 725 one minute surveys. Of the 725 survey minutes, monarchs were present in 39% (284) of surveys. The maximum number of monarchs counted at each site varied between 137 and 1417 monarchs/minute. There was a 9 day range for max surveys (5 Mar 2020 - 14 Mar 2020).
Weather Affects
During the 1000 hours survey, the average hourly temperature ranged 13.4 degrees (10.3° C to 23.7° C), but monarchs were sited in a narrower range (11.3° C to 20.6° C). Days with ≥ 10 monarchs present had a narrower range, but varied by distance. The temperatures at Sites 1-3 skewed cooler (11.9° C to 19.9° C) when ≥10 monarchs were present compared to Sites 4 -5 (13.6° C - 20.6° C).
Distance Affects
The number of monarchs present was correlated to the distance each site was from the main colony. From 4 March 2020 to 26 March 2020 Site 1 saw monarchs present 98% of surveys. Only one survey was absent of monarchs (22 Mar 2020, 1600 hrs). Site 1 also had the highest daily monarch counts 91% of days. The daily abundance of monarchs for Sites 1, 2, and 3 corresponded with distance 82% of days.
Sites 5 never had the highest daily count, but did have lowest daily counts 45% of time. These sites saw zero monarchs during 28 hourly counts (56% of the time).
Distance also affected the ratio of monarchs seen during the three daily survey hours. From 4 March to 26 March monarch counts at Site 1 were highest at 1000 hrs (relative to other survey times) 45.4% of days, but only 18% of days at Sites 5. Sites 4 saw highest numbers during the 1000 survey only once (5% of days), and on this day their combined daily monarch count was one.
Migration Timeline
Counts of monarchs reached their max at each site between 5 March and 14 March. The first day with no monarchs present at closest site was 28 March. Total counts of all houses dropped to <26 by 27 March.
DISCUSSION
Weather Effects
Preliminary analysis suggests a relationship between monarch abundance, cloud cover, and temperature. At Site 1, low monarchs numbers appear linked to lower 1000 hr temperatures coupled with increased cloud cover. This suggests that cold temperatures either prevented monarchs from flying or led them to stay close to the colony to avoid getting caught away from the colony if temperatures were to drop suddenly (a likely effect of increased cloud cover).
The flight threshold for monarchs is 12.7° C, but monarchs were present when temperatures were recorded at sites between 11.3° C to 20.6° C. The temperatures below 12.7° C could be due to the monarchs ability to warm up through shivering and flight or because our thermometers were measuring a cooler microclimate.
We used Kestrel Data Loggers to capture temperature. Despite following a protocol for logger placement, the microclimates of each location experienced different confounding variables (e.g., wind shadows, exposure, thermal sinks), which were likely not experienced by monarchs streaming meters off the ground. Attempts to standardize placement were made, but were complicated by having a limited number of placement options at each site and the need to hide loggers to avoid detection by passersby. Therefore, recorded temperatures must be seen as relative measurements rather than the specific ambient air temperatures experienced by streaming monarchs. We plan to add 7 Tempest Weather Flow stations (one for each cluster of sites) to select participant’s roofs to standardize placement, decrease possible theft, and collect data closer to where monarchs are found.
Our qualitative method to describe wind and cloud cover was too coarse (three levels each) and objective for finding subtle patterns. We were also unable to collect wind direction, which likely affected streaming routes, thermoregulation, and travel speed. While we did collect cloud cover data, this was not a good indicator for site specific incident radiation, which would more directly influence local monarch behavior than clouds cover across all visible sky. Again, we plan to add 7 Tempest Weather Flow stations to gather quantitative data to more finely express cloud cover (incident radiation) and wind speed (km/hr with directionality). Until we can provide a weather station at each site, windsocks will be used to collect wind direction and a more scaled and unobjective wind speed. Since wind direction and speed would affect monarchs flying against the wind differently than monarchs flying with the wind, we will also include monarch flight direction in our data collection.
Recording monarch flight direction will help determine thresholds for streaming. Streaming away from colony would be an indicator of favorable conditions, while returning monarchs would be associated with (or a precursor to) unfavorable streaming conditions. Furthermore, we should see monarchs begin to stream uphill earlier in the day at further sites than closer sites, as further distanced monarchs would need more time to return before unfavorable conditions were met.
For a third level of data confirmation, we will provide participants with cell phones to be used to take video of streaming and photos of the sky. This will help double check directionality, abundance, cloud cover, and add a timestamp to data.
We also propose increasing the number of daily monarch counts at selected sites (1-3 sites) from 3/day to 10/day. More frequent intervals will help better capture daily monarch patterns in relation to weather conditions.
The lack of correlation between abundance and humidity suggests that monarchs are streaming for reasons beyond getting water. Further work is necessary to understand motives for streaming.
Distance Effects
Sites closer (Sites 1, 2, 3) had higher daily monarch counts than further sites (Sites 4, 5), with some daily exceptions. This suggests that monarchs tend to limit the distance of streaming, possibly to conserve energy and better guarantee enough time to return to the colony before conditions are no longer suitable for travel.
The effect of distance was not completely linear, however. From 4 March to 26 March, Site 1 recorded the highest daily counts 91 % of days, 2nd highest count 5% of days, and third highest count 5% of days. Site 2’s counts were the 1st, 2nd, and 3rd highest monarch counts (compared to other sites) 5%, 77%, and 18% of days respectively, and Site 3 was 5%, 14%, and 41% respectively. This suggests that while distance is a major factor in monarch presence, other factors influence streaming abundance as discussed earlier.
Furthermore, Site 5 saw more monarchs than Site 4 32% of days. This might be explained by the influence of monarchs streaming from outlying colonies. Recording direction of travel and mapping outlying colonies will help test this hypothesis. If monarchs are coming from other locations besides the main El Rosario colony, then the direction of streaming monarchs will originate from these outlying colonies in the morning and point to them in the afternoon.
Distance also seemed to directly influence the ratio of monarchs by survey hour. Sites had different ratios of monarchs seen between the 3 survey hours (1000, 1300, and 1600). Sites closer (Sites 1,2, 3) to the El Rosario colony saw higher ratios of 1000 hours to other times than further sites (Sites 4, 5). This could be partially explained by monarch flight thresholds. The range of temperatures at 1000 hrs, when sites saw ≥10 monarchs, was skewed warmer at Sites 4 and 5 than Sites 1, 2, and 3. This suggests that suitable conditions for streaming were not met early enough for monarchs to travel to the furthest sites by 1000 hours, and therefore patterns of streaming are location dependent due to a correlation with temperature.
Our preliminary study had a small number of sites (9 houses, or 4.5 site pairs), and thus missed surveys were more likely to obscure patterns and underrepresent site. We propose increasing household clusters from 2 to 3, to help buffer these missed surveys. We also propose adding 2 additional sites (with 3 households participating at each) to better capture landscape level patterns. Both of these additions would increase participating households from 9 to 21.
Our preliminary study based distance from the El Rosario colony off of one GPS point taken in March 2020. Data is inconsistent as to the stability of colony location.Vidal and Rendón-Salinas (2014) didn’t see a shift in colony from 2004 to 2014. Calvert and Master (1986), however, noted that during exceptionally dry periods entire colonies moved downslope to reformed closer to water sources, and daily streaming monarchs return to the downslope edge of the colony creating a downslope “creep” effect. We propose taking weekly GPS points to better capture possible seasonal location changes. It is unlikely that small shifts would change the order (near to far) of the survey sites, but we might expect to see a comparative shift in monarch numbers as the colony moves closer to or further from individual sites.
Colony location is a function of sun exposure and water availability. The majority of colonies are on the SW slope (Calvert and Brower 1986, Vidal and Rendón-Salinas 2014). If topography dictates direction of travel, then we should expect to see monarchs streaming in greater numbers downslope of the colony rather than going over ridges. We propose adding topography to our analysis. Again, weekly GPS points of the colony’s location will be needed, especially if the colony moves between ridge lines.
Migration Timeline
Counts of monarchs reached their max at each site between 5 March 2020 and 14 March 2020. By 27 March combined daily totals were ≤ 25 monarchs/day. This points to the trailing edge of the migration having left before 27 March and the leading edge likely occurred in early March. Collecting this baseline data on the migration timeline and tracking trends is critical. Monarch departures must balance lipid reserves, water availability, and allow monarchs to time their arrival to their southern breeding range with the emergence of milkweed.
We propose not only increasing sites but also beginning surveys by no later than 1 February to capture the start of streaming and peak streaming. We hypothesize that because streaming is a precursor to the migration, our methods can capture the start of streaming and then the leading edge of the migration in addition to the trailing edge. This information will become a baseline to clarify the migration timeline and understand changes in timing related to climate change.
Further Investigations
Streaming has been identified as a way for monarchs to get water (Calvert and Brower 1986, Brower 1996), yet some monarchs travel much further than other monarchs during streaming, and many bypass available water to stream further. This suggests other drivers may be at work. One possible explanation for the variable distances traveled is variable reproductive timelines. If males need to break diapause sooner than females, then you could expect males to stream further as a way to increase metabolism and produce more juvenile hormone (JH). If males are unlikely to be able to survive the migration (e.g., small mass) we would expect them to travel further to jumpstart reproduction (e.g., by increasing temperature to increase metabolism, to increase JH) and increase odds of mating before death. If this is true, we would expect to see the smaller male monarchs traveling further, earlier in the season than larger males and females.
We were unable to identify the sex of streaming monarchs. Future work could include collecting road mortalities or photographing streaming (to sex later) to identify if there are changes in the sex ratio or size ratio over time or by distance. We also didn’t note the type of flight, which might be a useful measure of why monarchs are streaming. If streaming is only to get water and not to increase metabolism, then we would expect to see the same levels of gliding and powered fight across all sizes, sexes, and overwintering stages. If monarchs are flying further to generate heat to break diapause, then we would expect to see more powered flight than gliding, especially on colder days, when ambient temperature needs supplementing by metabolic heat production.
Participation
We found participants both willing and excited about participating. High trust data came from surveyors with high school degrees. They also tended to be younger (20-30 years old) and had young (3-10 years old) kids that kept them home. Several of these participants took detailed notes and made increasing numbers of additional observations. We would like to continue to train these women, and possibly offer them extra work training new participants, participating in weekly payment and quality check visits.
The household responsible for data we considered low trust was allowed to finish out the entire study, as we did not have any standard or protocols in place, and didn’t want to cause drama. We propose setting standards at the start, and being upfront with households about what happens if standards are not met. Care must be taken to not create ill will, jealousy, or resentment, as we need the community's support to collect reliable data safely.
We will provide participants with cell phones to be used to take a video of streaming. This will help confirm counts and add a timestamp to each survey for an extra measure of protection against falsified data.
REFERENCES
Alonso-Mejía, A., E. Rendon-Salinas, E. Montesinos-Patiño, and L. P. Brower. 1997. Use of lipid reserves by monarch butterflies overwintering in Mexico: implications for conservation. Ecological. Applications 7: 934-947.
Batalden, R.V., K. Oberhauser, and A. T. Peterson. 2007. Ecological. Niches in Sequential. Generations of Eastern North American Monarch Butterflies (Lepidoptera: Danaidae): The Ecology of Migration and Likely Climate Change Implications. Environmental. Entomology 36: 1365-1373.
Bonney, R., T. B. Phillips, H. L. Ballard, and J. W. Enck. 2015. Can citizen science enhance public understanding of science? Public Understanding of Science 25(1), 2–16.
Brower, L. P. 1997. Biological. necessities for monarch butterfly overwintering
in relation to the Oyamel forest ecosystem in Mexico. North American Conference on the Monarch Butterfly: 11- 28.
Brower, L. P., D. R. Kust, E. Rendón-Salinas, E. García-Serrano, K. R. Kust, J. Millar, C. Fernandez del Rey, and K. Pape. 2004. Catastrophic Winter storm mortality of monarch butterflies in Mexico during January 2002. Pages 151-999 in K. S. Oberhauser and M. J. Solensky, editor. The monarch butterfly: biology and conservation. Cornell University Press, Ithica, USA.
Brower, L. S. Fink, P. Walford. 2006. Fueling the fall migration of the monarch butterfly. Integrative and Comparative Biology 46 (6): 1123–114.
Brower, L. P., E. H. Williams, P. Jaramillo-López, D. R. Kust, D. A. Slayback, and M. Isabel Ramírez. 2017. Butterfly Mortality and Salvage Logging from the March 2016 Storm in the Monarch Butterfly Biosphere Reserve in Mexico. American Entomologist 63: 151–164.
Brower, L.P., O. R. Taylor, E. H. Williams, D. A. Slayback, R. R. Zubietas, and M. I. Ramírez. 2012. Decline of monarch butterflies overwintering in Mexico: is the migratory phenomenon at risk? Insect Conservation and Diversity 5: 95–100
Calvert, W. H. and L. P. Brower. 1986. The location of monarch butterfly (Danaus Plexippus L) overwintering colonies in México in relation to topography and climate. Journal of the Lepidopterists' Society 40(3):164-187.
Calvert, W. H., W. Zuchowski, and L. P. Brower. 1983. The Effect of Rain, Snow and Freezing Temperatures on Overwintering Monarch Butterflies in Mexico. Biotropica 15: 42–47.
Danielsen, F., N. D. Burgess, and A. Balmford. 2005. Monitoring Matters: Examining the Potential. of Locally-based Approaches. Biodiversity and Conservation 14: 2507–2542.
García-Serrano, E., J. L. Reyes, and B. X. Mora Alverez. 2004. Locations and area occupied by monarch butterflies overwintering in Mexico from 1993 to 2002. Pages 128-999 in K. S. Oberhauser and M. J. Solensky, editor. The monarch butterfly: biology and conservation. Cornell University Press, Ithica, USA.
Hahn, D. A., and D. L. Denlinger. 2011. Energetics of insect diapause. Annual Review of Entomology 56.1: 103–121.
Heinrich, B. 1986. Comparative thermoregulation of 4 montane butterflies of different mass. Physiol. Zool. 59:616–626.
Herman, W. S., L. P. Brower, and W. H. Calvert. 1989. Reproductive tract development in monarch butterflies overwintering in California and Mexico. Journal of the Lepidopterists' Society 43(1): 50-58.
Masters, A., S. Malcolm, and L. Brower. 1988. Monarch Butterfly (Danaus Plexippus) Thermoregulatory Behavior and Adaptations for Overwintering in Mexico. Ecology. 69 458-467.
McKinley, D. C., A. J. Miller-Rushing, H. L. Ballard, R. Bonney, H. Brown, D. M. Evans, R. A. French, J. K. Parrish, T. B. Phillips, S. F. Ryan, L. A. Shanley, J. L. Shirk, K. F. Stepenuck, J. F. Weltzin, A. Wiggins, O. D. Boyle, R. D. Briggs, S. F. Chapin III, D. A. Hewitt, P. W. Preuss, and M. A. Soukup. 2015. Investing in Citizen Science Can Improve Natural Resource Management and Environmental Protection Issues in Ecology. Ecological. Society of America. Report Number 19.
Monterrubio, C., and G. Muñoz. 2013. Social benefits of ecotourism: The Monarch Butterfly Reserve in Mexico. Enlightening Tourism. A Pathmaking Journal. 3. 105-124.
Oberhauser, K., and A. T. Peterson. 2003. Modeling current and future potential. wintering distributions of eastern North American monarch butterflies. Proceedings of the National. Academy of Science: 14063–14068.
Pandya, R. E. 2012. A framework for engaging diverse communities in citizen science in the US. Frontiers in Ecology and the Environment 10: 314-317.
Ries, L., and K. Oberhauser. 2015. A Citizen Army for Science: Quantifying the Contributions of Citizen Scientists to our Understanding of Monarch Butterfly Biology, BioScience 65(4): 419–430.
Root, T., J. Price, K.R. Hall, S. H. Schneider, C. Rosenzweig, and J. A. Pounds. 2003. Fingerprints of global warming on wild animals and plants. Nature 421: 57-60.
Sáenz-Romero, C., G. Rehfeldt, N. Crookston, P. Duval, R. Saint-Amant, J. Beaulieu, B. Richardson. 2010. Spline models of contemporary, 2030, 2060 and 2090 climates for Mexico and their use in understanding climate-change impacts on the vegetation. Clim. Change. 102. 595-623.
Ticheler, H., J. Kolding, and B. Chanda, B. 1998. Participation of local fishermen in scientific fisheries data collection: A case study from the Bangweulu Swamps, Zambia. Fisheries Management and Ecology 5(1): 81–92.
Van Hook, T. 1993. Non-random mating in monarch butterflies overwintering in Mexico. Biology and conservation of the monarch butterfly, Natural History Museum of Los Angeles County, Los Angeles.
Vidal, O., J. López-García, and E. Rendón-Salinas. 2013. Trends in Deforestation and Forest Degradation after a Decade of Monitoring in the Monarch Butterfly Biosphere Reserve in Mexico. Conservation Biology 28: 177-186.
Vidal, O., and E. Rendon-Salinas. 2014. Dynamics and trends of overwintering colonies of the monarch butterfly in Mexico. Biological. Conservation 180: 165-175.