Recap of past year & summer 2018 field season

It’s time to recap everything that’s been going on with the Echinacea Project for the last 12(ish) months – and trust me, it’s a lot! We report all of this info annually to our two major grant providers, CBG & UMN. This includes all of our lab and field activity.

Last spring the lab was busy as always. Led by Tracie, volunteer citizen scientists at the Chicago Botanic Garden finished cleaning heads harvested in summer 2016 and began cleaning heads harvested in 2017. These volunteers clean heads to remove all the achenes, which are then counted to give us an accurate metric of echinacea plant fitness. There were a lot of heads from 2017, and volunteers continued to clean them through the summer

A bunch of undergraduate students have worked on projects in the lab this past year, including Emily, Emma, Leah, Julie (joining Team Echinacea 2019!), Tris, Sarah, and Evan. It’s always great to have undergrads in the lab – they learn a lot from us, and we learn a lot from them! Of course, graduate students were hard at work as well. Lea not only analyzed her data regarding seed set in Liatris and Solidago, but also set up a whole new experimental plot in California. Kristen, along with volunteer Mike Humphrey, is making a collection from the hundreds of bees she caught this summer in her yellow pan traps and emergence tents.

[STUART – add something here about papers that have been written/ are currently being reviewed by journals?]

Now on to the big part of this report – our super-productive 2018 field season! The 2018 summer team (pictured) included three undergraduate students from Minnesota Colleges (Andy, Brigid, and Riley),  three undergraduates in the Ison Lab at the College of Wooster (Evan, Mia, and Zeke), two high-school students (Anna and Morgan), one high-school teacher (John), one graduate student (Kristen), two recent college grads (Michael and Will), and, of course, Stuart. Gretel and Amy also came to the field intermittently throughout the summer.

We summarized the progress we made on many summer projects this past year and made flog posts about the ones where considerable new progress was made. You’ll notice this part may look remarkably similar to previous years – we’ve been conducting many of these experiments for many years!

As always, we measured survival, growth, phenology, and flowering effort of our model plant, Echinacea angustifolia, in several experimental plots. The earliest was established in 1996 and the most recent in 2015. For many of these experiments it was business as usual, and if you’re interested in learning more about them we’ve linked to their background pages below. We spent quite a bit of time measuring plants in the qGen2 & qGen3 plot (exPt 8), and while many of the plants are doing well, we had almost 50% mortality from 2017 to now. In Amy Dykstra’s experiments, we continued to monitor plant survival and growth. While mortality is low, there are still no flowering plants!

Otherwise, here are new 2018 update flog posts about new data in the experiments that take place in our common garden experiments. Michael is currently working on a manuscript about the effects of pollen limitation in echinacea:

In addition to out common gardens, we make observations of Echinacea plants in natural prairie remnants in our study area. These observations include flowering phenology, survival, reproduction, and incidence of disease. Amy is currently investigating remnant flowering phenology for her PhD.

Echinacea angustifolia interacts with and shares space with many plant and insect species. Here are updates and flog posts about projects on species that are echinacea-adjacent. Kristen is using the data collected about pollinators on roadsides and ground nesting bees for her Master’s thesis.  Andy found this year that aphids have virtually no effect on the fitness of echinacea plants. While no one this year is specifically looking at Hesperostipa, its worth noting that we did go out and check! We found only a few seeds, but collected them anyway.

Also, we have some new projects that don’t necessarily fit into any of the above categories. Here are updates of their projects.

And finally, we are worried about non-native Echinacea plants that are used in restorations and how they impact populations of the native Echinacea angustifolia. We have several ongoing experiments that investigate a population of Echinacea pallida introduced within our study area. Riley used the plants in P7 to gather data for his senior thesis at Gustavus Adolphus College.

Team Echinacea 2018 at exPt2. From left to right: Gretel, Amy, Will, Evan, Morgan, Zeke, Mia, John, Anna, Kristen, Andy, Brigid, Riley, Michael (Stuart took the photo)

2018 update: Phenology in the remnants

In 2018, we collected data on the timing of flowering in 333 individual plants growing in our naturally occurring prairie remnants: 119 plants at Staffanson Preserve and 214 at others remnants. Flowering began on June 20th – four days earlier than last year. The last date of flowering was on August 9th – the latest bloomer was a roadside plant that had been mowed early in the season but put up another stem later in the season. Peak flowering for the remnants we observed in 2018 was on July 9th, which again was 4 days earlier than 2017. That day there were 257 individuals flowering. The figure below was generated with R package mateable, which was was developed by Team Echinacea to visualize and analyze phenology data.

From 2014-2016, determining flowering phenology was a major focus of the summer fieldwork, with Team Echinacea tracking phenology in all plants in all of our remnant populations. Stuart began studying phenology in remnant populations in 1996, but he didn’t know that keeping track of the dates was called “phenology.” In following years, several students & interns also studied phenology in certain populations. The motivation behind this study is to understand how timing of flowering affects the reproductive opportunities and fitness of individuals in natural populations.

Start year: 1996

Location: roadsides, railroad rights of way, and nature preserves in and near Solem Township, MN

Overlaps with: Phenology in experimental plots, demography in the remnants, reproductive fitness in remnants

Physical specimens:

  • Amy Waananen harvested some heads in fall 2018 and is germinating seeds right now at the U of MN. She is keeping track of which plant (mom) each seedling came from. She aims to use DNA fingerprinting techniques to identify the pollen donor (dad) of each seedling to get a sense of how far pollen moves in fragmented prairie habitat.

Data collected: We identify each plant with a numbered tag affixed to the base and give each head a colored twist tie, so that each head has a unique tag/twist-tie combination, or “head ID”, under which we store all phenology data. We monitor the flowering status of all flowering plants in the remnants, visiting at least once every three days (usually every two days) until all heads were done flowering to obtain start and end dates of flowering. We managed the data in the R project ‘aiisummer2018′ and will add it to the database of previous years’ remnant phenology records. Ask Amy Waananen for more specific data regarding phenology in the 2017 and 2018 seasons.

GPS points shot: We shot GPS points at all of the plants we monitored. The locations of plants this year will be aligned with previously recorded locations, and each will be given a unique identifier (‘AKA’). We will link this year’s phenology and survey records via the headID to AKA table. Ask Amy Waananen for more specific data regarding phenology in the 2017 and 2018 seasons.

You can find more information about phenology in the remnants and links to previous flog posts regarding this experiment at the background page for the experiment.

2018 update: Echinacea pallida flowering phenology

A pallida head. Notice the white pollen, which is the only 100% sure way you can be sure a head is pallida and not angustifolia

Echinacea pallida is an Echinacea species that is not native to Minnesota, but instead ranges East of the range of E angustifolia (and SE of our research site). In the summer of 2018, we identified 96 flowering E. pallida plants with over 200 heads that were planted in a restoration at Hegg Lake WMA. Every year for the past several years, we have visited the E. pallida plants, taken phenology data, and chopped off their heads. We do this to prevent E. pallida from being a bad pollen source or sink for native E. angustifolia populations. We were able to do this early this year, as E. pallida flowers significantly earlier than E. angustifolia.

We went back to check if we missed any heads on in September and found 3. They were done flowering, but hadn’t dropped seeds. We collected those heads, and they are currently stored at CBG. We hope that we might be able to germinate them for tissue. We want to analyze the ploidy of pallida compared to angustifolia. We have sneaking suspicions that pallida may be tetraploid where angustifolia is diploid.

Start year: 2011

Location: Hegg Lake Wildlife Management Area restoration

Overlaps with: Echinacea hybrids (exPt6, exPt7, exPt9),  flowering phenology in remnants

Physical specimens: 200+ heads were cut from E. pallida plants and removed then composted. We brought three heads back with us to Chicago Botanic Garden.

Data collected: All pallida data is in demap

GPS points shot: We shot points for all flowering E. pallida plants.

Products: In Fall 2013, Aaron and Grace, externs from Carleton College, investigated hybridization potential by analyzing the phenology and seed set of Echinacea pallida and neighboring Echinacea angustifolia that Dayvis collected in summer 2013. They wrote a report of their study. Pallida counts are being somewhat incorporated into demap.

Previous team members who have worked on this project include: Nicholas Goldsmith (2011), Shona Sanford-Long (2012), Dayvis Blasini (2013), and Cam Shorb (2014)

You can find more information about Echinacea pallida flowering phenology and links to previous flog posts regarding this experiment at the background page for the experiment.

2018 Update: Ground Nesting Bees

The tallgrass prairie once occupied vast expanses of land across America’s heartland. Today, it is among the most threatened and least protected habitats in the world. Each year, parts of the tallgrass prairie continue to be lost to agriculture and development making the conservation and protection of this system of utmost importance.

Native bees are the most abundant and most important pollinators in the tallgrass prairie. The bees that we study for this project are called solitary bees. They are different from honeybees in that they are native to North America. They are also different from bumblebees (where many genera are native to North America) in that they do not form a colony and build their nests individually.

We know a lot about the kinds of things bees like to eat (pollen and nectar) and their foraging behavior. However, most solitary bees spend the majority of their life in their nests, yet we know so little about what conditions are suitable for them to build their nests. In the tallgrass prairie, over 80% of bees are solitary, ground-nesting bees. We have a lot to learn about the kinds of habitat suitable for them to build their nests in.

We know some things about what ground-nesting bees may like. Evidence suggests they might like sandy soil, bare ground, and well-drained, south-facing slopes. However, we don’t know what bees in the tallgrass prairie may like for their nesting habitat conditions as most of these studies have been done across other ecosystems.

Much of the prairie has been changed from its original condition. We call the history of this condition “land-use history.” I am interested in how the history of the land may determine where bees build their nests in the ground. Some common types of land use history are remnant prairies which are pristine habitats with untilled soil, prairie restorations which are plantings of prairie plants with disturbed soil, and old fields which are fields leftover from agriculture that may have been tilled or grazed.

Using emergence traps, we moved traps everyday for a total of 1,440 across the season. We caught 110 ground-nesting bees in traps across 24 sites this summer. I placed traps at 8 different locations, each with three different land types at each location (remnant prairie, prairie restoration, and old fields). We found that the most bees nest in the prairie (40), while restorations and old fields have the same numbers of nesters (35). While land use is not good at determining bee nests, we did find that the location and land use when combined are both important in determining where bees nests.

I also placed pan traps at all 24 sites and caught 564 bees. Pan traps were colored blue, white, and yellow to attract a diversity of foraging bees at every site. We will use these bees to compare the foraging and nesting communities at each site.

I also measured many microhabitat characteristics of the soil and vegetation at some of the traps. We found that bare ground is a good predictor of where bees build their nests. We also found that the soil texture, especially the amount of silt and sand help determine where bees nest. A diverse plant community with lots of native plants is also a good predictor for bee nests.

We still have a lot more work to do to determine where bees are building their nests. Our next steps are to identify all the bee specimens caught in ground nests and in pan traps. Once specimens are identified, we can learn more about the species specific results for ground nesting bees.

Two of the tents used to capture bees out in the field

Start year: 2018

Location: Hegg Lake Wildlife Management Area restoration, Riley, Aanenson, East Elk Lake Road, and other non-project sites

Overlaps with: Pollinators on Roadsides

Physical specimens: 674 bees were brought back to CGB and are currently being pinned and photographed by Mike Humphrey. Soil samples were collected from every location where bees were caught + a random sample from other traps.

GPS points shot: We shot points for all trap locations. Ask/email Kristen for this data.

Products: This work is part of Kristen’s Master’s thesis

Previous team members who have worked on this project include: Anna Vold (2018)

Thanks so much to help from Team Echinacea 2018, especially Anna Vold who helped measure soil texture. Also many thanks to Emily Staufer from Lake Forest College who processed bees from HFW, and Mike Humphrey who has pinned some bees from this project.

2018 Update: Pollinators on Roadsides

A bumblebee on a yellow flower. We use yellow pan traps to mimic these Asteraceae

Pollinator diversity and abundance are declining due in part to land use change such as habitat destruction and fragmentation, pesticide contamination, among other numerous anthropogenic disturbances. The extent to which pollinator and native bee diversity and abundance is changing is not well understood, especially within tallgrass prairie ecosystems. Pollinators are important in the prairie and they provide valuable ecosystem services to native plants and to economically important plants used in agriculture.

In summer 2018, we collected bee specimens from 37 roadside sites using yellow pan traps. These sites are located within a gradient of various surrounding landscapes, some surrounded by natural areas, semi-natural areas, agricultural fields, development, or a mixture of the above. IN summer ’17 we sampled over 600+ bee specimens across 8 sampling weeks. IN summer ’18, we captured similar abundances of bees (~450 specimens) collected across 6 weeks. Once specimens are collected, they are stored in ethanol until we are able to pin them. Once specimens are processed, we catalog specimens and keep a record for later specimen identification. Identifying specimens to species requires specific, expert knowledge of the families and genera of native bees and pollinators in this ecosystem.

The goal of this experiment was to repeat a similar study done in 2004 by Wagenius and Lyon, in which they collected information on pollinator abundance and diversity. The aim of the project was to understand how landscape characteristics may influence bee community composition. The information from this project allows us to make comparisons between the pollinator communities collected in 2017, and a similar project from 2004. This information can inform diversity and abundance changes across the 13-14 years and provide valuable insight into native bee declines in this system.

Year started: 2004, rebooted in 2017

Location: Roadsides in and around Solem Township, Minnesota.

Overlaps with: Ground nesting bees (link to come)

Samples collected: Over 450 bee specimens, currently being pinned at CBG

GPS points shot: Locations for each of the pan trap sites

Team Members who have worked on this project include:  Steph Pimm Lyon (2004), Alex Hajek (2017), Kristen Manion (2017 & 2018), and John VanKempen (2018). Also, a big thank you to Mike Humphrey who has worked in the lab pinning, processing, and cataloging native bee specimens from the 2017 and 2018 field seasons.

You can find out more about the pollinators on roadsides project and links to previous posts regarding it on the background page for this experiment.

2018 Update: Echinacea hybrids– exPt 6

In 2018, we searched for 30 of the original 66 Echinacea hybrid plants. We found 29 Echinacea hybrids… which shows incredibly low mortality!  This means that 40% of the original cohort is still alive, with the survival rate this winter of more than 96%! Of the surviving plants, the average leaf count was 2.2 leaves, the longest basal leaf was 14.75cm. These plants are considerably smaller than their exPt9 counterparts, despite being several years older.

This big bluestem made finding these tiny plants pretty hard!

This plot was originally developed for Josh Drizin’s experiment with exotic grasses, but 66 hybrids of Echinacea angustifolia and Echinacea pallida were also planted in 2012. In 2011, Gretel and Nicholas Goldsmith performed reciprocal crosses between 5 non-native pallida plants found at Hegg Lake and 31 angustifolia plants in P1. These plants have been revisited each summer since then.

Year started: Crossing in 2011, planting in 2012

Location: Experimental Plot 6, on Tower Road

Overlaps with: Echinacea hybrids — ex Pt 7, Echinacea hybrids — ex Pt 9

Data collected: Status, rosette count, longest leaf measurement, and number of leaves for each plant. Exported to CGData.

Products: Nicholas Goldsmith wrote a summary of the crosses he conducted in 2011. A chapter of his dissertation, which he defended in December, reports on the fitness of hybrids compared to plants of either species.

You can find more information about experimental plot 6 and previous flog posts about it on the background page for the experiment.

2018 Update: Echinacea hybrids — exPt 9

In summer 2018, we again measured Echinacea plants in experimental plot 9 at Hegg Lake. These plants are from open-pollinated E. angustifolia plants near the restoration plot with flowering E. pallida plants. ExPt9 includes some hybrid plants, as determined by DNA fingerprinting techniques. The table below shows the number of plants found alive during each search since the experiment started in 2014. Much like last year, the average surviving plant had about 3 leaves. The average longest leaf was 21 cm, 4 cm shorter than in 2017. We suspect that leaves are shorter this year than last year (25cm in 2017 on average) because of a burn in the Hegg Lake WMA. This year we searched for plants once then rechecked every position where we didn’t find a plant during our first search. No plants flowered this year (no flowering plants yet!).

Year / Event Number Alive % Original remaining % Of previous year
Planting (2014) 746 100 N/A
2014 638 85.5 85.5
2015 521 69.8 81.7
2016 493 66.1 94.6
2017 401 53.8 81.3
2018 329 44.1 82.0

This experiment comparing the fitness of Echinacea hybrids with pure-bred E. angustifolia and E. pallida will give insight into the possible consequences of non-native E. pallida being planted in restorations in Minnesota, where E. angustifolia is the only native Echinacea to this area of MN.

Most exPt9 plants look like this!

Start year: 2014

Location: Hegg Lake Wildlife Management Area — experimental plot 9

Overlapping experiments: Echinacea hybrids — experimental plot 6Echinacea hybrids — experimental plot 7

Data collected: Rosette number, length of all leaves, herbivory for each plant collected electronically and exported to CGData. Recheck information for plants not found was also collected electronically and stored in CGData.

You can find out more information about experimental plot 9 and flog posts mentioning the experiment on the background page for the experiment.

2018 update: Demographic census in remnants

As always, demo was a huge job this year. This season we added 3868 demo records and about 1539 survey records to our database. After over 20 years of this method, we are starting to get a very good idea of the demography of Echinacea in Solem Township.

So how do we do demo? When we find a new flowering Echinacea plant, we give it a tag and get its location with a survey-grade GPS (better than 6 cm precision). Then, we can revisit this plant for years to come and monitor its survival and reproduction. We’re monitoring over 10,000 plants as of 2018. With this many records, organizing demap in the future is going to be a big task!

We made some small adjustments to how we do total demo and flowering demo (we do flowering demo in every remnant now). Now, in our big plots, we only do a sample of the plants instead of total demo. Check out the updated table below (remember we did flowering demo everywhere). We still managed to shoot thousands of points, take even more demography records, and, as always, follow the colored flags!

Total Demo BTG, Common Garden,  DOG, East of Town Hall, KJs, Krusemark, Landfill West, Loeffler Corner East, Near Town Hall, Nessman, NNWLF, NRRX,   recruit he, recruit hp, recruit hs, recruit el, recruit hw, recruit ke, recruit kw, RRX,  RRXDC, South of Golf Course, Steven’s Approach, transplant plot, Tower, Town Hall, West of Aanenson, Woody’s
Sample Demo Aanenson, Around LF, East Riley, Hegg Lake, Elk Lake Road East, Golf Course, Landfill, Loeffler Corner, North of Golf Course, NWLF, On 27,  Riley, Staffanson,  Yellow Orchid Hill

Evan and Will doing demo by the corn. While you might not be able to see the echinacea, we promise, it’s under there!

Year started: 1996

Location: Roadsides, railroad rights of way, and nature preserves in and near Solem Township, Minnesota.

Overlaps with: Flowering phenology in remnants, fire and flowering at SPP

Data collected: demo records include Flowering status, number of rosettes, number of heads, neighbors within a 12 cm radius of plants found. These are all taken with PDAs that sync with an MS Access database. They are all transferred to the demap R repository in bitbucket with git version control.

GPS points shot: Points for each flowering plant this year shot mostly in SURV records, stored in surv.csv. Each location should be either associated with a loc from prior years or a point shot this year.


  • Amy Dykstra’s dissertation included matrix projection modeling using demographic data
  • Project “demap” merges phenological, spatial and demographic data for remnant plants

You can find out more about the demographic census in the remnants and links to previous posts regarding it on the background page for this experiment.

2018 update: Fire and flowering in SPP

For her REU project, Brigid gathered data to study the relationship between flowering density and seed set. She worked at Staffanson Prairie Preserve, which appears to have higher flowering density in burn years than non-burn years. This year, 2018, was a burn year on the east side of the preserve. Brigid and Team Echinacea kept track of the style persistence of about ~150 individuals many of  which we have phenology and style persistence information from prior years. These individuals were harvested and their achene count and seed set will be assessed by volunteers  and  interns at the CBG.

Brigid also observed nearest neighbors for many of the plants that she tracked. It might be the cases that echinacea flowers are more successful if they have other flowering plants nearby. Synchrony is a large part of why fire is so important, and, since SPP is our largest remnant prairie, it’s the best place to test the relationship between fire and synchrony. Number of heads, phenology, and head size may also \ interact with fire — we’ll know once we look at the data!

Site: Staffanson Prairie Preserve

Start year: 1996

Overlaps with: Phenology in Remnants, Reproductive Fitness in Remnants

Data and Samples: We shot 90 GPS points for nearest neighbors, many of which were plants that flowered for the first time this year. We also harvested 22 heads that are awaiting cleaning at CBG

Products: None so far

2018 update: Common garden experiment–1996 cohort

In 2018 only 19% of the plants flowered, despite it being a burn year. Is the 1996 finally showing its age?

In 2018, 51 plants flowered of the surviving 269 plants in the 1996 cohort. That means that 41% of the original plants are surviving and 19% of the living individuals flowered. That’s up huge since last year, where only 2% flowered, and the year before where five percent of living individuals flowered. In contrast, however, 45% of living plants flowered in 2015, and  37%, 34%, and 40% flowered in 2014, 2013, and 2012 respectively. We found that of the original 646 individuals, 269 were alive in 2018, only 15 fewer than last year. We are not sure why so many more plants flowered this year. It’s probable that the fire in the plot in fall 2017 influenced flowering rates.

The 1996 cohort has the oldest Echinacea plants in experimental plot 1; they are 22 years old. They are part of a common garden experiment designed to study differences in fitness and life history characteristics among remnant populations. Every year, members of Team Echinacea assess survival and measure plant growth and fitness traits including plant status (i.e. if it is flowering or basal), plant height, leaf count, and number of flowering heads. We harvest all flowering heads in the fall, count all achenes, and estimate seed set for each head in the lab. As yet, these heads are still waiting to be cleaned April 2019.

Start year: 1996

Location: Experimental plot 1

Overlaps with: phenology in experimental plots, qGen3, pollen addition/exclusion

Physical specimens:

  • We harvested 59 heads. At present, they await processing in the lab to find their achene count and seed set.

Data collected:

  • We used Visors to collect plant growth and fitness traits—plant status, height, leaf count, number of flowering heads, presence of insects—these data have been added to the database
  • We used Visors to collect flowering phenology data—start and end date of flowering for all individual heads—which is ready to be added to the exPt1 phenology dataset
  • Eventually, we will have achene count and seed set data for all flowering plants (stay tuned)


  • See the exPt1 core dataset where yrPlanted == ‘1996’ for 1996 cohort fitness measurements
  • Amy Waananen’s paper, Mating opportunity increases with synchrony of flowering among years more than synchrony within years in a nonmasting perennial, published last year in The American Naturalist, was based on plants in this cohort.

You can find more information about the 1996 cohort and links to previous flog posts regarding this experiment at the background page for the experiment.