Landscape Ecology of Singing Insects 2: Human Influences

by Carl Strang

The previous post illustrated that the Chicago region has been a dynamically changing landscape through the recent millennia, but that now is overshadowed by the alterations our own species has made. Burgeoning human numbers have overwhelmed the planet’s ecosystems, and the native habitats described earlier mostly have been replaced by agriculture and urban growth in the Chicago region. One of the more dramatic changes is the loss of the Kankakee wetland, once described as the Everglades of the North. That vast wetland was drained for agriculture, and only a few pockets of it survive in preserves. Much of the Kankakee River in Indiana is now a straight channel with constructed high levee banks. Other smaller wetlands received similar treatment, with drainage ditches spreading across the agricultural portion of the region. This is not universally devastating to wetland species. Northern mole crickets, for instance, occasionally can be found along drainage ditches.

Drainage ditch, upper reaches of the Kankakee River, St. Joseph County, Indiana. Note farm fields on both sides.

Prairie mostly has been replaced by agricultural fields, and fire suppression has led to its invasion by woody plants. Specialists such as the prairie meadow katydid, prairie cicada and short-winged toothpick grasshopper are hard to find.

Prairie meadow katydid

My singing insects research has required a lot of driving to reach the relatively tiny surviving preserves and parks to which many of the species are now restricted. Much management effort is required in these little islands to maintain their habitats. There are exceptions, of course. Many other species have thrived under our influence. These are mainly weedy ones such as the striped ground cricket, short-winged meadow katydid and Carolina grasshopper, which do well in disturbed habitats, along with woodland edge species such as the greater angle-wing, snowy tree cricket, and jumping bush cricket, which can meet their needs in residential neighborhoods dominated by lawns and scattered trees and shrubs.

Jumping bush cricket

Habitat destruction is not the only human influence. Climate change is the probable cause of northward range expansions by several singing insect species, and it likely will lead to the extinction of the sphagnum ground cricket from the region as the sphagnum bogs dry up. Say’s cicada and some northern grasshoppers already appear to be pushed out.

Sphagnum ground cricket

Climate change isn’t simply a matter of rising temperatures, as the term “global warming” may seem to imply. Global warming is an accurate enough term, as the simplest way to measure climate change is to track the global average temperature. But the point is that our changes to the Earth’s thin skin of atmosphere are increasing its held solar energy. That energy alters patterns of atmospheric flow and the behavior of storms. Droughts, more frequent flood-causing rains, and seasonal increases or decreases in temperature that seem abnormal are examples of results we observe locally. The singing insects are forced to adjust as best they can. Droughts force sphagnum ground crickets into the wettest parts of their bogs. The severe drought of 2012 concentrated wetland meadow katydids and marsh coneheads into the small portions of the Great Marsh in the Indiana Dunes National Park that remained wet. Oblong-winged meadow katydids may be pre-adapted to such year-to-year variability. Blatchley (1920) observed that their eggs, laid in moist soil, can take 2-3 years to hatch. In my travels through the region I failed to hear a single individual in the years 2010 and 2019, but in other years they have been abundant and widespread. Some of the cicadas and other species may have similar flexibilities.

Oblong-winged katydid

People also have introduced plant species from other parts of the world which, released from the consumers and competitors which hold them in check in their native lands, have become invasive plants here. Their unfair competitive advantage has led to their displacing the region’s native vegetation in an increasing number of places. This is most evident in our wetlands. Wetland meadow katydids and other singing insects are limited to places where native wetland grasses have not been supplanted by reed canary grass, common reed, purple loosestrife, and hybrid cattails. These invasive plants are proving difficult to control, and the outlook is not good for species such as the dusky-faced meadow katydid and marsh conehead. To my knowledge the once relatively widespread stripe-faced meadow katydid now is confined to a single site, and the slender conehead, never known from many locations, apparently is gone from the region.

Stripe-faced meadow katydid

Introductions have not been limited to plants. Several species of singing insects also have been imported. Roesel’s katydid is the most common of these in our region. A European predaceous katydid, Roesel’s was introduced to the Montreal, Quebec, area several decades ago and expanded from there. They occur in open habitats with tall herbaceous vegetation throughout the Chicago region. Japanese burrowing crickets are thought to have arrived at the port of Mobile and spread out from there. They are abundant as far north as Indianapolis, and common in Rensselaer in the southern part of our region. With scattered new appearances each year occurring as far north as DuPage County in Illinois, so far, I expect them to become widespread and abundant here. The tropical house cricket represents the possibility of other, short-term introductions that are unlikely to persist in our climate.

Roesel’s katydid

Wetland Singing Insects Update

by Carl Strang

For years, now, my biggest conservation concern among the singing insects has been in the wetlands. Though our historically abundant prairies in the Chicago region were diminished nearly to nothing by 19th– and 20th-century agriculture, preservation and restoration projects across the region have halted and, to a small degree, reversed that trend. The same could be said for savannas, and our forests did not suffer as much.

Wetlands, like prairies, declined thanks to agriculture, but a new challenge continues to threaten their integrity: invasive wetland plants. Four of these are especially problematic: common reed (Phragmites australis), reed canary grass (Phalaris arundinacea), hybrid cattails (Typha x glauca), and purple loosestrife (Lythrum salicaria). These plants, released from consumers and competitors, have displaced the diverse native species in a large and increasing portion of our wetland acreage. The loss of native wetland grasses, especially, appears to account for the difficulty I am experiencing in finding wetland katydids.

Reed canary grass

Two species that were here historically, I have not found at all: the delicate meadow katydid (Orchelimum delicatum), and the slender conehead (Neoconocephalus lyristes). In the past these were known to occur in four and three, respectively, of the Chicago region’s 22 counties, and I am nearly out of places to check where they might still live. The stripe-faced meadow katydid (O. concinnum), once found in 8 of the counties, appears to be down to a single population at Illinois Beach State Park.

Stripe-faced meadow katydid

Two wetland katydids are doing well. Gladiator meadow katydids (O. gladiator) and black-legged meadow katydids (O. nigripes) are tolerant of the invasive plants, and remain common in every county.

That leaves an in-between category of wetland singing insects that apparently are limited to invasives-free wetlands, and are managing to hang on in a few to several sites. Northern mole crickets (Neocurtilla hexadactyla) occur in wet prairies as well as marshes. In 2017 I added records for two more sites, one of which represented an additional county record. To date I have found them in 10 counties, and remain optimistic that I can add more populations to the inventory.

Dusky-faced meadow katydids (O. campestre) historically were ubiquitous in our marshes. To date I have found them only in marshes with minimal impact by the invasive plants. These katydids seem able to persist in relatively small wetland areas, however, and each year I have been able to add new populations to my list. In 2017 I found them in the Indiana Kankakee Sands preserve, adding Newton County to the record, and in the Tefft Savanna preserve in Jasper County, also a county record. That brings to seven the number of counties where I have found the species, but there are seven more where it once lived, but where my search has been unsuccessful. Dusky-faced meadow katydids also proved this year to be abundant in the panne wetlands at West Beach in the Indiana Dunes National Lakeshore. That was a good find, but I had hoped for delicate meadow katydids there.

Female dusky-faced meadow katydid, Tefft Savanna

Finally, this year I added a third population and county for the nimble meadow katydid (O. volantum). They were singing from arrowheads (Sagittaria sp.) mixed with cattails along Grant Creek in the Midewin National Tallgrass Prairie. That find was made from a kayak, and that is the vehicle from which further searching for the species will need to happen, as this species likes plants growing in relatively deep water. Some places which historically held nimble meadow katydids no longer have them, but several other sites remain for me to check in future years.

Literature Review: A Surprisingly Carnivorous Plant

by Carl Strang

Among the most insidious of our invasive plants are two species of teasels, unfortunate imports from Europe.

Here’s one of them, the cut-leaved teasel.

One of the studies which caught my eye this past year demonstrated that the other species behaves like a carnivorous plant (Shaw PJA, Shackleton K (2011) Carnivory in the Teasel Dipsacus fullonum — The Effect of Experimental Feeding on Growth and Seed Set. PLoS ONE 6(3): e17935. doi:10.1371/journal.pone.0017935). Dipsacus fullonum, AKA D. sylvestris, the common teasel, has water-holding leaf bases. Shaw and Shackleton experimentally added insects to these little cups, and found that this led to improved seed set and a higher proportion of plant biomass in seeds. Plant growth and total biomass were connected more to rosette size, which in turn is indicative of first season, non-carnivorous growth in this biennial.

Garlic Mustard Study Continued (Seedlings)

by Carl Strang

In the previous post I looked at this year’s results in my study of alternative manual removal methods for garlic mustard, an invasive biennial. In summary, uprooting the plants is effective any time. Cutting them off at ground level also is effective, but more so in April than in March. Today I want to shift the focus to first year plants, or seedlings.

I have been following the development of seedlings in my study plots, in part out of general curiosity and in part to test the possibility that uprooting second-year plants stimulates additional seed germination. Last year, when I uprooted relatively small second-year plants in March, there was no increase in seedlings then or in the following year. This year I uprooted the older plants in April, when they were larger, so there was a greater possibility of impact.

Numbers of seedlings prior to treatments in study plot square meters were not different statistically, with median counts of 64 in clipped squares and 32 in uprooted squares (I did not count seedlings then in controls, as the process would have been too much of a disturbance). When I returned to assess the results at fruiting time I found median seedling counts of 65 in uprooting treatments, 67 in clipping treatments and 47 in controls, with enough overlap in ranges of counts that there were no statistically significant differences in pairwise comparisons.

Last year I found that seedlings in control squares were fewer at the end of the reproductive season, and had been suppressed in their growth. This year I found no such difference. This implies that the advantage of seedling suppression is achieved by waiting until April to remove second-year plants. Also, pulling second-year plants in April did not increase numbers of seedlings in the same year. I will need to return next spring to see if the soil disturbance has an effect the following spring.

Garlic Mustard Study Continued (Mature Plants)

by Carl Strang

I have the first results from this year’s study plots in my garlic mustard control study. This year I applied treatments of uprooting versus clipping second-year plants in mid-April, checking the effectiveness of these removal techniques when the plants were developing fruits a month later, and comparing to controls. I also wanted to see how the data would compare to last year’s study plots, where I applied the same treatments in mid-March.

First I compared the numbers of plants prior to treatments. Median counts in the 9 square-meters assigned to each option were 288 in clip treatment squares, 276 in pulled treatment squares, and 233 in controls. Pairwise comparisons of ranked counts resulted in no statistically significant differences, indicating an acceptable uniformity in starting conditions.

When I returned in May I found only a few tiny plants in treatment squares, plants which I had either missed or mistaken for seedlings when applying the treatments. Median counts in May were 4 plants in clipped squares, 1 plant in pulled squares. There was no statistically significant difference between treatments.

I also wanted to see whether there were differences between last year’s March treatments and this year’s April treatments. There was no difference in effectiveness of pulling plants in the two months. The median of 3 post treatment pulled plants last year was close to this year’s median of 1, and there was no statistical significance in the comparison of ranks. In other words, pulling is very effective early or late in the season. Clipping produced different results, however. March clipping resulted in a median of 24 recovering plants, compared to 4 after April clipping. The difference was statistically significant (no overlap in the ranges of counts produced a Mann-Whitney U statistic of 0, easily meeting my 1% error criterion). Clipping has to be regarded as effective even in March, because that median of 24 compared to an initial median count of 178 plants per square meter.

This clipping treatment was a slow process, with plants removed individually at the soil’s surface. Next year I want to test a more realistic mass removal process.

March Garlic Mustard Pulling Does Not Encourage Germination

by Carl Strang

I have resumed my tests of manual removal methods for controlling garlic mustard. This invasive plant has become a significant bane of native ecosystems, and considerable expense and effort go into knocking it back. Last year I set up my first experimental plots, comparing the removal methods of pulling (uprooting) second-year plants, versus clipping them at ground level in March. Untouched control plants I later clipped when they began to mature their fruits. This year I applied the same treatments to new study plots in April (the difference in timing is exaggerated by this year’s advanced phenology), and those results still are developing. For now I want to share some of the final results from last year’s plots.

One concern raised by restoration practitioners is that uprooting second-year garlic mustard plants may increase germination of garlic mustard seeds in the soil. Last year I found that this was not true for the year in which the pulling is done. There was no statistically significant difference between the pulling and clipping treatments. Seedling counts were lower in the control subplots, apparently because of suppression by the undisturbed second-year plants.

New Garlic Mustard Seedlings

Recently I returned to last year’s study plots to see if the pulling treatment produced increased seedling germination in the following year. I found plenty of seedlings in control (median count 198 seedlings per square meter), clipped (median 293 seedlings) and pulled (median 277 seedlings) treatment squares (9 squares total in each category, from three 3×3–meter study plots). Applying the Mann-Whitney U-test to the count ranks of the squares, I got U values of 30 for control vs. pulling treatment, 32 for control vs. clipping treatment, and 40 for pulling vs. clipping. None of these are anywhere close to statistically significant.

Older Garlic Mustard Seedlings

So far it appears that pulling the plants is OK. However, the second-year plants are relatively small in March. This year, pulling the larger (in flower) April plants has disturbed the soil to a greater degree, so I will be interested in seeing whether that makes a difference.

Garlic Mustard Seedling Survival

by Carl Strang

In the spring I began a study of how garlic mustard, a harmful exotic biennial, might best be controlled by manual means. In small areas where the plant is just beginning to invade, and where use of herbicides is undesirable, it is possible to uproot or clip the second-year plants. Results so far indicate that pulling is more effective than clipping, but there is a timing variable to investigate, and I need also to determine whether pulling stimulates an increased germination of seedlings in the following year.

GM October 1b

Last week I returned to my study plots to count seedlings at the end of their first season. As the above photo shows, some tree and shrub leaves had fallen, so I carefully removed these to make sure my seedling counts were complete.

GM October 2b

I had expected some attrition through competition, but was surprised at the numbers of seedlings that had died. Every single one of the 27 square meters in the study plots showed big drops in numbers of seedlings, even in cases where there were so few that competition between them would seem to be negligible. Where in May seedling counts ranged from 12 to 345 in the square meter areas, in October the counts were 0 to 55. Especially dramatic were the control squares, in which second-year plants had been allowed to proceed to fruiting before I clipped them. There, seedlings had looked weak, but plenty still remained in May. However, the total of 214 seedlings in May had dropped to only 3 seedlings in the 9 square meters of the control treatment by October. Apparently their inhibition by the second year plants had been too great for them to overcome. Attrition in pulled treatment squares had been from 747 to 236 between May and October, and the corresponding numbers for clipped treatment squares were 1002 and 107. Statistical computations supported the difference between controls and both treatments in October counts, but indicated no statistical significance between the two treatments.

GM October 3b

Now I wait for spring. I plan to set up new study plots next year, but will apply the same treatments a month later, to see what difference timing makes. I also will return to this year’s plots. I want to follow this year’s seedlings through to their fruiting times, and to see if the numbers of new seedlings in those squares support or reject the notion that pulling increases seed bank germination.

Garlic Mustard Study: Final Spring Results

by Carl Strang

One of my studies this spring has been an experimental comparison of removal techniques for garlic mustard, an invasive biennial that poses problems in our woodlands because it inhibits the growth of all other plants (including trees).

GM bolting b

A month ago I removed all the second-year plants from my study plots. As I reported then, it was clear that pulling the plants in March was more effective than was cutting them off at ground level, though both treatments killed most of the plants. In the following photo you can see an example of new side shoots springing up from where the main stem had been clipped.

GM clipped recovering b

In that May treatment I also clipped the control plants at ground level. After waiting four weeks I recently returned to see whether any second-year plants had recovered to send up new stalks. I also wanted to check progress of seedlings, which had been suppressed by the control plants but had been growing vigorously in treatment areas.

The results again were impressive. All but 6 of the 1482 control plants (99.6%) were killed by the May clipping treatment, in contrast to clipping in March. None of the few survivors of the pull treatment that I cut in May survived. More (17) second-year plants survived their second clipping.

GM 15JE 1b

At the moment it appears that early pulling and late clipping both are effective techniques, but I have more tests to do next year. The jury still is out on the question of whether (or under what conditions) pulling may increase germination from the seed bank in the soil.

After a month the seedlings in the control squares still were far from making up the ground they lost in comparison to seedlings in treatment squares.

GM 15JE 3b

The next step in this study will be a return to count the seedlings in the fall. I am interested in how much they thin themselves through competition with one another.

Garlic Mustard Study Update

by Carl Strang


Since treating the garlic mustard plots , I have returned weekly to check on them. As of April 16, the seedlings have barely progressed beyond the cotyledon stage.




Untreated (control) plants continue to add leaves and expand in size, though they remain rosettes.




Some plants have survived the treatments, either because they were so tiny that I missed them or because pinching them off did not in fact kill them.




Later I will examine those survivors in the pinch treatments to see which of these possibilities was the case. A few new plants have appeared each week since the treatment. As of April 16, a total of 16 plants had appeared in the 9 squares where I had pulled plants, and 76 plants had appeared in the 9 pinch treatment squares. Where I had counted 418 seedlings in pinch treatments and 450 in pull treatment squares at the start, by April 16 these numbers were 838 and 944, respectively. At the end of the study period I will apply statistical tests to determine whether these differences are due to chance or to real differences between the treatments. I also will return next year, to see if more seedlings emerge in pull treatment squares than in the others.

Garlic Mustard Removal Study

by Carl Strang


One of the challenges facing people trying to restore biodiversity to native woodlands is invasive plants. Earlier  I outlined the general problem in the context of shrubs. The herbaceous plant causing the most trouble in our woodlands is garlic mustard. Garlic mustard is a biennial. Seeds sprout in spring, grow into rosettes of leaves that survive the winter, then the plants grow up, flower and produce seeds in their second spring. At this point in the season the rosettes look like this.




When I was at Willowbrook I undertook to remove garlic mustard from the fenced area that contains the outdoor animal exhibit. When I was transferred to Fullersburg I did the same in the Wildflower Trail area. Both were high quality areas in terms of the native plants that were present, but garlic mustard was expanding, and suppressing the native wildflowers. I started pulling out the garlic mustard each spring, and was gratified by the quick recovery by diverse native species. Then I began trying fall and winter pulling, and got good results with that, too.


My annual review of the scientific literature last December turned up a study* describing success in controlling garlic mustard by clipping the plants rather than uprooting them. I decided to try a study of my own, that would compare results of uprooting plants versus breaking them off, early and late in the season. Yes, a study had been done. But replication is important in science. I wanted to see for myself that new plants can’t grow up from the decapitated roots. Also, the earliest the other researchers had clipped their plants was late April. I wanted to try it earlier. My move to Mayslake was timely, because the restoration program is relatively advanced there, and I can focus on research rather than rescue.


I have set up 3 study plots, each 3 by 3 meters. I used large nails to mark the corners.




When the time came to treat the plots, I temporarily outlined the plot with bright orange string.




Each of the 9 square meters in each plot gets one of 3 treatments: uprooting, breaking off below the lowest leaves in mid-March, and a control that will be left until the plants are about to flower. At that point I will cut them off below the lowest leaves. This last treatment follows a practice recommended by some experienced restoration specialists, who discourage uprooting because it may stimulate germination of garlic mustard seeds in the soil. I used a random number generator (easy to find on the Internet) to determine which squares got each treatment.


Here is an experimental square from which plants were removed.




And here is a control square with the plants still in place.




I counted the plants in each square as I treated them. In the three study plots combined the total number of pulled plants was 1395, pinched off total 1617, and control plants 1176. The overall average density was 155 plants per square meter. Now, I wait and see what will happen. I’ll provide an update later.


(It should be obvious, but I’ll state it anyway, that I could do this kind of manipulation on public land as a forest preserve district employee, but still had to get clearance to do so.)


*Here is the reference for the study I mentioned: Pardini, Eleanor A., Brittany J. Teller, and Tiffany M. Knight. 2008. Consequences of density dependence for management of a stage-structured invasive plant (Alliaria petiolata). Am. Midl. Nat. 160:310-322.

%d bloggers like this: