Taken by Ants in the Prairie Peninsula

Traipsing around in an open savanna under gnarled, seemingly ancient, Bur Oaks trees is like being in a little slice of heaven.  But why are these gigantic mounds of dirt lying about?  

"Solarium"  constructed by Allegheny Mound Ants, central Ohio

The denizens of these mounds quickly showed themselves; large two-colored ants, with red heads and thorax and  large black, shiny gaster or "metasoma".  Based on their appearance and distinctive mounds, I believe they were Allegheny Mound Ants (Formica exsectoides).  

Allegheny Mound Ants (presumably Formica exsectiodes), Crawford Co, Ohio 

Location of  Bur Oak savanna (bounded by the black box) amidst
active agriculture of central Ohio

The site is truly remarkable in the modern day landscape, and has been considered to be the "largest and best preserved remnant of the unplowed, deep soil prairies and savannahs that were present at the easternmost extension of the Prairie Peninsula prior to European settlement" (Mack & Boerner 2004, Castanea).

Bur Oak Savanna dominated by Quercus macrocarpa, Crawford County, Ohio.  

What ecological role could ants be serving in this habitat remnant?

Previously, at this site I had noticed ants clamoring on Sullivant's Milkweed (Asclepias sullivantii). Not being an ant taxonomist, I struggled with whether or not these were the same species as those from the nearby mounds. With the gracious assistance of Dr. Kaloyan Ivanov of the Virginia Museum of Natural History they are tentatively identified as Formica pallidefulva, not the Allegheny Mound Builders.  These ants seemed to be daintily sipping nectar from the flowers (see image below).  Could they be pollinating this prairie species?  

Ants (presumably Formica pallidefulva) sipping nectar from Sullivant's Milkweed in Bur Oak Savanna, Crawford County, Ohio. 

Perhaps, but after a bit of research I learned that ants rarely pollinate flowers.  More often than not, they simply steal nectar, and are often referred to as "nectar robbers".  For example, Wyatt (1980, Bulletin of Torrey Botanical Club) found that ants on a different milkweed species both reduced the amount of nectar available to butterflies and the overall effectiveness of the pollinators.  He termed ants "harmful to evolved pollination systems", and a "maladaptive force".   I looked at further images I had taken of the Sullivant's Milkweed with ants and noticed what appeared to be "maladaptive" behavior taken to a new level.  It appears that the ants were actively chewing away the flower structures, literally tearing open the stigmatic slit (where a pollinator leg would be expected to enter), impacting or even removing the ovary! (see image below).

Ants impacting Sullivant's Milkweed reproductive capacity(?);
arrow illustrates one example of torn flowers & ovary

Damaging milkweed!  Interrupting butterflies!  This behavior must stop! 

Although Allegheny Mound Ants weren't apparently responsible for the milkweed damage, much of  the readily available information on them involved recommendations like, "drenching a mound with liquid insecticide".  Michigan State Extension justifies doing so because, "the ants kill nearby vegetation including small tress and shrubs to keep shade off the mound. They do this by chewing a small hole in the bark and injecting formic acid into the wound" (http://msue.anr.msu.edu/news/mound_ants).  In contrast, Maine State Extension actually recommends ENCOURAGING their populations because the ant "plays a vital role....in a variety of ecosystems" (https://extension.umaine.edu/blueberries/factsheets/insects/195-allegheny-mound-ant/). 
Nice!

Henry McCook ~ 1877 with his beloved
Allegheny Mound Ants

Reverend Henry McCook (image left), was possibly the first and most careful observer of the Mound Ants. At one point he actually dug up an entire mound, frozen in the dead of winter, and transported it intact to the Academy of Natural Sciences in Philadelphia. His assumption that the colony "had gone down for winter-quarters into the deep galleries below the surface" was proven wrong when the mound thawed along with the "horde of hibernating ants within" which poured out all over the museum (1907, Nature's Craftsmen Popular Studies of Ants and other insects).     





In his earlier work, McCook (1877, Transactions American Entomological Society) described the most common size of mounds as 10-12 feet in circumference and 2.5 - 3 feet in height, but also mounds of tremendous size (over 6' in height and 50' in diameter). Andrews (1926, Pysche) suggested the Allegheny Mound Ants were the most common mound builder in North America,  ranging at least from Maine through Georgia.  However, he also suggested they were most often associated with places "left comparatively free from live stock and human interferences".  When encountered, mounds can be quite abundant. The incomparable Reverend McCook reported densities ranging from 25-59/acre.  My observations (see image below) suggest densities can be even higher in some locations.  High population densities, coupled with their reported tendency to kill nearby small trees and shrubs could certainly play a role in maintaining the open structure of prairies and savannas they inhabit.

Allegheny Mound Ant density, evident after prescribed fire, Daughmer Savanna, Ohio

Allegheny Mound Ant surfacing with soil aggregate
Adams County, Ohio

Overall, ants are probably the most consistent and important soil turning organisms worldwide (Tschinkle, 2015, PLoS ONE);
numerous papers describe their contributions to mixing, aerating, and altering nutrient dynamics of  soils they inhabit, and several document vegetation associations with mound ants (see Nemec, 2014, Journal of Insect Conservation).

Consider just the amount of surface area directly affected by the mounds in the image above. Baxter & Hole (1967, Proc. Soil Science Society Abstract, ) studied mound ants of a related species in a prairie remnant in southeastern Wisconsin. They showed the ants (which occupied 1.7% of the surface area) significantly altered the soil surface by bringing material upward and introducing organic material from their feeding habits. Overall, they believed the ants may be responsible for distinguishing the prairie soil from that of an adjacent and comparable forested area.

  In the end, it seems, the ants may create the prairie!

Hairy and Hoary Puccoon; Inbred Cousins & Implications for Restoration?

Hoary Puccoon (Lithospermum canescens), Halifax County, VA

Hairy Pucoon (Lithospermum caroliniense), Sussex County, VA

Distribution Maps (County-level)

from https://plants.usda.gov

Hairy Puccoon  (Lithospermum caroliniense) (top) has widely disjunct northern and southern populations, with an extreme outlier in southeastern Virginia.








Hoary Puccoon (Lithospermum canescens)  (bottom) is strongly centered in the midwestern U.S., gradually dribbling into reasonably adjacent areas of the southeastern U.S








Although apparently closely related "cousins" (Cohen, Cladistics 2011) these species rarely overlap except in parts of the midwestern U.S.  L. caroliniense is largely restricted to deep sandy soils. The outlier population in southeastern Virginia is found in historic longleaf pine savanna habitat (see image below), superficially similar to many of the sites for this species in the West Gulf Coastal Plain, East Gulf Coastal Plain of Florida & Alabama, and portions of Florida & South Carolina. 

Lithospermum caroliniense in longleaf pine habitat, Sussex Co, VA

Some of the northern populations of Hairy Puccoon are found in the dry, sandy "oak openings" region of Ohio, Michigan, and Indiana along with species such as Sundial Lupine (Lupinus perennis).  

Lithospermum caroliniense in Black Oak sand barren habitat, northwestern Ohio

Hoary Puccoon occurs a wide range of habitats, and a larger range of soil textures. An example includes heavy "vertic" clay derived from diabase rocks in the Piedmont region of the Mid-Atlantic.

Lithospermum canescens in central Piedmont Diabase Woodland, Durham County, NC, with Clematis ochroleuca a Mid-Atlantic endemic 

Both Lithospermum species produce distylous flowers, with two morphological variations potentially present in a given population. These "different forms of flowers on plants of the same species" (to copy the title of Charles Darwin's 1877 work) may ultimately affect the reproductive potential of both the individual and the population as a whole. These flower morphological types, essentially alternate arrangements of the anthers and stigma, can be seen in the the following examples of  L. canescens.  The uppermost image has "pin" or 'pin-eyed" flowers (stigma elevated above the anthers, which are not visible) while the lowermost has "thrum" flowers (anthers elevated above stigma which is not visible). Theoretically, this pattern should help ensure cross pollination and minimize self-fertilization. McCall (American Journal of Botany, 1996) indicates that little to no seed production occurs when distylous flowers are either self  pollinated or crossed with flowers of the same morph.

Pin flowers in Lithospermum canescens, Durham County, NC

Thrum flowers in Lithospermum canescens, Durham County, NC

This suggests the importance of a morphologically balanced population, consisting of relatively equal numbers of  pin and thrum flowers, to maintain successful reproduction. 

Molano-Flores (Proc. 17th Annual Prairie Conference, 2001) recognized the potential importance of the proportion of flower morphs and examined 17 different L. canescens populations in Illinois. In general, she found many skewed population ratios but also some with equal proportions; importantly she noted differences between "restored" and natural populations.

As suggested above, if one flower morph is disproportionately represented the result may well be low or no seed set. Molano-Flores suggest this problem can be compounded when populations are especially small and/or isolated.

Lithospermum canescens seedling; Granville County, NC
(this population supports approximately 50 individuals at last count)

Given the overall distribution maps (presented above) coupled with  localized patterns of habitat fragmentation, it is entirely possible this phenomenon could suggest a threat to remaining populations of both Lithospermum species.  Interestingly, Kirster & Levin (Genetics, 1968) inferred that colonies of L. caroliniense are highly inbred and exhibit pronounced gene flow restriction. Levin (American Journal of Botany, 1972) found a higher proportion of thrum flowers and overall low levels of seed production, minimal self-fertility, and minimal compatibility between pins and thrums. Weller (American Journal of Botany, 1980) showed that L. caroliniense has both an inefficient pollination system and intrinsically low fecundity. McCall (American Journal of Botany 1996) found that pins produced somewhat more flowers and fruits than thrums, but that thrums were slightly more common in the population of L. caroliniense she studied. 

Monarch feeding on Hoary Puccoon

How viable are the remaining, isolated populations?  Cross pollination appears to be important but Kirster & Levin (Genetics, 1968) found that pollinators (both bumblebees and Lepidopterans) rarely moved more than 10 meters from a given plant!

Given this observation and the apparent low reproduction reported by several authors it is probably especially important that at least one of the species, L. caroliniense, can live as long as suggested by Weller (Ecological Monographs, 1985).

Hmm?

  

A most marked new plant & reverse nyctinasty?

Kate's Mountain Clover (Trifolium virginicum) with inflorescence and  horizontal leaves in shaded portion of shale barren
Douthat State Park, VA (May 13, 2018)

Kate's Mountain Clover (Trifolium virginicum) was first discovered during a plant collecting trip to "southwestern Virginia" by famed botanist J.K Small in 1892.  Observed on the rocky slopes of Kate's Mountain, Greenbrier County, West Virginia, Small's published account (1893, Memoirs of the Torrey Botanical Club) indicated this species to be "the most marked new plant collected on the expedition", and distinct from all other eastern American species.  Today, we recognize the global range of Kate's Mountain Clover extending across the Valley & Ridge region of VA, WV, MD, and PA as shown on the map below.

Trifolium virginicum range,
https://plants.usda.gov/core/profile?symbol=TRVI3

A 1908 botanical monograph on North American Clovers by McDermott lumped Small's Clover with Trifolium reflexum and suggested that the species was "abundant throughout the Appalachian Mountains".  Small's opinion on this treatment was documented by Hunnewell in a brief note entitled, "A new station for three local Appalachian Plants" (Rhodora, 1923). Small was quoted as follows, "to say that the species is common in the Appalachian Mountains, may be prophecy, but such a statement certainly cannot be backed by good evidence".

Prophetic or not, many botanists have focused their attention on Trifolium virginicum since then. At one time or another, all Natural Heritage Programs within the natural range have tracked the species. Collectively, the species has been assigned a "G3 or vulnerable" rank by NatureServe. The West Virginia Native Plant Society is justifiably proud of both the plant and its discovery and has adopted the species as their official symbol. Kate's Mountain Clover has also been considered the "flagship species of the shale barrens" (Maryland Botanical Heritage Working Group, 2014).

Shale Barren habitat @ Douthat State Park, VA

Although Small did not refer to the habitat for his collection as a "shale barren" (the term did not come into common usage until approximately 1911), there are well known shale barrens on Kate's Mountain and some authors have indicated Trifolium virginicum as a shale barren endemic (Braunschweig, Nilsen & Wiebolt 1999).  The species has since been rarely encountered on related habitats, but is still most often associated with shale barrens (see http://vaplantatlas.org/index.php?do=plant&plant=3623). 

The typical shale barren habitat is a harsh environment for plant growth. The surface is often unstable with loose rock fragments and minimal soil development & the typically steep slopes tend to shed any litter accumulation down slope where it accumulates at the toe slopes (See images below).

Shale barren fractured rock surface consisting of thousands of "channers"

Toe slope of shale barren/woodland;
green line roughly indicates boundary of litter accumulation  

According to Braunschweig, Nilsen & Wiebolt (1999), surface soil temperatures and high solar irradiance distinguish shale barrens from non-shale barren sites. It has been suggested that air temperatures on the barrens are comparable to the desert regions of North America and several sources indicate soil surface temperatures at shale barrens throughout the mid-Appalachians routinely reach 50-60 degrees Celsius, or 120 degrees Farenheit.

However, such conditions do not prelude seedlings from establishing. I noticed many areas of exposed channery literally carpeted with seedlings (see image below), although I fully expect most of these seedlings to eventually succumb thereby maintaining the relatively open aspect typical of share barrens.

Exposed shale surface with abundant seedlings

What allows a species like Trifolium virginicum to adapt to the harsh environmental conditions of shale barrens?  Apparently noone really knows!  However, Clovers (Trifolium spp.) are generally known to exhibit "nyctinasty", a fancy word for plant movement caused by a stimulus, and others think of this as "sleep movement."  J. W. Harshberger devoted at least 15 years to the study of clovers and devised a contraption to measure the nyctinastic movements of various species (He published his work in the Proceedings of the American Philosophical Society, 1922). Charles Darwin also studied this phenomenon and wrote an entire book entitled "Power of Movement in Plants" in which he detailed many examples of nyctinasty. A classic example of this movement would be drooping leaves in the evening that open again in the morning, presumably to capture sunlight. However, Trifolium growing in extreme situations of high irradiance and surface temperature may need to do the opposite.  Widening and unfurling leaflets in midday would expose them to potentially dangerous heat levels. Compare the image at the top of this post taken mere moments before the image below. The former plant occurred in a patch of shade and had fully unfurled leaves/leaflets; the plant below shows vertically upright leaves in a sunnier and much hotter location.

Kate's Mountain Clover (Trifolium virginicum) with inflorescence and mostly vertical leaves in open shale barren

Douthat State Park, Bath County, VA (May 13, 2018)

I've got to go back to do more field research!

View from shale barren@Douthat State Park

Prairie Goldenrod & Its Unusual Habitats

Prairie Goldenrod (Oligoneuron album) doesn't look like your typical "goldenrod".  But it was once considered a member of the genus Solidago, lending some scientific basis to calling it a goldenrod. To some it is still referred to as Solidago ptarmicoides. This plant has been kicked around taxonomically an almost shocking number of times. But I digress....





Prairie Goldenrod is a legitimate rarity in North Carolina, and is either absent from or rare in most other southeastern states. Why in the world doesn't it occur in Virginia?

The apparent rarity or absence from much of the southeast may not be surprising since it is presumably PRAIRIE Goldenrod and relatively few such habitats occur in the region.  However, one site, perhaps the only extant site in North Carolina, is both botanically rich and very much "prairie-like".

Prairie-like habitat of Oligoneuron album in Granville Co, North Carolina; note faint white specks of Oligoneuron stems in flower., see inset below

Closer view of Oligoneuron album in Prairie-like habitat,in North Carolina; flanked by the basal rosettes of Prairie Dock. (9/18/2013)

The North Carolina site discussed here also includes abundant Prairie Dock (Silphium terebinthinaceum), Stiff Goldenrod (Solidago rigida) and many other species often associated with prairies elsewhere in the country.  This site may be as close to true prairie as it can get in NC and is certainly one of the only known locations resembling a prairie with reasonably intact vegetation on a soil typical of prairie. Broadly, most of the North American Prairie region is closely tied to the occurrence of Mollisols (see map below).

Soil from the North Carolina Oligoneuron "prairie" site,
classified by soil scientists as a Mollisol.
The dark color is one the characteristics of this soil order.

USDA Soil Map, Mollisols

     




It is hard to overstate how unusual this soil is in Carolina.  As the generalized USDA soil map shows, there are no mapped Mollisols of sufficient extent to appear on the map almost anywhere in the southeastern US outside Florida.  A retired University Soil Scientist once told me it took years for the existence of these soils to be accepted in North Carolina, even having the "dirt in hand", since historical convention suggested such soils couldn't develop in the North Carolina climate.  But again, I digress......

This post was supposed to be about Prairie Goldenrod.

Oligoneuerun album apparently thriving in remnant tallgrass prairies in Paintbrush Prairie, Missouri.(08/17/2015)
Rattlesnake Master (Eryngium yuccifolium) is barely visible as well.

Prairie Goldenrod does actually occur in undeniably classic "tallgrass prairies" in the midwestern US, which some refer to as the Central Tallgrass region.  I have observed it in several tallgrass prairie remnants in Missouri such as that shown above.

Taken together, my admittedly few observations seem to fit into a nice and tidy picture of the habitat for Prairie Goldenrod.









But, then I visited alvars along Lake Huron in Ontario Canada.

I was absolutely shocked to find Prairie Goldenrod on lakeshore pavements (on exposed rock I believe is dolomite), and even in loose sandy soil at the margins of Lake Huron!.  These were NOT PRAIRIES.

Prairie Goldenrod on alvar,lakeshore pavement, Misery Bay

Prairie Goldenrod along Lake Huron, Ontario Canada

I did finally find Prairie Goldenrod in one habitat much more akin to prairie near Lake Huron. Much of the ground surface was littered with small boulders that are likely glacial erratics. The matrix soil was primarily sand that apparently washed in under extreme tides or storm events, covering the underlying dolomite flat rock. Such disturbance events, coupled with the underlying pavement, help keep the habitat from becoming overgrown by shrubs and trees. The most visually dominant herb, other than Prairie Goldenrod was Limestone Calamint (Calamintha arkansana). 

Prairie Goldenrod amidst glacial erratics and loose sand near Lake Huron, 

Interestingly, NatureServe uses Oligoneuron album as a nominal in one of its described alvar natural community types, but rather than it being used to name a "grassland" type it is considered part of a shrubland:

(see http://explorer.natureserve.org/servlet/NatureServe?searchSystemUid=ELEMENT_GLOBAL.2.722949)

No matter what habitat seems to support Prairie Goldenrod they are all well worth the look!

Great Lakes Alvars

One of the interesting, mostly open aspect, natural community groups of eastern North America are the so-called "alvars" of the Great Lakes region. Concentrated in Ontario Canada where these images were taken, alvars include a range of  natural communities. Like many similar habitats in eastern North America they stand in striking contrast to the surrounding, often densely forested landscape. In the image below, these forests consist largely of Jack Pine (Pinus banksiana) and Northern White Cedar (Thuja occidentalis).

Alvar  Inland Pavement - dominated by non-vascular plants, primarily tussock mosses, Bruce Peninsula

It has been shown that the majority of known alvars occur with  approximately 1 km of shorelines (https://archive.epa.gov/solec/web/pdf/alvars_cobble_beaches.pdf).  Many authors, especially Catling and Brownell, 1995 and 1999, distinguish shoreline alvars from those occurring in more interior locations.  The most notable difference between these is the sparseness of the vegetation on shoreline or coastal examples (see images below).

Alvar  Coastal Pavement - sparse to very sparse plant cover, Bruce Peninsula

Alvar  Coastal Pavement - inundated by low water levels and wave actions, Bruce Peninsula

Lobelia kalmii in shoreline alvar@Misery Bay

Fringed Gentian (Gentianopsis sp) in shoreline alvar@Misery Bay

However, Catling and Brownell (1999) also list several plant species indicative of both types. Among their list of species "more or less restricted to shoreline alvars" is Kalm's Lobelia (image, right). Although not included on their list, I noticed fringed gentians (image, below left) only in shoreline alvars as well. 








All North American alvars are restricted to  limestone or dolomite in the Great Lakes region (see map below), stripped of surface material by glaciation action. Several examples I observed  had glacial erratics present. 

Alvar Distribution: https://archive.epa.gov/ecopage/web/pdf/alvar-technical-report-199903.pdf

Inland alvar pavement with glacial erratics, Misery Bay

The extent of vegetation cover across alvars is highly variable; NatureServe recognizes at least 13 different natural communities with different structure and composition. 

Inland alvar with very shallow soil, superficially dominated
 by Limestone Calamint (Clinopodium arkansanum), 6/17/17

Limestone Calamint, formerly Calamintha arkansana, was considered a dominant species on Lake Huron alvars according to Catling and Brownell (1999). This appeared to be the case in several examples I saw as well especially in very shallow soils around open inland pavements; NatureServe includes it as a nominal in one of their community types as well.

(Right: Limestone Calamint (Clinopodium arkansanum); 6/17/17)

Only one of the NatureServe described types is considered "sparse", much like the image at 

the top of this page. The majority of the remaining community types are considered grasslands

or shrublands (See NatureServe Explorer).  The following two images are examples I would 

consider alvar shrublands. The first shows an extensive colony of Creeping Juniper 

(Juniperus horizontalis) and could perhaps be considered a dwarf shrubland.  

The second, and perhaps regular 'ole shrubland, has numerous & prominent clumps of Common Juniper (Juniperus communis) with patches of bare soil, grasses, and Shrubby-Cinquefoil (Dasiphora fruticosa).

I conclude this entry with a series of images of herbaceous dominated alvars. The first was quite moist and dominated by Spikerush (Eleocharis sp.), the second was even wetter and dominated by what I believe was Tufted Hairgrass (Deschampsia  caespitosa). The last, is a dry grassland with emergent Common Juniper and Shrubby-Cinquefoil.