email alerts

To receive email alerts for new posts of this blog, enter your address below.

Saturday, October 14, 2023

A Way To Judge Ecosystem Health

Coefficients of Conservatism - measuring "plant quality" or "conservation significance"

To reverse the decline of biodiversity, there is a crucial need for more people who can see what is happening to our "protected" natural ecosystems. It's easy to notice a prairie being plowed or a woodland cut down. But most biodiversity loss today is on protected lands, and it happens gradually, by ecosystem quality decline. Many people who would otherwise come to the ecosystem's defense can't recognize that sort of change. 

By learning to recognize key species and their significance, many more people could contribute much more meaningfully. It may seem like a lot of work to learn to identify species and understand their roles and lifeways. But it's fun, rewarding, and critically important. Managers of conservation lands need to hear feedback from a knowledgeable and caring public. Good land management should be applauded. Losses and threats need to evoke concern. 

"Coefficients of conservatism" are numbers that help you evaluate the quality of plants in an ecosystem - in other words the health of the ecosystem itself.

How do I tell if a site is high quality - or, for that matter, if a site is recovering or degrading?
Knowledge of "coefficients of conservatism" is a big help.

For a quick over-simplification, a "conservative" or "high-quality" plant is the opposite of a "weed" or "invasive." Technically, the coefficients reflect the fidelity of species to high-quality natural areas, on a scale of 0 to 10. If most species at some site have coefficients in the 0 to 1 range - for example ragweed (C = 0) and tall goldenrod (C = 1) - the site has suffered gross degradation. It's probably low on the priority list for land management or restoration. High-quality natural areas are the only places likely to have plentiful high-quality species - those in the 8 to 10 range that most indicate health and integrity. These places deserve the best possible protection and care. For examples, see Endnote 1. 

The coefficients for the plant species of your state or region can be found at the Universal Floristic Quality Assessment Calculator website. For how to use it, see Endnote 2.

For information about using these coefficients mathematically in a Floristic Quality Assessment, click here. But you can get a great start by merely walking through a site, identifying the species you see the most of, and checking their average coefficients, as described below.   

High-quality natural areas typically have large numbers of diverse conservative plants and associated animals. But such areas are being lost or are already gone as a result of plowing, bulldozing, pollution, over-grazing, lack of fire and other human-caused changes that eliminate long-evolved species and relationships. Conservative plants are indicators. Less easily observed is the concomitant loss of now-rare animals, fungi, and other lifeforms interdependent with plants. If we aim to conserve the temperate world's biodiversity, two vital goals are 1) to maintain the health, quality, and size of natural species populations in original high-quality natural areas and 2) to restore populations to larger areas so that fragmentation and small habitat size don't impede evolution and other processes that maintain healthy natural ecosystems. 


For those seeking to learn the plants, there are countless books and other helpful resources, many freely available online. One of the best ways to learn is in a group that needs the knowledge, stewards collecting seeds for restoration. Relevance and repetition helps the details stick. A knowledgeable mentor can expedite the process. 

 

For more details and references, see here and here and here.

 

Endnote 1: Sample Coefficients for Woodlands and Prairies


These coefficients are from the Flora of the Chicago Region. 


        Sample Coefficients for Woodland Species

Scientific Name

C

Common Name

Brachyelytrum erectum

9

awned wood grass

Calystegia spithamaea

10

low bindweed

Carex blanda

1

common wood sedge

Carex davisii

4

awned graceful sedge

Carex gracillima

7

purple-sheathed graceful sedge

Galium aparine

0

cleavers

Galium concinnum

7

shining bedstraw

Lathyrus ochroleucus

10

pale vetchling

Solidago flexicaulis

7

zig-zag goldenrod

Symphyotrichum drummondii

3

Drummond’s aster

Symphyotrichum shortii

7

Short’s aster

Trillium grandiflorum

9

large-flowered trillium

Trillium recurvatum

5

prairie trillium

 


        Sample Coefficients for Prairie Species

Common Name

C

 Scientific Name

Nodding wild onion

7

 Allium cernuum

Big bluestem

5

 Andropogon gerardii

Purple prairie clover

9

 Dalea purpurea

Biennial gaura

0

 Gaura biennis

Wild bergamot

4

 Monarda fistulosa

Common evening primrose

0

 Oenothera biennis

Yellow coneflower

4

 Ratibida pinnata

Black-eyed susan

1

 Rudbeckia hirta

Compass plant

5

 Silphium laciniatum

Gray goldenrod

6

 Solidago decemflora

Smooth blue aster

9

 Symphyotrichum leave

Hairy aster

0

 Symphyotrichum pilosum

 


Below is a list of the "Highest Quality" nine species of the classic prairie:


Common Name

C

Scientific name
Wilhelm and Rericha

Scientific name
Swink and Wilhelm

cream false indigo

10

Baptisia leucophaea

Baptisia leucophaea

scarlet painted-cup

10

Castilleja coccinea

Castilleja coccinea

white prairie clover

10

Dalea candida

Petalostemum candidum

prairie gentian

10

Gentiana puberulenta

Gentiana puberulenta

prairie white-fringed orchid

10

Platanthera leucophaea

Habenaria leucophaea

prairie lily

10

Lilium philadelphicum 

Lilium philadelphicum

prairie panic grass

10

Dichanthelium leibergii

Panicum leibergii

prairie dropseed

10

Sporobolus heterolepis

Sporobolus heterolepis

heart-leaved Alexanders

10

Zizia aptera

Zizia aptera

 

For more comprehensive lists of the species that indicate prairie quality (and some history and detail), see Endnote 3 of the post: https://woodsandprairie.blogspot.com/2018/07/a-myth-coming-true.html


Endnote 2: Using the Universal Floristic Quality database in Seven Steps

1. Click on it. 

2. Log in. Doing so will provide access and will store any data you add.

3. Study details, if you want. Especially "FAQ" under "Help".

4. Or go directly to "FQA Databases". Here you'll find lists of species, coefficients, and other info for many states and regions. The most "up to date" list for the Chicago Region is found under "Flora of the Chicago Region".

5. You can learn to identify most plants with diligent study and a good book ... or by taking a class ... or with the coaching of any interested, supportive, knowledgeable professional or amateur botanist. Many states have a "Native Plant Society" that could help. 

6. One good start might be to map areas of conservation lands according to their priority. Areas with plentiful species in the C = 4 to 7 range are important. When such areas also have plentiful species in the C = 8 to 10 are the most important. 

7. To determine whether the quality of such an area is declining, recovering, or staying the same requires years of study. One approach is to walk a path and record numbers of key species. For example, along a narrow footpath through a prairie, you might count all plants of the "Highest Quality Species" from the list above within five feet of the trail and compare them from year to year as shown below:


    Numbers of plants along a transect - Example A

Common name

2023

2028

cream false indigo

17

20

white prairie clover

52

54

prairie gentian

12

31

prairie lily

5

8

heart-leaved Alexanders

16

15

TOTAL

102

128

 

The 2023 data suggests that this site is valuable and deserves good care. The 2028 results would suggest to me that the ecosystem was recovering lost quality as a result of good stewardship. 


On the other hand, if the results were: 


    Numbers of plants along a transect - Example B

Common name

2023

2028

cream false indigo

17

10

white prairie clover

52

34

prairie gentian

12

8

prairie lily

5

2

heart-leaved Alexanders

16

9

TOTAL

102

63


The results in Example B would suggest to me that something was wrong.  Concern is warranted. Action is needed. 


More sophisticated approaches can be found in the links to this blog. Collecting data on a longer (and more varied) list of species would be an improvement. But a little good data is a great deal better than no data. And for most sites today, no data is the norm. 


Acknowledgements


Thanks to Christos Economou for extensive helpful revisions. 


12 comments:

  1. In the Ecosphere article you linked to, Spyreas makes a good argument to not include abundance in FQA because it doesn’t significantly improve the assessment & adds complexity to monitoring. If the goal of assessment is to identify higher quality areas, then I think abundance of conservative plants is not as important. But your “walking a path” example suggests that if the goal is restoration & resilience assessment, abundance should be measured. Maybe there should be a “resilience FQA” that has a correction factor for conservative species that fall below a threshold for multi-year averages of spatial distribution. But I’d guess that properly defining such a thing may be too difficult to be practical.

    ReplyDelete
    Replies
    1. Don, thanks for the good comment. Greg Spyreas does indeed make good arguments, and he has studied the options carefully and made good judgments. But he also changes his mind in respect to new data and situations.

      In the paper that Karen Glennemeier and I wrote with him, we found that percent cover data made at least one FQI assessment more responsive to change. If something's going wrong, percent cover data might reveal the problem earlier.

      I do very much agree that FQA should be looked at carefully to make it more helpful for restoration progress evaluation.

      Delete
    2. Yeah, even in my graduate work with diverse prairie plantings, mean C didn't change much with age, because everything was there in low abundance early, but weighted mean C improved as plantings matured. There are many remnant sites that hold on very low numbers of conservative species simply because they haven't been completely extirpated yet (but its going that way), and in such sites it makes a big difference. Mean C gives you an idea of the potential (if problems are can be feasibly addressed)--basically tells you if the right parts are there but not if the system has ecological integrity. Weighted mean C is more the reality of the moment and if it diverges from unweighted mean C, it's usually because things getting better or worse. Even crude estimations of cover made upon completion of a meander survey work for weighting mean C (even cover Daubenmire-like categories), so it's not difficult. All of that said, especially in bright woodlands and savannas there fairly commonly comes a point in invasive shrub invasion and mesofication where only sparse herbaceous layer plants remain, quite a few of which are somewhat conservative and more shade-tolerant or avoiding members (a lot of spring ephemerals, shining bedstraw, etc.), and mean C (at least of that herbaceous stratum) stays reasonably good (above 4) but the system has obviously collapsed. So as with all things, the user needs to have some insight into the context--the site--and what the numbers mean in that setting.

      Delete
  2. Changes in FQA tell us that something is better or worse, but it does not help us decide what, if anything, we can do about it.

    Restorations have a tendency to decrease in mean C over time. Even many high-quality natural areas have seen a decrease in quality assessment numbers.

    What can we do about it?

    ReplyDelete
    Replies
    1. A single number doesn't, but based on the species at a site, it's possible to gain insight into why and what to do. Very few prairies are burned frequently enough (which is at least every other year, if not almost every year), and too many of the burns that happen aren't in the dormant season (ends around 4/1 most years on the WI-IL line)--bad for things like Hesperostipa, Hypoxis, Pedicularis canadensis, Primula, Castilleja, etc. and good for things like Andropogon, Sorghastrum, Solidago altissima, etc. Minimize harm to living tissues of conservative biota, minimize growing season nutrient pulses, minimize fire intensity (use other means for woodies), and keep thatch out. Facilitate opportunities for ecesis of more conservative community members (add seed in fragmented systems over the backdrop frequent dormant burning). At least that's what I tell people, because its the only thing I've seen work, but I also don't think it's coincidence that it lines up well with historical accounts.

      Delete
    2. Yes, indeed. Well put. With any data, interpretation is needed. As with a human patient, taking a temperature doesn't tell a doctor (or a parent for that matter) what to do, but it may provide helpful information. Other tests may be needed. But "insight into the context" and judgment are key. Both insight and judgment benefit from knowing which plants are conservative and which are not.

      Delete
    3. The use of Mean-C and Floristic Quality Index (FQI) is to some extent inadequate to evaluate progress made during botanical restoration projects. They are inadequate because they cannot distinguish between “good’ and “better.” For example (Case 1), imagine two plots of land, identical in area and with identical plant species lists. Plot A has 10 plants for each species and Plot B has 100 plants for each species. Clearly Plot B is “better” than Plot A, but for Mean-C and FQI there is no difference between the two. Numbers of plants should count somehow! Is there a metric which accounts for the numbers of plants for each species?

      Consider Case 2; a plot of land is monitored over many growing seasons. Initially only species with high C-values (8 to 10) are present yielding a high value for Mean-C. Over time ubiquitous native but low C-value (0 to 3) species appear; Mean-C will decrease (FQI may or may not change). Has degradation really occurred? Lowered Mean-C does not necessarily indicate degradation. Those native species will only have caused “degradation” if they depose higher C-value species (this is known to have happened, but this situation is not common).

      To me as a dilatant in botanical restoration, it seems that degradation is more the result of invasion by non-native species than by the “invasion” of ubiquitous native species with low C-values.

      Delete
    4. "Is there a metric that accounts for the numbers of plants of each species?" Metrics can be weighted by relative abundance, and I suppose by any aspect of quantity, relative or not. In most instances relative abundance should be adequate because the situation where there would be a lot of bare ground and isolated conservative plants here and there would be rare (exception maybe in some sandy situations, where intervening space may be lichens, bryophytes, and cyanobacteria. However, usually weighted and unweighted metrics give similar results--but not always. One instance I can think of where something like this can happen is with a degrading herbaceous community under oak and buckthorn, where sometimes a few widely scattered conservative plants remain, giving a deceptively high mean C. Those cases are obvious to those who are experienced with savannas--or maybe I should say to those who have experienced real remnants or the very rare cases of complete (and not overly canopy-focused) reconstructions of savanna.

      The second case would is not a realistic one unless one has cleared land of vegetation and planted plugs of conservative species only. However, metrics would not change much if weighted by relative abundance, unless there actually was degradation. If 0-3 species were relatively abundant where they had been absent before, it would be degradation. Usually most of the small scale richness in remnant sods consists of moderately to very conservative species.

      Degradation by native species is harder to see. Degraded remnant prairies may be dominated by big bluestem and have reduced small scale richness (and lower mean C vs. undegraded porcupine-Leiberg's panic grass--prairie dropseed)-dominated prairie. Many don't recognize the degradation of a savanna where oak structure is there but the herbaceous layer is mostly tall goldenrod or white snakeroot.

      When it comes to any floristic quality or ecological integrity metric, the context of the site really determines what is best for monitoring or assessment, and there is nothing that can be blindly used.

      I would recommend Justin Thomas' presentations on YouTube:
      https://www.youtube.com/watch?v=KBWK-Jdj1HA
      https://www.youtube.com/watch?v=3euqfbW3QPs&t=665s

      Delete
  3. The original (1979, S&W) C values had negative values (down to -3) for non-native species and super values for E&T species. This seems superior to me than the present non-inclusion of non-natives.
    There is no discussion of the area being assayed in the above. Comparing 0.5, 5 and 50 acres with the FQI metric will not result in evaluations most would regard as 'fair'.
    Population size is critically important. Populations require recruitment to be sustainable.
    Autochthonous populations can be enhanced by good management and but garden production of seed of rare (low abundance) species.
    I believe a focus on enhancing population size is more sucessful in the long run than worring about calculations based on C values.

    ReplyDelete
    Replies
    1. A tool must be suitable for the purpose for which it is created. Coefficient of Conservatism (C) was intended as an indication of a species’ ability to recover from a disturbance (C = 0 for almost no deleterious effect and C = 10 for a severe effect). Initially an element for “badness” (Rhamnus cathartica, C = -3, for example) and “goodness” (Cypripedium candidum, C = 20, for example) was included. Eventually both the elements of “badness” and “goodness” were deemed unnecessary to the original intent of describing the “effect of disturbance” upon a species. Intermediate degrees of disturbance effect were assigned values from 1 to 9.

      Mean-C and Floristic Quality Index (FQI) were tools created to determine and measure an indication, any indication, that a site retained some degree of pre-colonial, vegetative integrity. Even a single plant of a species would be sufficient evidence. No exhaustive, and time consuming, study was needed to evaluate what might be literally thousands of sites to be investigated. Ranking sites by FQI was a way of identifying sites for more detailed study and for restoration work.

      Once restoration projects began to add species (salting the mine, so to speak) FQI was not really able to distinguish between “good” and “better” as the numbers of both species and the numbers of their plants at a site increased and the recovery or the restoration or a recreation succeeded.

      There is an excellent article “Floristic Quality Assessment for Vegetation in Illinois…” in Erigena, Number 15, November 1997. It is available on-line at
      https://illinoisplants.org/images/pub/Erigenia_No_15_Nov1997_FQA.pdf

      Delete
  4. Dr. Nyberg's point about population size is a good one. However, I think the quality of an ecosystem Dr. Carter advocates for is important too.

    Let me give an example from Nachusa Grasslands. The prairie fringed orchid was restored to Nachusa Grasslands. I have not seen it personally since it is in the bison enclosure. Susan Kleiman said there are more prairie fringed orchids in previously disturbed areas than in the higher-quality area of fen because the high-quality area has more competition.

    If only population was considered, the disturbed areas with more orchids would be considered more successful. However, I think we would all agree a high-quality area, even with less orchids, has an added value for the irreplaceable natural community it possesses.

    I think both populations (advocated by Dr. Nyberg) and quality of habitat (advocated by Dr. Carter) are important even though the two objectives do not always align.

    ReplyDelete
    Replies
    1. This is true of a population in Jefferson County, WI, more fringed orchids following brush work in a low quality area. Of course, maintaining minimal thatch with very frequent dormant fire can maintain a viable population, but perhaps not a maximal population. We run into that with invertebrates too--are we trying to maintain viable populations in healthy communities or maximal populations in more degraded communities, where short term population and long term viability may diverge. A site that that historically supported a rare butterfly comes to mind. Very frequent burning there stopped when the butterfly was discovered, which might have good for the butterfly in a numbers sense (or at least a variation in population size sense) for a couple years, but now most of that prairie is shrub thicket and that species has not been observed for several years and is likely extirpated. I can think of situations where maximizing a species in a place is warranted, and fringed orchids at Nachusa may be one, but too often, I think, we fail to take the long view.

      Delete