In West Australia the Herbarium specimens have proved invaluable in determining species status for hundreds of species.Are the conclusions that concern your questioner to do with reliability of data,i.e. misnaming of type specimens and thus inability to search for matches with current field collections or reliance on collections to the exclusion of fieldwork?

I definitely see the need for herbarium records and type specimen's. Particularly for uncommon species with a narrow geographic range. However, I guess an argument could be made that more common and widespread species may have a tendency to be "split" into new species because of sub species variation that occurs across their natural range. This could occur because a specimen recorded in the herbarium may have a slightly different form than that collected from another area within its range. A split may occur when one may not be needed. This may be getting into the depths of taxonomy (I am not a taxonomist) and may have implications from a ring species perspective. 

I mention this as a hypothetical and have no first hand experience of this actually occurring however I have heard it discussed by colleagues.

Yes, I agree with the question very much, because we can only find morphological two-dimensional connections via dried specimens, and the future of botany, taxonomy and ecology will be with two "new" three dimensional concepts, ecotypes and genetics.  

Ecotypes make the connections between the plant species and the environment that creates fixed genetic forms called ecotypes, that you can read about at http://www.ecoseeds.com/juicy.gossip.three.html.  

We will have to invent a new language to be able to discuss the fixed ecotypes, and when Gote Turesson coined the word ecotype in 1922 he also tried to invent a language to discuss and describe the various ecotypes.  When Greene split the California poppy into 116 new species, people thought he was crazy, but he was just a decade ahead of discussing the ecotype questions, and only had 100 years ago the old taxonomic "species, subspecies and variety" tools to work with.

Genetics will open the door for us to sort out the thousands of cryptic species that are hiding amongst all of  the species that we thought were solid taxonomic entities, but there are look-alikes that cannot breed with each other, so should be separated and described as new species.

Before Dr. Ledyard Stebbins passed away in 2000, we co-authored a paper in the journal Grasslands (Native Grass Association, Davis CA) in 1998, "Jepson Manual chromosome numbers may indicate new 'cryptic' native grass species." GRASSLANDS 8 (3) 4-5.  

What our article talked about, is that there are 300 described native grass species in California based on their physical/taxonomic traits, and our paper is recommending that there are hidden species via genetics among the 300, another 100 species that must be teased out and described that will take lifetimes of future geneticists and taxonomists to resolve.

Hi folks. Thanks very much for your thoughtful responses to my question. I have been ruminating on these issues for a couple of months. I am working with a group of succulents and trying to define an appropriate suite of morphological characters to revise the taxonomic key using herbarium specimen. I suppose my biggest concern is that herbarium specimen quality varies, and the preserved plants don't look exactly like what's encountered in the field. If I measure parts of pressed specimen, how useful will my revised key be for field identification? On the one hand, morphology has been conducted like this for centuries. Surely, a great deal of plant taxonomy relies on herbarium specimen. These are invaluable to our work. However, some studies on the preservation process indicate up to 20% of biomass is lost in the drying process. So, how reliable is this information?

Craig, the work you published in Grasslands is similar to what I'm looking at here. Molecular data suggests we're looking at 4 species, but these are currently classified as subspecies. Now, I'm testing whether the molecular differences are reflected in morphology. Thanks so much for the references regarding ecotypes. I think this is something I must consider. Greg and Harley, I am also worried about (1) historical misclassification, (2) splits where they aren't needed and the opposite, and (3) the quality of older specimen may contribute to a wide deviation of measurements that will be difficult to resolve statistically. 

Thanks again for taking some time to respond!

Thanks for your email.  Genetics should be our first cut when describing species in the future, followed by finding morphological traits we can use to describe each new species as separate from the others.  

This can be especially difficult if the difference is in their chromosome numbers with polyploidy being the fact that causes them to to reproduce with each other, then here may be few if any morphological traits to hang a new separation of two species onto.  

What I would do with the succulents, and what they are doing with sorting the polyploid fescues from each other, is microscopic cross sections of the fresh leaves, and see if there is any difference in the vascular structure?

You may have to only used the herbarium specimens for locations to collect new specimens, and to collect holotypes and isotypes for the new species you are going to describe, then see if you find traits to match your four species with the herbarium material that exists.  

In order of sort, genetics first, then look for morphological keys in the fresh material, then see if you can match the dried material to sort it into one of your four species.

Last year I discovered a six foot tall new native shrub species in central California on a cattle ranch, just because nobody ever ask the rancher to botanize the ranch in the last 100 years, 

A purely herbarium taxonomist is like a desk ecologist. They both will be lost when they go to the actual field situation. The plants growing in the field in their natural habitats have a lot to tell the taxonomists which herbarium specimens cannot. Therefore, it is better for you to combine herbarium work with fieldwork to get the full flavour of modern plant taxonomy and be a competent plant taxonomist who can work not only within four walls but the open space.

Sorry for the error in my last answer, which should have said that ..."polyploidy being the fact that causes them to NOT reproduce with each other..."  For centuries we were classifying plants with half of our eyes closed, only looking at their external description, which many times did not tell us anything about what was going on inside with the genetics.  Now that we know about genetics, we can get a clearer picture that actually represents what is happening in nature, in terms of who can reproduce with whom, and sort genera of plants more properly into individual species that way. 

i think the better way is the combination of herbarium specimens identification work with the field work.

A great deal of taxonomy is now based on genetics--molecular botany--so the characteristics used to identify plant species in herbaria or in the field may not always be a useful guide. Check out the Angiosperm Phylogeny Group's web site at  http://www.mobot.org/MOBOT/research/APWeb/

Plus, it has been demonstrated that epigenetic traits can be passed down to subsequent generations (See Bond D.M. and E.J. Finnegan (2007) Passing the message on: inheritance of epigenetic traits. Trends in Plant Science 12:211-216.) I must confess that I prefer to differentiate species by characteristics I can actually see in the field.

I suppose that becacuse there might be damage to the specimen and changes in floral coloration, some specimens'leaves tend to coil or change morphology, and in some specimen that possess pappus, it might not be as present as it was after collection.

In addition, the inclusion of environmental conditions could be lacking or limited from studying herbarium specimens. Moreso, I believe that geological information in particular could explain a lot about the biogeographical and evolutionary history. That you might not find at first hand from the herbarium specimen and might not stabilize your judgement.

Yes, thank you Thamsanqa for brining up the huge ecotype question, which in California and its varied terrain and rainfall extremes,  is a fierce modifier of physical types based on the environment that a species of plant has adapted itself to.  You can see my studies on the ecotype question plus a bibliography at http://www.ecoseeds.com/juicy.gossip.three.html

Thank you Craig. I am doing a similar study as Kalliopi, using herbarium specimens to revise whether Prismatocarpus subgenus afrotrachelium deserves a generic status. please see my question in my profile. Craig, I must admit it is the first time that I am interested in the word ecotype. I will definitely visit the link.

ECOTYPES & COMMON GARDENS. I forgot to mention the value of conducting a Common Garden study to sort out ecotypes from within species, which is what is discussed at http://www.ecoseeds.com/juicy.gossip.three.html.  

What you do is choose a species and collect seed from a wide range of habitats and environments.  Like the high and low end of the rainfall range for the species, then at least three in between.  

Then collect from the extremes in elevation, and then some in between,and  from the north, south, east, west extremes for the species, and then on a checkerboard pattern evenly throughout the range, like at least one sample for every 250 square miles, is what I used. 

And then if you have populations along sea coasts, or from windswept hilltops, anywhere like that with weather or wind conditions should be sampled.

Then you need to sample from soil extremes.  We are generally in a sandstone sedimentary area, but there serpentine soils, saline soils, seasonal wetlands, and sand dunes, and limestone soils and even a few volcanics.  

Then I planted each seed collection in it own wood box measuring one by two feet and 4 inches deep, filled each box with the same potting soil, sowed the same number of seeds of each collection in each pot, and then wait.  If these collections are very closely genetically related, then they will grow and develop and flower at the same speed and height--look like someone sowed them all out of the same seed sack.

But if you use grasses for this experiment, you can actually detect evolution in action.  At the extreme edges, like the lowest annual rainfall area for the species, the species starts morphing into new species, in order to survive in that extreme.  

You may also see this happening at  the other extremes, like elevation extremes, windswept or coastal populations, extreme soil types, rainfall extremes, etc.  The problem with all of these ecotypes, is that it will be very difficult to shoe-horn all of them into one species description, and then there is no method invented yet, to describe and name the wide range of ecotypes you will find.  

And then for the extreme ecotypes, are they sometimes reasonably so different from the described type for the species, that they really should be granted new species status, because it is too different to try and include it under the umbrella of a single species?

 Until we develop a method within plant taxonomy to describe and name the ecotypic variations within a single species, the question of what is an ecotype and which of those ecotypes fall within a single species, and which really deserve their own species description and name, will never be clearly resolved.

I hope this information is useful.

Kalliopi, I have recently been frustrated by Sedum keys based on characters that are not visible on herbarium specimens, or colors of flower parts that change over phenological time.  Sedum specimens often consist of brown sticks, perhaps with a packet of brown shriveled leaves and perhaps flower parts.  In this case, the solution has been to hire a group of field systematists to spend several years measuring and photographing live material, writing the keys based on the live material, and developing a method of preparing herbarium specimens that retains the leaves and also mounting photos of the fresh flowers and leaves on the herbarium sheet.  The work also has resulted in two new taxa published so far, and several others in the manuscript stage.  There will also be a field guide to Sedum sect. Gormania with full color photos.  The new herbarium material will be much more useful than the old stuff; but even the old material may be useful for genetic analysis.  In most cases (sedum is an anomaly here) I think herbarium specimens, while never a complete picture of the species in the wild, are essential for anchoring data to a real place and time.  And many new taxa are found in herbaria by people looking through specimens.

There will be no 'one-size-fits-all' answer to this question but as you have noted it is worth careful consideration. Well done to whoever posed that to you originally.

The sampling for any project should be designed with the question in mind. This includes taxonomy as well as ecology.

Considerations include where the specimens already available in herbaria are from geographically, how the spatial spread relates to the entire species range and ecological gradients, and whether the herbarium specimen actually retains the relevant traits. Determining this is an outcome in itself.

If there is a comprehensive and representative collection of  specimens for a group then the bias may be minimal. But the onus is on the researcher to demonstrate every time they conduct an experiment, or undertake comparisons for taxonomic  that they have sampled appropriately. 

Where possible, collect from field populations (at least for some specimens). Also take notes on population size, variability within populations and how that relates to the traits

You can also take an experimental approach to answering your question.

Imagine using specimens available at say the year 1990 (just to pick one time) and answer your question with that dataset. Now repeat with all specimens available in 2014. Did the conclusions change? Was the change due to bias? Was it just due to smaller sample size? Test that by randomly sampling from the later date, but choosing the same number of specimens as the earlier period would have provided. 

Examine the spatial spread in the samples. Has that changed? Do we still only collect near populated areas and near roads? Does that matter for your group of interest?

Best wishes for your study.

@ Herbarium specimens are the base to understand the taxonomic status of a particular species. Because even the dried specimen can provide a morphological details up to certain extent. The changes in color of the dried specimens can be verified with the written matter of the Herbarium sheet. By getting an idea from the Herbarium specimen, ONE can go for the cytological studies, biochemical studies, anatomical studies and even the molecular studies to establish the identity of the species. Therefore, I don't think that Herbarium specimens are clouding our conclusions.