Topic: Proposed updates to IHS coding

I have been gathering together some of the new IHS codes in use in the South East that are not officially in the SERC versions but possibly useful nationally (i.e. not relating to very localised habitats types).

Hope this links work for the IHS codes already in use and the grassland proposed ones. I've made it editable by everyone with this link so you can update with your ideas and suggested new IHS codes:
https://docs.google.com/document/d/1jTJ … y=CM-_s70M

As there is a big project in the South East to standardise our habitat datasets it would be great to get a steer and hopefully other recommendations from the LRC community on proposed updates to IHS. Also, with SERC's permission, look at options for how ALERC or a sub group of this forum could consider and collate major suggested changes and maintain version control?

What do the ALERC Directors think, is this somethign ALERC could in the future support/fund SERC to maintain?

2 (edited by Rob Large 17-03-2011 09:17:48)

Re: Proposed updates to IHS coding

Hi Melanie

Thanks for posting this. There's quite a lot of stuff in there, but I will try to work through some of the issues.

Regarding the proposed neutral grassland changes by Graham Hawker, I have seen various versions of this idea originating from different projects and I understand (as a surveyor), how frustrating the perceived lack of certain codes can appear. A year or more ago I was consulted on how the next version of IHS should deal with this situation. As far as I am aware version 3 is more or less finalised, but I cannot comment on how far SERC has got with releasing it. Perhaps Tony Price will be able to shed some light.

The changes for grassland presented in your document violate a number of the basic principles of IHS, I will try to explain how.

Firstly the existing GNZ code is what we call an inverse code. A fundamental principle of IHS is that at any given level of the hierarchy the codes should exhaustively cover all of the possibilities encompassed by the parent. The role of the inverse code (or Z-code) is to stand for all the habitats which are not covered by the other items at the same level. That is to say that GNZ represents all neutral grasslands which are neither Lowland meadows, nor Upland hay meadow (two well defined habitats representing a very small range of NVC communities). As such GNZ is a ragbag of assorted grassland types on circumneutral, mesotrophic soils which remain semi-natural enough not to be called improved. IHS does not allow for inverse categories to be subdivided in this way (e.g. GNZ1, GNZ2 etc.) for this reason and the following.

Clearly the three proposed sub-divisions (Wet, Rough and Inundation grassland) do not exhaust all the possibilities embodied by GNZ. What about dry "good semi-improved", but well managed grass? So you would need to have another category GNZZ Other other neutral grassland (which is daft) and given time I could come up with twenty other subdivisions all of which would be useful to someone.

Also the three categories (Wet, Rough and Inundation) are not mutually exclusive, how would you code wet rough grassland? Remember you can only use one habitat code.

Next the three categories also include the possibility of overlap with the priority grassland types. Some lowland meadows (especially GN11) are wet. Unmanaged for a couple of years they become rough (but remain priority habitat).

A related issue is how you deal with wet calcareous grassland (All CG types are priority grassland), or wet acid grassland?

I recognise the attempt made by the definitions supplied to justify what has been done, but they do not work.

Our solution to this problem was to add this kind of item as secondary codes rather than as primary habitat codes. I don't have access to the final structure we decided on for version 3 right now, but I have an early draft in which we added a new category section at the start of the formation codes as follows

Grassland descriptors (GD)
GD1 Wet grassland
GD2 Dry grassland
GD3 Neutral grassland with calcicoles.

Thus your GNZ2 Wet grassland becomes GNZ.GD1 but you can also have GN11.GD1 (a wet MG4 meadow). And so on.

We also discussed adding a Rough grassland code in the same series, but didn't think it was worth it. Rough grassland is to my mind just unmanaged grassland, thus GNZ.GM4 is a good description of MG1 in most circumstances. Remember this is not a community classification like NVC. You could also have GAZ.GM4 rough acid grassland etc.

Let's look at another of your suggestions GNZ12 Calcareous rough grassland as a subcategory of Neutral grassland? I laughed when I first read this, are we talking about neutral grassland or calcareous grassland here? If it is truly calcareous it belongs in GC1 and is priority. But I see from the definition that what we are talking about is MG1d a community with which Wiltshire is very well supplied. This is not a form of calcareous grassland, it is mesotrophic as demonstrated by the dominance of coarse grasses such as Arrhenatherum and Dactylis. It may include remnant calcicoles but for the most part these are deep-rooted taxa such as Pastinaca, Daucus and Tragopogon or hardier types with catholic tastes such as Lotus, or anthill specialists like Helianthemum, Thymus etc surviving on disturbed patches. Over time this community becomes more mesotrophic and eventually loses almost all calcicoles. How much more elegant is GNZ.GD3.GM4 Unmanaged neutral grassland with calcicoles?

Inundation grassland could be added as a code in this section if desired but I would think it is covered by EM22 Inundation vegetation which has been in IHS from its inception.

As for Recently reseeded meadow communities, that could cover a huge range of community types and degrees, it is certainly not a single well defined habitat. There is massive overlap between this and several of the other categories. My version of IHS has this instead, we have subdivided the grassland land-use code GL2 Non-amenity grassland as follows.

GL2 Non-amenity grassland
GL21 Permanent agricultural grassland
GL211 Arable reversion grassland
GL2111 Species-rich conservation grassland
GL211Z Other arable reversion grassland
GL21Z Other permanent agricultural grassland
GL2Z Other grassland use

So your GNZ4 could be GNZ.GL211, but you could also have GN1.GL2111 (or GC1.GL111) to describe a maturing successful reversion. I agree that this does not cover re-diversified reseeded grassland very well, but it would only require a tweak of the wording to fix that, perhaps Re-seeded or reverted grassland would suffice as an alternative.

Now the improved grassland. Firstly a technical point, everywhere else in IHS the children of (e.g.) GN0 are GN1, GN2...GNZ not GN01 GN02 etc. . So GI01 is just plain wrong. No matter... Any subdivision is incomplete without a Z-code, so you would need to add one. Ask yourself a couple of questions though.
Dou you really need to subdivide GI0? If so would NVC not be a better tool?
Are your surveyors capable of making the distinction between Typical permanent pasture and Poor semi-improved reliably? Probably not because there is a continuous spectrum of variation in improved grassland and while we may like to think we can tell the difference in the field, I question whether many can actually decide where one ends and the other begins without resorting to NVC. But give them the option and it is human nature to try. Let

Re: Proposed updates to IHS coding

Hi Rob

Thanks for the detailed response. The problem of having sub categories of an inverse category was pointed out to me when I came up with idea but I struggled to see another way of doing it. I can see the potential of using formation codes though I'm not entirely convinced. My approach was take fairly familiar NVC categories and use them to define the wet, rough and inundation categories though neutral grassland is much harder than acid and calcareous. The would need to be another inverse category essentially "other other neutral grassland" and the need to have another name at the top this bit of the hierarchy. When habitat mapping I find I always want to define these categories and at present all we can do it use the comments field and where available the NVC information within the source information. Perhaps this enough.

Calcareous rough grassland is neutral grassland NVC MG1d. I agree it sounds odd perhaps neutral rough grassland with calcicoles is better name  and perhaps this will be covered by the formation code mentioned especially coupled with unmanaged grassland code. But then rough grasslands. especially on road verges can be mown.

When it comes to phase 1, I agree with the approach of ignoring all those useless grassland categories. I've always disliked them, avoided using them, would like to see them consigned to history and it's certainly a significant advantage of IHS. However the one category than did make sense to me was "poor semi-improved". It was properly defined and fitted with grasslands I was seeing. Considering how poorly terms like semi-improved are used, and this is a major problem, then perhaps it is wrong of me to think it still has a place but I think it says things that are useful to know even amongst the variation within improved grasslands. It could be argued that improved grassland as a whole belongs in GNZ because it is after all it's just another neutral grassland type.

I think the frustration is that with most IHS habitat codes they give a good idea of what the habitat is but in the mish mash of neutral grasslands they don't. It will be interesting to see the updated version. The question I should ask myself is do we really need to capture all this detail in the IHS codes? I would agree that we need to try harder with the application of the non-habitat codes which might help much more than adding habitat codes.

Graham Hawker
Thames Vallley Environmental Records Centre