Quantifying the unquantifiable: How to measure the success of habitat creation or restoration? Dr Alistair Headley PlantEcol Mobile: 07884 490494 Email: headleyplantecol@btinternet.com
Phase 1 habitat surveys Can Phase 1 surveys be used to assess the success of habitat restoration or re-creation? Possibly for distinguishing between some categories, e.g. wet modified blanket bog and dry modified blanket bog? But one surveyors assessment of wet modified blanket bog may be different from another s. Does a plantation of native broad-leaved trees in a former improved pasture make it a native broad-leaved woodland? Is the presence of associated woodland herbs, insects and other animals necessary before it can be called native broad-leaved woodland?
They showed that the average consistency between any two surveyors doing a Phase 1 survey of the same farm in Northumberland was only 26%. The most consistent pair of surveyors had the same categories for 39% of the farm and the worst was 17%.
What about assessing habitat quality using the National Vegetation Classification (NVC) system to map the different vegetation types? Advantages Certain plant communities are associated with better quality habitats. They give greater detail. Disadvantages Low level of consistency between surveyors in assigning areas to different communities. A plant community does not equal a habitat (despite the correspondence tables provided at the back of Phase 1 manual).
The average proportion of the area that had the same NVC community for each pair-wise comparison of the maps produced by the seven surveyors was 34%. The level of agreement ranged from 5 to 70%. The average level of spatial agreement at the sub-community level was 19% (9-29%). The more experienced professional surveyors had greater levels of agreement.
The mapping of vegetation using the NVC sytem does not work for the following reasons: The NVC was not constructed for the mapping of vegetation and it was originally meant to be a communication tool. The separation of the different NVC communities is based largely on frequency and not on abundance/dominance of species (look at the range of Domin values of the constant and preferential species), but stands are usually separated visually on areas of similar texture and composition of the dominants! Many plant communities can vary enormously in their diversity and using matching programs can give false matches. They are flawed for a variety of reasons. A particular plant community is NOT necessarily equivalent to a particular habitat. For example heather hare s-tail cotton-grass mire (M19) can be found on peat <0.5 m thick.
An example of why NVC plant communities alone should not be used for mapping habitats. A comparison of three different survey methods for the same 511 hectares of upland in Scotland Birks & Ratcliffe Phase 1 survey Classification NVC survey
So how do we measure the success of habitat creation and restoration schemes? The system should be reproducible The implementation of the method of assessment should be feasible, realistic and not onerous Fit for purpose Based on ecosystem function and not biodiversity or other measures of habitat quality. - How do you measure ecosystem function? Assess whether the majority of energy flow and nutrient cycling is being performed by species characteristic of that habitat (biotope). The abiotic and biotic attributes will vary between habitat types and thresholds will need some adjustment. Surrogates (biological indicators) may be needed for some physical attributes.
Example The restoration of bogs by blocking grips and drains What is the aim of the restoration work? Is it a) the restoration of peat-forming conditions? or b) the restoration of typical bog vegetation? or c) the restoration of plant and/or animal diversity For aim a) Monitor the cover of Sphagnum along transects at 90 to the drain and/or Assess the frequency and depth of the acrotelm Achieving aim b) does not necessarily achieve aim a) Do not monitor water-levels unless you have a lot of spare kit or money to spend as water-levels fluctuate considerably depending on rainfall events and evapotranspiration rates and therefore detecting a statistically significant change takes a lot of data.
Habitat (Biotope) Creation Which habitats can or cannot be created? Some habitats will take many hundreds of years to create 1) Native broad-leaved woodland on former improved pasture (herb paris and many woodland herbs spread laterally very slowly by rhizomes) 2) Bog (it would take a minimum of 1,000 years to build up 0.5 metre of ombrotrophic peat) 3) Limestone pavement (another ice-age or very sophisticated quarrying techniques) Where does that leave DEFRA s biodiversity offsetting?
Woodland creation Plantings should be at low densities Mixture of trees and shrubs with c. 50% gaps in tree canopy Random clusters to produce naturalistic plantings
Some suggested targets for the creation of native broad-leaved woodland For all native broad-leaved woodlands the targets would be the area covers at least 0.5 ha with a minimum width at any one point of 20 m; the tree (one or two stemmed woody perennials and abandoned coppiced trees) canopy is at least 5m high; canopy cover from native species of trees is between 30 and 90% and the combined canopy cover from trees, understorey (trees at least 2 m tall) and shrub layers (multistemmed woody plants at least 1 m tall) is at least 50%; at least 80% of the canopy cover is composed of native woodland species relevant to the appropriate BAP woodland category; at least 50% of the cover of the field layer (herbaceous perennials) is composed of native woodland plant species relevant to the BAP woodland category; at least one species of native woodland bird species is breeding; more than 90% of each of the canopy, understorey and shrub canopy cover should be of native species of broad-leaved tree or shrub. For example Lowland Beech and Yew woodlands the targets could be: at least 25% of the canopy cover is from beech and/or yew; at least 50% of the cover of the field layer (herbaceous perennials) composed of native woodland plant species characteristic of lowland beech and yew woods.
Native species (69) of herbaceous (mostly non-woody) plants (field layer species) indicative of beech and yew woodland: Agrostis capillaris Ajuga reptans Alliaria petiolata Allium ursinum Anemone nemorosa Arctium minus agg. Arum maculatum Asplenium scolopendrium Blechnum spicant Brachypodium sylvaticum Bromopsis ramosa Calluna vulgaris Campanula trachelium Carex pilulifera Carex remota Carex sylvatica Cephalanthera damasonium Circaea lutetiana Clematis vitalba Cynoglossum offiicinale Daphne laureola Deschampsia cespitosa Deschampsia flexuosa Digitalis purpurea Dryopteris dilatata Dryopteris filix-mas Epilobium montanum Epipactis helleborine Euphorbia amygdaloides Festuca gigantea Fragaria vesca Galium odoratum Geranium robertianum Geum urbanum Glechoma hederacea Hedera helix Holcus lanatus Holcus mollis Hyacinthoides non-scripta Hypericum pulchrum Iris foetidissima Lamiastrum galeobdolon Lonicera periclymenum Luzula pilosa Luzula sylvatica Melampyrum pratense Melica uniflora Mercurialis perennis Milium effusum Moehringia trinervia Mycelis muralis Neottia ovata Oxalis acetosella Poa nemoralis Primula vulgaris Ranunculus ficaria Ranunculus repens Rubus fruticosus agg. Rubus idaeus Rumex sanguineus Ruscus aculeatus Sanicula europaea Silene dioica Stachys sylvatica Tamus communis Teucrium scorodonia Vaccinium myrtillus Viola reichenbachiana Viola riviniana
Potential bird indicators of lowland mixed deciduous woodland Blue tit Coal tit Great tit Long-tailed tit Marsh tit Willow tit Blackbird Mistle thrush Song thrush Nightingale Redstart Robin Blackcap Chiffchaff Garden warbler Willow warbler Wood warbler Firecrest Goldcrest Wren Pied flycatcher Spotted flycatcher Bullfinch Chaffinch Hawfinch Lesser redpoll Siskin Treecreeper Nuthatch Great spotted woodpecker Green woodpecker Lesser spotted woodpecker Jay Stock dove Wood pigeon
Some suggested targets for the creation of calcareous grassland The canopy of trees should be no more than 5%. Shrubs should cover no more than 30% of the area. Grasses should make up between 10 and 70% of the herbaceous vegetation cover. At least one strict calcicolous perennial should be at least frequent (in >40% of sample quadrats). At least three characteristic perennials should be constant (i.e. in more than 60% of sample quadrats). At least one characteristic annual/biennial at a frequency of at least 20%. At least one species of butterfly typical of calcareous grassland should be present (e.g. common blue). Cover of agriculturally favoured species should be <5% and each species have a frequency of less than 40%.
Some key points to consider in assessing the success of habitat restoration and creation schemes Phase 1 habitat and NVC mapping are not sufficiently robust to monitor change (these off-the-shelf classifications need revisiting). Clearly define the objectives of the scheme. Set targets for those objectives that can be achieved within the life-time of the restoration period. Set up a reproducible monitoring scheme that can give unambiguous answers as to whether the restoration objectives have been met. Planning officers should raise big red flags for schemes with vague objectives and poorly-defined monitoring regimes that will give ambiguous results. There is only thing worse than not having any data and that is: having data that is of no use. You need data from before as well as after the scheme has started.
How do you produce a restoration and/or habitat creation scheme with an appropriate monitoring regime that will work? Do you need me to tell you the answer? Use a plant-ecologist