Thursday, 21 August 2014

Towards a new national plant monitoring scheme

How can we monitor changes in our wildflower populations more effectively? 

This question has recently been on the minds of many of Britain’s botanical movers and shakers, across the various societies who take an interest in such things. 

The group assembles to start work; Kevin Walker on right
Image: M. Pocock 
These organisations are looking at a range of survey methods that might be used in a new national plant monitoring scheme, similar to ones already up and running for birds, bats and butterflies.

I asked BSBI botanist Oli Pescott, based within the Biological Records Centre/Centre for Ecology and Hydrology, to tell us a bit more about one trial that he recently helped to organise, and to give us a bit of an insight into the scientific rationale behind the methods being tested.

LM: What’s your involvement with this, Oli, and how are these trials connected to the proposed national plant monitoring scheme?

Woodland trials: Oli Pescott (right); Bob Ellis (centre)
Image: M. Pocock 
OP: "My employer, the Centre for Ecology and Hydrology (CEH), has long been involved in botanical monitoring through the links between the Biological Records Centre (based at CEH) and the BSBI. CEH also co-authored a review of the need for a new plant monitoring scheme, jointly published with BSBI, Plantlife and the British Trust for Ornithology in 2010. 

The reason for the recent field tests was to try and see if methods utilising randomly selected plots would be practical for the national plant monitoring scheme currently being developed.  We wanted to field test survey methods, which have been designed by BSBI in collaboration with CEH, Plantlife and JNCC.".

Felicity Harris (Plantlife) and WFC volunteers
Image: M. Pocock
LM: So you held a workshop to trial some methods? What were you hoping to achieve?

OP: "We wanted to find out how different ways of selecting plots for monitoring might affect the practicality and see how volunteers experience the survey. We decided to field test three different methods of plot selection (the scheme will be based around volunteers recording up to 5 plots within a 1 km square)".

LM: Why was that so important?

OP: "It’s very important to have randomisation at the heart of any monitoring scheme. The benefits of randomised plots are not only for the quality of the information provided, but they can also mean that surveyors have fewer choices to make in the field. 

The team, with Pete Stroh, Bob & Oli
Image: M. Pocock
We often found that self-selecting plots for recording can be fraught with difficulty: should you include or exclude the scrub from your chalk grassland plot? Do you put your woodland plot in the nice open glade one side of the ditch or the empty shady area where you are standing? 

One idea is that by giving a largish (25-30) selection of random plots to volunteers, some accessible plots can be found and surveyed, but the unknown bias inherent in surveyor’s self-selecting locations is minimised as far as possible".

LM: So you all held a workshop at Juniper Hall to trial some of the different methods currently on the drawing board. Who was there?

OP: "Yes, we had a very successful two days at Juniper Hall in Surrey last month (thanks to the FSC for hosting us!). Kevin Walker, Bob Ellis and Pete Stroh from BSBI and Felicity Harris from Plantlife were there, as were volunteers from Plantlife’s Wildflowers Count".

Pete, Kevin and the volunteers
Image: M. Pocock
LM: And what happened when you went out in the field to test the three methods?

OP: "Well, we spilt into three groups and each group was assigned one of three methods to try out on a grid square. So, essentially we were all traipsing around a 1 km square trying to work out how many of the plot locations on our maps were accessible! We also tried out laying our plots in different habitat types, and attempted to see how different plot sizes would affect how many ‘target’ species were recorded in any one habitat.

Gridded plots
Image: O. Pescott
We had prepared maps for a number of 1 km squares around Box Hill. For any one square, three different plot selection methods were used to generate 1 km square OS maps overlaid with the potential plot positions for any one method. All the groups had a chance to try each one of the methods trialled. For example, one of the methods trialled uses gridded plots. 

Here (on left) is an example of a systematic plot selection method with the plots laid out in a regular grid, so you can see what I mean".

LM: So, what do you see as the main challenge with this approach?

OP: "Well, the main challenge when volunteers self-select plots is that they may choose the nicer locations containing the species that we are asking them to monitor. This means that the starting point of the scheme is a set of fairly species-rich plots, and so any indicator that might be produced and used to inform on the state of the countryside is far more likely to go down (i.e. most plots are better than average and are therefore more likely to get worse than better). 

Volunteers try out one of the survey methods
Image: M. Pocock 
We need plots in all type of situations so that we can detect increases in plant populations as well as declines. Ultimately it’s about knowing that the approach will provide information that is as accurate as possible, but still remains fun and interesting for volunteers".

LM: Did you find anything interesting while you were out in the field?

OP: "Yes! One of the squares had good populations of one of Britain’s rarest woodland plants, Cynoglossum germanicum (Green Hound’s-tongue)! Box Hill is one of it strongholds but we had no idea we would find it in the woods we were surveying. This came to light as we were walking to a random plot location in the evening gloom".

Cynoglossum germanicum
Image: K. Walker
LM: And here is BSBI's distribution map for C. germanicum. Oli, can you close by giving us an idea of what the next step is and tell us a bit more about how these trials are connected to ideas for a new national plant monitoring scheme?

OP: "The next step is to review all of the results from these trials. This year Wildflowers Count volunteers have also been given the option to use plots in their surveys, although these were all self-selected. We have to review the feedback from those volunteers as well, and then finally recommend a particular approach to JNCC. The scheme is currently out to tender, but the organisation (or organisations!) that are successful in their bid will roll-out the scheme for an initial three year period using the methods that we have developed.

It’s a big challenge, but it would be fantastic if the plant world finally had something to rival the Breeding Bird Survey or the UK Butterfly Monitoring Scheme!"

Thanks Oli, I couldn't agree more! Let's close with a few words from Kevin Walker, BSBI's Head of Science:

Pete, Felicity, Kevin and one of the WFC volunteers
Image: M. Pocock
"The absolute key for this scheme is to make it as straightforward and enjoyable as possible for volunteers. We therefore wanted to road-test the different methods to help us select which approach worked best, as well as to iron out as many 'bugs' with the methods as possible. 

"It was great to be able to discuss these issues in the field with colleagues whom we've been working with for a number of years, developing the methods and producing the guidance. It was also very sobering to see how things did or didn't work, but great to be able to make decisions on the spot when it was clear that aspects of the scheme weren't practical. We are now much more confident that the scheme will work, and that it will provide an enjoyable and rewarding experience for volunteers!"