Adder Dispersal Distances

calumma wrote: The process of calculating range using the database may take some computation time though. If I remember correctly you mentioned that it takes 20 seconds to undertake the nearest neighbour based risk analysis. If there are 2015 squares in the county, thats an 11+ hour calculation

I think there's an easier way to do this. If the 95% nearest neighbour distance is, say 1800m, that's about 2 grids.

For each grid where adder occurs (already listed in the database), I can add the grids within 1800m as potential range, then subtract unsuitable habitat grids. This will be much faster, and considering the assumptions...probably good enough.

This should take less than a minute to calculate.


Edited by Vicar

Steve, how about a single map to bring all three elements together?

Red squares to indicate areas that have suitable habitat and are within
predicted range, yellow squares to indicate areas that have suitable habitat
and are outside of predicted range - or similar?

More work has been done on this, particularly re habitat suitability, which has reduced the number of grids deemed to be suitable.

Probability of occurrence across Surrey for adder, based on curve fitting of nearest neighbour analysis to a hyperbolic logistic offset function looks like:


Where Y = Probability of occurrence(%) and X = range from nearest neighbour (km).  This type of function fits all the nearest neighbour analysis well, using different a,b & c coefficients for different species (d = 100: 100% probability at zero range, i.e. at a known location.).

The surface plot looks like:



The above makes no compensation for habitat, just works on spatial distribution. When you add the suitable habitat filter you get:



And when you put it all together you get a map like:



Where black grids are known occupancy, red grids are predicted range, and yellow grids are suitable habitat outside of the predicted range.


Edited by Vicar

calumma wrote: A conservation conundrum rears its head though. Is it better to expend limited resources protecting large blocks of habitat around core populations and accept that outlying populations may be lost. Or expend resources creating linkages between outliers and accept that core populations may be further eroded.

I may be at odds with the rest of the conservation world here...

I accept that large meta-populations, linked by suitable habitat are great for stirring up the gene pool and improving fitness. However; they are also a great means for ensuring that a pathogen gets your whole population.

I'm thinking along the lines of Natterjacks and chytrid etc.

The 'happy mix' as far as I'm concerned is a mix of both, where we have large linked populations, but also a few 'reserves' from which re-population can be seeded in the event of fire or disease.

I'm happy with some isolated outliers .

In connection with the issue of big blocks of habitat, there is the paradoxical negative consequence of securing more habitat such as heathland and consequent large scale inappropriate mismanagement.  I heard a programme on R4 the other day in which the RSPB representative stated that the implication of securing more heathland from Forestry Commission land around the Poole Bournemouth conurbation was that they could start burning the heath again rather than using more intensive kinds of management.  Not sure what NE would say about that ?..