the online meeting place for all who love our amphibians and reptiles |
|
GCN terrestrial habitat range |
Post Reply | Page <12 |
Author | |
Matt Harris
Senior Member Joined: 03 Jun 2003 Location: United Kingdom Status: Offline Points: 233 |
Post Options
Thanks(0)
|
As you double the radius of search area, the area of search is quadrupled so detectability would be quartered, assuming even distribution? |
|
Local Authority Ecologist
|
|
GemmaJF
Admin Group Joined: 25 Jan 2003 Location: Essex Status: Offline Points: 4359 |
Post Options
Thanks(0)
|
That is how I would figure it Matt.
It seems that people are looking at the linear distance but not allowing for increase in search area and subsequent decrease in detectability simply due to it being a significantly larger area. Getting away from the uniform habitat model, in the real world one could also easily have a 'hot spot' due to significantly better habitat or a barrier beyond the easy to search 50m circle that is not considered.
|
|
Noodles
Senior Member Joined: 05 Dec 2010 Location: United Kingdom Status: Offline Points: 534 |
Post Options
Thanks(0)
|
One of the of the findings of the Cresswell report is that capture rates do increase beyond 50 or 100 m if foci exist e.g. hedgerows. Since it was a study of past mitigation schemes, capture figures were obtained from trap densities spaced evenly throughout sites. The findings therefore do show a typical spatial distribution in relation to the pond, regardless of the size of search area.
|
|
GemmaJF
Admin Group Joined: 25 Jan 2003 Location: Essex Status: Offline Points: 4359 |
Post Options
Thanks(0)
|
I would consider that to be mathematically a contradiction of terms, as the trap density would have to be increased the further away from the pond you were to give a true indication of spatial distribution as detectability reduces by a factor of 4 each time the radius of the search area is doubled as discussed. If the trap density is kept constant than all one is showing is that detectability reduces with increase in search area, which was the point I was making. That aside it's not to say that I don't think the number of animals reduces the further away from the pond one gets, as at most sites it is logical that one would find that. I'm just dubious of so called 'scientific' data that contain flaws in basic reasoning being quoted as proof of anything! At the very least this methodology would exaggerate the results significantly, the bulk of the population could well have been further out than the results suggest for example.
Edited by GemmaJF - 08 Nov 2012 at 4:10pm |
|
MancD
Member Joined: 05 Jun 2011 Status: Offline Points: 38 |
Post Options
Thanks(0)
|
It is difficult to compare trapping data between sites for a host of reasons. There is a minimum trap density required based on the level of population at the site (small/medium/large), and in some cases ecologists deliberately lay more traps around and close to ponds because they believe there will be more newts there. I've been on sites where they are a 2 metres apart. On the flip side, you could argue that trapping density doesn't really matter. If a newt has to walk 5 metres along a fence line before it hits a trap, over the course of that evening the likelihood is that it would walk an extra 5 metres and fall in the next one so it would be captured anyway.
Drift fencing is also a problem when comparing sites. Some people put huge amounts of drift fencing in, others very little, some put them in parallel lines or "X's" across sites, whereas others target specific habitat areas.
Time of year is also important, the spring migration to the pond is probably much stronger and shorter than the late summer/autumn migration to hibernation sites and of course, the habitat quality as always is important.
As I'm a bit of a geek, I love wading through trapping data to try and figure out what is going on at a site. But using that data or its interpretation on another site is pretty much a waste of time due to the number of variables.
A case I'm working on at the moment has caught around 50 individuals, all of them in the areas you would expect, but all of them were female. What are the chances of that?
|
|
Noodles
Senior Member Joined: 05 Dec 2010 Location: United Kingdom Status: Offline Points: 534 |
Post Options
Thanks(0)
|
I believe all those variables were taken into account and investigated during data analysis in the Cresswell report. Other similar projects nowadays would also take that approach to analysis. Actually i think the Cresswell study is extremely valid and useful and one that we should be building upon with continued research. As Jon C suggests the original distances quoted now seem very arbitrary.
|
|
GemmaJF
Admin Group Joined: 25 Jan 2003 Location: Essex Status: Offline Points: 4359 |
Post Options
Thanks(0)
|
I would say the chances of an all female 'population' were pretty slim, one would assume the males left earlier in the season. Why? Perhaps they know another pond full of females! I would have thought the seasonal dynamic would be the main issue when collecting data. No doubt it would have been a case that mitigations would have been targeted when newts were migrating. So how far out the majority of captures were is not really indicative of distance they may travel to terrestrial habitat in a season. Simply indicative of where the migrating animals were intercepted. Edited by GemmaJF - 08 Nov 2012 at 4:40pm |
|
GemmaJF
Admin Group Joined: 25 Jan 2003 Location: Essex Status: Offline Points: 4359 |
Post Options
Thanks(0)
|
Trying to take all the variables into account when dealing with ecology is simply impossible. We have trouble mathematically taking all the variables into account for an apparently 'simple' natural phenomenon such as flight. To think this is achievable during a field study in the real world involving the distribution of wild animals with variables we don't even know about can't be valid? There are so many published scientific reports in ecology with flaws in basic reasoning. At best they relate to the dynamics of the study site, providing a 'snap-shot' of what might be happening. Often they do not even give a true or valid representation even of that! Perhaps these people should actually study maths before writing the papers! You state above that the report is valid, but what about the glaring flaw in the methodology let alone all the other variables? Here is analogy of what they actually did: A shop keeper opens a shop in the middle of town. After a couple of months he decides to do a survey to see if people know about his shop. He asks five random people in the street that the shop is in if they know about his shop. 4 of them say yes. He then travels 1km out into the suburbs and asks five random people if they know about his shop. They all reply NO. He concludes that his customer base in mainly in the street where the shop is. He totally misses the point that 1km out 5000 people know about his shop, he's just far less likely to meet one of them while taking his random sample.... Edited by GemmaJF - 08 Nov 2012 at 5:06pm |
|
Caleb
Senior Member Joined: 11 Apr 2011 Status: Offline Points: 660 |
Post Options
Thanks(0)
|
Surely if the trap density is uniform, it would be more like going 1km out and asking 500 people? Or am I missing something?
|
|
GemmaJF
Admin Group Joined: 25 Jan 2003 Location: Essex Status: Offline Points: 4359 |
Post Options
Thanks(0)
|
I see your point Caleb a better analogy would ask more people further out, but my point remains that detectability has dropped by a factor of 4 each time the radius doubles.
Lets say the trapping was a ring fence with pitfall traps around a pond just after the breeding season. At 50m out one can expect a lot of animals to intercept the fence and fall in the traps. At 100m out they have 4 times the area (3 times if you subtract the first 50m circle) to wander about in, so would be less likely to intercept the ring fence however many traps are in place. One just can't make the assumption that they walk away from the pond in straight lines until they hit the fence and then hope to extrapolate from that population distribution. As the data comes from mitigation work, I doubt they released animals caught at 50m to the other side of the fence either to see if they would have gone further, so the results just end up very skewed towards large numbers nearer the pond. The main point I'm really making is that we should all be careful when interpreting results of studies. They can 'look' very conclusive at times but even a gentle analysis often reveals flaws. I would not for example let the findings of this report influence any assessment I made about the impacts of a potential development on a GCN population for example. Others will take the findings and interpret it as proof, this is how we often end up with things working down to the lowest common denominator.
Edited by GemmaJF - 09 Nov 2012 at 5:07pm |
|
Post Reply | Page <12 |
Tweet
|
Forum Jump | Forum Permissions You cannot post new topics in this forum You cannot reply to topics in this forum You cannot delete your posts in this forum You cannot edit your posts in this forum You cannot create polls in this forum You cannot vote in polls in this forum |