0

I am interested in testing whether species with trait A have wider niche breadths than species with trait B. I have been using Maxent species distribution modeling. After generating 100 bootstrap replicate Maxent species distribution maps for each species, I used a function in ENMtools to calculate niche breadth using Levins' B, giving me 100 estimates of niche breadth for each species.

What statistical test can I use for 1) a single pair of species with trait A and trait B and 2) a group of species pairs consisting of species with trait A and trait B

Also, am I really dealing with pseudoreplicate data?

Any assistance is appreciated.

Edit, Data is pasted below. I am using the stand alone versions of Maxent and ENMTools, so I do not have code. I have used R to try and analyze the data using t-tests and sign tests, otherwise there is no code to add.

Sample data:

species A   species B
0.168581128 0.092476942
0.138468258 0.097536175
0.16412658  0.107661982
0.16371685  0.103260373
0.162732326 0.121757497
0.129050784 0.113096489
0.126583781 0.114981732
0.176855441 0.099364014
0.257605323 0.103776886
0.182453389 0.108293928
0.175081388 0.106879389
0.223292158 0.112947306
0.254094648 0.127991856
0.147219108 0.108193268
0.215448167 0.092778064
0.103140149 0.098756118
0.176492044 0.101350818
0.17121499  0.085262687
0.173945262 0.123863121
0.187958506 0.103502695
0.14655381  0.129826611
0.217358822 0.097517987
0.129096849 0.107359879
0.19682274  0.096810086
0.138933825 0.098270337
0.165596467 0.106029483
0.132607982 0.112006155
0.195556231 0.104323653
0.117660212 0.099375585
0.203652419 0.095314923
0.148629883 0.102969753
0.134932182 0.095842915
0.170387616 0.10520627
0.155080925 0.11811477
0.16431599  0.094490674
0.150336814 0.127094893
0.152238428 0.101582407
0.20186197  0.106561491
0.148248253 0.102802711
0.187076952 0.112530243
0.136677706 0.105465642
0.163433556 0.109735202
0.199990899 0.114393721
0.137892935 0.101238746
0.134763941 0.106226666
0.209942031 0.100820784
0.217706637 0.108691033
0.20036096  0.096355532
0.164807164 0.092417125
0.15371586  0.114739325
0.219866044 0.106098684
0.206539001 0.108628272
0.143972195 0.100132843
0.141272459 0.102480243
0.147306813 0.101444936
0.122265342 0.105942924
0.180917047 0.105722483
0.120716411 0.106555329
0.18800414  0.113761784
0.174485389 0.106940597
0.136967932 0.115609256
0.214880862 0.116237716
0.152007642 0.099203843
0.171732613 0.106697919
0.16261444  0.100392708
0.211592418 0.111231976
0.185334669 0.121699976
0.205723283 0.101651487
0.177553405 0.103634347
0.255510009 0.130654929
0.233807884 0.105441321
0.134768798 0.101113701
0.199416249 0.091038829
0.151712987 0.105402425
0.209823704 0.115964528
0.181800387 0.100236837
0.13240936  0.112243762
0.163134833 0.102686995
0.133972129 0.099741739
0.167130799 0.099626019
0.201614387 0.127056303
0.162793942 0.106440826
0.115234276 0.105374317
0.187650165 0.105206766
0.157882045 0.106361622
0.158200651 0.097265822
0.16939029  0.103107653
0.186945284 0.109809801
0.21049161  0.108904949
0.12530485  0.107502039
0.17948561  0.112303038
0.171934415 0.113321492
0.216209636 0.092409221
0.133687523 0.121289466
0.170516709 0.095395971
0.123796452 0.097354942
0.152966332 0.102279392
0.173328517 0.127320454
0.218387925 0.088878651
0.213022848 0.119487608
IRTFM
  • 258,963
  • 21
  • 364
  • 487
Alannie-Grace Grant
  • 139
  • 1
  • 1
  • 7
  • 1
    Need more code and datasets for this to be meaningful coding question. It's way too nebulous at the moment. – IRTFM Apr 12 '16 at 23:37
  • Thanks. I pasted in some of my data, but since the programs I used to generate the data are stand alone, there is not any code to add. I am just seeking recommendations for data analysis. – Alannie-Grace Grant Apr 13 '16 at 02:35
  • The program I use to search for function that do "new" things is `sos::findFn`. I get 128 hits in 24 packages for a search on "maxent". There is also an R-SIG-epi mailing list. – IRTFM Apr 13 '16 at 02:45
  • Maxent is a general term. I used Maxent species distribution modeling. – Alannie-Grace Grant Apr 13 '16 at 17:00
  • A single vector for each species is offered but what connection to reality is presumed? Are these measurements of some observable quantity at simultaneous sequential times with constant interval? You are not doing a good job of explaining your domain in terms that a computer scientist might understand. Using `sos::findFn("Maxent species distribution")` I still find a great many hits. – IRTFM Apr 13 '16 at 17:13

0 Answers0