Docstoc

splash

Document Sample
splash Powered By Docstoc
					The Donut Selectivity Model for Fish Fee

Start Model
Read Background

el for Fish Feeding

Enter Predator Name

Predator

Enter Number of Prey

7

Note Major Factors or Considerations of the Predator

none

Done

Redo

Please use your mouse to move from textbox to textbox Then click on the appropriate button

Once you are entirely finished, you will want to save your spreadsheet with a new name

Please Enter Prey Names

And Relative Abundances (Decimal Format; e.g If used, the sum should =1. If not known, just le
1 1 1 1
1

Prey 1 Prey 2 Prey 3 Prey 4
Prey 5

Prey 6 Prey 7

1 1

Please use your mouse to m Then click on the appropria

If you have more than 25 prey items, you'll need to aggregate them

s (Decimal Format; e.g. 0.5) 1. If not known, just leave all as 1

Done

Redo

Back To Start

Please use your mouse to move from textbox to textbox Then click on the appropriate button

Predator Name Predator Location, details, etc.

Predator none

Rank Matrix Taxa Prey 1 Prey 2 Prey 3 Prey 4 Prey 5 Prey 6 Prey 7 Rel Abundance 1 1 1 1 1 1 1 Overlap (Oij) 1 1 1 1 1 1 1

(Rijm) Detection

Number of Prey

7

Summation Sum of ranks should =

0 28

Finish Ranking

Redo Ranking

PLEASE FILL IN RANKS, with 1 as the highest
Reaction Capture Ingestion Icing Taxa Prey 1 Prey 2 Prey 3 Prey 4 Prey 5 Prey 6 Prey 7

Proportion Matrix Overlap 1 1 1 1 1 1 1 Detection #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0!

0

0

0

0

(Pijm) Reaction Capture #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! Ingestion Icing #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! Product (Sij) #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! Taxa Prey 1 Prey 2 Prey 3 Prey 4 Prey 5 Prey 6 Prey 7 RPA Model #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0! #DIV/0!

Summation

#DIV/0!

Summation

#DIV/0!

View Graph

Null Ambient (Ai) 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Null Selectivity 14.3% 14.3% 14.3% 14.3% 14.3% 14.3% 14.3%

700.0%

100.0%

Back to Start

Back to Ranks

If I told you we could predict fish prey preference, and maybe even diet compos with limited data and information, would you believe me? Actually, that is exactly what this model is intended to do.
Start Model

Description of the Rank Proportion Algorithm (RP
1 2 3 4

Assess general characteristics of a predator j Determine all possible prey N If feasible, determine relative abundance A (numerical or mass, ranging from 0-1) If feasible, determine spatio-temporal (x, y, z, & t) overlap Oij between the predato
If steps 3 & 4 possible, then can do diet composition If not possible, then can only prey selectivity, set all abundance and overlaps to 1

5 6 7 8

Determine number of factors M to evaluate Usually four steps of the predation process plus an “icing” factor Rank each prey i for each factor m, ranging from 1 to N (1 being the highest and s Calculate an adjusted (inverse) rank R’ijm to account for ranking highest prey as #

Rijm  ( N  1)  Rijm 
9 Generate rank proportions for each prey and factor, Pijm

P ijm 

 R ijm



N

i1

 R ijm

10 Multiply rank proportions across all M factors for each prey i to develop a preferen

S ij 



M

Pijm

m 1

S ij 


Dij

M

Pijm

m 1

11 Calculate diet composition proxy Dij if abundance and overlap are available (Sij* A 12 Calculate RPA estimate of diet composition (or, using just S’s, for preference), D’i

D 'ij 

D
i 1

N

ij

aybe even diet composition,

Start Model

See Example

n Algorithm (RPA )

or mass, ranging from 0-1) of all prey ap Oij between the predator j and each prey i (scaled from 0-1)

nd overlaps to 1

(1 being the highest and so forth), Rijm r ranking highest prey as #1

rey i to develop a preference proxy, Sij

verlap are available (Sij* Ai* Oij) st S’s, for preference), D’ij

Return

Start Model

Let's say we have a big gaped, visually oriented piscivore. Let's also say that we have two fish prey, where one is less savvy at predator avoidance and the other is much better at it.
Then we would fill in the following rank matrix as such:
Taxa Fish 1 (The wimp) Fish 2 (The savvy fish) Summation Rank Matrix (Rijm ) Overlap (Oij ) Detection Reaction 1 2 1 1 3 Capture 1 2 3 1 2 3 Ingestion 1 2 3 Icing 1 2 3

Which would produce the following proportion matrix:
Proportion Matrix Taxa Fish 1 (The wimp) Fish 2 (The savvy fish) Overlap (Pijm) Capture Ingestion Icing Detection Reaction 1 0.333333 0.666666667 1 0.666667 0.333333333

0.66666667 0.666667 0.666667 0.33333333 0.333333 0.333333

And finally, given our spatial overlap and an equal relative abundance (Null Ambient), we would get the following RPA Model output for predicted diet composition:
Taxa Fish 1 (The wimp) Fish 2 (The savvy fish) Product (Sij ) 0.03292181 0.00411523 RPA Model 88.9% 11.1% Null Ambient (Ai) Selectivity Null 50% 50% 50% 50%

Summation 0.03703704

See Link (in review), xxxx; for further de

uch better at it.

See Link (in review), xxxx; for further details


				
DOCUMENT INFO
Shared By:
Tags: splas
Stats:
views:10
posted:11/28/2009
language:English
pages:22
Description: splash