Display sequence analysis aug-14

http://marceloarayasalas.weebly.com/

---

---

Site map

---

STATISTICS

Repertoire

## HC1 repertoire:
##  BK  BP  CH  DP  FL  GD  HU  PD  PE  SQ  SS  TF 
##  19  88  15  17 177  36  28  23 117  32   7  72 
## ######  
##  
## TR2 repertoire:
## BK BP CH DP FL GD HU PD PE SQ SS TF 
##  8 10  5  9 90 26  9 15 32 13  7 16 
## ######  
##  
## LOC repertoire:
##  BK  BP  CH  DP  FL  GD  HU  PD  PE  SQ  SS  TF 
##   6  15   4  12 152  23   8  15  38  14   6  18 
## ######  
##  
## SJA repertoire:
##  BK  BP  CH  DP  FL  GD  HU  PD  PE  SQ  SS  TF 
##   1  47  10  20 162  19  21  10  39  16  10  23 
## ######  
##  
## SUR repertoire:
##  BK  BP  CH  DP  FL  GD  HU  PD  PE  SQ  SS  TF 
##  31  55  31  55 241  28  44  46 161  21  60 139 
## ######  
##  
## CCL repertoire:
## BK BP CH DP FL GD HU PD PE SQ SS TF 
##  9  4  2 10 25  1  9  7 22  2  4 15 
## ######  
## 

Unique elements per site

## TR2 VS HC1character(0)
## LOC VS TR2character(0)
## SJA VS LOCcharacter(0)
## SUR VS SJAcharacter(0)
## CCL VS SURcharacter(0)

## [1] unidis i     
## <0 rows> (or 0-length row.names)

Relation between # of elements and sampling effort

Elements common to all sites

##  [1] "BK" "BP" "CH" "DP" "FL" "GD" "HU" "PD" "PE" "SQ" "SS" "TF"

transition charts per lek

transition charts per lek only for common display elements

(Remove displays with less than 4 elements)

(Remove males with only 1 sequence)

---

Descriptive statistics

Vocal dialects results

Songtypes per lek

Non conservative

## CCL HC1 LOC SJA SUR TR2 
##   1   3   2   7   4   4

## [1] 3.5

Conservative

## CCL HC1 LOC SJA SUR TR2 
##   1   2   1   2   2   2

## [1] 1.667

Number of males sound-recorded per lek

## CCL HC1 LOC SJA SUR TR2 
##   8   9   9  14  15   9

## [1] "mean= 10.67"

Visual display variation

Number of displays elements per site

## [1] 12

## [1] "min= 12"

## [1] "max= 12"

Total number of males

## [1] 29

Number of males per lek

## CCL HC1 LOC SJA SUR TR2 
##   4   4   4   4   9   4

## [1] "mean # of males= 4.833"

## [1] "min= 4"

## [1] "max= 9"

Number of hours of video per male

## [1] "mean minutes of video per male for all males= 29.7625"

## [1] "number of males= 56"

## [1] "mean minutes of video per male for focal males= 40.44"

## [1] "min minutes of video per male for focal males= 9.93"

## [1] "max minutes of video per male for focal males= 100.22"

## [1] "number of males= 29"

Common displays to all leks

## [1] 12

Uncommon displays (Not found at all leks)

## [1] 0

total number of display elements

## [1] 12

## [1] unidis i     
## <0 rows> (or 0-length row.names)

Display element composition

Per lek:

All:

Hypothesis testing

Repertoire composition analysis (contingency table chi.square)

##    Disp.elem SUR HC1 TR2 LOC SJA CCL
## 1         BK  31  19   8   6   1   9
## 2         BP  55  88  10  15  47   4
## 3         CH  31  15   5   4  10   2
## 4         DP  55  17   9  12  20  10
## 5         FL 241 177  90 152 162  25
## 6         GD  28  36  26  23  19   1
## 7         HU  44  28   9   8  21   9
## 8         PD  46  23  15  15  10   7
## 9         PE 161 117  32  38  39  22
## 10        SQ  21  32  13  14  16   2
## 11        SS  60   7   7   6  10   4
## 12        TF 139  72  16  18  23  15

## 
##  Pearson's Chi-squared test
## 
## data:  as.matrix(cont.tab.unc[, 2:ncol(cont.tab.unc)])
## X-squared = 291.7, df = 55, p-value < 2.2e-16

Excluding uncommon display elements

##    Disp.elem SUR HC1 TR2 LOC SJA CCL
## 1         BK  30  19   8   6   1   9
## 2         BP  53  88  10  15  47   4
## 3         CH  31  14   4   4   9   1
## 4         DP  51  15   9  12  20   6
## 5         FL 217 169  88 152 160  20
## 6         GD  28  36  26  23  19   1
## 7         HU  39  24   7   7  16   4
## 8         PD  46  21  15  15  10   6
## 9         PE 153 115  32  37  38  14
## 10        SQ  18  32  13  14  16   2
## 11        SS  58   6   7   5  10   4
## 12        TF 131  65  15  16  21  11

## 
##  Pearson's Chi-squared test
## 
## data:  as.matrix(cont.tab.com[, 2:ncol(cont.tab.com)])
## X-squared = 303.7, df = 55, p-value < 2.2e-16

Compare display composition per SITE (Combine leks from same site)

## 
##  Pearson's Chi-squared test
## 
## data:  as.matrix(site.cont.tab)
## X-squared = 101.8, df = 22, p-value = 3.096e-12

##        comps    pvalue chisquared
## 1  LS vs HC1 2.997e-10      65.42
## 2  LS vs TR2 6.123e-05      35.79
## 3 HC1 vs TR2 1.128e-05      40.14

##         comps    pvalue chisquared
## 1  SUR vs HC1 3.797e-14      82.14
## 2  SUR vs TR2 4.651e-10      60.66
## 3  SUR vs LOC 7.101e-16      90.99
## 4  SUR vs SJA 2.727e-16      93.11
## 5  SUR vs CCL 1.088e-02      15.44
## 6  HC1 vs TR2 2.256e-06      40.14
## 7  HC1 vs LOC 3.306e-10      61.46
## 8  HC1 vs SJA 1.898e-08      51.90
## 9  HC1 vs CCL 1.259e-06      41.61
## 10 TR2 vs LOC 3.711e-02       9.71
## 11 TR2 vs SJA 9.119e-06      36.55
## 12 TR2 vs CCL 2.477e-05      33.93
## 13 LOC vs SJA 2.408e-04      27.69
## 14 LOC vs CCL 1.914e-08      51.88
## 15 SJA vs CCL 5.184e-11      65.75

Temporal parameters:

##                  Display.Duration TFR.duration  PEbreak FLduration
## Display.Duration           1.0000      0.30175  0.19298     0.1386
## TFR.duration               0.3018      1.00000 -0.09649    -0.1491
## PEbreak                    0.1930     -0.09649  1.00000     0.1982
## FLduration                 0.1386     -0.14912  0.19825     1.0000
## FLnumber                   0.5380     -0.02984  0.32032     0.3624
##                  FLnumber
## Display.Duration  0.53796
## TFR.duration     -0.02984
## PEbreak           0.32032
## FLduration        0.36244
## FLnumber          1.00000

## 'data.frame':    19 obs. of  6 variables:
##  $ lekmale         : Factor w/ 19 levels "CCL 184","HC1 Male2",..: 1 2 3 4 5 6 7 8 9 10 ...
##  $ Display.Duration: num  9 8.93 27.56 13.98 24.73 ...
##  $ TFR.duration    : num  3.37 2.73 2.02 3.36 8.4 ...
##  $ PEbreak         : num  3.7 15.63 5.27 2.7 5.78 ...
##  $ FLduration      : num  2.83 2.95 6.2 5.51 5.61 ...
##  $ FLnumber        : num  1.5 1.333 2.561 0.922 3.012 ...

MANTEL TESTS:

all are based on similarity or distance matrix. Jaccard similarity index for presence absence data. Morisita index for composition data. Euclidean distance for log-transformed acoustic/visual parameters and geographic distance

Visual temporal parameters VS LEK identity

## Monte-Carlo test
## Observation: 0.1255 
## Call: mantel.rtest(m1 = distemplekmale, m2 = lekorgmat, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.04601

## [1] "number of males (n)= 19"

## [1] "males per lek"

## LEK
## CCL HC1 LOC SJA SUR TR2 
##   1   3   3   4   5   3

Visual temporal parameters VS SITE identity

## Monte-Carlo test
## Observation: 0.2327 
## Call: mantel.rtest(m1 = dispsitelmale, m2 = siteorgmat, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.02563

## [1] "number of males (n)= 19"

## [1] "males per SITE"

## SITE
## HC1  LS TR2 
##   3  13   3

Visual repertoire proportions VS LEK identity

## Monte-Carlo test
## Observation: 0.08346 
## Call: mantelnoneuclid(m1 = m1, m2 = m2, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.0802

## [1] "number of males (n)= 21"

## [1] "males per lek"

## LEK
## CCL HC1 LOC SJA SUR TR2 
##   3   2   4   4   5   3

Visual repertoire proportions VS LEK identity for individuals with more than 5 display elements

## Monte-Carlo test
## Observation: 0.154 
## Call: mantelnoneuclid(m1 = m1, m2 = m2, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.02837

## [1] "number of males (n)= 18"

## [1] "males per lek"

## LEK
## CCL HC1 LOC SJA SUR TR2 
##   1   2   4   4   4   3

Visual element presence/absense VS LEK identity

## Monte-Carlo test
## Observation: -0.008026 
## Call: mantelnoneuclid(m1 = m1, m2 = m2, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.5134

## [1] "number of males (n)= 21"

## [1] "males per lek"

## SITE
## CCL HC1 LOC SJA SUR TR2 
##   3   2   4   4   5   3

Visual repertoire proportions VS SITE identity

## Monte-Carlo test
## Observation: -0.1129 
## Call: mantelnoneuclid(m1 = m1, m2 = m2, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.7143

## [1] "number of males (n)= 21"

## [1] "males per lek"

## SITE
## HC1  LS TR2 
##   2  16   3

Visual element presence/absense VS SITE identity

## Monte-Carlo test
## Observation: -0.2516 
## Call: mantelnoneuclid(m1 = m1, m2 = m2, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.9773

## [1] "number of males (n)= 21"

## [1] "males per lek"

## SITE
## HC1  LS TR2 
##   2  16   3

Sequence similarity VS BIRD identity

## Monte-Carlo test
## Observation: 0.1072 
## Call: mantelnoneuclid(m1 = m1, m2 = m2, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 1e-05

## [1] "number of sequences (n)= 160"

## [1] "sequences per male"

## 
##   CCL.146   CCL.185   CCL.187 HC1.Male4 HC1.Male7   LOC.108   LOC.117 
##         2         5         2        15        12         2         4 
##   LOC.174   LOC.208   SJA.124   SJA.155   SJA.228   SJA.243   SUR.143 
##         6         4         7        11         3         5         2 
##   SUR.152   SUR.176   SUR.177   SUR.179 TR2.Male1 TR2.Male2 TR2.Male9 
##         5        34         4        26         3         4         4

## [1] "number of males"

## [1] 21

## [1] "sequences per lek"

## LEK
## CCL HC1 LOC SJA SUR TR2 
##   9  27  16  26  71  11

Sequence similarity VS LEK identity

## Monte-Carlo test
## Observation: 0.05828 
## Call: mantelnoneuclid(m1 = m1, m2 = m2, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.1817

## [1] "number of males (n)= 21"

## [1] "Males per lek"

## LEK
## CCL HC1 LOC SJA SUR TR2 
##   3   2   4   4   5   3

Sequence similarity VS SITE identity

## Monte-Carlo test
## Observation: -0.01591 
## Call: mantelnoneuclid(m1 = m1, m2 = m2, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.5573

## [1] "number of males (n)= 21"

## [1] "Males per site"

## SITE
## HC1  LS TR2 
##   2  16   3

Sequence similarity VS SONG TYPE identity

## Monte-Carlo test
## Observation: -0.05966 
## Call: mantelnoneuclid(m1 = m1, m2 = m2, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.8277

## [1] "number of males (n)= 28"

## [1] "Males per song type"

## 
## A HC1 A LOC A TR2 B HC1 B SJA B TR2 D SJA E CCL L SUR O SUR 
##     1     4     1     2     3     3     1     4     7     2

Acoustic similarity VS SONG TYPE identity

## Monte-Carlo test
## Observation: 0.25 
## Call: mantel.rtest(m1 = dispmean.acST, m2 = storgmatprop6, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 1e-05

## [1] "number of males (n)= 60"

## [1] "Males per lek"

## LEK
## CCL HC1 LOC SJA SUR TR2 
##   6  10   8  13  15   8

Acoustic similarity VS LEK identity

##  num [1:60, 1:60] 0 0 0 0 0 0 1 1 1 1 ...
##  - attr(*, "dimnames")=List of 2
##   ..$ : chr [1:60] "HC1 Male14 1" "HC1 Male15 2" "HC1 Male16 3" "HC1 Male17 4" ...
##   ..$ : chr [1:60] "HC1 Male14 1" "HC1 Male15 2" "HC1 Male16 3" "HC1 Male17 4" ...

## Monte-Carlo test
## Observation: 0.1919 
## Call: mantel.rtest(m1 = dispmean.acL, m2 = storgmatprop7, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 1e-05

## [1] "number of males (n)= 60"

## [1] "Males per lek"

## LEK
## CCL HC1 LOC SJA SUR TR2 
##   6  10   8  13  15   8

Acoustic similarity VS SITE identity

## Monte-Carlo test
## Observation: 0.05844 
## Call: mantel.rtest(m1 = dispmean.acL, m2 = storgmatprop8, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.216

## [1] "number of males (n)= 60"

## [1] "Males per lek"

## SITE
## HC1  LS TR2 
##  10  42   8

Acoustic similarity VS geographic distance

## Monte-Carlo test
## Observation: -0.05455 
## Call: mantel.rtest(m1 = dispmeansong, m2 = dispmeancoor, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.7042

## [1] "number of males (n)= 54"

## [1] "Males per lek"

## LEK
## CCL HC1 LOC SJA SUR TR2 
##   6   9   8  12  13   6

Acoustic similarity VS geographic distance REMOVING HC1

## Monte-Carlo test
## Observation: 0.2446 
## Call: mantel.rtest(m1 = dispmeansong, m2 = dispmeancoor, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.01645

## [1] "number of males (n)= 45"

## [1] "Males per lek"

## LEK
## CCL LOC SJA SUR TR2 
##   6   8  12  13   6

Acoustic similarity VS geographic distance for SUR ONLY

## Monte-Carlo test
## Observation: 0.2334 
## Call: mantel.rtest(m1 = dispmeansong, m2 = dispmeancoor, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 9e-05

## [1] "number of males (n)= 39"

## [1] "Males per lek"

## LEK
## CCL LOC SJA SUR 
##   6   8  12  13

Visual element proportions VS geographic distance

## Monte-Carlo test
## Observation: -0.1551 
## Call: mantelnoneuclid(m1 = m1, m2 = m2, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.8021

Visual element proportions VS geographic distance REMOVING HC1

## Monte-Carlo test
## Observation: -0.2065 
## Call: mantelnoneuclid(m1 = m1, m2 = m2, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.8087

Visual element proportions VS geographic distance for SUR ONLY

## Monte-Carlo test
## Observation: 0.9876 
## Call: mantelnoneuclid(m1 = m1, m2 = m2, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.04197

Acoustic similarity VS geographic distance by LEK

## [1] "TR2"
## [1] "number of males (n)= 6"
## Monte-Carlo test
## Observation: 0.1211 
## Call: mantel.rtest(m1 = surdispmeansong, m2 = surdispmeancoor, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.2733 
## [1] "SUR"
## [1] "number of males (n)= 13"
## Monte-Carlo test
## Observation: 0.05414 
## Call: mantel.rtest(m1 = surdispmeansong, m2 = surdispmeancoor, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.3131 
## [1] "CCL"
## [1] "number of males (n)= 6"
## Monte-Carlo test
## Observation: -0.3971 
## Call: mantel.rtest(m1 = surdispmeansong, m2 = surdispmeancoor, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.867 
## [1] "HC1"
## [1] "number of males (n)= 9"
## Monte-Carlo test
## Observation: 0.6246 
## Call: mantel.rtest(m1 = surdispmeansong, m2 = surdispmeancoor, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.01164 
## [1] "LOC"
## [1] "number of males (n)= 8"
## Monte-Carlo test
## Observation: 0.0798 
## Call: mantel.rtest(m1 = surdispmeansong, m2 = surdispmeancoor, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.33 
## [1] "SJA"
## [1] "number of males (n)= 12"
## Monte-Carlo test
## Observation: 0.5125 
## Call: mantel.rtest(m1 = surdispmeansong, m2 = surdispmeancoor, nrepet = nrepet)
## Based on 100000 replicates
## Simulated p-value: 0.00161