Birds_Batha_analysis_by_plot
Ron Chen
2023-06-12
## Loading required package: readxl## Loading required package: lubridate##
## Attaching package: 'lubridate'## The following objects are masked from 'package:data.table':
##
## hour, isoweek, mday, minute, month, quarter, second, wday, week,
## yday, year## The following objects are masked from 'package:base':
##
## date, intersect, setdiff, union## Loading required package: solartime## - Winter data excluded.
## - The counts of all passing / wintering species (not resident nor breeding in the sampled unit) were set to zero. Abreed_and_non_breed contains the non-breeders as well.Herbaceous and Dwarf-Shrub Vegetation
Factors are proximity to agriculture (plantations and orchards) and time. Total 4 campaigns. 2 regions: herbaceous batha (grassland) and dwarf-shrub batha (low shrubland). 5 sites per region (10 sites total) with 6 plots per site (60 plots total). Raw data Total abundance: 14138 Number of observations: 3561 Total richness: 122
Filtered data Total abundance: 6741 Number of observations: 2521 Total richness: 57
Model gma, abundance and richness
Richness done with rare species. Abundance and mean abundance done without rare species. Full models include cosine and sine of the time difference from June 21st (in radians).
richness
Explore data and plot mean-variance plot. There is a strong relationship, indicating that employing GLMs is the proper way to analyze, rather than OLS (assumption of homogeneity is violated).
## [1] "RICHNESS WITH RARE SPECIES"## ℹ SHA-1 hash of file is "eef55df1963ba201f57cf44513f566298756fd4f"
| richness | year_ct | habitat | site | agriculture | td_sc | cos_td_rad | sin_td_rad | h_from_sunrise | cos_hsun | sin_hsun | monitors_name | wind | precipitation | temperature | clouds | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Min. : 1.000 | Min. :2.0 | Basalt :120 | Batot Taninim :24 | Far :120 | Min. :-2.17579 | Min. :-0.2498 | Min. :-0.9995 | Length:240 | Min. :-0.1175 | Min. :-0.1771 | Other : 53 | 0 : 0 | 0 : 13 | 0 : 0 | 0 : 6 | |
| 1st Qu.: 6.000 | 1st Qu.:3.5 | Limestone:120 | Gamla :24 | Near:120 | 1st Qu.:-0.76251 | 1st Qu.: 0.3757 | 1st Qu.:-0.9267 | Class :difftime | 1st Qu.: 0.7043 | 1st Qu.: 0.2034 | Eyal Shochat: 7 | 1 : 0 | 3 : 0 | 1 : 0 | 1 : 1 | |
| Median : 8.000 | Median :5.0 | NA | Hatzor :24 | NA | Median :-0.51689 | Median : 0.4758 | Median :-0.8792 | Mode :numeric | Median : 0.8880 | Median : 0.4598 | Adi Domer : 0 | 2 : 0 | NA’s:227 | 2 : 6 | 2 : 4 | |
| Mean : 8.108 | Mean :5.0 | NA | Karei Deshe :24 | NA | Mean : 0.09807 | Mean : 0.5626 | Mean :-0.6066 | NA | Mean : 0.8063 | Mean : 0.4517 | Asaf Mayrose: 0 | 3 : 1 | NA | 3 : 0 | 3 : 0 | |
| 3rd Qu.:10.000 | 3rd Qu.:6.5 | NA | Natur :24 | NA | 3rd Qu.: 1.11539 | 3rd Qu.: 0.9048 | 3rd Qu.:-0.3233 | NA | 3rd Qu.: 0.9791 | 3rd Qu.: 0.7099 | Eliraz Dvir : 0 | NA’s:239 | NA | NA’s:234 | NA’s:229 | |
| Max. :16.000 | Max. :8.0 | NA | Ramat Hashofet:24 | NA | Max. : 2.28504 | Max. : 0.9905 | Max. : 0.1882 | NA | Max. : 1.0000 | Max. : 0.9998 | (Other) : 0 | NA | NA | NA | NA | |
| NA | NA | NA | (Other) :96 | NA | NA | NA | NA | NA | NA’s :60 | NA’s :60 | NA’s :180 | NA | NA | NA | NA |
no observation for all 4 weather variables. many NAs for sampling time of day variables. exclude from model.
## Warning in cor(P.anal[, lapply(X = .SD, FUN = as.numeric), .SDcols =
## IVs[1:12]], : the standard deviation is zero| richness | year_ct | habitat | site | agriculture | td_sc | cos_td_rad | sin_td_rad | h_from_sunrise | cos_hsun | sin_hsun | monitors_name | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| richness | 1.0000000 | 0.2277410 | 0.1538342 | 0.0664861 | 0.4137611 | -0.2439739 | -0.1946559 | -0.2720706 | -0.1340734 | 0.1471056 | -0.1232836 | 0.1321602 |
| year_ct | 0.2277410 | 1.0000000 | 0.0000000 | 0.0000000 | 0.0000000 | -0.7463144 | -0.6558516 | -0.7587452 | -0.0861080 | 0.0846278 | -0.0719069 | NA |
| habitat | 0.1538342 | 0.0000000 | 1.0000000 | -0.0348155 | 0.0000000 | -0.0427597 | -0.0632391 | -0.0406794 | 0.1656632 | -0.1716399 | 0.1548456 | 0.2595871 |
| site | 0.0664861 | 0.0000000 | -0.0348155 | 1.0000000 | 0.0000000 | -0.0384624 | -0.0368278 | -0.0572513 | -0.0170782 | -0.0178346 | -0.0435386 | 0.4247701 |
| agriculture | 0.4137611 | 0.0000000 | 0.0000000 | 0.0000000 | 1.0000000 | -0.0030733 | -0.0053199 | -0.0014405 | -0.0292304 | 0.0494225 | -0.0202046 | 0.0519174 |
| td_sc | -0.2439739 | -0.7463144 | -0.0427597 | -0.0384624 | -0.0030733 | 1.0000000 | 0.9542304 | 0.9752788 | -0.1904036 | 0.1801172 | -0.1861300 | -0.5897244 |
| cos_td_rad | -0.1946559 | -0.6558516 | -0.0632391 | -0.0368278 | -0.0053199 | 0.9542304 | 1.0000000 | 0.8674071 | -0.1918124 | 0.1793379 | -0.1892847 | -0.5202428 |
| sin_td_rad | -0.2720706 | -0.7587452 | -0.0406794 | -0.0572513 | -0.0014405 | 0.9752788 | 0.8674071 | 1.0000000 | -0.1677980 | 0.1708620 | -0.1544942 | -0.7612499 |
| h_from_sunrise | -0.1340734 | -0.0861080 | 0.1656632 | -0.0170782 | -0.0292304 | -0.1904036 | -0.1918124 | -0.1677980 | 1.0000000 | -0.9486862 | 0.9809251 | -0.0373316 |
| cos_hsun | 0.1471056 | 0.0846278 | -0.1716399 | -0.0178346 | 0.0494225 | 0.1801172 | 0.1793379 | 0.1708620 | -0.9486862 | 1.0000000 | -0.8726227 | 0.0445516 |
| sin_hsun | -0.1232836 | -0.0719069 | 0.1548456 | -0.0435386 | -0.0202046 | -0.1861300 | -0.1892847 | -0.1544942 | 0.9809251 | -0.8726227 | 1.0000000 | -0.0380718 |
| monitors_name | 0.1321602 | NA | 0.2595871 | 0.4247701 | 0.0519174 | -0.5897244 | -0.5202428 | -0.7612499 | -0.0373316 | 0.0445516 | -0.0380718 | 1.0000000 |
Fit Poisson glm, check for existence of overdispersion
## [1] "Estimating overdispersion parameter phi: (Res. Dev.)/(n-p) where n=number of observations; p=number of parameters in the model."## [1] "od = 0.967198263346182"Overdispersion parameter is <1 (underdispersion) and therefore Poisson is preferable to negative binomial.
| Observations | 180 (60 missing obs. deleted) |
| Dependent variable | richness |
| Type | Generalized linear model |
| Family | poisson |
| Link | log |
| χ²(7) | 58.03 |
| Pseudo-R² (Cragg-Uhler) | 0.28 |
| Pseudo-R² (McFadden) | 0.06 |
| AIC | 890.92 |
| BIC | 916.46 |
| Est. | S.E. | z val. | p | |
|---|---|---|---|---|
| (Intercept) | 0.79 | 0.55 | 1.43 | 0.15 |
| agricultureNear | 0.19 | 0.20 | 0.96 | 0.34 |
| year_ct | -0.00 | 0.02 | -0.10 | 0.92 |
| cos_td_rad | 0.43 | 0.21 | 1.99 | 0.05 |
| sin_td_rad | -0.92 | 0.40 | -2.33 | 0.02 |
| cos_hsun | 0.24 | 0.24 | 0.99 | 0.32 |
| sin_hsun | 0.03 | 0.17 | 0.15 | 0.88 |
| agricultureNear:year_ct | 0.03 | 0.03 | 0.85 | 0.39 |
| Standard errors: MLE |
cosine and sine of hours from sunrise - remove them because of many NAs.
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: poisson ( log )
## Formula: richness ~ agriculture * year_ct + cos_td_rad + sin_td_rad +
## (1 | habitat)
## Data: P.anal
##
## AIC BIC logLik deviance df.resid
## 1166.4 1190.7 -576.2 1152.4 233
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.26707 -0.70417 -0.03776 0.56600 2.74584
##
## Random effects:
## Groups Name Variance Std.Dev.
## habitat (Intercept) 0.002287 0.04782
## Number of obs: 240, groups: habitat, 2
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 1.62519 0.16271 9.988 < 2e-16 ***
## agricultureNear 0.14763 0.11840 1.247 0.21244
## year_ct -0.01015 0.01957 -0.519 0.60393
## cos_td_rad 0.22649 0.13994 1.618 0.10556
## sin_td_rad -0.34819 0.11766 -2.959 0.00308 **
## agricultureNear:year_ct 0.03398 0.02110 1.611 0.10727
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) agrclN yer_ct cs_td_ sn_td_
## agricultrNr -0.411
## year_ct -0.422 0.582
## cos_td_rad -0.809 -0.002 -0.023
## sin_td_rad 0.581 0.019 0.391 -0.758
## agrcltrNr:_ 0.380 -0.922 -0.639 0.005 -0.022Mixed model with site as random variable did not converge. Mixed model with habitat as random variable converged. perform stepwise model selection of poisson mixed model.
## Single term deletions
##
## Model:
## richness ~ agriculture * year_ct + cos_td_rad + sin_td_rad +
## (1 | habitat)
## npar AIC
## <none> 1166.4
## cos_td_rad 1 1167.0
## sin_td_rad 1 1173.2
## agriculture:year_ct 1 1167.0drop cosine.
## Single term deletions
##
## Model:
## richness ~ agriculture * year_ct + sin_td_rad + (1 | habitat)
## npar AIC
## <none> 1167.0
## sin_td_rad 1 1172.0
## agriculture:year_ct 1 1167.6drop agriculture : year.
## Single term deletions
##
## Model:
## richness ~ agriculture + year_ct + sin_td_rad + (1 | habitat)
## npar AIC
## <none> 1167.6
## agriculture 1 1215.8
## year_ct 1 1166.1
## sin_td_rad 1 1172.5drop year.
## Single term deletions
##
## Model:
## richness ~ agriculture + sin_td_rad + (1 | habitat)
## npar AIC
## <none> 1166.1
## agriculture 1 1214.2
## sin_td_rad 1 1185.7agriculture and time of year remain. Final model:
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: poisson ( log )
## Formula: richness ~ agriculture + sin_td_rad + (1 | habitat)
## Data: P.anal
##
## AIC BIC logLik deviance df.resid
## 1166.1 1180.0 -579.0 1158.1 236
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4158 -0.7297 -0.0456 0.5709 2.7045
##
## Random effects:
## Groups Name Variance Std.Dev.
## habitat (Intercept) 0.001993 0.04464
## Number of obs: 240, groups: habitat, 2
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 1.76597 0.05832 30.280 < 2e-16 ***
## agricultureNear 0.32310 0.04593 7.035 1.99e-12 ***
## sin_td_rad -0.23971 0.05263 -4.554 5.25e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) agrclN
## agricultrNr -0.456
## sin_td_rad 0.589 0.002## $habitat
## (Intercept)
## Basalt -0.03602694
## Limestone 0.03647298
##
## with conditional variances for "habitat"## $habitat
## Registered S3 methods overwritten by 'broom':
## method from
## tidy.glht jtools
## tidy.summary.glht jtools| Observations | 240 |
| Dependent variable | richness |
| Type | Mixed effects generalized linear model |
| Family | poisson |
| Link | log |
| AIC | 1166.057 |
| BIC | 1179.980 |
| Pseudo-R² (fixed effects) | 0.244 |
| Pseudo-R² (total) | 0.257 |
| exp(Est.) | S.E. | z val. | p | |
|---|---|---|---|---|
| (Intercept) | 5.847 | 0.058 | 30.280 | 0.000 |
| agricultureNear | 1.381 | 0.046 | 7.035 | 0.000 |
| sin_td_rad | 0.787 | 0.053 | -4.554 | 0.000 |
| Group | Parameter | Std. Dev. |
|---|---|---|
| habitat | (Intercept) | 0.045 |
| Group | # groups | ICC |
|---|---|---|
| habitat | 2 | 0.002 |
## Confidence intervals for merMod models is an experimental feature. The
## intervals reflect only the variance of the fixed effects, not the random
## effects.
There is a statistically significant effect for proximity to agriculture.
Near plots have on average 2.5792208 more species than far plots, which is 38.1407177 percent higher.
No statistically significant effect for the other factors tested.
geometric mean of abundance
Explore data. Exclude time of day because of high number of NAs.
## [1] "GEOMETRIC MEAN ABUNDANCE WITHOUT RARE SPECIES"
| gma | year_ct | habitat | site | agriculture | td_sc | cos_td_rad | sin_td_rad | |
|---|---|---|---|---|---|---|---|---|
| Min. :1.072 | Min. :2.0 | Basalt :120 | Batot Taninim :24 | Far :120 | Min. :-2.17579 | Min. :-0.2498 | Min. :-0.9995 | |
| 1st Qu.:2.000 | 1st Qu.:3.5 | Limestone:120 | Gamla :24 | Near:120 | 1st Qu.:-0.76251 | 1st Qu.: 0.3757 | 1st Qu.:-0.9267 | |
| Median :2.465 | Median :5.0 | NA | Hatzor :24 | NA | Median :-0.51689 | Median : 0.4758 | Median :-0.8792 | |
| Mean :2.683 | Mean :5.0 | NA | Karei Deshe :24 | NA | Mean : 0.09807 | Mean : 0.5626 | Mean :-0.6066 | |
| 3rd Qu.:3.014 | 3rd Qu.:6.5 | NA | Natur :24 | NA | 3rd Qu.: 1.11539 | 3rd Qu.: 0.9048 | 3rd Qu.:-0.3233 | |
| Max. :8.367 | Max. :8.0 | NA | Ramat Hashofet:24 | NA | Max. : 2.28504 | Max. : 0.9905 | Max. : 0.1882 | |
| NA | NA | NA | (Other) :96 | NA | NA | NA | NA |
Fit glm, compare gamma, gaussian (poisson inappropriate because response is not discrete)
Gamma seems better than gaussian. Fit fixed and mixed models.
## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00342479 (tol = 0.002, component 1)Mixed model with site as random variable did not converge. Mixed model with habitat as random variable converged:
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( inverse )
## Formula: gma ~ agriculture * year_ct + cos_td_rad + sin_td_rad + (1 |
## habitat)
## Data: P.anal
##
## AIC BIC logLik deviance df.resid
## 596.1 623.8 -290.0 580.1 230
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.6410 -0.6505 -0.1671 0.4522 4.3563
##
## Random effects:
## Groups Name Variance Std.Dev.
## habitat (Intercept) 0.0008619 0.02936
## Residual 0.1203319 0.34689
## Number of obs: 238, groups: habitat, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## (Intercept) 0.358281 0.069384 5.164 2.42e-07 ***
## agricultureNear -0.012325 0.040567 -0.304 0.761
## year_ct 0.026391 0.006516 4.050 5.11e-05 ***
## cos_td_rad 0.002725 0.046505 0.059 0.953
## sin_td_rad 0.186770 0.037079 5.037 4.73e-07 ***
## agricultureNear:year_ct 0.007124 0.007563 0.942 0.346
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) agrclN yer_ct cs_td_ sn_td_
## agricultrNr -0.289
## year_ct -0.419 0.499
## cos_td_rad -0.679 0.007 0.167
## sin_td_rad 0.470 -0.032 0.298 -0.697
## agrcltrNr:_ 0.266 -0.929 -0.529 -0.007 0.032
## optimizer (Nelder_Mead) convergence code: 0 (OK)
## Model failed to converge with max|grad| = 0.00342479 (tol = 0.002, component 1)perform stepwise model selection of Gamma model.
## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00412004 (tol = 0.002, component 1)## Single term deletions
##
## Model:
## gma ~ agriculture * year_ct + cos_td_rad + sin_td_rad + (1 |
## habitat)
## npar AIC
## <none> 596.07
## cos_td_rad 1 594.07
## sin_td_rad 1 618.99
## agriculture:year_ct 1 594.95drop cosine.
## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.00412004 (tol = 0.002, component 1)## Single term deletions
##
## Model:
## gma ~ agriculture * year_ct + sin_td_rad + (1 | habitat)
## npar AIC
## <none> 594.07
## sin_td_rad 1 638.35
## agriculture:year_ct 1 592.96drop agriculture : year.
## Single term deletions
##
## Model:
## gma ~ agriculture + year_ct + sin_td_rad + (1 | habitat)
## npar AIC
## <none> 592.96
## agriculture 1 593.32
## year_ct 1 619.93
## sin_td_rad 1 636.87drop distance from agriculture
## Single term deletions
##
## Model:
## gma ~ year_ct + sin_td_rad + (1 | habitat)
## npar AIC
## <none> 593.32
## year_ct 1 619.99
## sin_td_rad 1 636.90year and sampling-time-of-year remain. This is the final model:
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( inverse )
## Formula: gma ~ year_ct + sin_td_rad + (1 | habitat)
## Data: P.anal
##
## AIC BIC logLik deviance df.resid
## 593.3 610.7 -291.7 583.3 233
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.5605 -0.6313 -0.2148 0.4428 4.4426
##
## Random effects:
## Groups Name Variance Std.Dev.
## habitat (Intercept) 0.0008809 0.02968
## Residual 0.1239210 0.35202
## Number of obs: 238, groups: habitat, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## (Intercept) 0.355511 0.046982 7.567 3.82e-14 ***
## year_ct 0.029617 0.005462 5.422 5.89e-08 ***
## sin_td_rad 0.187937 0.026769 7.021 2.21e-12 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) yer_ct
## year_ct -0.295
## sin_td_rad -0.021 0.718## $habitat
## (Intercept)
## Basalt -0.04700188
## Limestone 0.03723469
##
## with conditional variances for "habitat"## $habitat
| Observations | 238 |
| Dependent variable | gma |
| Type | Mixed effects generalized linear model |
| Family | Gamma |
| Link | inverse |
| AIC | 593.318 |
| BIC | 610.680 |
| Pseudo-R² (fixed effects) | NA |
| Pseudo-R² (total) | NA |
| Est. | S.E. | t val. | p | |
|---|---|---|---|---|
| (Intercept) | 0.356 | 0.047 | 7.567 | 0.000 |
| year_ct | 0.030 | 0.005 | 5.422 | 0.000 |
| sin_td_rad | 0.188 | 0.027 | 7.021 | 0.000 |
| Group | Parameter | Std. Dev. |
|---|---|---|
| habitat | (Intercept) | 0.030 |
| Residual | 0.352 |
| Group | # groups | ICC |
|---|---|---|
| habitat | 2 | 0.007 |
## Confidence intervals for merMod models is an experimental feature. The
## intervals reflect only the variance of the fixed effects, not the random
## effects.
## Confidence intervals for merMod models is an experimental feature. The
## intervals reflect only the variance of the fixed effects, not the random
## effects.
significant temporal decrease in gma. No effect found for distance from agriculture.
GMA is decreasing on average in 0.2065628 individuals per species per year.
abundance
Explore data
## [1] "ABUNDANCE WITHOUT RARE SPECIES"
| abundance | year_ct | habitat | site | agriculture | td_sc | cos_td_rad | sin_td_rad | |
|---|---|---|---|---|---|---|---|---|
| Min. : 3.00 | Min. :2.0 | Basalt :120 | Batot Taninim :24 | Far :120 | Min. :-2.17579 | Min. :-0.2498 | Min. :-0.9995 | |
| 1st Qu.:15.00 | 1st Qu.:3.5 | Limestone:120 | Gamla :24 | Near:120 | 1st Qu.:-0.76251 | 1st Qu.: 0.3757 | 1st Qu.:-0.9267 | |
| Median :23.00 | Median :5.0 | NA | Hatzor :24 | NA | Median :-0.51689 | Median : 0.4758 | Median :-0.8792 | |
| Mean :26.25 | Mean :5.0 | NA | Karei Deshe :24 | NA | Mean : 0.09807 | Mean : 0.5626 | Mean :-0.6066 | |
| 3rd Qu.:35.00 | 3rd Qu.:6.5 | NA | Natur :24 | NA | 3rd Qu.: 1.11539 | 3rd Qu.: 0.9048 | 3rd Qu.:-0.3233 | |
| Max. :82.00 | Max. :8.0 | NA | Ramat Hashofet:24 | NA | Max. : 2.28504 | Max. : 0.9905 | Max. : 0.1882 | |
| NA | NA | NA | (Other) :96 | NA | NA | NA | NA |
PHI>1, hence choose negative binomial. Fit fixed and mixed models. Choose mixed model if possible, otherwise choose a model with fixed-effects only.
Mixed model converged:
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(5.6356) ( log )
## Formula: abundance ~ agriculture * year_ct + cos_td_rad + sin_td_rad +
## (1 | habitat)
## Data: P.anal
##
## AIC BIC logLik deviance df.resid
## 1850.6 1878.5 -917.3 1834.6 232
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.8794 -0.7330 -0.1364 0.5294 4.6896
##
## Random effects:
## Groups Name Variance Std.Dev.
## habitat (Intercept) 0.005891 0.07675
## Number of obs: 240, groups: habitat, 2
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.98398 0.20800 14.346 < 2e-16 ***
## agricultureNear 0.27315 0.15007 1.820 0.0687 .
## year_ct -0.09937 0.02495 -3.983 6.80e-05 ***
## cos_td_rad 0.07756 0.17278 0.449 0.6535
## sin_td_rad -0.88290 0.14375 -6.142 8.16e-10 ***
## agricultureNear:year_ct 0.00714 0.02729 0.262 0.7936
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) agrclN yer_ct cs_td_ sn_td_
## agricultrNr -0.377
## year_ct -0.477 0.517
## cos_td_rad -0.813 0.008 0.082
## sin_td_rad 0.523 -0.001 0.383 -0.701
## agrcltrNr:_ 0.339 -0.915 -0.562 -0.002 -0.001Perform stepwise model selection of mixed model.
## Single term deletions
##
## Model:
## abundance ~ agriculture * year_ct + cos_td_rad + sin_td_rad +
## (1 | habitat)
## npar AIC
## <none> 1850.6
## cos_td_rad 1 1848.8
## sin_td_rad 1 1885.8
## agriculture:year_ct 1 1848.7drop agriculture : year.
## Single term deletions
##
## Model:
## abundance ~ agriculture + year_ct + cos_td_rad + sin_td_rad +
## (1 | habitat)
## npar AIC
## <none> 1848.7
## agriculture 1 1872.8
## year_ct 1 1867.8
## cos_td_rad 1 1846.9
## sin_td_rad 1 1883.8drop cosine.
## Single term deletions
##
## Model:
## abundance ~ agriculture + year_ct + sin_td_rad + (1 | habitat)
## npar AIC
## <none> 1846.9
## agriculture 1 1870.9
## year_ct 1 1866.8
## sin_td_rad 1 1911.6agriculture, year and sampling time of year remain. The final model:
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Negative Binomial(5.627) ( log )
## Formula: abundance ~ agriculture + year_ct + sin_td_rad + (1 | habitat)
## Data: P.anal
##
## AIC BIC logLik deviance df.resid
## 1846.9 1867.8 -917.4 1834.9 234
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.8915 -0.7491 -0.1417 0.5098 4.5994
##
## Random effects:
## Groups Name Variance Std.Dev.
## habitat (Intercept) 0.005933 0.07703
## Number of obs: 240, groups: habitat, 2
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 3.04140 0.09867 30.823 < 2e-16 ***
## agricultureNear 0.30863 0.06041 5.109 3.24e-07 ***
## year_ct -0.09659 0.02051 -4.708 2.50e-06 ***
## sin_td_rad -0.83754 0.10255 -8.167 3.16e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) agrclN yer_ct
## agricultrNr -0.313
## year_ct -0.563 0.003
## sin_td_rad -0.132 0.008 0.746## $habitat
## (Intercept)
## Basalt 0.06745415
## Limestone -0.06726207
##
## with conditional variances for "habitat"## $habitat
Interpretation of abundance model:
| Observations | 240 |
| Dependent variable | abundance |
| Type | Mixed effects generalized linear model |
| Family | Negative Binomial(5.627) |
| Link | log |
| AIC | 1846.887 |
| BIC | 1867.771 |
| Pseudo-R² (fixed effects) | 0.680 |
| Pseudo-R² (total) | 0.723 |
| exp(Est.) | S.E. | z val. | p | |
|---|---|---|---|---|
| (Intercept) | 20.935 | 0.099 | 30.823 | 0.000 |
| agricultureNear | 1.362 | 0.060 | 5.109 | 0.000 |
| year_ct | 0.908 | 0.021 | -4.708 | 0.000 |
| sin_td_rad | 0.433 | 0.103 | -8.167 | 0.000 |
| Group | Parameter | Std. Dev. |
|---|---|---|
| habitat | (Intercept) | 0.077 |
| Group | # groups | ICC |
|---|---|---|
| habitat | 2 | 0.026 |
## Confidence intervals for merMod models is an experimental feature. The
## intervals reflect only the variance of the fixed effects, not the random
## effects.
## Confidence intervals for merMod models is an experimental feature. The
## intervals reflect only the variance of the fixed effects, not the random
## effects.
## Confidence intervals for merMod models is an experimental feature. The
## intervals reflect only the variance of the fixed effects, not the random
## effects.
Significant higher abundance near agriculture, and overall decrease in total abundance
Near plots have on average 7.7613477 more individuals than far plots, which is 36.1552262 percent higher.
On average, total abundance decreases 2.1025166 individuals per year.
community analysis using package MVabund
## Overlapping points were shifted along the y-axis to make them visible.##
## PIPING TO 2nd MVFACTOR## Only the variables Galerida.cristata, Streptopelia.decaocto, Curruca.melanocephala, Carduelis.carduelis, Pycnonotus.xanthopygos, Emberiza.calandra, Prinia.gracilis, Passer.domesticus, Alectoris.chukar, Chloris.chloris, Corvus.cornix, Parus.major were included in the plot
## (the variables with highest total abundance).
start model specification:
## nb po
## 462.6697 685.9955## [1] "POISSON"
## [1] "NEGATIVE BINOMIAL"
negative binomial model is better than poisson according to residuals and AIC comparison.
## nb po nb1 nb2
## 462.6697 685.9955 430.8902 450.6422The addition of the explanatory variable ‘site’ is improving the AIC of the model the most. stepwise selection of model:
## Single term deletions
##
## Model:
## spp_no_rare ~ agriculture * year_ct + cos_td_rad + sin_td_rad +
## site
## Df AIC
## <none> 13358
## cos_td_rad 31 13373
## sin_td_rad 31 13453
## site 279 14343
## agriculture:year_ct 31 13327drop agriculture:year
## Single term deletions
##
## Model:
## spp_no_rare ~ agriculture + year_ct + cos_td_rad + sin_td_rad +
## site
## Df AIC
## <none> 13327
## agriculture 31 13526
## year_ct 31 13331
## cos_td_rad 31 13347
## sin_td_rad 31 13426
## site 279 14305final model includes agriculture, year, sampling time of year and site.
##
## Test statistics:
## wald value Pr(>wald)
## (Intercept) 10.144 0.001 ***
## agricultureNear 17.030 0.001 ***
## year_ct 8.740 0.001 ***
## cos_td_rad 8.960 0.002 **
## sin_td_rad 12.608 0.001 ***
## siteGamla 10.016 0.001 ***
## siteHatzor 10.675 0.001 ***
## siteKarei Deshe 8.592 0.001 ***
## siteNatur 10.064 0.001 ***
## siteRamat Hashofet 9.798 0.001 ***
## siteSde Eliezer 12.411 0.001 ***
## siteShaal 10.103 0.001 ***
## siteYavneel 10.584 0.001 ***
## siteYiron 11.603 0.001 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Test statistic: 40.34, p-value: 0.001
## Arguments:
## Test statistics calculated assuming response assumed to be uncorrelated
## P-value calculated using 999 resampling iterations via pit.trap resampling (to account for correlation in testing).## Analysis of Deviance Table
##
## Model: spp_no_rare ~ agriculture + year_ct + cos_td_rad + sin_td_rad + site
##
## Multivariate test:
## Res.Df Df.diff Dev Pr(>Dev)
## (Intercept) 239
## agriculture 238 1 150.1 0.01 **
## year_ct 237 1 84.4 0.01 **
## cos_td_rad 236 1 136.0 0.01 **
## sin_td_rad 235 1 165.3 0.01 **
## site 226 9 1535.6 0.01 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Univariate Tests:
## Alectoris.chukar Burhinus.oedicnemus
## Dev Pr(>Dev) Dev Pr(>Dev)
## (Intercept)
## agriculture 0.467 1.00 0.708 1.00
## year_ct 1.128 1.00 0.143 1.00
## cos_td_rad 0.224 0.97 0.699 0.97
## sin_td_rad 3.683 0.71 0.704 1.00
## site 50.77 0.01 22.157 0.10
## Carduelis.carduelis Cecropis.daurica Cettia.cetti
## Dev Pr(>Dev) Dev Pr(>Dev) Dev
## (Intercept)
## agriculture 0.802 1.00 1.137 0.97 2.871
## year_ct 0.644 1.00 2.66 0.91 3.083
## cos_td_rad 0.436 0.97 1.586 0.95 14.062
## sin_td_rad 0.028 1.00 0.947 1.00 2.469
## site 51.057 0.01 16.71 0.13 59.697
## Chloris.chloris Cinnyris.osea
## Pr(>Dev) Dev Pr(>Dev) Dev Pr(>Dev)
## (Intercept)
## agriculture 0.84 35.448 0.01 4.29 0.56
## year_ct 0.77 0.182 1.00 0.29 1.00
## cos_td_rad 0.02 1.331 0.95 6.775 0.29
## sin_td_rad 0.86 3.212 0.84 6.23 0.25
## site 0.01 47.492 0.01 74.209 0.01
## Cisticola.juncidis Clamator.glandarius
## Dev Pr(>Dev) Dev Pr(>Dev)
## (Intercept)
## agriculture 0.009 1.00 0.011 1.00
## year_ct 6.504 0.16 0.854 1.00
## cos_td_rad 1.613 0.95 1.387 0.95
## sin_td_rad 0.229 1.00 0.94 1.00
## site 32.302 0.01 29.807 0.02
## Corvus.cornix Curruca.communis Curruca.curruca
## Dev Pr(>Dev) Dev Pr(>Dev) Dev
## (Intercept)
## agriculture 3.995 0.62 0.005 1.00 0.539
## year_ct 0.21 1.00 6.235 0.18 8.487
## cos_td_rad 5.43 0.42 18.584 0.01 10.165
## sin_td_rad 0.673 1.00 22.056 0.01 20.143
## site 36.927 0.01 39.874 0.01 23.408
## Curruca.melanocephala Dendrocopos.syriacus
## Pr(>Dev) Dev Pr(>Dev) Dev
## (Intercept)
## agriculture 1.00 2.748 0.87 0.311
## year_ct 0.07 0.082 1.00 0.099
## cos_td_rad 0.05 19.653 0.01 0.732
## sin_td_rad 0.01 8.936 0.11 0.001
## site 0.09 162.311 0.01 53.462
## Emberiza.calandra Falco.naumanni
## Pr(>Dev) Dev Pr(>Dev) Dev Pr(>Dev)
## (Intercept)
## agriculture 1.00 1.887 0.92 0.306 1.00
## year_ct 1.00 7.956 0.07 2.252 0.93
## cos_td_rad 0.97 3.526 0.70 0.726 0.97
## sin_td_rad 1.00 8.345 0.11 0.184 1.00
## site 0.01 61.528 0.01 57.733 0.01
## Falco.tinnunculus Galerida.cristata
## Dev Pr(>Dev) Dev Pr(>Dev)
## (Intercept)
## agriculture 0.451 1.00 16.671 0.01
## year_ct 1.158 1.00 2.965 0.81
## cos_td_rad 0.087 0.97 0.171 0.97
## sin_td_rad 5.597 0.39 16.318 0.01
## site 32.451 0.01 98.456 0.01
## Garrulus.glandarius Hirundo.rustica
## Dev Pr(>Dev) Dev Pr(>Dev)
## (Intercept)
## agriculture 5.943 0.27 0.768 1.00
## year_ct 1.034 1.00 0.423 1.00
## cos_td_rad 3.434 0.70 7.003 0.25
## sin_td_rad 1.344 0.98 0.234 1.00
## site 60.46 0.01 18.257 0.13
## Lanius.senator Linaria.cannabina Parus.major
## Dev Pr(>Dev) Dev Pr(>Dev) Dev
## (Intercept)
## agriculture 0.147 1.00 1.433 0.95 5.211
## year_ct 16.679 0.01 0.021 1.00 6.291
## cos_td_rad 6.29 0.34 3.979 0.63 4.484
## sin_td_rad 0.745 1.00 2.361 0.86 9.573
## site 23.892 0.06 27.576 0.02 68.967
## Passer.domesticus Prinia.gracilis
## Pr(>Dev) Dev Pr(>Dev) Dev Pr(>Dev)
## (Intercept)
## agriculture 0.41 3.757 0.65 7.265 0.15
## year_ct 0.18 5.265 0.33 0.548 1.00
## cos_td_rad 0.59 0.814 0.96 0.797 0.97
## sin_td_rad 0.11 8.425 0.11 4.38 0.58
## site 0.01 66.086 0.01 68.361 0.01
## Psittacula.krameri Pycnonotus.xanthopygos
## Dev Pr(>Dev) Dev Pr(>Dev)
## (Intercept)
## agriculture 2.416 0.90 30.215 0.01
## year_ct 1.202 1.00 1.157 1.00
## cos_td_rad 1.721 0.94 0.021 0.97
## sin_td_rad 2.75 0.85 7.306 0.17
## site 40.069 0.01 56.799 0.01
## Streptopelia.decaocto Streptopelia.turtur
## Dev Pr(>Dev) Dev Pr(>Dev)
## (Intercept)
## agriculture 10.294 0.04 1.876 0.92
## year_ct 0.697 1.00 0.203 1.00
## cos_td_rad 1.35 0.95 7.086 0.25
## sin_td_rad 9.246 0.11 9.263 0.11
## site 42.477 0.01 26.85 0.02
## Turdus.merula Vanellus.spinosus
## Dev Pr(>Dev) Dev Pr(>Dev)
## (Intercept)
## agriculture 7.219 0.15 0.932 1.00
## year_ct 5.73 0.26 0.21 1.00
## cos_td_rad 9.661 0.06 2.184 0.88
## sin_td_rad 6.885 0.20 2.078 0.87
## site 53.565 0.01 31.887 0.01
## Arguments:
## Test statistics calculated assuming uncorrelated response (for faster computation)
## P-value calculated using 99 iterations via PIT-trap resampling.All factors (agriculture, year, time of year and site) have a statistically significant effect on community composition.
## Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
## ℹ Please use `linewidth` instead.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.## [[1]]
##
## [[2]]
##
## [[3]]
## [[1]]
##
## [[2]]
##
## [[3]]
Significant effect of plantation agriculture proximity: synanthrope / invasive species near agriculture. Most species showing negative coefs, consistent with overall decrease in gma and total abundance.
Significant negative temporal effect on corn bunting and lesser whitethroat (but see species-specific plots below). Positive significant trend for woodchat shrike, but the magnitude is very small.
Worth noting that the following have consistent results across several units:
crested lark, negative, in Sfar and Batha. lesser whitethroat, negative, in Batha and Maquis. house sparrow, negative, in Sfar, Batha and Maquis. red-rumped swallow, positive, in sfar, batha and maquis. palestine sunbird, negative, in sfar, batha and maquis. goldfinch, negative, in maquis and batha.
Below plots for crested lark, lesser whitethroat, house sparrow and corn bunting
## [1] "Galerida cristata"## [1] "manyglm coefficients"## (Intercept) agricultureNear year_ct cos_td_rad
## 0.1520779 -0.7592007 -0.0731277 0.7692822
## sin_td_rad siteGamla siteHatzor siteKarei Deshe
## -1.8545331 0.9115167 -1.0991985 0.3222154
## siteNatur siteRamat Hashofet siteSde Eliezer siteShaal
## 1.1731718 -0.6210582 -1.0092375 0.7693465
## siteYavneel siteYiron
## -0.9695270 -1.4923494## [1] "species-specific glm"## (Intercept) agricultureNear year_ct cos_td_rad
## 0.15208103 -0.75920009 -0.07312734 0.76927779
## sin_td_rad siteGamla siteHatzor siteKarei Deshe
## -1.85452759 0.91151810 -1.09919768 0.32221281
## siteNatur siteRamat Hashofet siteSde Eliezer siteShaal
## 1.17317199 -0.62105611 -1.00923659 0.76934814
## siteYavneel siteYiron
## -0.96952666 -1.49234746
## [1] "Curruca curruca"## [1] "manyglm coefficients"## (Intercept) agricultureNear year_ct cos_td_rad
## -20.49734676 0.42482008 -0.02686165 -0.63045292
## sin_td_rad siteGamla siteHatzor siteKarei Deshe
## -19.82445963 -7.88455397 0.48323713 -10.31007440
## siteNatur siteRamat Hashofet siteSde Eliezer siteShaal
## 1.33424086 -0.42614332 2.19018775 2.26385067
## siteYavneel siteYiron
## 0.48153107 2.33172040## [1] "species-specific glm"## (Intercept) agricultureNear year_ct cos_td_rad
## -20.61839642 0.42474006 -0.02702414 -0.59991852
## sin_td_rad siteHatzor siteNatur siteRamat Hashofet
## -19.94699265 0.48019463 1.33292357 -0.42619501
## siteSde Eliezer siteShaal siteYavneel siteYiron
## 2.18798032 2.26154006 0.47795899 2.32960699
## [1] "Passer domesticus"## [1] "manyglm coefficients"## (Intercept) agricultureNear year_ct cos_td_rad
## -19.96014323 1.49241185 -0.49805791 7.54082785
## sin_td_rad siteGamla siteHatzor siteKarei Deshe
## -5.04075189 15.35464525 12.31878831 13.37700418
## siteNatur siteRamat Hashofet siteSde Eliezer siteShaal
## 14.42194958 -0.07832843 -0.76381509 12.72438970
## siteYavneel siteYiron
## 15.22977429 -0.55955882## [1] "species-specific glm"## (Intercept) agricultureNear year_ct cos_td_rad sin_td_rad
## -4.6055325 1.4924224 -0.4980637 7.5408654 -5.0407946
## siteHatzor siteKarei Deshe siteNatur siteShaal siteYavneel
## -3.0358365 -1.9776337 -0.9326851 -2.6302507 -0.1248692
## [1] "Emberiza calandra"## [1] "manyglm coefficients"## (Intercept) agricultureNear year_ct cos_td_rad
## -0.64160316 -0.79613036 -0.05601435 -1.19122249
## sin_td_rad siteGamla siteHatzor siteKarei Deshe
## -1.42714754 1.96787355 -13.54797201 0.28693655
## siteNatur siteRamat Hashofet siteSde Eliezer siteShaal
## 1.51812736 1.57155269 -1.22080035 2.18924457
## siteYavneel siteYiron
## 1.53390365 -1.53045976## [1] "species-specific glm"## (Intercept) agricultureNear year_ct cos_td_rad
## -0.64165224 -0.79611605 -0.05601625 -1.19120333
## sin_td_rad siteGamla siteKarei Deshe siteNatur
## -1.42719124 1.96789227 0.28694631 1.51814160
## siteRamat Hashofet siteSde Eliezer siteShaal siteYavneel
## 1.57155713 -1.22078848 2.18926918 1.53392503
## siteYiron
## -1.53044741
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge
## Warning: glm.fit: algorithm did not converge## Warning in glm.nb(formula = Linaria.cannabina ~ agriculture + year_ct + :
## alternation limit reached## [1] "Linaria cannabina"## [1] "manyglm coefficients"## (Intercept) agricultureNear year_ct cos_td_rad
## -9.48950163 1.81980017 -0.45753970 -5.45679431
## sin_td_rad siteGamla siteHatzor siteKarei Deshe
## 2.51918366 14.67288170 0.84240156 14.23392030
## siteNatur siteRamat Hashofet siteSde Eliezer siteShaal
## 13.87788395 0.09500468 12.94479100 14.84973861
## siteYavneel siteYiron
## 0.34784324 14.02102499## [1] "species-specific glm"## (Intercept) agricultureNear year_ct cos_td_rad sin_td_rad
## 4.4833623 2.2262363 -0.5857269 -4.6620185 1.4637443
## siteKarei Deshe siteNatur siteSde Eliezer siteShaal siteYiron
## -0.0926939 -0.8071995 -1.8725924 0.3897458 -0.4102873
Lesser whitethroat does not seem like true trend in species specific plots.
