-
Notifications
You must be signed in to change notification settings - Fork 0
/
5_last_Occ_jackal.R
465 lines (317 loc) · 14.5 KB
/
5_last_Occ_jackal.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
##
#=========== OCCUPANCY ANALYSIS OF PAPER:
# The influence of the landscape context around reserves on black backed jackal occupancy across South Africa
# Lain E. Pardo, Lourens Swanepoel, Goncalo Curveira-Santos, and Jan A. Venter
# NOTES
# Data preparation is not shown in this repository
library(readr)
library(unmarked)
library(tidyverse)
library(AICcmodavg)
---------------------------
#### 2. occupancy analysis ------------------------------------------------
----------------------------
# data
site_covs <- read.csv("data_in/siteCovs_No_scaled_fin.csv")
jackal_DH <- read_csv("data_in/+jackal_DH_without_sites_no_temp_rep_no_outli.csv")
### scale covariates ------------------------------------------------------
names(site_covs)
site_covs$Reserve.Area <-scale(site_covs$Reserve.Area)
site_covs$Tree.Cover.Cam <-scale(site_covs$Tree.Cover.Cam)
site_covs$Human.Density.Buffer <-scale(site_covs$Human.Density.Buffer)
site_covs$Agriculture.Buffer <-scale(site_covs$Agriculture.Buffer)
site_covs$Build.up.Buffer <- scale(site_covs$Build.up.Buffer)
site_covs$Area.PA.Buffer <-scale(site_covs$Area.PA.Buffer)
site_covs$Dist.Border.Res <-scale(site_covs$Dist.Border.Res)
site_covs$Cam.Days <-scale(site_covs$Cam.Days)
#before convert establish year to age
site_covs$Stablishment.Year <-2021- (site_covs$Stablishment.Year)
site_covs$Stablishment.Year <-scale(site_covs$Stablishment.Year)
site_covs$Season <-as.factor(site_covs$Season)
site_covs$Code.Loc <-as.factor(site_covs$Code.Loc)
str(site_covs)
siteCovs <- site_covs[,-1]
str(siteCovs)
# naive quick
naive <- apply(X = jackal_DH,
MARGIN = 1,
FUN = "max", na.rm = TRUE)
mean(naive) #[1] 0.32
# Create unmarkedFrame
y <- jackal_DH
data.umf = unmarkedFrameOccu(y = y,
siteCovs = siteCovs)
summary(data.umf) #
unique(siteCovs$Code.Loc)
unique(siteCovs$RSA.Vegetation)
unique(siteCovs$RSA.Biome)
# Data summaries
yy <- rowSums(y, na.rm=TRUE) # Number of times spp were detected per site (row):sum of 1?s basically
yy
nn <- rowSums(!is.na(y)) # basically the total occassions per site,
nn
cbind(yy, nn) #just to check y and n, per site
plot(data.umf)
-----------------------------------------------
# 3. model fitting --------------------------------------------------------
-----------------------------------------------
#Modeling selection approach = selecting for psi with general model for p (Mackenzie)...
# 1. constant MODEL
M0 <- occu(~1 ~1, data.umf)
backTransform(M0, type="state") # 0.381 SE 0.0335
backTransform(M0, type="det") # 0.27 SE 0.019
# null
psi_nul <-occu(~Flash + Cam.Days + Tree.Cover.Cam
~ 1, data.umf)
## 2. reserve level (macro scale): reserve features - same covariate value to all cameras within a reserve ---------------
# season is acting as a random factor
# global out: disturbance
class(siteCovs$Season)
class(siteCovs$Code.Loc)
names(siteCovs)
global_out <-occu(~Flash + Cam.Days + Tree.Cover.Cam
~ Human.Density.Buffer + Agriculture.Buffer + Build.up.Buffer +
(1 | Season), data.umf)
global_out
#global inside: if variables inside are more important than outside
global_in_res <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~ Reserve.Area + Stablishment.Year + Tree.Cover.Res + (1 | Season), data.umf)
global_in_res
# inside: reserve level individual
psi_area_res <-occu(~Flash + Cam.Days + Tree.Cover.Cam
~ Reserve.Area + (1 | Season), data.umf)
psi_area_res
#
psi_year <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~ Stablishment.Year + (1 | Season), data.umf)
psi_year
#
psi_tree.res <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~ Tree.Cover.Res + (1 | Season), data.umf)
psi_tree.res
# Buffer outside: reserve level individual
psi_human <-occu(~ Flash + Cam.Days +Tree.Cover.Cam
~ Human.Density.Buffer + (1 | Season), data.umf)
psi_human
#
psi_agri <-occu(~ Flash + Cam.Days +Tree.Cover.Cam
~ Agriculture.Buffer + (1 | Season), data.umf)
psi_agri
#
psi_build <-occu(~ Flash + Cam.Days +Tree.Cover.Cam
~ Build.up.Buffer + (1 | Season), data.umf)
psi_build
#
psi_area_PA <-occu(~ Flash + Cam.Days +Tree.Cover.Cam
~ Area.PA.Buffer + (1 | Season), data.umf)
psi_area_PA
# 3. site level: features at camera scale - different value per site
global_in_site <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~ Tree.Cover.Cam + Dist.Border.Res + (1 | Season), data.umf)
global_in_site
# individual factors
psi_tree_site <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~ Tree.Cover.Cam + (1 | Season), data.umf)
psi_tree_site
#
psi_border <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~ Dist.Border.Res + (1 | Season), data.umf)
psi_border
# 4. carnivores models at site level ---------------------------------------
#
psi_lion_site <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~ Lion.Site + (1 | Season), data.umf)
psi_lion_site
#
psi_leopard_site <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~ Leopard.Site + (1 | Season), data.umf)
psi_leopard_site
#
psi_hyena_site <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~ Hyena.Site + (1 | Season), data.umf)
psi_hyena_site
# psi_all_pred_site
psi_all_pred_site <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~Lion.Site + Leopard.Site + Hyena.Site + (1 | Season), data.umf)
psi_all_pred_site
# pred richness site
psi_richness_site <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~ Pred.Rich.Site + (1 | Season), data.umf)
summary(psi_richness_site)
# 5. carnivore models at reserve level ------------------------------------
psi_lion_reserve <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~ Lion.Reserve + (1 | Season), data.umf)
psi_lion_reserve
#
psi_leopard_reserve <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~ Leopard.Reserve + (1 | Season), data.umf)
psi_leopard_reserve
#
psi_hyena_reserve <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~ Hyena.Reserve + (1 | Season), data.umf)
psi_hyena_reserve
# psi_all_pred_reserve
psi_all_pred_res <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~Lion.Reserve + Leopard.Reserve + Hyena.Reserve + (1 | Season), data.umf)
psi_all_pred_res
# richness reserve
psi_richness_reserve <-occu(~ Flash + Cam.Days + Tree.Cover.Cam
~ Pred.Rich.Reserve + (1 | Season), data.umf)
summary(psi_richness_reserve)
# model selection all models -----------------------
#reserve features: models 1-9
# site level features: models 10-12
# community site 13-16
# community reserve 17-20
Cand.mods1 <- list("p(Flash + Cam.Days + tree)psi(global.out)" = global_out,
"p(Flash + Cam.Days + tree)psi(global.in.res)" = global_in_res,
"p(Flash + Cam.Days + tree)psi(human.out)" = psi_human,
"p(Flash + Cam.Days + tree)psi(agri.out)" = psi_agri,
"p(Flash + Cam.Days + tree)psi(buildup.out)" = psi_build,
"p(Flash + Cam.Days + tree)psi(area.res.out)" = psi_area_PA,
"p(Flash + Cam.Days + tree)psi(area.res)" = psi_area_res,
"p(Flash + Cam.Days + tree)psi(year.res)" = psi_year,
"p(Flash + Cam.Days + tree)psi(tree.cov.res)" = psi_tree.res,
"p(Flash + Cam.Days + tree)psi(global.in.site)" = global_in_site,
"p(Flash + Cam.Days + tree)psi(tree.cov.site)" = psi_tree_site,
"p(Flash + Cam.Days + tree)psi(dist.border.site)" = psi_border,
"p(Flash + Cam.Days + tree)psi(lion.site)" = psi_lion_site,
"p(Flash + Cam.Days + tree)psi(leopard.site)" = psi_leopard_site,
"p(Flash + Cam.Days + tree)psi(hyena.site)" = psi_hyena_site,
"p(Flash + Cam.Days + tree)psi(pred.rich.site)" = psi_richness_site,
"p(Flash + Cam.Days + tree)psi(lion.res)" = psi_lion_reserve,
"p(Flash + Cam.Days + tree)psi(leopard.res)" = psi_leopard_reserve,
"p(Flash + Cam.Days + tree)psi(hyena.res)" = psi_hyena_reserve,
"p(Flash + Cam.Days + tree)psi(pred.rich.res)" = psi_richness_reserve,
"p(Flash + Cam.Days + tree)psi(.)" = psi_nul,
"p(.)psi(.)" = M0)
(model_sel_all <- aictab(Cand.mods1))
(model_sel_all <- aictab(Cand.mods1,rank="QAICc", chat=1.34))
(model_sel_table <- as.data.frame(model_sel_all))
#write.csv(model_sel_all, "data_out/final_results/model_sel_psi_last_QAICc.csv")
# MODEL SELECTION FOR P BASED ON THESE RESULTS only two models below 2AIC
flash_p <-occu(~Flash
~ Tree.Cover.Cam + (1 | Season) , data.umf)
camdays_p <- occu(~Cam.Days
~ Tree.Cover.Cam + (1 | Season) , data.umf)
tree_p <- occu(~Tree.Cover.Cam
~ Tree.Cover.Cam + (1 | Season) , data.umf)
null_p <- occu(~1
~ Tree.Cover.Cam + (1 | Season) , data.umf)
#for second mod
flash_p2 <- occu(~ Flash
~ Tree.Cover.Cam + Dist.Border.Res + (1 | Season), data.umf)
camdays_p2 <- occu(~ Cam.Days
~ Tree.Cover.Cam + Dist.Border.Res + (1 | Season), data.umf)
tree_p2 <- occu(~ Tree.Cover.Cam
~ Tree.Cover.Cam + Dist.Border.Res + (1 | Season), data.umf)
null_p2 <- occu(~ 1
~ Tree.Cover.Cam + Dist.Border.Res + (1 | Season), data.umf)
Cand.mods_p <- list("p(Flash)psi(tree.cover.site)" = flash_p,
"p(Cam.Days)psi(tree.cover.site)" =camdays_p,
"p(tree)psi(tree.cover.site)" = tree_p,
"p(Flash)psi(global.in.site)" = flash_p2,
"p(Cam.Days)psi(global.in.site)" =camdays_p2,
"p(tree)psi(global.in.site)" = tree_p2,
"p(.)psi(tree.cover.site)" = null_p,
"p(.)psi(global.in.site)" = null_p2,
"p(.)psi(.)" = M0)
(model_sel_all2 <- aictab(Cand.mods_p))
(model_sel_all2 <- aictab(Cand.mods_p,rank="QAICc", chat=1.34 ))
(model_sel_table <- as.data.frame(model_sel_all))
#write.csv(model_sel_all2, "data_out/model_sel_with_p.csv")
#godness of fit ---------------
coef(global_out)
global_out@TMB$par
global_out@TMB$starts_order
#gofdet1 <- mb.gof.test(global_out, nsim = 10000, plot.hist=FALSE) #
#gofdet1
# Chi-square statistic = 3905.671
# Number of bootstrap samples = 10000
# P-value = 0.1238
#
# Quantiles of bootstrapped statistics:
# 0% 25% 50% 75% 100%
# 558 1177 1534 2304 499110
#
# Estimate of c-hat = 1.34
------------------------------------
# 4. PREDICTIONS -------------------------------------------------------------
-----------------------------------
# BEST MODEL NOW:
flash_p <- occu(~ Flash
~ Tree.Cover.Cam + (1 | Season), data.umf)
range(siteCovs$Tree.Cover.Cam)
newData <- data.frame(Tree.Cover.Cam = seq(-0.7123054, 4.6376593, length = 8),
Season = c("autunm","autunm-winter", "spring", "spring-summer", "summer", "summer-autunm", "winter",
"winter-spring"))
pred <- predict(flash_p, type = "state", newdata = newData, appendData = TRUE)
pred
# PLOT IT
f1 <- ggplot(pred,aes(Tree.Cover.Cam, Predicted, ymin=lower,ymax=upper))
f1 + geom_line(colour = "blue") +
geom_ribbon(alpha=0.2,colour=NA) +
labs(x = "Tree cover (std)", y = "Prob. occupancy") +
theme_bw() +
theme(axis.text=element_text(size=15),axis.line=element_line(colour="black"))+
ylim(0,1)
#ggsave("figures/pred_plot_with_season.jpg")
# alternative figure without random factor season
flash_NS <-occu(~Flash
~ Tree.Cover.Cam, data.umf)
newData1 <- data.frame(Tree.Cover.Cam = seq(-0.7123054, 4.6376593, by = 0.1))
pred2 <- predict(flash_NS, type = "state", newdata = newData1, appendData = TRUE)
pred2
# PLOT IT
f2 <- ggplot(pred2,aes(Tree.Cover.Cam, Predicted, ymin=lower,ymax=upper))
f2 + geom_line(colour = "blue") +
geom_ribbon(alpha=0.2,colour=NA) +
labs(x = "Tree cover (std)", y = "Prob. occupancy") +
theme_bw() +
theme(axis.text=element_text(size=15),axis.line=element_line(colour="black"))+
ylim(0,1)
#ggsave("figures/pred_plot_NO_season_std.jpg")
-------------------------------
# average prediction per reserve ---------------------------------------------------
-------------------------------
#predict again but put all columns with appendData
#using season as random, model flash_p
class(siteCovs)
siteCovs <- as.data.frame(siteCovs)
predmap <- predict(flash_p, newdata = siteCovs, type = "state", appendData=T)
pred_all <- as.data.frame(predmap)
library("plotrix") #to calc se
#add col with code
avg_pred <- pred_all %>%
group_by(Code.Loc) %>%
summarise(psi_mean = mean(Predicted),
psi_sd = sd(Predicted),
psi_se = std.error(Predicted),
min = psi_mean - psi_sd,
max = psi_mean + psi_sd)
avg_pred
# write.csv(avg_pred,"data_out/avg_occ_persite.csv")
# ordered catterpillar plot
str(avg_pred)
occgraph <- avg_pred
occgraph$Code.Loc <- factor(occgraph$Code.Loc, levels = occgraph$Code.Loc, ordered = TRUE)
str(occgraph)
rr4 <- transform(occgraph,Reserve = reorder(Code.Loc,psi_mean))
theme_set(theme_bw())
ggplot(rr4,aes(Reserve,psi_mean,ymin=min,ymax=max))+
geom_pointrange(data = rr4, aes(ymin =psi_mean - psi_se, ymax = psi_mean + psi_se),
alpha = 2, width = 4, color = "blue") +
#c +
# xlim=c(-1.5,1.5) +
xlab("Reserves") +
ylab("Average estimated Occupancy probability") +
theme(axis.text=element_text(size=11, color = "black")) +
coord_flip() +
theme(legend.position='none') +
geom_hline(yintercept= 0.5, linetype="dashed",
color = "grey", size=0.7)
#ggsave("figures/catterpillar_avg_psi_res1.jpg")
# END ---------------------------------------------------------------------
# knitr::opts_knit$set(root.dir = rprojroot::find_rstudio_root_file())
# rmarkdown::render("5_last_Occ_jackal.R") not working...Error in -`*tmp*` : invalid argument to unary operator
# (Ctrl+Shift+K)