Ple.27.3a.21-32 Benchmark Data Investigation
Elliot J. Brown
Casper W. Berg
Sven Stötera
16/09/2024
1 Introduction
1.1 Contents
This script generates a working document describing the data available for an assessment of ple.27.3a.21-32; Plaice in the Kattegat, The Belt Seas and the Baltic Sea. The report generated covers the following topics:
- Assessment Input Data
- Catches and Landings
- Survey Indices
- Biological Data
- Reference Points
- Background Data and ad-hoc Requests are dealt with in the relevant sections.
Most of the code used to process data and create visualisations are hidden in this HTML. To view each chunk, simply click on the button to the right of the relevant section. To download the entire R-markdown (.rmd) file including all text and code chunks, use the code button at the top of the page, with the dropdown arrow.
1.2 Preparation
To run this script you require the following packages and functions:
require(mgcv)
require(data.table)
require(reshape2)
require(knitr)
require(scales)
require(bookdown)
require(ggplot2)
require(ggthemes)
require(stringr)
require(lattice)
require(plyr)
require(plotly)
require(icesAdvice)
require("stockassessment") # available from https://github.com/fishfollower/SAM
require(parallel)
require(sf)
myprop <- function(df, nomo = "nomo", NoAtLngt = "CANoAtLngt") {
t <- sum(df$nomo)/sum(df$CANoAtLngt)
}
catchFrac <- function(x, nm, w, frac) {
F <- getF(x)
Z <- F + nm
N <- getN(x)
C <- F/Z * (1 - exp(-Z)) * N
return(sum(frac * w * C))
}
save_as_lowestoft <- function(df, file_path, title, rando_numbers = paste(c("1",
"6"), collapse = "\t"), year_range = paste(c(2002, DataYear + 1), collapse = "\t"),
age_range = paste(c(1, 10), collapse = "\t"), timeseries_type = paste(c(1, "# 1 indicates annually varying, 2 indicates fixed values"),
collapse = "\t")) {
# Open a connection to the file
con <- file(file_path, open = "wt")
# Write the header (customize as needed)
writeLines(paste0(title), con)
writeLines(paste0(rando_numbers), con)
writeLines(paste0(year_range), con)
writeLines(paste0(age_range), con)
writeLines(paste0(timeseries_type), con)
# Write the dataframe to the file
write.table(df, con, sep = "\t", row.names = FALSE, col.names = FALSE, quote = FALSE)
# Close the connection
close(con)
}
## Create sequences for moving window average:
an = function(n, len) c(seq.int(n), rep(n, len - n))We also need to set a few variables manually.
The first is the DataYear, which is the year immediately preceeding the assessment year or the year for which the most recent data is available.
The second is the catch we (ICES) advised as TAC for the last assessment LastIcesAdvice, in tonnes. NOTE this value is taken from the advice sheet, to compare changes in advice, not changes in advice relative to actual Allowable Catches.
The third is whether this run is exploratory, or to be used to generate the final advice. This option will determine the number of iterations used to determine the short-term forecasts.
DataYear <- 2023
LastIcesAdvice <- 21735 # Advised catch for areas 21-32 (from stock splitting table in advice)
isFinal <- FALSE
# isFinal <- TRUE2 Assessment Input Data
2.1 Landings and Catches
Catch data, including break downs by landings/discards, are taken directly from inter-catch for each of the historic stocks (namely ple.27.21-23 and ple.27.24-32), where discards and sampled measures are raised according to the prioritisation described in the corresponding stock annexes, before the resultant catch data are combined. The imported table is the cumulative extracts from InterCatch, specifically Table 1 from the AllCatches extraction and file CatchAndSampleDataTables.txt.
Furthermore, recreational catches from Germany have been introduced late in the process. Here we describe the data that are available for incorporation into an assessment model, but also describe how these data will be further integrated to the whole assessment procedure in the near future.
## Read in the combined Intercatch output from the two old stocks.
ic_raw <- read.csv(file = "Data/CatchData/Ple.27.21-32_IntercatchTable1_2002-2023.csv",
header = T)
## Read in historic TAC values by area
tac_raw <- read.csv(file = "Data/ple.27.21_22-32_TAC-Histories_2000-DataYear.csv",
header = T)These data are cleaned according to data formats and naming conventions (hidden chunk).
ic_raw$Year <- as.factor(as.character(ic_raw$Year))
# ic_raw$CatchCategory <- as.factor(gsub(pattern = 'BMS landing', replacement =
# 'BMS', x = ic_raw$CatchCategory))
ic_raw$Season <- as.factor(as.character(ic_raw$Season))
names(ic_raw)[names(ic_raw) == "CATON..Tons."] <- "CATON"
ic_raw$Fleet <- gsub(pattern = " ", replacement = "", x = ic_raw$Fleet, fixed = TRUE)
### Allocate some strange fleet types according to logic, or 'All' to 'Active'
### as the most likely.
ic_raw[ic_raw$Fleet %in% c("OTB_CRU_90-119_0_0_all", "MIS_MIS_0_0_0_HC", "Acitve ",
"Trawl", "All", "Fleet-All"), "Fleet"] <- "Active"
ic_raw[ic_raw$Fleet %in% c("GNS_DEF_all_0_0_all", "Passive ", "Gillnet", "Trapnet",
"Longline"), "Fleet"] <- "Passive"
ic_clean <- droplevels(ic_raw)
## Create novel fleet by management area
ic_clean$Fleet2 <- ifelse(ic_clean$Area == "27.3.a.21" & ic_clean$Fleet == "Active",
"NS_Active", ifelse(ic_clean$Area == "27.3.a.21" & ic_clean$Fleet == "Passive",
"NS_Passive", ifelse(ic_clean$Area != "27.3.a.21" & ic_clean$Fleet == "Active",
"BS_Active", ifelse(ic_clean$Area != "27.3.a.21" & ic_clean$Fleet ==
"Passive", "BS_Passive", NA))))
### Clean TAC data
tac_raw$Year <- as.factor(as.character(tac_raw$Year))
k <- c("k", "ic_clean", "myprop", "DataYear", "ebpal", "isFinal", "LastIcesAdvice",
"save_as_lowestoft", "an", "tac_raw")
rm(ic_raw)Historic landings and TAC values are imported from legacy records (hidden chunk). These need to be collated for the new “super stock”.
HistLand_raw <- read.csv(file = "Data/ple.27.21-23_HistoricLandings_1970-2017.csv",
header = T)
HistLand_raw$Year <- as.factor(as.character(HistLand_raw$Year))
# === Update HistLand ====
landings <- aggregate(CATON ~ Area + Country + Year, data = ic_clean[as.numeric(as.character(ic_clean$Year)) >=
2018 & ic_clean$CatchCategory != "Discards", ], FUN = sum)
colnames(landings)[colnames(landings) == "CATON"] <- "Landings"
# landings2018 <- aggregate(CATON~Area+Country+Year, data =
# ic_clean[ic_clean$Year == '2018' & ic_clean$CatchCategory != 'Discards', ],
# FUN = sum) landings2018$Year <- as.factor(as.character(2018))
# names(landings2018)[names(landings2018) == 'CATON'] <- 'Landings' #
# landings2018$Landings <- landings2018$Landings/1000 landings2019 <-
# aggregate(CATON~Area+Country, data = ic_clean[ic_clean$Year == '2019' &
# ic_clean$CatchCategory != 'Discards', ], FUN = sum) landings2019$Year <-
# as.factor(as.character(2019)) names(landings2019)[names(landings2019) ==
# 'CATON'] <- 'Landings' # landings2019$Landings <- landings2019$Landings/1000
# landings2020 <- aggregate(CATON~Area+Country, data = ic_clean[ic_clean$Year
# == '2020' & ic_clean$CatchCategory != 'Discards', ], FUN = sum)
# landings2020$Year <- as.factor(as.character(2020))
# names(landings2020)[names(landings2020) == 'CATON'] <- 'Landings'
# landings2021 <- aggregate(CATON~Area+Country, data = ic_clean[ic_clean$Year
# == '2021' & ic_clean$CatchCategory != 'Discards', ], FUN = sum)
# landings2021$Year <- as.factor(as.character(2021))
# names(landings2021)[names(landings2021) == 'CATON'] <- 'Landings'
# HistLand <- rbind(HistLand_raw, landings2018, landings2019, landings2020,
# landings2021)
HistLand <- rbind(HistLand_raw, landings)
getwd()
rm(list = c("HistLand_raw", "landings"))
# =====
k <- append(k, c("HistLand"))Catch cohort matrices are set up externally in excel, according to the annually updated intercatch output (hidden chunk). This section needs to be updated so that the cohort matrices used to structure the figures are generated automagically with R, to remove the excel step
catchcohorts <- read.csv(file = "Data/ple.27.21-32_CohortMatrix_1998-WorkingYear.csv",
header = T)
catchcohorts$Year <- as.factor(as.character(catchcohorts$Year))
# #=== # Attempts to create cohort matrix in R #==== for(i in
# 1:(length(1992:DataYear)+1)){ assign(x = paste0('cohort_',i), value =
# c(rep(NA, i-1), c(1:10),
# c(rep(NA,((length(1992:DataYear)+1)-min(10+(i-1),23)))) ) ) } catchcohorts <-
# do.call(cbind, mget(ls(pattern = 'cohort_'))) catchcohorts$Year <-
# 1992:(DataYear+1) #====
catchNum <- read.csv(file = "Data/ple.27.21-32_CatchNumAge_2002_DataYear.csv", header = T)
catchNum$Year <- as.factor(as.character(catchNum$Year))
k <- append(k, c("catchcohorts", "catchNum"))2.1.1 Landings - Results
In more recent history, observers and other sampling programmes have been able to provide estimates of discarded portions of catch. Below we consider only the landings for the period when catch was divided into landings & discards.
# === Figure 3 Short term Landings by country ====
IC_agg_temp <- aggregate(CATON ~ Year + Country, data = ic_clean[ic_clean$CatchCategory ==
"Landings", ], FUN = sum)
IC_agg_temp$Country <- factor(IC_agg_temp$Country, levels = c("Sweden", "Germany",
"Denmark", "Poland", "Latvia", "Lithuania", "Estonia", "Finland"))
# IC_agg_temp$CATON <- IC_agg_temp$CATON.kg/1000
plt_land_ctry <- ggplot(data = IC_agg_temp, aes(x = Year, y = CATON, group = Country,
fill = Country)) + stat_summary(fun.y = sum, geom = "area", position = "stack",
na.rm = TRUE) + ylab("Landings (t)") + theme_clean() + theme(axis.text.x = element_text(angle = 90,
vjust = 0.5, hjust = 1)) + scale_fill_manual(values = c("#004B87", "#000000",
"#C60C30", ebpal))
ggplotly(plt_land_ctry)Figure 2.1: Recent history landings by country
# ====We can also interrogate the landings relative to the TACs for the respective management areas that this stock straddles. Namely the North Sea management area in SD21 (Kattegat), and the Baltic Sea management area covering the remainder of the stock; SD22-32.
# === Figure 4 Landings in 21 ====
IC_agg_temp <- aggregate(CATON ~ Year + Country, data = ic_clean[ic_clean$CatchCategory ==
"Landings" & ic_clean$Area == "27.3.a.21", ], FUN = sum)
IC_agg_temp$Country <- factor(IC_agg_temp$Country, levels = c("Sweden", "Germany",
"Denmark"))
# IC_agg_temp$CATON <- IC_agg_temp$CATON.kg/1000
TAC21 <- tac_raw[tac_raw$Area == "27.3.a.21" & tac_raw$Year %in% c(2002:DataYear),
]
plt_land_21 <- ggplot() + stat_summary(fun.y = sum, geom = "area", position = "stack",
na.rm = TRUE, data = IC_agg_temp, mapping = aes(x = Year, y = CATON, group = Country,
fill = Country)) + geom_line(data = TAC21, mapping = aes(x = Year, y = TAC,
group = Area, colour = Area), size = 2) + ylab("Landings (kg)") + theme(axis.text.x = element_text(angle = 90,
vjust = 0.5, hjust = 1)) + theme_clean() + scale_fill_manual(values = c("#004B87",
"#000000", "#C60C30")) + scale_colour_manual(name = "TAC", values = "#7570b3",
)
style(ggplotly(plt_land_21), showlegend = FALSE)Figure 2.2: Recent history landings for the North Sea Management area in subdivision 21
# =====# === Figure 5 Landings in 22 ====
IC_agg_temp <- aggregate(CATON ~ Year + Country, data = ic_clean[ic_clean$CatchCategory !=
"Discards" & ic_clean$Area == "27.3.c.22", ], FUN = sum)
IC_agg_temp$Country <- factor(IC_agg_temp$Country, levels = c("Sweden", "Germany",
"Denmark"))
# IC_agg_temp$CATON <- IC_agg_temp$CATON.kg/1000
TAC22 <- tac_raw[tac_raw$Area == "27.3.c.22" & tac_raw$Year %in% c(2002:DataYear),
]
plt_land_22 <- ggplot() + stat_summary(fun.y = sum, geom = "area", position = "stack",
na.rm = TRUE, data = IC_agg_temp, mapping = aes(x = Year, y = CATON, group = Country,
fill = Country)) + geom_line(data = TAC22, mapping = aes(x = Year, y = TAC,
group = Area), size = 2, colour = "#7570b3") + ylab("Landings (kg)") + theme(axis.text.x = element_text(angle = 90,
vjust = 0.5, hjust = 1)) + scale_fill_manual(values = c("#004B87", "#000000",
"#C60C30"))
# ====# === Figure 6 Landings in 23 ====
IC_agg_temp <- aggregate(CATON ~ Year + Country, data = ic_clean[ic_clean$CatchCategory !=
"Discards" & ic_clean$Area == "27.3.b.23", ], FUN = sum)
IC_agg_temp$Country <- factor(IC_agg_temp$Country, levels = c("Sweden", "Germany",
"Denmark"))
# IC_agg_temp$CATON <- IC_agg_temp$CATON.kg/1000
TAC23 <- tac_raw[tac_raw$Area == "27.3.b.23" & tac_raw$Year %in% c(2002:DataYear),
]
plt_land_23 <- ggplot() + stat_summary(fun.y = sum, geom = "area", position = "stack",
na.rm = TRUE, data = IC_agg_temp, mapping = aes(x = Year, y = CATON, group = Country,
fill = Country)) + geom_line(data = TAC23, mapping = aes(x = Year, y = TAC,
group = Area), size = 2, colour = "#7570b3") + ylab("Landings (kg)") + theme(axis.text.x = element_text(angle = 90,
vjust = 0.5, hjust = 1)) + scale_fill_manual(values = c("#004B87", "#C60C30"))
# ====# === Figure 5 Landings in 22:32 ====
IC_agg_temp <- aggregate(CATON ~ Year + Country, data = ic_clean[ic_clean$CatchCategory ==
"Landings" & ic_clean$Area != "27.3.a.21", ], FUN = sum)
IC_agg_temp$Country <- factor(IC_agg_temp$Country, levels = c("Sweden", "Germany",
"Denmark", "Poland", "Latvia", "Lithuania", "Estonia", "Finland"))
# IC_agg_temp$CATON <- IC_agg_temp$CATON.kg/1000
TAC22 <- tac_raw[tac_raw$Area == "ple.27.22-32" & tac_raw$Year %in% c(2002:DataYear),
]
plt_land_22_32 <- ggplot() + stat_summary(fun.y = sum, geom = "area", position = "stack",
na.rm = TRUE, data = IC_agg_temp, mapping = aes(x = Year, y = CATON, group = Country,
fill = Country)) + geom_line(data = TAC22, mapping = aes(x = Year, y = TAC,
group = Area, colour = Area), size = 2) + ylab("Landings (kg)") + theme(axis.text.x = element_text(angle = 90,
vjust = 0.5, hjust = 1)) + theme_clean() + scale_fill_manual(values = c("#004B87",
"#000000", "#C60C30", ebpal)) + scale_colour_manual(values = "#7570b3")
style(ggplotly(plt_land_22_32), showlegend = FALSE)Figure 2.3: Recent history landings for the Baltic Sea Management area in subdivisions 22-32
# ====#===
# Figure 7 Landings in all areas by country
#====
IC_agg_temp <- aggregate(CATON~Year+Country+Area,
data = ic_clean[ic_clean$CatchCategory != "Discards", ],
FUN = sum)
IC_agg_temp$Country <- factor(IC_agg_temp$Country, levels = c("Sweden", "Germany", "Denmark", "Poland", "Latvia", "Lithuania", "Estonia", "Finland"))
# IC_agg_temp$CATON <- IC_agg_temp$CATON.kg/1000
# TACall <- tac_raw[tac_raw$Year %in% c(2002:DataYear),]
TAC21 <- tac_raw[tac_raw$Area== "27.3.a.21" & tac_raw$Year %in% c(2002:DataYear+2), ]
TAC22 <- tac_raw[tac_raw$Area == "ple.27.22-32" & tac_raw$Year %in% c(2002:DataYear+2), ]
plt_land_all <- ggplot() +
stat_summary(fun.y = sum,
geom = "area",
position = "stack",
na.rm = TRUE,
data = IC_agg_temp[IC_agg_temp$Area %in% c("27.3.a.21", "27.3.b.23", "27.3.c.22", "27.3.d.24", "27.3.d.25"),],
mapping = aes(x = Year,
y = CATON,
group = Country,
fill = Country)) +
geom_line(data = TAC21[TAC21$Area %in% c("27.3.a.21", "27.3.b.23", "27.3.c.22", "27.3.d.24", "27.3.d.25"),],
mapping = aes(x = Year,
y = TAC),
# size = 2,
colour = ebpal[length(ebpal)-1]) +
# geom_line(data = TAC22,
# mapping = aes(x = Year,
# y = TAC),
# # size = 2,
# colour = ebpal[length(ebpal)-2]) +
ylab("Landings (t)") +
theme_clean()+
theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust = 1)) +
scale_fill_manual(values = c("#004B87","#000000", "#C60C30", ebpal)) +
facet_grid(rows = vars(Area))#, scales = "free_y")
ggplotly(plt_land_all)
#====# === Figure of landings by gear type ====
IC_agg_temp <- aggregate(CATON ~ CatchCategory + Fleet + Year, data = ic_clean, FUN = sum)
plt_land_gear <- ggplot(data = IC_agg_temp[IC_agg_temp$CatchCategory == "Landings",
], aes(x = Year, y = CATON, group = Fleet, fill = Fleet)) + stat_summary(fun.y = sum,
geom = "area", position = "stack", na.rm = TRUE) + ylab("Landings (Tons)") +
theme_clean() + theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust = 1)) +
scale_fill_manual(values = ebpal[1:length(unique(IC_agg_temp$Fleet))])
ggplotly(plt_land_gear)Figure 2.4: Recent history landings split by active and passive gears
2.1.2 Commercial Catches - Results
2.1.2.1 Catch Volume (Mass)
Catches for the year 2023 are summarised across different factors below and are given in tonnes.
IC_agg_temp <- aggregate(CATON ~ Area, data = ic_clean[ic_clean$Year == DataYear &
ic_clean$CatchCategory %in% c("Landings", "Discards"), ], FUN = sum)
Proportion <- numeric()
for (i in IC_agg_temp$Area) {
Proportion <- append(x = Proportion, values = IC_agg_temp[IC_agg_temp$Area ==
i, "CATON"]/sum(IC_agg_temp[, "CATON"]))
}
IC_agg_temp <- cbind(IC_agg_temp, Proportion)
kable(IC_agg_temp, caption = paste0("Total commercial catch and proportion of catch of plaice, per area in ",
DataYear, " (tonnes)."), digits = 2)| Area | CATON | Proportion |
|---|---|---|
| 27.3.a.21 | 669.44 | 0.21 |
| 27.3.b.23 | 63.16 | 0.02 |
| 27.3.c.22 | 1808.06 | 0.57 |
| 27.3.d.24 | 551.52 | 0.17 |
| 27.3.d.25 | 96.80 | 0.03 |
| 27.3.d.26 | 3.67 | 0.00 |
| 27.3.d.27 | 0.00 | 0.00 |
| 27.3.d.28 | 0.00 | 0.00 |
| 27.3.d.29 | 0.00 | 0.00 |
| 27.3.d.31 | 0.00 | 0.00 |
kable(aggregate(CATON ~ Area + Fleet, data = ic_clean[ic_clean$Year == DataYear &
ic_clean$CatchCategory %in% c("Landings", "Discards"), ], FUN = sum), caption = paste0("Total commercial catch of plaice, per area and by fleet, in ",
DataYear, " (tonnes)."), digits = 2)| Area | Fleet | CATON |
|---|---|---|
| 27.3.a.21 | Active | 635.02 |
| 27.3.b.23 | Active | 2.09 |
| 27.3.c.22 | Active | 1225.14 |
| 27.3.d.24 | Active | 514.82 |
| 27.3.d.25 | Active | 82.57 |
| 27.3.d.26 | Active | 0.48 |
| 27.3.d.27 | Active | 0.00 |
| 27.3.d.28 | Active | 0.00 |
| 27.3.d.31 | Active | 0.00 |
| 27.3.a.21 | Passive | 34.42 |
| 27.3.b.23 | Passive | 61.07 |
| 27.3.c.22 | Passive | 582.92 |
| 27.3.d.24 | Passive | 36.70 |
| 27.3.d.25 | Passive | 14.23 |
| 27.3.d.26 | Passive | 3.19 |
| 27.3.d.27 | Passive | 0.00 |
| 27.3.d.28 | Passive | 0.00 |
| 27.3.d.29 | Passive | 0.00 |
kable(aggregate(CATON ~ Fleet2, data = ic_clean[ic_clean$Year == DataYear & ic_clean$CatchCategory %in%
c("Landings", "Discards"), ], FUN = sum), caption = paste0("Total commercial catch of plaice, by new fleet concept (management area * fleet), in ",
DataYear, " (tonnes)."), digits = 2)| Fleet2 | CATON |
|---|---|
| BS_Active | 1825.10 |
| BS_Passive | 698.11 |
| NS_Active | 635.02 |
| NS_Passive | 34.42 |
kable(aggregate(CATON ~ Area + Fleet2, data = ic_clean[ic_clean$Year == DataYear &
ic_clean$CatchCategory %in% c("Landings", "Discards"), ], FUN = sum), caption = paste0("Total commercial catch of plaice, per area and new fleet concept (management area * fleet), in ",
DataYear, " (tonnes)."), digits = 2)| Area | Fleet2 | CATON |
|---|---|---|
| 27.3.b.23 | BS_Active | 2.09 |
| 27.3.c.22 | BS_Active | 1225.14 |
| 27.3.d.24 | BS_Active | 514.82 |
| 27.3.d.25 | BS_Active | 82.57 |
| 27.3.d.26 | BS_Active | 0.48 |
| 27.3.d.27 | BS_Active | 0.00 |
| 27.3.d.28 | BS_Active | 0.00 |
| 27.3.d.31 | BS_Active | 0.00 |
| 27.3.b.23 | BS_Passive | 61.07 |
| 27.3.c.22 | BS_Passive | 582.92 |
| 27.3.d.24 | BS_Passive | 36.70 |
| 27.3.d.25 | BS_Passive | 14.23 |
| 27.3.d.26 | BS_Passive | 3.19 |
| 27.3.d.27 | BS_Passive | 0.00 |
| 27.3.d.28 | BS_Passive | 0.00 |
| 27.3.d.29 | BS_Passive | 0.00 |
| 27.3.a.21 | NS_Active | 635.02 |
| 27.3.a.21 | NS_Passive | 34.42 |
# === Figure Catch by catch category (all areas, all countries) ====
IC_agg_temp <- aggregate(CATON ~ Year + CatchCategory + Area, data = ic_clean[!ic_clean$CatchCategory %in%
c("BMS", "BMS landing", "Logbook Registered Discard", "Catch"), ], FUN = sum)
plt_catchcat <- ggplot() + stat_summary(fun.y = sum, geom = "area", position = "stack",
na.rm = TRUE, data = IC_agg_temp, mapping = aes(x = Year, y = CATON, group = CatchCategory,
fill = CatchCategory)) + ylab("Catch (Tons)") + theme_clean() + theme(axis.text.x = element_text(angle = 90,
vjust = 0.5, hjust = 1), strip.text = element_text(face = "bold", size = rel(0.75)),
strip.background = element_blank()) + scale_fill_manual(values = c("#1b9e77",
"#d95f02", "#7570b3", "grey")) + ylab("Catch (t)") + facet_wrap(facets = vars(Area),
ncol = 2) #, scales = 'free_y') +
style(ggplotly(plt_catchcat), showlegend = FALSE)Figure 2.5: Commercial catch by catch category for all areas and all countries (tonnes).
# ====2.1.2.2 Catch Numbers
We can also investigate stock structure by looking at numbers at age.
catchNumLng <- reshape(data = catchNum, varying = list(colnames(catchNum)[2:length((colnames(catchNum)))]),
v.names = c("Number"), timevar = "Age", direction = "long")
catchNumLng$Age <- as.factor(as.character(catchNumLng$Age))
catchNumLng$Age <- factor(catchNumLng$Age, levels = c("1", "2", "3", "4", "5", "6",
"7", "8", "9", "10"))
# catchNumLng$Year <- as.factor(as.character(catchNumLng$Year))
catchNumLng$Year <- as.numeric(as.character(catchNumLng$Year))
ggplotly(ggplot(catchNumLng) + geom_line(mapping = aes(x = Year, y = Number, colour = Age,
group = Age)) + theme_few() + scale_colour_manual(values = ebpal))2.1.2.3 Discards
Landings below minimum size (BMS landing) are estimated as part of the discards, due to expected poor reporting of official BMS landings. This means that in the assessment, we do not use reports of BMS. However, when reporting on BMS catch components, we must subtract official BMS landings from Discards. Prior to 2019 (<=2018), BMS landings were reported as part of landings.
## Aggregate catches by year and area
IC_agg_temp <- aggregate(CATON ~ Year + CatchCategory + Area, data = ic_clean[ic_clean$CatchCategory %in%
c("Landings", "Discards"), ], FUN = sum)
disrat <- as.data.frame(reshape(IC_agg_temp, idvar = c("Year", "Area"), timevar = "CatchCategory",
direction = "wide"))
disrat$DiscardRatio <- disrat$CATON.Discards/(disrat$CATON.Landings + disrat$CATON.Discards)
disrat$Area <- factor(disrat$Area, levels = c("27.3.a.21", "27.3.b.23", "27.3.c.22",
"27.3.d.24", "27.3.d.25", "27.3.d.26", "27.3.d.27", "27.3.d.28", "27.3.d.28.2",
"27.3.d.29", "27.3.d.30", "27.3.d.31", "27.3.d.32"))
## Aggregate catches by year and fleet
IC_agg_temp <- aggregate(CATON ~ Year + CatchCategory + Fleet2, data = ic_clean[ic_clean$CatchCategory %in%
c("Landings", "Discards"), ], FUN = sum)
disratFleet <- as.data.frame(reshape(IC_agg_temp, idvar = c("Year", "Fleet2"), timevar = "CatchCategory",
direction = "wide"))
disratFleet$DiscardRatio <- disratFleet$CATON.Discards/(disratFleet$CATON.Landings +
disratFleet$CATON.Discards)
rownames(disrat) <- NULLkable(disrat[disrat$Year == DataYear, c("Area", "DiscardRatio")], caption = paste0("Discard ratios by subdivision for ",
DataYear, "."), digits = 3)| Area | DiscardRatio | |
|---|---|---|
| 22 | 27.3.a.21 | 0.732 |
| 44 | 27.3.b.23 | 0.242 |
| 66 | 27.3.c.22 | 0.353 |
| 88 | 27.3.d.24 | 0.240 |
| 110 | 27.3.d.25 | 0.211 |
| 132 | 27.3.d.26 | 0.224 |
| 154 | 27.3.d.27 | NA |
| 172 | 27.3.d.28 | NA |
| 194 | 27.3.d.29 | NA |
| 222 | 27.3.d.31 | NA |
# === Figures of discard ratios ==== ## Bar chart plt_disrat_area <- ggplot() +
# geom_col(data = disrat, mapping = aes(x = Year, y = DiscardRatio, fill =
# Area), position = 'dodge') + scale_fill_manual(values = ebpal) + theme_bw() +
# theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5))
## Line Chart
disrat$Year <- as.numeric(as.character(disrat$Year))
plt_disrat_area <- ggplot() + geom_line(data = disrat, mapping = aes(x = Year, y = DiscardRatio,
colour = Area), size = 1.5, position = "dodge") + scale_colour_manual(values = ebpal) +
theme_clean() + theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5))
ggplotly(plt_disrat_area)Figure 2.6: Discard ratios (proportion of total commercial catch discarded) by subdivision over time.
# =====Plaice is considered a bycatch species in the Baltic Sea Management plan. In SD21 and 22 this remains to be the case where plaice are apparently caught in nephrops (SD21) and cod (SD21&22) fisheries, with some targeted coastal fisheries (especially in SD22&23). To further investigate the discard rates relative to landings by different fleets in the different managment areas, we can plot the two figures together. In doing so we make the discard ratios relative to the landings scale, thus the lines in the below figure show the relative contribution of discards to the catch compared to the maximum value (i.e. not absolute discard ratios, nor linked to the y-axis units).
# === Prepare data ====
IC_agg_temp <- droplevels(aggregate(CATON ~ Year + CatchCategory + Fleet2, data = ic_clean[ic_clean$CatchCategory ==
"Landings", ], FUN = sum))
# disrat$Year <- as.factor(as.character(disrat$Year))
disratFleet$Year <- as.numeric(as.character(disratFleet$Year))
IC_agg_temp$Year <- as.numeric(as.character(IC_agg_temp$Year))
ylim.prim <- c(0, max(range(IC_agg_temp$CATON)))
ylim.sec <- c(0, max(range(disratFleet$DiscardRatio, na.rm = T)))
b <- diff(ylim.prim)/diff(ylim.sec)
a <- ylim.prim[1] - (b * ylim.sec[1])
# =====
# === Figure for advice after 2021: Discard Ratios and Landings over time ====
plt_catchcat <- ggplot() + geom_col(data = IC_agg_temp, mapping = aes(x = Year, y = CATON),
fill = "#002b5f") + geom_line(data = disratFleet, mapping = aes(x = Year, y = a +
DiscardRatio * b), colour = "#ed5f26", size = 1.5) + scale_y_continuous("Landings (tonnes)",
sec.axis = sec_axis(~(. - a)/b, name = "Discard Ratio")) + theme_clean() + theme(axis.text.x = element_text(angle = 90,
vjust = 0.5, hjust = 1), axis.line.y.right = element_line(color = "#ed5f26"),
axis.ticks.y.right = element_line(color = "#ed5f26"), axis.text.y.right = element_text(color = "#ed5f26"),
axis.title.y.right = element_text(color = "#ed5f26"), axis.line.y.left = element_line(color = "#002b5f"),
axis.ticks.y.left = element_line(color = "#002b5f"), axis.text.y.left = element_text(color = "#002b5f"),
axis.title.y.left = element_text(color = "#002b5f")) + facet_grid(rows = vars(Fleet2))
ggplotly(plt_catchcat)Figure 2.7: Commercial catch as Landings (bars) and Discard Ratios as relative proportions (lines) over time by fleet and management area
# =====In subdivision 22, the collapse of the cod fishery from the period 2015 onward, appears to show a greater retention of Plaice, while fishers were still targetting cod. Subsequently, as fishing opportunities for cod collapsed, the fisher for plaice also collapsed. This could be due to the lower overall effort, with a specific interest in what cod is available, or it could be due to the explosion in the plaice population. In the latter case, plaice are both caught more easily, while remaining unwanted, and their condition has been deteriorating, meaning a higher proportion of the catch is not marketable. Below we focus on SD22 and how catches and discard ratios have developed together over time.
# === Prepare data ====
IC_agg_temp <- droplevels(aggregate(CATON ~ Year + CatchCategory + Area, data = ic_clean[ic_clean$Area ==
"27.3.c.22" & ic_clean$CatchCategory == "Landings", ], FUN = sum))
# disrat$Year <- as.factor(as.character(disrat$Year))
disrat$Year <- as.numeric(as.character(disrat$Year))
IC_agg_temp$Year <- as.numeric(as.character(IC_agg_temp$Year))
disrat <- disrat[disrat$Area == "27.3.c.22", ]
ylim.prim <- c(0, max(range(IC_agg_temp$CATON)))
ylim.sec <- c(0, max(range(disrat$DiscardRatio)))
b <- diff(ylim.prim)/diff(ylim.sec)
a <- ylim.prim[1] - (b * ylim.sec[1])
# =====
# === Figure of landings and discard ratios over time ====
plt_catchcat <- ggplot() + geom_col(data = IC_agg_temp, mapping = aes(x = Year, y = CATON),
fill = "#002b5f") + geom_line(data = disrat, mapping = aes(x = Year, y = a +
DiscardRatio * b), colour = "#ed5f26", size = 1.5) + scale_y_continuous("Landings (tonnes)",
sec.axis = sec_axis(~(. - a)/b, name = "Discard Ratio")) + theme_clean() + theme(axis.text = element_text(size = 20),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust = 1), axis.line.y.right = element_line(color = "#ed5f26"),
axis.ticks.y.right = element_line(color = "#ed5f26"), axis.text.y.right = element_text(color = "#ed5f26"),
axis.title.y.right = element_text(color = "#ed5f26"), axis.line.y.left = element_line(color = "#002b5f"),
axis.ticks.y.left = element_line(color = "#002b5f"), axis.text.y.left = element_text(color = "#002b5f"),
axis.title.y.left = element_text(color = "#002b5f"))
ggplotly(plt_catchcat)Figure 2.8: Discard Ratios (lines) and Landings (bars) over time in SD22 - the SW Baltic
# =====These same trends are mirrored in the discard rates by country, where vessels mainly operating in 21 (i.e. Sweden) have the highest rates, those operating mainly in the Baltic (i.e. Poland and Germany) the lowest, and those operating in both have an intermediate level of discards (i.e. Denmark).
# === Prep data ====
IC_agg_temp <- aggregate(CATON ~ Year + CatchCategory + Country, data = ic_clean[ic_clean$CatchCategory %in%
c("Landings", "Discards"), ], FUN = sum)
disrat <- as.data.frame(reshape(IC_agg_temp, idvar = c("Year", "Country"), timevar = "CatchCategory",
direction = "wide"))
# disrat[is.na(disrat$CATON.BMS), 'CATON.BMS'] <- 0
disrat$DiscardRatio <- disrat$CATON.Discards/(disrat$CATON.Landings + disrat$CATON.Discards)
disrat$Country <- factor(disrat$Country, levels = c("Sweden", "Germany", "Denmark",
"Poland", "Latvia", "Lithuania", "Estonia", "Finland"))
disrat$Year <- as.numeric(as.character(disrat$Year))
# =====
# === Figure Discard Ratios by Country ====
plt_disrat_country <- ggplot() + geom_line(data = disrat, mapping = aes(x = Year,
y = DiscardRatio, colour = Country), size = 1.5, position = "dodge") + theme_clean() +
scale_colour_manual(values = c("#004B87", "#000000", "#C60C30", ebpal))
ggplotly(plt_disrat_country)Figure 2.9: Discard ratios (proportion of total catch discarded) by country over time
# ====Drilling down into a bit more detail we can see where (subdivision), when (quarter) and by which general gear types the majority of catches (discards & landings) are taken.
#===
# Figure Discards/landings of DataYear by quarter and gear
#====
IC_agg_temp <- aggregate(CATON~Season+CatchCategory+Area+Fleet,
# data = ic_clean[ic_clean$Year == DataYear & ic_clean$Fleet %in% c("Active", "Passive"), ],
data = ic_clean[ic_clean$Year == DataYear & ic_clean$CatchCategory %in% c("Discards", "Landings"), ],
FUN = sum)
IC_agg_temp <- droplevels(IC_agg_temp)
#=====
#===
# Figure Discards/landings of DataYear by quarter and gear
#====
plt_disrat_Q_area <- ggplot() +
geom_col(data = IC_agg_temp,
mapping = aes(x = Season,
y = CATON,
fill = CatchCategory)) +
facet_grid(rows = vars(Area),
cols = vars(Fleet),
scales = "free_y") +
theme_clean() +
theme(strip.background = element_blank(),
strip.text = element_text(size = rel(0.75))) +
scale_fill_manual(values = ebpal) +
xlab("Quarter") +
ylab("Catch (t)")
style(ggplotly(plt_disrat_Q_area), showlegend = FALSE)Figure 2.10: Discards and landings from Current year by quarter and gear
#====#===
# Figure Discards/landings of DataYear by quarter and gear
#====
IC_agg_temp <- aggregate(CATON~Area+CatchCategory+Fleet+Season,
# data = ic_clean[ic_clean$Year == DataYear & ic_clean$Fleet %in% c("Active", "Passive"), ],
data = ic_clean[ic_clean$Year == DataYear & ic_clean$CatchCategory %in% c("Discards", "Landings"), ],
FUN = sum)
IC_agg_temp <- droplevels(IC_agg_temp)
#=====
#===
# Reshape to make Quarters wide
#====
IC_agg_temp <- reshape(IC_agg_temp,
direction = "wide",
idvar = c("Area", "CatchCategory", "Fleet"),
timevar = "Season")
colnames(IC_agg_temp) <- c("Subdivision", "CatchCategory", "Fleet", "Q1", "Q2", "Q3", "Q4")
IC_agg_temp <- IC_agg_temp[order(IC_agg_temp$Subdivision, IC_agg_temp$CatchCategory, IC_agg_temp$Fleet), ]
#=====
kable(IC_agg_temp,
caption = paste0("Commercial catch from ple.27.21-23 in ", DataYear, " by catch category, by fleet and over quarters (tonnes)."),
digits = 0)| Subdivision | CatchCategory | Fleet | Q1 | Q2 | Q3 | Q4 | |
|---|---|---|---|---|---|---|---|
| 1 | 27.3.a.21 | Discards | Active | 82 | 82 | 194 | 126 |
| 14 | 27.3.a.21 | Discards | Passive | 1 | 1 | 3 | 1 |
| 6 | 27.3.a.21 | Landings | Active | 42 | 21 | 44 | 45 |
| 19 | 27.3.a.21 | Landings | Passive | 3 | 7 | 12 | 6 |
| 27 | 27.3.b.23 | Discards | Active | NA | 1 | NA | 1 |
| 15 | 27.3.b.23 | Discards | Passive | 4 | 7 | 1 | 1 |
| 7 | 27.3.b.23 | Landings | Active | 0 | 0 | 0 | 0 |
| 20 | 27.3.b.23 | Landings | Passive | 13 | 14 | 15 | 5 |
| 2 | 27.3.c.22 | Discards | Active | 327 | 129 | 14 | 154 |
| 16 | 27.3.c.22 | Discards | Passive | 3 | 3 | 6 | 2 |
| 8 | 27.3.c.22 | Landings | Active | 152 | 172 | 27 | 251 |
| 21 | 27.3.c.22 | Landings | Passive | 237 | 173 | 101 | 58 |
| 3 | 27.3.d.24 | Discards | Active | 1 | 18 | 75 | 35 |
| 17 | 27.3.d.24 | Discards | Passive | 0 | 0 | 2 | 2 |
| 9 | 27.3.d.24 | Landings | Active | 54 | 51 | 237 | 45 |
| 22 | 27.3.d.24 | Landings | Passive | 2 | 14 | 15 | 2 |
| 4 | 27.3.d.25 | Discards | Active | 4 | 8 | 5 | 3 |
| 18 | 27.3.d.25 | Discards | Passive | 0 | 0 | 0 | 0 |
| 10 | 27.3.d.25 | Landings | Active | 14 | 22 | 16 | 11 |
| 23 | 27.3.d.25 | Landings | Passive | 2 | 6 | 4 | 0 |
| 5 | 27.3.d.26 | Discards | Active | 0 | NA | NA | NA |
| 44 | 27.3.d.26 | Discards | Passive | NA | 0 | 0 | 1 |
| 11 | 27.3.d.26 | Landings | Active | 0 | 0 | 0 | 0 |
| 24 | 27.3.d.26 | Landings | Passive | 0 | 0 | 2 | 1 |
| 12 | 27.3.d.27 | Landings | Active | 0 | 0 | 0 | 0 |
| 25 | 27.3.d.27 | Landings | Passive | 0 | 0 | 0 | 0 |
| 13 | 27.3.d.28 | Landings | Active | 0 | 0 | 0 | 0 |
| 52 | 27.3.d.28 | Landings | Passive | NA | 0 | 0 | 0 |
| 53 | 27.3.d.29 | Landings | Passive | NA | 0 | 0 | 0 |
| 94 | 27.3.d.31 | Landings | Active | NA | NA | NA | 0 |
Furthermore, we can split the discard rates by our new fleet concept, that of fleets by management area. This way different discard rates in different managemetn areas can be taken into consideration when setting TACs and other management devices.
# === Prep data ====
IC_agg_temp <- aggregate(CATON ~ Year + CatchCategory + Fleet2, data = ic_clean[ic_clean$CatchCategory %in%
c("Landings", "Discards"), ], FUN = sum)
disrat <- as.data.frame(reshape(IC_agg_temp, idvar = c("Year", "Fleet2"), timevar = "CatchCategory",
direction = "wide"))
# disrat[is.na(disrat$CATON.BMS), 'CATON.BMS'] <- 0
disrat$DiscardRatio <- disrat$CATON.Discards/(disrat$CATON.Landings + disrat$CATON.Discards)
disrat$Fleet2 <- factor(disrat$Fleet2, levels = c("NS_Active", "NS_Passive", "BS_Active",
"BS_Passive"))
disrat$Year <- as.numeric(as.character(disrat$Year))
# =====
# === Figure Discard Ratios by Country ====
plt_disrat_country <- ggplot() + geom_line(data = disrat, mapping = aes(x = Year,
y = DiscardRatio, colour = Fleet2), size = 1.5, position = "dodge") + theme_clean() +
# facet_grid(Country~.) +
scale_colour_manual(values = ebpal)
ggplotly(plt_disrat_country)Figure 2.11: Discard ratios (proportion of total commercial catch discarded) by management area and fleet, over time
# ====# === Data agreggation ====
IC_agg_temp <- aggregate(CATON ~ Year + CatchCategory, data = ic_clean, FUN = sum)
disrat <- as.data.frame(reshape(IC_agg_temp, idvar = c("Year"), timevar = "CatchCategory",
direction = "wide"))
disrat[is.na(disrat$CATON.BMS), "CATON.BMS"] <- 0
disrat$DiscardRatio <- disrat$CATON.Discards/(disrat$CATON.Landings + disrat$CATON.Discards)
mdr <- median(disrat$DiscardRatio, na.rm = T)
mdr3 <- mean(disrat[disrat$Year %in% c((DataYear - 2):(DataYear)), "DiscardRatio"],
na.rm = T)
# =====The general time trends for the whole stock are of less interest but are included in the report as a time series. The median of the time series is 0.3677652. The mean discard rate for the prior three years is 0.3589337.
# === Discard Ratios all together ====
plt_disrat <- ggplot() + geom_point(data = disrat, mapping = aes(x = Year, y = DiscardRatio),
colour = ebpal[1]) + geom_line(data = disrat, mapping = aes(x = Year, y = DiscardRatio,
group = 1), colour = ebpal[2]) + geom_hline(mapping = aes(yintercept = mdr),
colour = ebpal[3]) + theme_clean() + theme(axis.text.x = element_text(angle = 90,
hjust = 1, vjust = 0.5))
ggplotly(plt_disrat)Figure 2.12: Annual discard ratio for the total stock over time. Orange line represents median of the time series.
# ====2.1.3 Catches - advice sheet requirements
The advice sheets require specific information which is presented below
# === Aggregate catches in a format for use in the stock splitting table ====
IC_agg_temp <- aggregate(CATON ~ CatchCategory + Area, data = ic_clean[ic_clean$Year ==
DataYear & ic_clean$CatchCategory %in% c("Landings", "Discards"), ], FUN = "sum")
IC_agg_temp <- reshape(IC_agg_temp, direction = "wide", idvar = "Area", timevar = "CatchCategory")
IC_agg_temp[is.na(IC_agg_temp$CATON.Discards), "CATON.Discards"] <- 0
IC_agg_temp$Catch <- IC_agg_temp$CATON.Discards + IC_agg_temp$CATON.Landings
# =====
kable(IC_agg_temp, caption = paste0("Catches by catch category and subdivision in ",
DataYear, " (tonnes). _Used for the stock splitting table_."), digits = 2)- The catch from subdivisions 21, 22 & 23 combined in 2023 was 3192.66578 tonnes.
- The landings from subdivisions 21, 22 & 23 combined in 2023 was 1894.59435 tonnes.
- The percentage of landings that were taken in active gears was 63.4407624%.
- The percentage of landings that were taken in passive gears was 36.5592376%.
- The discards from subdivisions 21, 22 & 23 combined in 2023 was 1298.05243 tonnes.
The History of commercial catch and landings table in the advice sheet requires that BMS official landings are subtracted from the discard values. Need to determine the advice sheet resolution requirements (by stock, by management area, by subdivision etc.) before reworking this code to fit.
# === Aggregate and format Landings and Discards ====
IC_agg_temp <- aggregate(CATON ~ Area + CatchCategory + Country, data = ic_clean[ic_clean$Year ==
DataYear, ], FUN = "sum")
adctchtble <- reshape(data = IC_agg_temp[IC_agg_temp$CatchCategory %in% c("Landings",
"Discards"), ], direction = "wide", timevar = "Area", idvar = c("CatchCategory",
"Country"))
# adctchtble[is.na(adctchtble$CATON.27.3.a.21), 'CATON.27.3.a.21'] <- 0
# adctchtble[is.na(adctchtble$CATON.27.3.b.23), 'CATON.27.3.b.23'] <- 0
# adctchtble[is.na(adctchtble$CATON.27.3.c.22), 'CATON.27.3.c.22'] <- 0
adctchtble <- data.table::transpose(adctchtble)
colnames(adctchtble) <- c("DK_Discards", "DK_Landings", "DE_Discards", "DE_Landings",
"SK_Discards", "SK_Landings")
adctchtble <- adctchtble[3:5, ]
adctchtble <- as.data.frame(lapply(adctchtble, FUN = "as.numeric"))
# ====
# === Aggregate and format BMS Landings ====
IC_agg_temp1 <- aggregate(OfficialLandings.kg ~ Area + CatchCategory + Country, data = ic_clean[ic_clean$Year ==
DataYear, ], FUN = "sum")
adbmstble <- reshape(data = IC_agg_temp1[IC_agg_temp1$CatchCategory %in% c("BMS landing"),
], direction = "wide", timevar = "Area", idvar = c("CatchCategory", "Country"))
# adbmstble[is.na(adbmstble$OfficialLandings.kg.27.3.a.21),
# 'OfficialLandings.kg.27.3.a.21'] <- 0
# adbmstble[is.na(adbmstble$OfficialLandings.kg.27.3.b.23),
# 'OfficialLandings.kg.27.3.b.23'] <- 0
# adbmstble[is.na(adbmstble$OfficialLandings.kg.27.3.c.22),
# 'OfficialLandings.kg.27.3.c.22'] <- 0
adbmstble <- data.table::transpose(adbmstble)
colnames(adbmstble) <- c("DK_BMS", "DE_BMS", "SK_BMS")
adbmstble <- adbmstble[3:5, ]
adbmstble <- as.data.frame(lapply(adbmstble, FUN = "as.numeric"))
## Convert from kg to tonnes
adbmstble <- adbmstble/1000
# ===== === Combine, subtract BMS from Discards & Restructure ==== Combine
adctchtble <- cbind(adctchtble, adbmstble)
## Replace NAs with zeros for accurate arithmetic
adctchtble[is.na(adctchtble)] <- 0
## Subtract BMS from discards
adctchtble$DK_Discards <- adctchtble$DK_Discards - adctchtble$DK_BMS
adctchtble$DE_Discards <- adctchtble$DE_Discards - adctchtble$DE_BMS
adctchtble$SK_Discards <- adctchtble$SK_Discards - adctchtble$SK_BMS
## Calculate total catch
IC_agg_temp <- aggregate(CATON ~ Area, data = ic_clean[ic_clean$Year == DataYear &
ic_clean$CatchCategory %in% c("Landings", "Discards"), ], FUN = "sum")
adctchtble$TotalCatch <- IC_agg_temp$CATON
## Add subdivision information
adctchtble$Subdivision <- unique(IC_agg_temp$Area)
## Aggregate BMS across all countries
adctchtble$BMS <- rowSums(adctchtble[, c("DK_BMS", "DE_BMS", "SK_BMS")], na.rm = TRUE)
## Reformat
adctchtble <- adctchtble[, c("Subdivision", "DK_Discards", "DK_Landings", "DE_Discards",
"DE_Landings", "SK_Discards", "SK_Landings", "BMS", "TotalCatch")]
# =====
# === Add Annual Totals ====
tots <- adctchtble[0, ]
for (i in 2:ncol(tots)) {
tots[1, i] <- sum(adctchtble[, i], na.rm = TRUE)
}
tots$Subdivision <- as.character(DataYear)
adctchtble <- rbind(tots, adctchtble)
# =====
kable(adctchtble, caption = paste0("Official catch numbers by country and subdivision for ",
DataYear, ". Official landings of BMS have been subtracted from Discards. _can be copy-pasted into the advice sheet_."),
digits = 0)2.1.4 Commercial Sampling Coverage
Not all trips have observers or electronic monitoring to enable measurements of discard rates, or the sampling of all strata of fisheries for biological information. Therefore, raising is carried out to match unsampled strata with those where we have observations (this is currently done in intercatch). Each year, different portions of the fishery are sampled and to varying degrees, therefore there is no hard and fast rule about which strata should be matched. Rather, there are a set of priorities by which we try to match unsampled strata to sampled strata. These can be found in the stock annex. Below we explore the level of sampling coverage across the different strata for 2023 (using tonnage per category to derive proportions). Generally, this plaice stock is well covered by sampling and there are a minority of catches for which data are matched and assumptions are made.
# === Data wrangling and calculation of Raised/imported sampled/Estimated
# catches for Current Year ====
datacoverage_DY <- aggregate(CATON ~ CatchCategory + CATONRaisedOrImported + SampledOrEstimated,
data = ic_clean[ic_clean$Year == DataYear, ], FUN = sum)
lnd_DY <- sum(datacoverage_DY[datacoverage_DY$CatchCategory == "Landings", "CATON"])
disc_DY <- sum(datacoverage_DY[datacoverage_DY$CatchCategory %in% c("Discards"),
"CATON"])
datacoverage_DY$Proportion <- NA
datacoverage_DY[datacoverage_DY$CatchCategory == "Landings" & datacoverage_DY$SampledOrEstimated ==
"Sampled_Distribution", "Proportion"] <- datacoverage_DY[datacoverage_DY$CatchCategory ==
"Landings" & datacoverage_DY$SampledOrEstimated == "Sampled_Distribution", "CATON"]/lnd_DY
datacoverage_DY[datacoverage_DY$CatchCategory == "Landings" & datacoverage_DY$SampledOrEstimated ==
"Estimated_Distribution", "Proportion"] <- datacoverage_DY[datacoverage_DY$CatchCategory ==
"Landings" & datacoverage_DY$SampledOrEstimated == "Estimated_Distribution",
"CATON"]/lnd_DY
datacoverage_DY[datacoverage_DY$CatchCategory == "Discards" & datacoverage_DY$SampledOrEstimated ==
"Sampled_Distribution" & datacoverage_DY$CATONRaisedOrImported == "Imported_Data",
"Proportion"] <- datacoverage_DY[datacoverage_DY$CatchCategory == "Discards" &
datacoverage_DY$SampledOrEstimated == "Sampled_Distribution" & datacoverage_DY$CATONRaisedOrImported ==
"Imported_Data", "CATON"]/disc_DY
datacoverage_DY[datacoverage_DY$CatchCategory == "Discards" & datacoverage_DY$SampledOrEstimated ==
"Estimated_Distribution" & datacoverage_DY$CATONRaisedOrImported == "Imported_Data",
"Proportion"] <- datacoverage_DY[datacoverage_DY$CatchCategory == "Discards" &
datacoverage_DY$SampledOrEstimated == "Estimated_Distribution" & datacoverage_DY$CATONRaisedOrImported ==
"Imported_Data", "CATON"]/disc_DY
datacoverage_DY[datacoverage_DY$CatchCategory == "Discards" & datacoverage_DY$SampledOrEstimated ==
"Estimated_Distribution" & datacoverage_DY$CATONRaisedOrImported == "Raised_Discards",
"Proportion"] <- datacoverage_DY[datacoverage_DY$CatchCategory == "Discards" &
datacoverage_DY$SampledOrEstimated == "Estimated_Distribution" & datacoverage_DY$CATONRaisedOrImported ==
"Raised_Discards", "CATON"]/disc_DY
datacoverage_DY$Source <- paste(datacoverage_DY$SampledOrEstimated, datacoverage_DY$CATONRaisedOrImported,
sep = "\n")
datacoverage_DY$Source <- as.factor(datacoverage_DY$Source)
datacoverage_DY$Source <- factor(datacoverage_DY$Source, levels = c("Estimated_Distribution\nRaised_Discards",
"Estimated_Distribution\nImported_Data", "Sampled_Distribution\nImported_Data"))
datacoverage_DY <- droplevels(datacoverage_DY)
# =====# === Figure of sampling coverage for DataYear ====
plt_dataCover <- ggplot() + geom_col(data = datacoverage_DY[!datacoverage_DY$CatchCategory %in%
c("BMS", "BMS landing", "Logbook Registered Discard", "Catch"), ], mapping = aes(x = CatchCategory,
y = Proportion, fill = Source)) + theme_clean() + scale_fill_manual(values = c("#1b9e77",
"#d95f02", "#7570b3"))
ggplotly(plt_dataCover)Figure 2.13: Proportions of catch components that were sampled for biological data, and the proportion of discards that were raised by pairing with reported discard data. Proportions are made with catch tonnage by strata (fleet, area, country, quarter) for the year 2023
# =====# === Tabulate for presentation ====
dcDYtab <- datacoverage_DY[datacoverage_DY$CatchCategory %in% c("Landings", "Discards"),
c("CatchCategory", "Source", "CATON", "Proportion")]
dcDYtab$InfoSource <- gsub(pattern = "\\\n", replacement = " and ", dcDYtab$Source)
dcDYtab <- dcDYtab[dcDYtab$CatchCategory %in% c("Landings", "Discards"), c("CatchCategory",
"InfoSource", "CATON", "Proportion")]
rownames(dcDYtab) <- NULL
dcDYtab$CATON <- round(dcDYtab$CATON, digits = 0)
dcDYtab$Proportion <- icesRound(dcDYtab$Proportion * 100, percent = TRUE, sign = FALSE)
kable(dcDYtab, caption = paste0("Proportions of catch components that were sampled for biological data, and the proportion of discards that were raised by pairing with reported discard data. Proportions are made with catch tonnage by strata (fleet, area, country, quarter) for the year ",
DataYear))| CatchCategory | InfoSource | CATON | Proportion |
|---|---|---|---|
| Discards | Estimated_Distribution and Imported_Data | 9 | 0.70% |
| Landings | Estimated_Distribution and Imported_Data | 400 | 21% |
| Discards | Estimated_Distribution and Raised_Discards | 201 | 15.5% |
| Discards | Sampled_Distribution and Imported_Data | 1088 | 84% |
| Landings | Sampled_Distribution and Imported_Data | 1495 | 79% |
# =====# === Data wrangling ====
IC_agg_temp <- aggregate(cbind(CATON, No.of.length.sanples, No.length.mesures, No.of.age.samples,
No.of.age.readings) ~ Area + CatchCategory + Country + Fleet, data = ic_clean[ic_clean$CatchCategory %in%
c("Landings", "Discards") & ic_clean$Year == DataYear, ], FUN = "sum")
colnames(IC_agg_temp) <- c("Subdivision", "CatchCategory", "Country", "Fleet", "Catch(tonnes)",
"LengthSamples", "LengthsMeasured", "AgeSamples", "AgesRead")
samptable <- IC_agg_temp[order(IC_agg_temp$Subdivision, IC_agg_temp$CatchCategory,
IC_agg_temp$Country, IC_agg_temp$Fleet), ]
samptable$`Catch(tonnes)` <- round(samptable$`Catch(tonnes)`, digits = 3)
# =====
# === ====
rownames(samptable) <- NULL
kable(samptable, caption = paste0("Sampling effort in ", DataYear, " by strata."))| Subdivision | CatchCategory | Country | Fleet | Catch(tonnes) | LengthSamples | LengthsMeasured | AgeSamples | AgesRead |
|---|---|---|---|---|---|---|---|---|
| 27.3.a.21 | Discards | Denmark | Active | 388.544 | 39 | 3072 | 39 | 551 |
| 27.3.a.21 | Discards | Denmark | Passive | 3.525 | 0 | 0 | 0 | 0 |
| 27.3.a.21 | Discards | Germany | Active | 7.554 | 0 | 0 | 0 | 0 |
| 27.3.a.21 | Discards | Germany | Passive | 0.348 | 0 | 0 | 0 | 0 |
| 27.3.a.21 | Discards | Sweden | Active | 88.018 | 27 | 2496 | 30 | 1121 |
| 27.3.a.21 | Discards | Sweden | Passive | 2.235 | 3 | 163 | 15 | 671 |
| 27.3.a.21 | Landings | Denmark | Active | 136.852 | 17 | 3050 | 17 | 616 |
| 27.3.a.21 | Landings | Denmark | Passive | 25.286 | 17 | 3050 | 17 | 616 |
| 27.3.a.21 | Landings | Germany | Active | 1.771 | 0 | 0 | 0 | 0 |
| 27.3.a.21 | Landings | Germany | Passive | 1.436 | 0 | 0 | 0 | 0 |
| 27.3.a.21 | Landings | Sweden | Active | 12.279 | 0 | 0 | 0 | 0 |
| 27.3.a.21 | Landings | Sweden | Passive | 1.592 | 0 | 0 | 0 | 0 |
| 27.3.b.23 | Discards | Denmark | Active | 1.584 | 0 | 0 | 0 | 0 |
| 27.3.b.23 | Discards | Denmark | Passive | 9.726 | 4 | 208 | 4 | 37 |
| 27.3.b.23 | Discards | Sweden | Passive | 3.987 | 0 | 0 | 0 | 0 |
| 27.3.b.23 | Landings | Denmark | Active | 0.506 | 0 | 0 | 0 | 0 |
| 27.3.b.23 | Landings | Denmark | Passive | 39.614 | 0 | 0 | 0 | 0 |
| 27.3.b.23 | Landings | Sweden | Passive | 7.744 | 0 | 0 | 0 | 0 |
| 27.3.c.22 | Discards | Denmark | Active | 265.661 | 6 | 609 | 6 | 86 |
| 27.3.c.22 | Discards | Denmark | Passive | 7.596 | 10 | 210 | 10 | 54 |
| 27.3.c.22 | Discards | Germany | Active | 358.402 | 4 | 2703 | 4 | 406 |
| 27.3.c.22 | Discards | Germany | Passive | 7.077 | 15 | 113 | 15 | 45 |
| 27.3.c.22 | Landings | Denmark | Active | 303.055 | 45 | 9032 | 45 | 1706 |
| 27.3.c.22 | Landings | Denmark | Passive | 247.115 | 45 | 9032 | 45 | 1706 |
| 27.3.c.22 | Landings | Germany | Active | 298.019 | 4 | 1372 | 4 | 536 |
| 27.3.c.22 | Landings | Germany | Passive | 321.134 | 21 | 9744 | 21 | 1289 |
| 27.3.d.24 | Discards | Denmark | Active | 44.533 | 0 | 0 | 0 | 0 |
| 27.3.d.24 | Discards | Denmark | Passive | 0.075 | 0 | 0 | 0 | 0 |
| 27.3.d.24 | Discards | Germany | Active | 58.850 | 5 | 1275 | 5 | 221 |
| 27.3.d.24 | Discards | Germany | Passive | 2.911 | 3 | 8 | 3 | 7 |
| 27.3.d.24 | Discards | Poland | Active | 25.384 | 0 | 0 | 0 | 0 |
| 27.3.d.24 | Discards | Poland | Passive | 0.748 | 0 | 0 | 0 | 0 |
| 27.3.d.24 | Discards | Sweden | Passive | 0.068 | 0 | 0 | 0 | 0 |
| 27.3.d.24 | Landings | Denmark | Active | 165.153 | 0 | 0 | 0 | 0 |
| 27.3.d.24 | Landings | Denmark | Passive | 1.091 | 0 | 0 | 0 | 0 |
| 27.3.d.24 | Landings | Germany | Active | 146.961 | 5 | 1026 | 5 | 239 |
| 27.3.d.24 | Landings | Germany | Passive | 16.144 | 6 | 594 | 6 | 204 |
| 27.3.d.24 | Landings | Poland | Active | 73.935 | 0 | 0 | 0 | 0 |
| 27.3.d.24 | Landings | Poland | Passive | 15.055 | 0 | 0 | 0 | 0 |
| 27.3.d.24 | Landings | Sweden | Passive | 0.609 | 0 | 0 | 0 | 0 |
| 27.3.d.25 | Discards | Denmark | Active | 2.148 | 4 | 358 | 4 | 60 |
| 27.3.d.25 | Discards | Denmark | Passive | 0.062 | 0 | 0 | 0 | 0 |
| 27.3.d.25 | Discards | Germany | Active | 2.130 | 1 | 5 | 1 | 5 |
| 27.3.d.25 | Discards | Poland | Active | 15.361 | 0 | 0 | 0 | 0 |
| 27.3.d.25 | Discards | Poland | Passive | 0.567 | 0 | 0 | 0 | 0 |
| 27.3.d.25 | Discards | Sweden | Passive | 0.138 | 0 | 0 | 0 | 0 |
| 27.3.d.25 | Landings | Denmark | Active | 0.757 | 4 | 123 | 0 | 0 |
| 27.3.d.25 | Landings | Denmark | Passive | 2.143 | 0 | 0 | 0 | 0 |
| 27.3.d.25 | Landings | Germany | Active | 1.260 | 0 | 0 | 0 | 0 |
| 27.3.d.25 | Landings | Poland | Active | 60.918 | 1 | 152 | 1 | 50 |
| 27.3.d.25 | Landings | Poland | Passive | 10.246 | 0 | 0 | 0 | 0 |
| 27.3.d.25 | Landings | Sweden | Passive | 1.074 | 0 | 0 | 0 | 0 |
| 27.3.d.26 | Discards | Germany | Active | 0.000 | 0 | 0 | 0 | 0 |
| 27.3.d.26 | Discards | Poland | Passive | 0.821 | 0 | 0 | 0 | 0 |
| 27.3.d.26 | Landings | Denmark | Active | 0.000 | 0 | 0 | 0 | 0 |
| 27.3.d.26 | Landings | Denmark | Passive | 0.000 | 0 | 0 | 0 | 0 |
| 27.3.d.26 | Landings | Germany | Active | 0.479 | 0 | 0 | 0 | 0 |
| 27.3.d.26 | Landings | Lithuania | Active | 0.000 | 0 | 0 | 0 | 0 |
| 27.3.d.26 | Landings | Lithuania | Passive | 0.000 | 0 | 0 | 0 | 0 |
| 27.3.d.26 | Landings | Poland | Active | 0.000 | 0 | 0 | 0 | 0 |
| 27.3.d.26 | Landings | Poland | Passive | 2.367 | 0 | 0 | 0 | 0 |
| 27.3.d.27 | Landings | Denmark | Active | 0.000 | 0 | 0 | 0 | 0 |
| 27.3.d.27 | Landings | Denmark | Passive | 0.000 | 0 | 0 | 0 | 0 |
| 27.3.d.27 | Landings | Sweden | Passive | 0.000 | 0 | 0 | 0 | 0 |
| 27.3.d.28 | Landings | Lithuania | Active | 0.000 | 0 | 0 | 0 | 0 |
| 27.3.d.28 | Landings | Sweden | Passive | 0.000 | 0 | 0 | 0 | 0 |
| 27.3.d.29 | Landings | Sweden | Passive | 0.000 | 0 | 0 | 0 | 0 |
| 27.3.d.31 | Landings | Sweden | Active | 0.000 | 0 | 0 | 0 | 0 |
2.2 Survey Indices
This stock utilises data from four surveys. A combination of 1st quarter NS-IBTS and the 1st quarter BITS on the one hand, and the combination of 3rd quarter NS-IBTS and 4th quarter BITS on the other.
Currently, these indices are generated by Casper Berg from DTU Aqua, using a method he published in 2014. An adaptation of Casper’s method is in progress. Description of new model and updated indices need to be provided in this section once ready.
# ## 2019 Q3/4 data excluded from calculation of indices
survTun <- read.csv(file = "Data/ple.27.21-23_SurveyTuning_IBTS-BITS_1992-DataYear_2019Q34Excluded.csv",
header = T)
k <- append(k, "survTun")# === Data prep ====
survTunLng <- reshape(data = survTun, varying = list(colnames(survTun)[3:length((colnames(survTun)))]),
v.names = c("Index"), timevar = "Age", direction = "long")
survTunLng$Age <- as.factor(as.character(survTunLng$Age))
survTunLng$Year <- as.factor(as.character(survTunLng$Year))
k <- append(k, "survTunLng")
# =====
# === Figure 12 Survey Tuning Indices over time ====
set.seed(5)
plt_survTun <- ggplot() + geom_line(data = survTunLng, mapping = aes(x = Year, y = Index,
group = Age, colour = Age), size = 1.25) + facet_grid(rows = vars(Survey), scales = "free_y") +
theme_clean() + # scale_colour_manual(values = sample(ebpal, size = length(levels(survTunLng$Age)))) + theme_clean()
theme_clean() + # scale_colour_manual(values = sample(ebpal, size = length(levels(survTunLng$Age)))) + +
theme_clean() + # scale_colour_manual(values = sample(ebpal, size = length(levels(survTunLng$Age)))) + #
theme_clean() + # scale_colour_manual(values = sample(ebpal, size = length(levels(survTunLng$Age)))) + scale_colour_manual(values
theme_clean() + # scale_colour_manual(values = sample(ebpal, size = length(levels(survTunLng$Age)))) + =
theme_clean() + # scale_colour_manual(values = sample(ebpal, size = length(levels(survTunLng$Age)))) + sample(ebpal,
theme_clean() + # scale_colour_manual(values = sample(ebpal, size = length(levels(survTunLng$Age)))) + size
theme_clean() + # scale_colour_manual(values = sample(ebpal, size = length(levels(survTunLng$Age)))) + =
theme_clean() + # scale_colour_manual(values = sample(ebpal, size = length(levels(survTunLng$Age)))) + length(levels(survTunLng$Age))))
theme_clean() + # scale_colour_manual(values = sample(ebpal, size = length(levels(survTunLng$Age)))) + +
scale_colour_manual(values = ebpal) + theme(axis.text.x = element_text(angle = 90,
vjust = 1))
ggplotly(plt_survTun)Figure 2.14: NS-IBTS & BITS derived tuning indices for ple.27.21-23 Q1 and Q3/4.
# ====# include_graphics(path = paste0('../', DataYear+1,
# '/SurveyIndices/CasperCalculations/ICQ1.png'))
# === Alternate code for including multiple plots ====
include_graphics(path = paste0("../", DataYear + 1, "/SurveyIndices/CasperCalculations/",
list.files(path = paste0("../", DataYear + 1, "/SurveyIndices/CasperCalculations"),
pattern = "Q1.png")))
# =====# === Alternate code for including multiple graphics ====
xmatches <- intersect(grep(pattern = ".png", x = list.files(path = paste0("../",
DataYear + 1, "/SurveyIndices/CasperCalculations"))), grep(pattern = "2019",
x = list.files(path = paste0("../", DataYear + 1, "/SurveyIndices/CasperCalculations"))))
include_graphics(path = paste0("../", DataYear + 1, "/SurveyIndices/CasperCalculations/",
list.files(path = paste0("../", DataYear + 1, "/SurveyIndices/CasperCalculations"))[xmatches]))
# =====2.3 Biological Data
To begin with we need survey data. These can be downloaded muliple ways from DATRAS but we need the DATRAS Exchange files (HH, CA, HL). The script 01_DatrasDownloads.R downloads all exchange data for NS-IBTS and BITS surveys and saves the results as .CSV files which we can use here.
2.3.1 Reading, Subsetting and Cleaning
## Reading layer `ICES_Areas_20160601_cut_dense_3857' from data source
## `C:\Users\elbr\OneDrive - Danmarks Tekniske Universitet\Advisory Requests\ICES\WGBFAS\WKBPLAICE 2024\Data\ICES_areas'
## using driver `ESRI Shapefile'
## Simple feature collection with 66 features and 10 fields
## Geometry type: MULTIPOLYGON
## Dimension: XY
## Bounding box: xmin: -4898058 ymin: 4300621 xmax: 7625385 ymax: 30240970
## Projected CRS: WGS 84 / Pseudo-Mercator2.3.2 Maturity Ogives (MO)
The maturity ogives utilised in the assessment are the mean maturity at age from 2002 until the present data year. Values are mean proportion mature at ages one through ten.
The data used to derive these maturity ogives come from quarter one of the NS-IBTS and BITS surveys, where the subsetting of relevant areas and species, as well as the merging of the two data series is carried out in the above section. These data come from the raw DATRAS Exchange products, not the SMALK products. This is important because we are subsetting and combining data sources, so the raising procedure needs to be done independently.
This section contains hidden chunks of code that prepare the data, generate the current ogives for the assessment and plot these out in a figure for presentations/reporting.
# === First prepare maturity data for calculating MOs ---- === Select data for
# use in this script by removing rows without relevant observations
ca <- subset(ca_hh_fin_bits_ibts, Quarter == 1 & Age != -9 & Age != 0)
hl <- subset(hl_hh_fin_bits_ibts, Quarter == 1)
# Re-code the spawning status so that they are all uniform and simplified from
# maturity stage to binary
ca$spawner <- NA
ca$nonspawner <- NA
ca$spawner <- ifelse(ca$Maturity %in% c("61", "I", "1", "A", "B", "Bb"), 0, ifelse(ca$Maturity %in%
c("II", "III", "IV", "IX", "M", "62", "63", "64", "65", "2", "3", "4", "5", "Ba",
"C", "Ca", "Cb", "D", "Da", "Db", "E"), 1, "NA"))
ca$nonspawner <- ifelse(ca$Maturity %in% c("61", "I", "1", "A", "B", "Bb"), 1, ifelse(ca$Maturity %in%
c("II", "III", "IV", "IX", "M", "62", "63", "64", "65", "2", "3", "4", "5", "Ba",
"C", "Ca", "Cb", "D", "Da", "Db", "E"), 0, "NA"))
ca$spawner <- as.numeric(as.character(ca$spawner))
ca$nonspawner <- as.numeric(as.character(ca$nonspawner))
# Sort the data
ca <- ca[order(ca$Year, ca$SubDivisio, ca$Sex, ca$Age, ca$LngtClass, ca$spawner,
ca$nonspawner), ]
cax <- ca[is.na(ca$spawner) & is.na(ca$nonspawner), ]
# Calculate number of fish (mature or immature) by year+sex+age+LngtClass
maturity_lngt <- aggregate(cbind(spawner, nonspawner) ~ Year + Sex + Age + LngtClass,
data = ca, FUN = sum)
# ====
# === Calculate length frequencies from at sea observations ---- === Establish
# parameters for calculations
lencl <- c(4:60)
age <- c(1:10)
year <- c(1999:(DataYear + 1))
# Make table for length freq data
length_freq_tmp <- by(hl, list(hl$Year, round(hl$LngtClass)), function(x) {
sum(x$HLNoAtLngt_1)
})
length_freq_tab <- t(as.table(length_freq_tmp))
Lfreq_sea <- data.frame(year = rep(year, each = length(lencl)), LngtClass = rep(lencl,
length(year)), HLNoLngt = NA)
# Fill the table
for (i in c(1:nrow(Lfreq_sea))) {
year_i <- Lfreq_sea[i, names(Lfreq_sea) == as.character("year")]
lencl_i <- Lfreq_sea[i, names(Lfreq_sea) == as.character("LngtClass")]
if (lencl_i %in% c(as.numeric(dimnames(length_freq_tab)[[1]]))) {
data_i <- length_freq_tab[as.numeric(dimnames(length_freq_tab)[[1]]) == lencl_i,
dimnames(length_freq_tab)[[2]] == year_i]
Lfreq_sea[i, names(Lfreq_sea) == as.character("HLNoLngt")] <- data_i
} else {
Lfreq_sea[i, names(Lfreq_sea) == as.character("HLNoLngt")] <- 0
}
}
# Replace NA with 0
Lfreq_sea[which(is.na(Lfreq_sea$HLNoLngt)), names(Lfreq_sea) == as.character("HLNoLngt")] <- 0
# ====
# === Calculate proportions from sampled data (observations) ---- === Make a
# table for number of fish sampled, for spawner and non-spawner
nr_fish_samp_sp_F <- data.frame(year = rep(year, each = length(lencl)), lencl = rep(lencl,
times = length(year)), a1 = NA, a2 = NA, a3 = NA, a4 = NA, a5 = NA, a6 = NA,
a7 = NA, a8 = NA, a9 = NA, a10 = NA)
nr_fish_samp_nonsp_F <- nr_fish_samp_sp_F
nr_fish_samp_sp_M <- nr_fish_samp_sp_F
nr_fish_samp_nonsp_M <- nr_fish_samp_sp_F
# Fill the table with how many fish have been sampled (mature, immature) Female
for (i in 1:nrow(nr_fish_samp_sp_F)) {
for (j in 3:ncol(nr_fish_samp_sp_F)) {
lencl_i <- nr_fish_samp_sp_F[i, names(nr_fish_samp_sp_F) == as.character("lencl")]
age_i <- j - 2
year_i <- nr_fish_samp_sp_F[i, names(nr_fish_samp_sp_F) == as.character("year")]
if (lencl_i %in% maturity_lngt[which(maturity_lngt$Year == year_i & maturity_lngt$Sex ==
as.character("F") & maturity_lngt$Age == age_i), names(maturity_lngt) ==
as.character("LngtClass")]) {
show_data <- maturity_lngt[which(maturity_lngt$Year == year_i & maturity_lngt$LngtClass ==
lencl_i & maturity_lngt$Sex == as.character("F") & maturity_lngt$Age ==
age_i), ]
nr_fish_samp_sp_F[i, j] <- sum(show_data$spawner)
nr_fish_samp_nonsp_F[i, j] <- sum(show_data$nonspawner)
} else {
nr_fish_samp_sp_F[i, j] <- 0
nr_fish_samp_nonsp_F[i, j] <- 0
}
}
}
## Male
for (i in 1:nrow(nr_fish_samp_sp_M)) {
for (j in 3:ncol(nr_fish_samp_sp_M)) {
lencl_i <- nr_fish_samp_sp_M[i, names(nr_fish_samp_sp_M) == as.character("lencl")]
age_i <- j - 2
year_i <- nr_fish_samp_sp_M[i, names(nr_fish_samp_sp_M) == as.character("year")]
if (lencl_i %in% maturity_lngt[which(maturity_lngt$Year == year_i & maturity_lngt$Sex ==
as.character("M") & maturity_lngt$Age == age_i), names(maturity_lngt) ==
as.character("LngtClass")]) {
show_data <- maturity_lngt[which(maturity_lngt$Year == year_i & maturity_lngt$LngtClass ==
lencl_i & maturity_lngt$Sex == as.character("M") & maturity_lngt$Age ==
age_i), ]
nr_fish_samp_sp_M[i, j] <- sum(show_data$spawner)
nr_fish_samp_nonsp_M[i, j] <- sum(show_data$nonspawner)
} else {
nr_fish_samp_sp_M[i, j] <- 0
nr_fish_samp_nonsp_M[i, j] <- 0
}
}
}
# Total number of fish sampled (immature+mature) by length
nr_fish_sampled_tot <- nr_fish_samp_sp_F[, 1:5]
nr_fish_sampled_tot[, 3:ncol(nr_fish_sampled_tot)] <- NA
names(nr_fish_sampled_tot)[3:ncol(nr_fish_sampled_tot)] <- c("F_sp_nonsp", "M_sp_nonsp",
"F_M_tot")
for (i in 1:nrow(nr_fish_sampled_tot)) {
f_sp <- sum(nr_fish_samp_sp_F[i, 3:ncol(nr_fish_samp_sp_F)])
f_nonsp <- sum(nr_fish_samp_nonsp_F[i, 3:ncol(nr_fish_samp_nonsp_F)])
m_sp <- sum(nr_fish_samp_sp_M[i, 3:ncol(nr_fish_samp_sp_M)])
m_nonsp <- sum(nr_fish_samp_nonsp_M[i, 3:ncol(nr_fish_samp_nonsp_M)])
nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) == as.character("F_sp_nonsp")] <- f_sp +
f_nonsp
nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) == as.character("M_sp_nonsp")] <- m_sp +
m_nonsp
nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) == as.character("F_M_tot")] <- f_sp +
f_nonsp + m_sp + m_nonsp
}
# Total number of fish sampled, by age
sum_nr_sampl_age_sp_F <- data.frame(year = year, a1 = NA, a2 = NA, a3 = NA, a4 = NA,
a5 = NA, a6 = NA, a7 = NA, a8 = NA, a9 = NA, a10 = NA)
sum_nr_sampl_age_nonsp_F <- sum_nr_sampl_age_sp_F
sum_nr_sampl_age_sp_M <- sum_nr_sampl_age_sp_F
sum_nr_sampl_age_nonsp_M <- sum_nr_sampl_age_sp_F
## Female
for (i in 1:nrow(sum_nr_sampl_age_sp_F)) {
for (j in 2:ncol(sum_nr_sampl_age_sp_F)) {
i_year <- sum_nr_sampl_age_sp_F[i, names(sum_nr_sampl_age_sp_F) == as.character("year")]
sum_nr_sampl_age_sp_F[i, j] <- sum(nr_fish_samp_sp_F[which(nr_fish_samp_sp_F$year ==
i_year), names(nr_fish_samp_sp_F) == names(sum_nr_sampl_age_sp_F)[j]])
}
}
#
for (i in 1:nrow(sum_nr_sampl_age_nonsp_F)) {
for (j in 2:ncol(sum_nr_sampl_age_nonsp_F)) {
i_year <- sum_nr_sampl_age_nonsp_F[i, names(sum_nr_sampl_age_nonsp_F) ==
as.character("year")]
sum_nr_sampl_age_nonsp_F[i, j] <- sum(nr_fish_samp_nonsp_F[which(nr_fish_samp_nonsp_F$year ==
i_year), names(nr_fish_samp_nonsp_F) == names(sum_nr_sampl_age_nonsp_F)[j]])
}
}
## Male
for (i in 1:nrow(sum_nr_sampl_age_sp_M)) {
for (j in 2:ncol(sum_nr_sampl_age_sp_M)) {
i_year <- sum_nr_sampl_age_sp_M[i, names(sum_nr_sampl_age_sp_M) == as.character("year")]
sum_nr_sampl_age_sp_M[i, j] <- sum(nr_fish_samp_sp_M[which(nr_fish_samp_sp_M$year ==
i_year), names(nr_fish_samp_sp_M) == names(sum_nr_sampl_age_sp_M)[j]])
}
}
#
for (i in 1:nrow(sum_nr_sampl_age_nonsp_M)) {
for (j in 2:ncol(sum_nr_sampl_age_nonsp_M)) {
i_year <- sum_nr_sampl_age_nonsp_M[i, names(sum_nr_sampl_age_nonsp_M) ==
as.character("year")]
sum_nr_sampl_age_nonsp_M[i, j] <- sum(nr_fish_samp_nonsp_M[which(nr_fish_samp_nonsp_M$year ==
i_year), names(nr_fish_samp_nonsp_M) == names(sum_nr_sampl_age_nonsp_M)[j]])
}
}
# ====
# === Raise proportions to unsampled strata ---- === Make tables
nr_fish_sea_sp_F <- nr_fish_samp_sp_F
nr_fish_sea_sp_F[, 3:ncol(nr_fish_sea_sp_F)] <- NA
nr_fish_sea_nonsp_F <- nr_fish_sea_sp_F
nr_fish_sea_sp_M <- nr_fish_sea_sp_F
nr_fish_sea_nonsp_M <- nr_fish_sea_sp_F
# Numbers at sea by maturity class by length Female spawner
for (i in 1:nrow(nr_fish_sea_sp_F)) {
for (j in 3:ncol(nr_fish_sea_sp_F)) {
if (nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) == as.character("F_sp_nonsp")] >
0) {
mat_tot_ratio <- nr_fish_samp_sp_F[i, j]/nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) ==
as.character("F_sp_nonsp")]
female_tot_ratio <- nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) ==
as.character("F_sp_nonsp")]/nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) ==
as.character("F_M_tot")]
tot_len <- Lfreq_sea[i, names(Lfreq_sea) == as.character("HLNoLngt")]
nr_fish_sea_sp_F[i, j] <- mat_tot_ratio * female_tot_ratio * tot_len
} else {
nr_fish_sea_sp_F[i, j] <- 0
}
}
}
## Female nonspawner
for (i in 1:nrow(nr_fish_sea_nonsp_F)) {
for (j in 3:ncol(nr_fish_sea_nonsp_F)) {
if (nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) == as.character("F_sp_nonsp")] >
0) {
mat_tot_ratio <- nr_fish_samp_nonsp_F[i, j]/nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) ==
as.character("F_sp_nonsp")]
female_tot_ratio <- nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) ==
as.character("F_sp_nonsp")]/nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) ==
as.character("F_M_tot")]
tot_len <- Lfreq_sea[i, names(Lfreq_sea) == as.character("HLNoLngt")]
nr_fish_sea_nonsp_F[i, j] <- mat_tot_ratio * female_tot_ratio * tot_len
} else {
nr_fish_sea_nonsp_F[i, j] <- 0
}
}
}
## Male spawner
for (i in 1:nrow(nr_fish_sea_sp_M)) {
for (j in 3:ncol(nr_fish_sea_sp_M)) {
if (nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) == as.character("M_sp_nonsp")] >
0) {
mat_tot_ratio <- nr_fish_samp_sp_M[i, j]/nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) ==
as.character("M_sp_nonsp")]
male_tot_ratio <- nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) ==
as.character("M_sp_nonsp")]/nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) ==
as.character("F_M_tot")]
tot_len <- Lfreq_sea[i, names(Lfreq_sea) == as.character("HLNoLngt")]
nr_fish_sea_sp_M[i, j] <- mat_tot_ratio * male_tot_ratio * tot_len
} else {
nr_fish_sea_sp_M[i, j] <- 0
}
}
}
## Male non-spawner
for (i in 1:nrow(nr_fish_sea_nonsp_M)) {
for (j in 3:ncol(nr_fish_sea_nonsp_M)) {
if (nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) == as.character("M_sp_nonsp")] >
0) {
mat_tot_ratio <- nr_fish_samp_nonsp_M[i, j]/nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) ==
as.character("M_sp_nonsp")]
male_tot_ratio <- nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) ==
as.character("M_sp_nonsp")]/nr_fish_sampled_tot[i, names(nr_fish_sampled_tot) ==
as.character("F_M_tot")]
tot_len <- Lfreq_sea[i, names(Lfreq_sea) == as.character("HLNoLngt")]
nr_fish_sea_nonsp_M[i, j] <- mat_tot_ratio * male_tot_ratio * tot_len
} else {
nr_fish_sea_nonsp_M[i, j] <- 0
}
}
}
# Sum of numbers at sea, by maturity class, by ages (sum over length-groups) -
# generate tables
sum_nr_sea_age_sp_F <- sum_nr_sampl_age_sp_F
sum_nr_sea_age_sp_F[, -1] <- NA
sum_nr_sea_age_nonsp_F <- sum_nr_sea_age_sp_F
sum_nr_sea_age_sp_M <- sum_nr_sea_age_sp_F
sum_nr_sea_age_nonsp_M <- sum_nr_sea_age_sp_F
## Female spawner
for (i in 1:nrow(sum_nr_sea_age_sp_F)) {
for (j in 2:ncol(sum_nr_sea_age_sp_F)) {
i_year <- sum_nr_sea_age_sp_F[i, names(sum_nr_sea_age_sp_F) == as.character("year")]
sum_nr_sea_age_sp_F[i, j] <- sum(nr_fish_sea_sp_F[which(nr_fish_sea_sp_F$year ==
i_year), names(nr_fish_sea_sp_F) == names(sum_nr_sea_age_sp_F)[j]])
}
}
## Female non-spawner
for (i in 1:nrow(sum_nr_sea_age_nonsp_F)) {
for (j in 2:ncol(sum_nr_sea_age_nonsp_F)) {
i_year <- sum_nr_sea_age_nonsp_F[i, names(sum_nr_sea_age_nonsp_F) == as.character("year")]
sum_nr_sea_age_nonsp_F[i, j] <- sum(nr_fish_sea_nonsp_F[which(nr_fish_sea_nonsp_F$year ==
i_year), names(nr_fish_sea_nonsp_F) == names(sum_nr_sea_age_nonsp_F)[j]])
}
}
## Male spawner
for (i in 1:nrow(sum_nr_sea_age_sp_M)) {
for (j in 2:ncol(sum_nr_sea_age_sp_M)) {
i_year <- sum_nr_sea_age_sp_M[i, names(sum_nr_sea_age_sp_M) == as.character("year")]
sum_nr_sea_age_sp_M[i, j] <- sum(nr_fish_sea_sp_M[which(nr_fish_sea_sp_M$year ==
i_year), names(nr_fish_sea_sp_M) == names(sum_nr_sea_age_sp_M)[j]])
}
}
## Male non-spawner
for (i in 1:nrow(sum_nr_sea_age_nonsp_M)) {
for (j in 2:ncol(sum_nr_sea_age_nonsp_M)) {
i_year <- sum_nr_sea_age_nonsp_M[i, names(sum_nr_sea_age_nonsp_M) == as.character("year")]
sum_nr_sea_age_nonsp_M[i, j] <- sum(nr_fish_sea_nonsp_M[which(nr_fish_sea_nonsp_M$year ==
i_year), names(nr_fish_sea_nonsp_M) == names(sum_nr_sea_age_nonsp_M)[j]])
}
}
# ====
# === Calculate the sex ratio ---- ===
sex_ratio_female <- sum_nr_sea_age_sp_F
sex_ratio_female[, -1] <- NA
for (j in 2:ncol(sex_ratio_female)) {
sex_ratio_female[, j] <- (sum_nr_sea_age_nonsp_F[, j] + sum_nr_sea_age_sp_F[,
j])/(sum_nr_sea_age_nonsp_F[, j] + sum_nr_sea_age_sp_F[, j] + sum_nr_sea_age_nonsp_M[,
j] + sum_nr_sea_age_sp_M[, j])
}
# ====
# === Calculate Maturity Ogives ---- === Generate tables for female, male and
# combined ogives
mat_ogive_F <- sum_nr_sampl_age_sp_F
mat_ogive_F[, -1] <- NA
mat_ogive_M <- mat_ogive_F
mat_ogive_combsex <- mat_ogive_F
# Female mature
for (i in 1:nrow(mat_ogive_F)) {
for (j in 2:ncol(mat_ogive_F)) {
mat_ogive_F[i, j] <- sum_nr_sea_age_sp_F[i, j]/(sum_nr_sea_age_sp_F[i, j] +
sum_nr_sea_age_nonsp_F[i, j])
}
}
# Male mature
for (i in 1:nrow(mat_ogive_M)) {
for (j in 2:ncol(mat_ogive_M)) {
mat_ogive_M[i, j] <- sum_nr_sea_age_sp_M[i, j]/(sum_nr_sea_age_sp_M[i, j] +
sum_nr_sea_age_nonsp_M[i, j])
}
}
# Combsex mature
for (i in 1:nrow(mat_ogive_combsex)) {
for (j in 2:ncol(mat_ogive_combsex)) {
mat_ogive_combsex[i, j] <- (mat_ogive_F[i, j] * sex_ratio_female[i, j]) +
(mat_ogive_M[i, j] * (1 - sex_ratio_female[i, j]))
}
}
# ====
# === Combined Sexes reshape and calculate means +/- 95%CI for plotting ----
# === Complete timeseries of annual maturity ogives ----
mo <- reshape(mat_ogive_combsex, varying = colnames(mat_ogive_combsex)[-1], v.names = "PropMat",
timevar = "Age", times = colnames(mat_ogive_combsex)[-1], idvar = "year", direction = "long")
## Mean of period 2002 - DataYear
rmprop <- aggregate(PropMat ~ Age, data = mo[mo$year %in% c(2002:(DataYear + 1)),
], FUN = mean)
## Std. error of period 2002 - DataYear
rmstdr <- aggregate(PropMat ~ Age, data = mo[mo$year %in% c(2002:(DataYear + 1)),
], FUN = function(x) {
sd(x, na.rm = TRUE)/sqrt(length(na.omit(x)))
})
rmstdr$AgeN <- as.numeric(sub(pattern = "a", replacement = "", x = rmstdr$Age))
colnames(rmstdr)[colnames(rmstdr) %in% c("PropMat")] <- "StdErr"
rmstdr$PropMat <- rmprop$PropMat
rmstdr$ymax <- rmstdr$PropMat + 1.96 * rmstdr$StdErr
rmstdr$ymin <- rmstdr$PropMat - 1.96 * rmstdr$StdErr
## Combine with raw data df
rmprop$year <- as.factor(rep("Running Mean", n = nrow(rmprop)))
rownames(rmstdr) <- NULL
mo <- rbind(mo, rmprop)
mo$AgeN <- as.numeric(sub(pattern = "a", replacement = "", x = mo$Age))
mo$yearF <- as.factor(as.character(mo$year))
# =====
# === Female Only reshape and calculate means +/- 95%CI for plotting ---- ===
# Complete timeseries of annual maturity ogives ----
fmo <- reshape(mat_ogive_F, varying = colnames(mat_ogive_F)[-1], v.names = "PropMat",
timevar = "Age", times = colnames(mat_ogive_F)[-1], idvar = "year", direction = "long")
## Mean of period 2002 - DataYear ----
rfmprop <- aggregate(PropMat ~ Age, data = fmo[fmo$year %in% c(2002:(DataYear + 1)),
], FUN = mean)
## Std. error of period 2002 - DataYear ----
rfmstdr <- aggregate(PropMat ~ Age, data = fmo[fmo$year %in% c(2002:(DataYear + 1)),
], FUN = function(x) {
sd(x, na.rm = TRUE)/sqrt(length(na.omit(x)))
})
rfmstdr$AgeN <- as.numeric(sub(pattern = "a", replacement = "", x = rfmstdr$Age))
colnames(rfmstdr)[colnames(rfmstdr) %in% c("PropMat")] <- "StdErr"
rfmstdr$PropMat <- rfmprop$PropMat
rfmstdr$ymax <- rfmstdr$PropMat + 1.96 * rfmstdr$StdErr
rfmstdr$ymin <- rfmstdr$PropMat - 1.96 * rfmstdr$StdErr
## Combine with raw data df ----
rfmprop$year <- as.factor(rep("Running Mean", n = nrow(rfmprop)))
rownames(rfmstdr) <- NULL
fmo <- rbind(fmo, rfmprop)
fmo$AgeN <- as.numeric(sub(pattern = "a", replacement = "", x = fmo$Age))
fmo$yearF <- as.factor(as.character(fmo$year))
# saveRDS(fmo, file =
# 'Data/BiologicalData/Ple.27.21-32_MaturitiesFemaleLong_1999-AssYear.RDS')
# saveRDS(fmstdr,
# file('Data/BiologicalData/Ple.27.21-32_MaturitiesFemaleAveStdErr_15yr.RDS'))
# saveRDS(rfmstdr,
# file('Data/BiologicalData/Ple.27.21-32_MaturitiesFemaleAveStdErr_2002-AssYear.RDS'))
# =====
# === Male Only reshape and calculate means +/- 95%CI for plotting ---- ===
# Complete timeseries of annual maturity ogives ----
mmo <- reshape(mat_ogive_M, varying = colnames(mat_ogive_M)[-1], v.names = "PropMat",
timevar = "Age", times = colnames(mat_ogive_M)[-1], idvar = "year", direction = "long")
## Mean of period 2002 - DataYear ----
rmmprop <- aggregate(PropMat ~ Age, data = mmo[mmo$year %in% c(2002:(DataYear + 1)),
], FUN = mean)
## Std. error of period 2002 - DataYear ----
rmmstdr <- aggregate(PropMat ~ Age, data = mmo[mmo$year %in% c(2002:(DataYear + 1)),
], FUN = function(x) {
sd(x, na.rm = TRUE)/sqrt(length(na.omit(x)))
})
rmmstdr$AgeN <- as.numeric(sub(pattern = "a", replacement = "", x = rmmstdr$Age))
colnames(rmmstdr)[colnames(rmmstdr) %in% c("PropMat")] <- "StdErr"
rmmstdr$PropMat <- rmmprop$PropMat
rmmstdr$ymax <- rmmstdr$PropMat + 1.96 * rmmstdr$StdErr
rmmstdr$ymin <- rmmstdr$PropMat - 1.96 * rmmstdr$StdErr
## Combine with raw data df ----
rmmprop$year <- as.factor(rep("Running Mean", n = nrow(rmmprop)))
rownames(rmmstdr) <- NULL
mmo <- rbind(mmo, rmmprop)
mmo$AgeN <- as.numeric(sub(pattern = "a", replacement = "", x = mmo$Age))
mmo$yearF <- as.factor(as.character(mmo$year))
# saveRDS(mmo, file =
# 'Data/BiologicalData/Ple.27.21-32_MaturitiesMaleLong_1999-AssYear.RDS')
# saveRDS(mmstdr,
# file('Data/BiologicalData/Ple.27.21-32_MaturitiesMaleAveStdErr_15yr.RDS'))
# saveRDS(rmmstdr,
# file('Data/BiologicalData/Ple.27.21-32_MaturitiesMaleAveStdErr_2002-AssYear.RDS'))
# =====
## Calculate 3y sliding window average for whole timeseries ---- Calculate the
## mean ----
ada <- an(3, nrow(mat_ogive_combsex))
mo3sw <- frollmean(x = mat_ogive_combsex[, -1], n = ada, na.rm = T, align = "right",
adaptive = TRUE)
mo3sw <- as.data.frame(do.call(cbind, mo3sw))
mo3sw$year <- mat_ogive_combsex$year
#### Wide to long format
mo3swL <- reshape(mo3sw, varying = colnames(mo3sw)[-11], v.names = "PropMat", timevar = "Age",
times = colnames(mo3sw)[-11], idvar = "year", direction = "long")
mo3swL$Age <- as.numeric(gsub(pattern = "V", replacement = "", x = mo3swL$Age))
### Calcualte the standard error ----
sw3stdr <- frollapply(mat_ogive_combsex[, -1], 3, FUN = function(x) {
sd(x, na.rm = TRUE)/sqrt(length(na.omit(x)))
})
sw3stdr <- as.data.frame(do.call(cbind, sw3stdr))
sw3stdr$year <- mat_ogive_combsex$year
#### Wide to long format
sw3stdrL <- reshape(sw3stdr, varying = colnames(sw3stdr)[-11], v.names = "StdErrPropMat",
timevar = "Age", times = colnames(sw3stdr)[-11], idvar = "year", direction = "long")
sw3stdrL$Age <- as.numeric(gsub(pattern = "V", replacement = "", x = sw3stdrL$Age))
### Calculate 95% confidence intervals ----
mo3swL <- merge(x = mo3swL, y = sw3stdrL, by = c("year", "Age"), all.x = TRUE)
mo3swL$ymax <- mo3swL$PropMat + (1.96 * mo3swL$StdErrPropMat)
mo3swL$ymin <- mo3swL$PropMat - (1.96 * mo3swL$StdErrPropMat)
mo3swL$AgeN <- mo3swL$Age
mo3swL$Age <- as.character(mo3swL$Age)
mo3swL <- mo3swL[!mo3swL$year %in% c(1999:2001), ]
## Calculate 5y sliding window average for whole timeseries ---- Calculate the
## mean ----
ada <- an(5, nrow(mat_ogive_combsex))
mo5sw <- frollmean(x = mat_ogive_combsex[, -1], n = ada, na.rm = T, align = "right",
adaptive = TRUE)
mo5sw <- as.data.frame(do.call(cbind, mo5sw))
#### Make the first year equal to the first calculable 5y average
mo5sw$year <- mat_ogive_combsex$year
# mo5sw[mo5sw$year == 2002, -ncol(mo5sw)] <- mo5sw[mo5sw$year == 2003,
# -ncol(mo5sw)]
mo5sw <- mo5sw[!mo5sw$year %in% c(1999:2001), ]
#### Wide to long format
mo5swL <- reshape(mo5sw, varying = colnames(mo5sw)[-11], v.names = "PropMat", timevar = "Age",
times = colnames(mo5sw)[-11], idvar = "year", direction = "long")
mo5swL$Age <- as.numeric(gsub(pattern = "V", replacement = "", x = mo5swL$Age))
### Calcualte the standard error ----
sw5stdr <- frollapply(mat_ogive_combsex[, -1], 5, FUN = function(x) {
sd(x, na.rm = TRUE)/sqrt(length(na.omit(x)))
})
sw5stdr <- as.data.frame(do.call(cbind, sw5stdr))
sw5stdr$year <- mat_ogive_combsex$year
#### Make the first year equal to the first calculable 5y average
sw5stdr[sw5stdr$year == 2002, -ncol(sw5stdr)] <- sw5stdr[sw5stdr$year == 2003, -ncol(sw5stdr)]
sw5stdr <- sw5stdr[!sw5stdr$year %in% c(1999:2001), ]
#### Wide to long format
sw5stdrL <- reshape(sw5stdr, varying = colnames(sw5stdr)[-11], v.names = "StdErrPropMat",
timevar = "Age", times = colnames(sw5stdr)[-11], idvar = "year", direction = "long")
sw5stdrL$Age <- as.numeric(gsub(pattern = "V", replacement = "", x = sw5stdrL$Age))
### Calculate 95% confidence intervals ----
mo5swL <- merge(x = mo5swL, y = sw5stdrL, by = c("year", "Age"), all.x = TRUE)
mo5swL$ymax <- mo5swL$PropMat + (1.96 * mo5swL$StdErrPropMat)
mo5swL$ymin <- mo5swL$PropMat - (1.96 * mo5swL$StdErrPropMat)
mo5swL$AgeN <- mo5swL$Age
mo5swL$Age <- as.character(mo5swL$Age)
mo5swL <- mo5swL[!mo5swL$year %in% c(1999:2001), ]
## Calculate annual values from a smoothing spline ---- Build the spline
mo2 <- mo
mo2$Age <- as.factor(mo2$Age)
mo2$year <- as.numeric(mo2$year)
moss <- gam(formula = PropMat ~ s(year, bs = "cr", by = Age), data = mo2, family = gaussian(link = "identity"))
### Predict with the smoothed spline
predmo <- predict.gam(moss, newdata = mo2, type = "response", se.fit = TRUE)
mo2$smoothPropMat <- unname(predmo$fit)
mo2$seSmoothPropMat <- unname(predmo$se.fit)
# ====
# === Clean Data Environment ====
k <- append(k, c("mo", "mstdr", "rmstdr", "mat_ogive_F", "mat_ogive_M", "mat_ogive_combsex",
"mo3swL", "mo5swL", "mo2", "sex_ratio_female"))
rm(list = ls()[!ls() %in% k])
# =====There are various ways of looking a the Maturity Ogives. The previous Ple.27.21-23 stock used running means from the whole time-series to get single values of maturity at age, which were then applied to every year in the assessment.
# kable(mo[mo$year == "Running Mean", c("AgeN", "PropMat") ],
rmstdrder <- rmstdr[order(rmstdr$AgeN), c("AgeN", "PropMat") ]
rownames(rmstdrder) <- NULL
### save_as_lowestoft Mean 2002-DataYear
mmA <- data.frame(lapply(colMeans(mat_ogive_combsex[mat_ogive_combsex$year %in% c(2002:DataYear), !colnames(mat_ogive_combsex) %in% ("year")], na.rm = TRUE), type.convert), stringsAsFactors=FALSE)
save_as_lowestoft(df = mmA, #[, !colnames(mWAA) %in% "Age"],
file_path = "Assessment Input/mo.dat",
title = "Stock weights at age for Ple.27.21-32. Mean values for combined sexes 2002 to Data year, from survey data",
rando_numbers = paste(c(1, 4), collapse = "\t"),
year_range = paste(c(2002, DataYear), collapse = "\t"),
age_range = paste(c(1, 10), collapse = "\t"),
timeseries_type = 2)
kable(rmstdrder,
caption = paste0("Maturity Ogives of Ple.27.21-32, using the stock annex method from the previous Ple.27.21-23. Values are timeseries means of both sexes for the period 2002-", DataYear),
digits = 2)| AgeN | PropMat |
|---|---|
| 1 | 0.30 |
| 2 | 0.70 |
| 3 | 0.84 |
| 4 | 0.91 |
| 5 | 0.95 |
| 6 | 0.97 |
| 7 | 0.98 |
| 8 | 0.98 |
| 9 | 0.98 |
| 10 | 0.94 |
An alternate approach is to use smoothed or sliding window averages to allow for change in the age at maturity over time (different values for every year in the assessment), while smoothing out some of the noise introduced from annual observations of maturity. This noise can be significant due to the relatively low sample sizes of maturity level, relative to the population.
mo3order <- mo3swL[order(mo3swL$year, mo3swL$AgeN), c("year", "AgeN", "PropMat")]
rownames(mo3order) <- NULL
mo3orderw <- dcast(mo3order, year ~ AgeN, value.var = "PropMat")
## Save maturity data for assessment input ### Make 'NA's into '-9'
## mo3orderwSAM <- mo3orderw[mo3orderw$year %in% c(2002:(DataYear+1)), ] for(i
## in 1:ncol(mo3orderwSAM)){ for(j in 1:nrow(mo3orderwSAM)){ mo3orderwSAM[j, i]
## <- ifelse(is.na(mo3orderwSAM[j, i]), -9, mo3orderwSAM[j, i]) } }
### Make early timeseries 'NA's into plausible value
mo3orderwSAM <- mo3orderw[mo3orderw$year %in% c(2002:(DataYear + 1)), ]
for (i in 1:ncol(mo3orderwSAM)) {
for (j in 1:nrow(mo3orderwSAM)) {
mo3orderwSAM[j, i] <- ifelse(is.na(mo3orderwSAM[j, i]), median(c(mo3orderwSAM[,
i]), na.rm = T), mo3orderwSAM[j, i])
}
}
save_as_lowestoft(mo3orderwSAM[, !colnames(mo3orderwSAM) %in% "year"], file_path = "Assessment Input/AlternateModel_3ysw/mo.dat",
title = paste0("Combined sex maturity ogives for ple.27.21-32. Three year sliding window means from 2002:",
DataYear + 1), rando_numbers = paste(c("1", "6", "#I don't know what this information stands for, it's just made up."),
collapse = "\t"), year_range = paste(c(2002, (DataYear + 1)), collapse = "\t"),
age_range = paste(c(1, 10), collapse = "\t"), timeseries_type = paste(c(1, "# 1 indicates annually varying, 2 indicates fixed values"),
collapse = "\t"))
kable(mo3orderw, caption = paste0("Maturity Ogives of Ple.27.21-32, using a 3 year sliding window mean approach. Values are timeseries means of the whole timeseries, 2002:",
DataYear + 1), digits = 2)| year | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
|---|---|---|---|---|---|---|---|---|---|---|
| 2002 | 0.00 | 0.88 | 0.94 | 0.97 | 0.97 | 1.00 | 1.00 | NaN | NaN | NaN |
| 2003 | 0.37 | 0.80 | 0.88 | 0.96 | 0.97 | 1.00 | 1.00 | NaN | NaN | NaN |
| 2004 | 0.38 | 0.75 | 0.81 | 0.93 | 0.97 | 1.00 | 1.00 | NaN | NaN | NaN |
| 2005 | 0.38 | 0.59 | 0.74 | 0.91 | 0.96 | 0.99 | 1.00 | 1.00 | NaN | NaN |
| 2006 | 0.22 | 0.61 | 0.71 | 0.90 | 0.96 | 0.94 | 0.91 | 1.00 | 1.00 | NaN |
| 2007 | 0.08 | 0.41 | 0.61 | 0.83 | 0.89 | 0.94 | 0.91 | 1.00 | 1.00 | 1.00 |
| 2008 | 0.17 | 0.53 | 0.66 | 0.83 | 0.89 | 0.93 | 0.89 | 0.98 | 0.98 | 1.00 |
| 2009 | 0.14 | 0.45 | 0.69 | 0.84 | 0.90 | 0.98 | 0.98 | 0.98 | 0.98 | 0.99 |
| 2010 | 0.39 | 0.63 | 0.82 | 0.89 | 0.95 | 0.98 | 0.98 | 0.98 | 0.97 | 0.99 |
| 2011 | 0.63 | 0.71 | 0.86 | 0.90 | 0.93 | 0.99 | 1.00 | 1.00 | 1.00 | 0.99 |
| 2012 | 0.52 | 0.82 | 0.91 | 0.91 | 0.94 | 0.99 | 1.00 | 1.00 | 1.00 | 1.00 |
| 2013 | 0.35 | 0.76 | 0.90 | 0.91 | 0.96 | 0.99 | 0.99 | 1.00 | 0.97 | 1.00 |
| 2014 | 0.16 | 0.70 | 0.83 | 0.89 | 0.98 | 0.99 | 0.99 | 1.00 | 0.97 | 1.00 |
| 2015 | 0.23 | 0.65 | 0.81 | 0.88 | 0.96 | 0.97 | 0.99 | 0.94 | 0.97 | 1.00 |
| 2016 | 0.40 | 0.75 | 0.83 | 0.89 | 0.93 | 0.94 | 0.99 | 0.92 | 0.97 | 0.94 |
| 2017 | 0.50 | 0.82 | 0.90 | 0.92 | 0.94 | 0.95 | 0.99 | 0.92 | 0.97 | 0.94 |
| 2018 | 0.46 | 0.86 | 0.94 | 0.95 | 0.96 | 0.97 | 0.99 | 0.97 | 0.97 | 0.94 |
| 2019 | 0.42 | 0.86 | 0.95 | 0.97 | 0.99 | 0.98 | 1.00 | 1.00 | 1.00 | 1.00 |
| 2020 | 0.32 | 0.76 | 0.95 | 0.98 | 0.99 | 0.99 | 1.00 | 1.00 | 1.00 | 1.00 |
| 2021 | 0.22 | 0.72 | 0.95 | 0.96 | 0.99 | 0.99 | 0.99 | 1.00 | 1.00 | 0.95 |
| 2022 | 0.16 | 0.71 | 0.95 | 0.95 | 0.98 | 0.98 | 0.98 | 1.00 | 0.98 | 0.73 |
| 2023 | 0.15 | 0.75 | 0.93 | 0.94 | 0.95 | 0.97 | 0.94 | 0.98 | 0.94 | 0.82 |
| 2024 | 0.18 | 0.74 | 0.92 | 0.94 | 0.94 | 0.95 | 0.95 | 0.96 | 0.94 | 0.75 |
mo5order <- mo5swL[order(mo5swL$year, mo5swL$AgeN), c("year", "AgeN", "PropMat")]
rownames(mo5order) <- NULL
mo5orderw <- dcast(mo5order, year ~ AgeN, value.var = "PropMat")
# ## Save maturity data for assessment input mo5orderwSAM <-
# mo5orderw[mo5orderw$year %in% c(2002:(DataYear+1)), ] for(i in
# 1:ncol(mo5orderwSAM)){ for(j in 1:nrow(mo5orderwSAM)){ mo5orderwSAM[j, i] <-
# ifelse(is.na(mo5orderwSAM[j, i]), -9, mo5orderwSAM[j, i]) } }
### Make early timeseries 'NA's into plausible value
mo5orderwSAM <- mo5orderw[mo5orderw$year %in% c(2002:(DataYear + 1)), ]
for (i in 1:ncol(mo5orderwSAM)) {
for (j in 1:nrow(mo5orderwSAM)) {
mo5orderwSAM[j, i] <- ifelse(is.na(mo5orderwSAM[j, i]), median(c(mo5orderwSAM[,
i]), na.rm = T), mo5orderwSAM[j, i])
}
}
save_as_lowestoft(mo5orderwSAM[, !colnames(mo5orderwSAM) %in% "year"], file_path = "Assessment Input/AlternateModel_5ysw/mo.dat",
title = paste0("Combined sex maturity ogives for ple.27.21-32. Five year sliding window means from 2002:",
DataYear + 1), rando_numbers = paste(c("1", "6", "#I don't know what this information stands for, it's just made up."),
collapse = "\t"), year_range = paste(c(2002, (DataYear + 1)), collapse = "\t"),
age_range = paste(c(1, 10), collapse = "\t"), timeseries_type = paste(c(1, "# 1 indicates annually varying, 2 indicates fixed values"),
collapse = "\t"))
kable(mo5orderw, caption = paste0("Maturity Ogives of Ple.27.21-32, using a 5 year sliding window mean approach. Values are timeseries means of the whole timeseries, 2002:",
DataYear + 1), digits = 2)| year | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
|---|---|---|---|---|---|---|---|---|---|---|
| 2002 | 0.00 | 0.88 | 0.94 | 0.97 | 0.97 | 1.00 | 1.00 | NaN | NaN | NaN |
| 2003 | 0.37 | 0.80 | 0.88 | 0.96 | 0.97 | 1.00 | 1.00 | NaN | NaN | NaN |
| 2004 | 0.38 | 0.80 | 0.86 | 0.95 | 0.98 | 1.00 | 1.00 | NaN | NaN | NaN |
| 2005 | 0.29 | 0.70 | 0.82 | 0.94 | 0.97 | 0.99 | 1.00 | 1.00 | NaN | NaN |
| 2006 | 0.28 | 0.65 | 0.75 | 0.91 | 0.96 | 0.96 | 0.95 | 1.00 | 1.00 | NaN |
| 2007 | 0.28 | 0.54 | 0.68 | 0.87 | 0.92 | 0.96 | 0.95 | 1.00 | 1.00 | 1.00 |
| 2008 | 0.19 | 0.54 | 0.68 | 0.86 | 0.92 | 0.95 | 0.94 | 0.98 | 0.98 | 1.00 |
| 2009 | 0.13 | 0.47 | 0.68 | 0.86 | 0.92 | 0.95 | 0.94 | 0.99 | 0.99 | 0.99 |
| 2010 | 0.26 | 0.56 | 0.72 | 0.86 | 0.91 | 0.96 | 0.94 | 0.99 | 0.99 | 0.99 |
| 2011 | 0.38 | 0.60 | 0.78 | 0.86 | 0.90 | 0.98 | 0.99 | 0.99 | 0.99 | 0.99 |
| 2012 | 0.46 | 0.72 | 0.86 | 0.91 | 0.94 | 0.98 | 0.99 | 0.99 | 0.99 | 0.99 |
| 2013 | 0.39 | 0.70 | 0.87 | 0.90 | 0.95 | 1.00 | 0.99 | 1.00 | 0.98 | 0.99 |
| 2014 | 0.35 | 0.75 | 0.86 | 0.89 | 0.95 | 0.99 | 0.99 | 1.00 | 0.98 | 1.00 |
| 2015 | 0.35 | 0.73 | 0.86 | 0.90 | 0.95 | 0.98 | 0.99 | 0.97 | 0.98 | 1.00 |
| 2016 | 0.30 | 0.73 | 0.85 | 0.89 | 0.96 | 0.97 | 0.99 | 0.95 | 0.97 | 0.96 |
| 2017 | 0.34 | 0.75 | 0.86 | 0.90 | 0.95 | 0.96 | 0.99 | 0.95 | 0.97 | 0.96 |
| 2018 | 0.41 | 0.80 | 0.88 | 0.92 | 0.95 | 0.96 | 0.99 | 0.95 | 0.98 | 0.96 |
| 2019 | 0.45 | 0.82 | 0.92 | 0.94 | 0.96 | 0.96 | 0.99 | 0.95 | 0.98 | 0.96 |
| 2020 | 0.39 | 0.81 | 0.94 | 0.96 | 0.97 | 0.97 | 0.99 | 0.98 | 0.98 | 0.95 |
| 2021 | 0.31 | 0.78 | 0.95 | 0.97 | 0.99 | 0.99 | 1.00 | 1.00 | 1.00 | 0.97 |
| 2022 | 0.25 | 0.76 | 0.95 | 0.96 | 0.98 | 0.98 | 0.99 | 1.00 | 0.99 | 0.86 |
| 2023 | 0.20 | 0.74 | 0.94 | 0.95 | 0.97 | 0.98 | 0.97 | 0.99 | 0.96 | 0.86 |
| 2024 | 0.17 | 0.71 | 0.93 | 0.95 | 0.96 | 0.97 | 0.97 | 0.98 | 0.96 | 0.79 |
We could also attempt to fit smoothers to these data, in place of the sliding window means. However, instead of doing this in data prep, we can simply provide the raw annual estimates to SAM, and estimate this smoother during the model fit.
moRawOrder <- mat_ogive_combsex[mat_ogive_combsex$year %in% c(2002:(DataYear + 1)),
]
rownames(moRawOrder) <- NULL
# ## Save maturity data for assessment input moRawOrderSAM <-
# moRawOrder[moRawOrder$year %in% c(2002:(DataYear+1)), ] for(i in
# 1:ncol(moRawOrderSAM)){ for(j in 1:nrow(moRawOrderSAM)){ moRawOrderSAM[j, i]
# <- ifelse(is.na(moRawOrderSAM[j, i]), -9, moRawOrderSAM[j, i]) } }
### Make early timeseries 'NA's into plausible value
moRawOrderSAM <- moRawOrder[moRawOrder$year %in% c(2002:(DataYear + 1)), ]
for (i in 1:ncol(moRawOrderSAM)) {
for (j in 1:nrow(moRawOrderSAM)) {
moRawOrderSAM[j, i] <- ifelse(is.na(moRawOrderSAM[j, i]), median(c(moRawOrderSAM[,
i]), na.rm = T), moRawOrderSAM[j, i])
}
}
save_as_lowestoft(moRawOrderSAM[, !colnames(moRawOrderSAM) %in% "year"], file_path = "Assessment Input/AlternateModel_AnnualRawBio/mo.dat",
title = paste0("Combined sex maturity ogives for ple.27.21-32. Annual estimates from 2002:",
DataYear + 1), rando_numbers = paste(c("1", "6", "#I don't know what this information stands for, it's just made up."),
collapse = "\t"), year_range = paste(c(2002, (DataYear + 1)), collapse = "\t"),
age_range = paste(c(1, 10), collapse = "\t"), timeseries_type = paste(c(1, "# 1 indicates annually varying, 2 indicates fixed values"),
collapse = "\t"))
kable(moRawOrder, caption = paste0("Maturity Ogives of Ple.27.21-32, using annual estimates from surveys for the whole timeseries, 2002:",
DataYear + 1), digits = 2)| year | a1 | a2 | a3 | a4 | a5 | a6 | a7 | a8 | a9 | a10 |
|---|---|---|---|---|---|---|---|---|---|---|
| 2002 | 0.00 | 0.80 | 0.87 | 0.94 | 0.94 | 1.00 | 1.00 | NaN | NaN | NaN |
| 2003 | 0.74 | 0.63 | 0.77 | 0.93 | 0.96 | 1.00 | 1.00 | NaN | NaN | NaN |
| 2004 | 0.41 | 0.82 | 0.78 | 0.92 | 1.00 | 1.00 | 1.00 | NaN | NaN | NaN |
| 2005 | 0.00 | 0.30 | 0.67 | 0.89 | 0.93 | 0.97 | 1.00 | 1.00 | NaN | NaN |
| 2006 | 0.25 | 0.70 | 0.68 | 0.89 | 0.95 | 0.85 | 0.74 | 1.00 | 1.00 | NaN |
| 2007 | 0.00 | 0.23 | 0.49 | 0.71 | 0.78 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| 2008 | 0.27 | 0.66 | 0.81 | 0.90 | 0.94 | 0.93 | 0.94 | 0.93 | 0.94 | 0.99 |
| 2009 | NaN | 0.46 | 0.77 | 0.91 | 0.98 | 1.00 | 1.00 | 1.00 | 1.00 | 0.97 |
| 2010 | 0.50 | 0.76 | 0.87 | 0.88 | 0.92 | 1.00 | 1.00 | 1.00 | NaN | 1.00 |
| 2011 | 0.76 | 0.90 | 0.94 | 0.91 | 0.89 | 0.98 | 1.00 | 1.00 | 1.00 | 1.00 |
| 2012 | 0.30 | 0.81 | 0.92 | 0.94 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| 2013 | 0.00 | 0.57 | 0.84 | 0.87 | 0.99 | 1.00 | 0.97 | 1.00 | 0.92 | 1.00 |
| 2014 | 0.17 | 0.71 | 0.73 | 0.87 | 0.95 | 0.97 | 1.00 | 1.00 | 1.00 | 1.00 |
| 2015 | 0.53 | 0.68 | 0.86 | 0.89 | 0.93 | 0.94 | 0.99 | 0.83 | 1.00 | 1.00 |
| 2016 | 0.49 | 0.86 | 0.90 | 0.90 | 0.90 | 0.92 | 0.97 | 0.94 | 0.91 | 0.81 |
| 2017 | 0.49 | 0.91 | 0.95 | 0.96 | 0.98 | 0.98 | 1.00 | 1.00 | 1.00 | 1.00 |
| 2018 | 0.40 | 0.82 | 0.95 | 0.99 | 0.98 | 1.00 | 1.00 | 0.99 | 1.00 | 1.00 |
| 2019 | 0.37 | 0.85 | 0.95 | 0.96 | 0.99 | 0.98 | 1.00 | 1.00 | 1.00 | 1.00 |
| 2020 | 0.19 | 0.61 | 0.96 | 0.97 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | NaN |
| 2021 | 0.11 | 0.70 | 0.94 | 0.95 | 0.98 | 1.00 | 0.98 | 1.00 | 1.00 | 0.89 |
| 2022 | 0.17 | 0.81 | 0.94 | 0.94 | 0.96 | 0.94 | 0.95 | 1.00 | 0.94 | 0.56 |
| 2023 | 0.18 | 0.73 | 0.92 | 0.94 | 0.91 | 0.98 | 0.89 | 0.94 | 0.87 | 1.00 |
| 2024 | 0.20 | 0.68 | 0.89 | 0.95 | 0.95 | 0.95 | 1.00 | 0.95 | 1.00 | 0.70 |
plt_mo_rm <- ggplot() +
geom_line(data = mo[!mo$year %in% c("Running Mean", DataYear) & as.numeric(as.character(mo$year)) >= 2002 & as.numeric(as.character(mo$year)) != (DataYear+1), ],
mapping = aes(x = AgeN,
y = PropMat,
group = yearF,
colour = yearF)) +
geom_line(data = mo[mo$yearF == "Running Mean",],
mapping = aes(x = AgeN,
y = PropMat,
group = yearF),
colour = "Black",
linetype = 3,
size = 1.5) +
geom_errorbar(data = rmstdr,
mapping = aes(x = AgeN,
ymax = ymax,
ymin = ymin),
colour = "Black",
linetype = 3,
size = 1) +
geom_line(data = mo[mo$yearF == DataYear, ],
mapping = aes(x = AgeN,
y = PropMat,
group = yearF),
colour = "Black",
# linetype = 9,
size = 1.2) +
theme_clean()+
# scale_color_manual(values = ebpal) +
scale_x_continuous(breaks = function(x) pretty(x, n = 5))
ggplotly(plt_mo_rm)Figure 2.15: Maturity ogives for individual previous years (colours), the running mean based on Ple.27.21-23 stock annex (black dotted +/- 95% CI), and for 2023 (solid black)
Utilising an average of maturity ogives from the whole data series seems to mask any recent developments in rates of maturity. Therefore we were adivised by WGBFAS and by the ADG to consider switching to a sliding window approach. Below we see plots similar to the above, but this time we substitute in the results for annually varying, 3 and 5 year sliding window means.
# === Figure showing 3 year sliding window MO vs others ==== Make ordered
# mo3swL, with ymax max 1
mo3swLO <- mo3swL[order(mo3swL$year, mo3swL$AgeN), c("year", "AgeN", "PropMat", "ymax",
"ymin")]
mo3swLO$ymax <- ifelse(mo3swLO$ymax > 1, 1, mo3swLO$ymax)
mo3swLO$ymin <- ifelse(mo3swLO$ymin < 0, 0, mo3swLO$ymin)
mo3swLO$yearF <- as.factor(as.character(mo3swLO$year))
plt_mo3swL <- ggplot() + geom_line(data = mo3swLO, mapping = aes(x = AgeN, y = PropMat,
group = yearF, colour = yearF)) + geom_ribbon(data = mo3swLO, mapping = aes(x = AgeN,
ymax = ymax, ymin = ymin, group = yearF, fill = yearF), alpha = 0.05) + geom_line(data = mo[mo$yearF ==
"Running Mean", ], mapping = aes(x = AgeN, y = PropMat, group = yearF), colour = "Black",
linetype = 3, size = 1.5) + geom_errorbar(data = rmstdr, mapping = aes(x = AgeN,
ymax = ymax, ymin = ymin), colour = "Black", linetype = 3, size = 1) + theme_clean() +
scale_x_continuous(breaks = function(x) pretty(x, n = 5))
ggplotly(plt_mo3swL)Figure 2.16: Maturity ogives for three year sliding window means applied as annually varying maturity values. The running mean utilised in the Ple.27.21-23 annex is overlain for comparison (black dotted +/- 95% CI).
# =====# === Figure showing 5 year sliding window MO vs others ==== Make ordered
# mo5swL, with ymax max 1
mo5swLO <- mo5swL[order(mo5swL$year, mo5swL$AgeN), c("year", "AgeN", "PropMat", "ymax",
"ymin")]
mo5swLO$ymax <- ifelse(mo5swLO$ymax > 1, 1, mo5swLO$ymax)
mo5swLO$ymin <- ifelse(mo5swLO$ymin < 0, 0, mo5swLO$ymin)
mo5swLO$yearF <- as.factor(as.character(mo5swLO$year))
plt_mo5swL <- ggplot() + geom_line(data = mo5swLO, mapping = aes(x = AgeN, y = PropMat,
group = yearF, colour = yearF)) + geom_ribbon(data = mo5swLO, mapping = aes(x = AgeN,
ymax = ymax, ymin = ymin, group = yearF, fill = yearF), alpha = 0.05) + geom_line(data = mo[mo$yearF ==
"Running Mean", ], mapping = aes(x = AgeN, y = PropMat, group = yearF), colour = "Black",
linetype = 3, size = 1.5) + geom_errorbar(data = rmstdr, mapping = aes(x = AgeN,
ymax = ymax, ymin = ymin), colour = "Black", linetype = 3, size = 1) + theme_clean() +
scale_x_continuous(breaks = function(x) pretty(x, n = 5))
ggplotly(plt_mo5swL)Figure 2.17: Maturity ogives for five year sliding window means applied as annually varying maturity values. The running mean utilised in the Ple.27.21-23 annex is overlain for comparison (black dotted +/- 95% CI).
# =====mo2w <- reshape(mo2[, c("year", "Age", "smoothPropMat")], v.names = "smoothPropMat",
timevar = "Age", idvar = "year", direction = "wide")
# for(i in 2:ncol(mo2w)){ for(j in 2:nrow(mo2w)){ mo2w[j, i] <-
# ifelse(is.na(mo2w[j, i]), -9, mo2w[j, i]) mo2w[j, i] <- ifelse((mo2w[j,
# i]>1), 1, mo2w[j, i]) } }
### Make early timeseries 'NA's into plausible value
mo2wSAM <- mo2w[mo2w$year %in% c(2002:(DataYear + 1)), ]
for (i in 1:ncol(mo2wSAM)) {
for (j in 1:nrow(mo2wSAM)) {
mo2wSAM[j, i] <- ifelse(is.na(mo2wSAM[j, i]), median(c(mo2wSAM[, i]), na.rm = T),
mo2wSAM[j, i])
mo2w[j, i] <- ifelse((mo2w[j, i] > 1), 1, mo2w[j, i])
}
}
rownames(mo2w) <- NULL
save_as_lowestoft(df = mo2w[mo2w$year %in% c(2002:(DataYear + 1)), ], file_path = "Assessment Input/AlternateModel_SSbio/mo.dat",
title = "Maturity Ogive for Ple.27.21-32. Smooth Spline for combined sexes, from survey data",
rando_numbers = paste(c(1, 4), collapse = "\t"), year_range = paste(c(2002, (DataYear +
1)), collapse = "\t"), age_range = paste(c(1, 10), collapse = "\t"))
kable(mo2w[mo2w$year >= 2002, ], caption = paste0("Maturity Ogives of Ple.27.21-32, using spline smoothed estimates from survey observations for the whole timeseries, 2002:",
DataYear + 1), digits = 2)| year | smoothPropMat.a1 | smoothPropMat.a2 | smoothPropMat.a3 | smoothPropMat.a4 | smoothPropMat.a5 | smoothPropMat.a6 | smoothPropMat.a7 | smoothPropMat.a8 | smoothPropMat.a9 | smoothPropMat.a10 | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 24 | 2022 | 0.6747897 | 0.9124604 | 0.9262226 | 0.8663733 | 0.8484428 | 0.8341142 | 0.8394752 | 0.8865159 | 0.9073891 | 0.7925157 |
| 25 | 2023 | 0.6922145 | 0.9190812 | 0.9344047 | 0.8677655 | 0.8478008 | 0.8318467 | 0.8378158 | 0.8901934 | 0.9131888 | 0.7712150 |
| 26 | 2024 | 0.7279639 | 0.9257020 | 0.9425868 | 0.8691578 | 0.8471588 | 0.8295791 | 0.8361565 | 0.8938708 | 0.9189848 | 0.7497751 |
| NA | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA | NA |
# swalba <- melt(setDT(WAA_combsex), id.vars = "year", variable.name = "Age", value.name = "MeanWeight")
ggplotly(ggplot() +
geom_line(data = mo2,
mapping = aes(x = year,
y = PropMat)) +
geom_line(data = mo2,
mapping = aes(x = year,
y = smoothPropMat,
colour = Age)) +
geom_ribbon(data = mo2,
mapping = aes(x = year,
ymin = smoothPropMat-(1.96*seSmoothPropMat),
ymax = smoothPropMat+(1.96*seSmoothPropMat),
# colour = Age,
fill = Age),
alpha = 0.15) +
facet_wrap(.~Age)+
theme_few()+
theme(axis.text.x = element_text(angle = 45, hjust = 1, vjust = 1)) +
scale_x_continuous(breaks = function(x) pretty(x, n = 5)) +
# scale_y_continuous(breaks = c(0, 0.2, 0.4, 0.6)) +
# coord_cartesian(ylim = c(0,0.95)) +
scale_color_manual(values = c(ebpal)) +
scale_fill_manual(values = c(ebpal))
)Figure 2.18: Maturity Ogives, actual values (black) and predicted from smoother (colours); ribbons = 95% CI.
In plenary at the Benchmark Workshop, it was decided that the levels of maturity in the early ages of the data do not look appropriate. This could be due to bias in teh sampling of the surveys, where larger fish from a given age group are more likely to be sampled in the deeper operating areas of the survey vessels. This may also be due to the difficulty in establishing the first winter ring when aging, leading to some larger two to three year olds being read as a year younger than they are.
Because of this we can produce some ogives with ages 1, or 1 & 2 set to zero, based on expert opinion.
mo5swL_1zero <- mo5swL
mo5swL_1zero[mo5swL_1zero$AgeN %in% c(1), c("PropMat", "StdErrPropMat", "ymax", "ymin")] <- 0
mo5swL_1zero$yearF <- as.factor(as.character(mo5swL_1zero$year))
# mo5swL_1zero$Age <- factor(as.character(mo5swL_1zero$AgeN), levels =
# as.character(sort(unique(mo5swL_1zero$AgeN))))
mo5swW_1zero <- reshape(mo5swL_1zero[, c("year", "Age", "PropMat")], timevar = "Age",
idvar = "year", direction = "wide")
colnames(mo5swW_1zero) <- gsub(pattern = "PropMat.", replacement = "", x = colnames(mo5swW_1zero))
### Make early timeseries 'NA's into plausible value
mo5swW_1zeroSAM <- mo5swW_1zero[mo5swW_1zero$year %in% c(2002:(DataYear + 1)), c(-1)]
mo5swW_1zeroSAM <- mo5swW_1zeroSAM[, as.character(sort(as.numeric(names(mo5swW_1zeroSAM))))]
for (i in 1:ncol(mo5swW_1zeroSAM)) {
for (j in 1:nrow(mo5swW_1zeroSAM)) {
mo5swW_1zeroSAM[j, i] <- ifelse(is.na(mo5swW_1zeroSAM[j, i]), median(c(mo5swW_1zeroSAM[,
i]), na.rm = T), mo5swW_1zeroSAM[j, i])
}
}
save_as_lowestoft(df = mo5swW_1zeroSAM, file_path = "Assessment Input/MaturityVariants/5ysw_1zeromo.dat",
title = "Maturity Ogive for Ple.27.21-32. Five year sliding window with age 1 set to 0, combined sexes, from survey data",
rando_numbers = paste(c(1, 4), collapse = "\t"), year_range = paste(c(2002, (DataYear +
1)), collapse = "\t"), age_range = paste(c(1, 10), collapse = "\t"), timeseries_type = 1)
plt_mo5swL_1zero <- ggplot() + geom_line(data = mo5swL_1zero, mapping = aes(x = AgeN,
y = PropMat, group = yearF, colour = yearF)) + geom_ribbon(data = mo5swL_1zero,
mapping = aes(x = AgeN, ymax = ymax, ymin = ymin, group = yearF, fill = yearF),
alpha = 0.05) + geom_line(data = mo[mo$yearF == "Running Mean", ], mapping = aes(x = AgeN,
y = PropMat, group = yearF), colour = "Black", linetype = 3, size = 1.5) + geom_errorbar(data = rmstdr,
mapping = aes(x = AgeN, ymax = ymax, ymin = ymin), colour = "Black", linetype = 3,
size = 1) + theme_clean() + scale_x_continuous(breaks = function(x) pretty(x,
n = 5))
ggplotly(plt_mo5swL_1zero)Figure 2.19: Five-year sliding window mean of maturity at age, with age 1 set to 0, post-hoc.
mo5swL_2zero <- mo5swL
mo5swL_2zero[mo5swL_2zero$AgeN %in% c(1, 2), c("PropMat", "StdErrPropMat", "ymax",
"ymin")] <- 0
mo5swL_2zero$yearF <- as.factor(as.character(mo5swL_2zero$year))
plt_mo5swL_2zero <- ggplot() + geom_line(data = mo5swL_2zero, mapping = aes(x = AgeN,
y = PropMat, group = yearF, colour = yearF)) + geom_ribbon(data = mo5swL_2zero,
mapping = aes(x = AgeN, ymax = ymax, ymin = ymin, group = yearF, fill = yearF),
alpha = 0.05) + geom_line(data = mo[mo$yearF == "Running Mean", ], mapping = aes(x = AgeN,
y = PropMat, group = yearF), colour = "Black", linetype = 3, size = 1.5) + geom_errorbar(data = rmstdr,
mapping = aes(x = AgeN, ymax = ymax, ymin = ymin), colour = "Black", linetype = 3,
size = 1) + theme_clean() + scale_x_continuous(breaks = function(x) pretty(x,
n = 5))
ggplotly(plt_mo5swL_2zero)Figure 2.20: Five-year sliding window mean of maturity at age, with ages 1 & 2 set to 0, post-hoc.
2.3.3 Sex Ratio
The sex ratio in the stock has
sr_lng <- reshape(sex_ratio_female, varying = colnames(sex_ratio_female)[-1], v.names = "PropFemale",
timevar = "Age", times = colnames(sex_ratio_female)[-1], idvar = "year", direction = "long")
sr_lng$AgeN <- as.numeric(as.character(gsub(pattern = "a", replacement = "", x = sr_lng$Age)))
ggplotly(ggplot() + geom_line(data = sr_lng, mapping = aes(x = year, y = PropFemale,
colour = Age)) + facet_wrap(. ~ AgeN) + theme_few() + theme(axis.text.x = element_text(angle = 45,
hjust = 1, vjust = 1)) + scale_x_continuous(breaks = function(x) pretty(x, n = 5)) +
scale_color_manual(values = c(ebpal)) + scale_fill_manual(values = c(ebpal)) +
guides(colour = "none", fill = "none"))Figure 2.21: Sex ratio over time, faceted by age.
ggplotly(ggplot(sr_lng) + geom_line(mapping = aes(x = AgeN, y = PropFemale)) + geom_hline(yintercept = 0.5,
linetype = 2) + facet_wrap(facets = "year") + scale_x_continuous(breaks = function(x) pretty(x,
n = 5)) + theme_few())Figure 2.22: Sex ratios. Ages faceted by time.
2.3.4 Stock Weight At Age (SWAA)
The stock mean weights by age were calculated from the two first quarter surveys for each individual year. BITS data only exist for the period since 2008 and NS-IBTS only for the period since 2003. Therefore, prior to 2019 the BITS series was extended backwards to 2003 based on the average of 2008 to 2012. However, in 2019 it was found out that the procedure used for computing this average was erroneous, computing only a simple average across all length classes without weighting by the number of individuals within each length class. This lead to a very high estimate of the mean weight of the older fish, being driven up by very few observations. A more standard procedure with weighted average was implemented in 2019 (the same procedure as used for Western Baltic cod). The common series is finally extended backwards to 1999 based on the average of 2003 to 2007.
As for the maturity ogives, the SWAA must be calculated from the raw DATRAS Exchange products. These were read-in and cleaned in a previous section.
The below chunks (hidden) prepare the data, generate up-to-date SWAA for the assessment and plots a these out in figures for presentations/reporting. The fluctuating stock mean weights of the older age classes is caused by the small number of individuals caught at the surveys and the extremely high variability of weight for these age classes.
#===
# Subset processed exchange data ----
#===
# Select the SD and quarter
ca_q1<-subset(ca_hh_fin_bits_ibts,Quarter==1&Age!=-9&Age!=0&Age<15)
hl_q1<-subset(hl_hh_fin_bits_ibts,Quarter==1)
#=====
#===
# Summarise and aggregate data ----
#===
# Add 1 to all rows just to be able to count the fish
ca_q1$Nr_fish<-1
# Round the length class
ca_q1$RLngtClass<-round(ca_q1$LngtClass)
# How many fish we have
nr_fish<-aggregate(ca_q1$CANoAtLngt,
by=list(ca_q1$Country,ca_q1$Quarter,ca_q1$Year),
FUN="length")
nr_fish<- aggregate(ca_q1$Nr_fish,
by=list(ca_q1$Year,ca_q1$Sex,ca_q1$RLngtClass,ca_q1$Age),
FUN="sum",
na.rm=T)
names(nr_fish)<-c("year","sex","LngtClass","age","nr_fish")
# Aggregate number of fish and average weight in samples
weight<-aggregate(ca_q1$IndWgt,
by=list(ca_q1$Year,ca_q1$Sex,round(ca_q1$RLngtClass),
ca_q1$Age),
FUN="mean",na.rm=T)
names(weight)<-c("year","sex","LngtClass","age","weight")
#=====
#===
# Length Frequencies "at sea" ----
#===
# Set parameters for data
lencl<-c(4:60)
age<-c(1:10)
year<-c(1999:(DataYear+1))
# Make tables for length freq data
length_freq_tmp<-by(hl_q1,list(hl_q1$Year,round(hl_q1$LngtClass)),function (x){
sum(x$HLNoAtLngt_1)
})
length_freq_tab<-t(as.table(length_freq_tmp))
Lfreq_sea<-data.frame(year=rep(year, each=length(lencl)),LngtClass=rep(lencl,length(year)),HLNoLngt=NA)
# Fill the table
for (i in c(1:nrow(Lfreq_sea))){
year_i<-Lfreq_sea[i,names(Lfreq_sea)==as.character("year")]
lencl_i<-Lfreq_sea[i,names(Lfreq_sea)==as.character("LngtClass")]
if(lencl_i%in%c(as.numeric(dimnames(length_freq_tab)[[1]]))){
data_i<-length_freq_tab[as.numeric(dimnames(length_freq_tab)[[1]])==lencl_i,dimnames(length_freq_tab)[[2]]==year_i]
Lfreq_sea[i,names(Lfreq_sea)==as.character("HLNoLngt")]<-data_i
}else{
Lfreq_sea[i,names(Lfreq_sea)==as.character("HLNoLngt")]<-0
}
}
# replace NA with 0
Lfreq_sea[which(is.na(Lfreq_sea$HLNoLngt)),names(Lfreq_sea)==as.character("HLNoLngt")]<-0
#=====
#===
# Summary of sampled fish: weight by length and by age ----
#===
# Number of samples
## Make tables
nr_fish_samp_F<-data.frame(year=rep(year, each=length(lencl)),LngtClass=rep(lencl,length(year)), a1=NA,a2=NA,a3=NA,a4=NA,a5=NA,a6=NA,a7=NA,a8=NA,a9=NA,a10=NA)
nr_fish_samp_M<-nr_fish_samp_F
wgt_samp_F<-nr_fish_samp_F
wgt_samp_M<-nr_fish_samp_M
## Filling the tables:
### Female
for (i in 1:nrow( nr_fish_samp_F)){
for (j in 3:ncol( nr_fish_samp_F)){
year_i<-nr_fish_samp_F[i,names(nr_fish_samp_F)==as.character("year")]
lencl_i<-nr_fish_samp_F[i,names(nr_fish_samp_F)==as.character("LngtClass")]
age_i<-j-2
if(lencl_i%in% weight[which(weight$year==year_i&weight$sex==as.character("F")& weight$age==age_i),names(weight)==as.character("LngtClass")]){
show_wgt<-weight[which(weight$year==year_i&weight$LngtClass==lencl_i&weight$sex==as.character("F")&weight$age==age_i),]
show_nr<-nr_fish[which(nr_fish$year==year_i&nr_fish$LngtClass==lencl_i&nr_fish$sex==as.character("F")&nr_fish$age==age_i),]
nr_fish_samp_F[i,j]<-sum(show_nr$nr_fish)
wgt_samp_F[i,j]<-mean(show_wgt$weight, na.rm=T)
} else{
nr_fish_samp_F[i,j]<-0
wgt_samp_F[i,j]<-NA
}
}
}
### Male
for (i in 1:nrow( nr_fish_samp_M)){
for (j in 3:ncol( nr_fish_samp_M)){
year_i<-nr_fish_samp_M[i,names(nr_fish_samp_M)==as.character("year")]
lencl_i<-nr_fish_samp_M[i,names(nr_fish_samp_M)==as.character("LngtClass")]
age_i<-j-2
if(lencl_i%in% weight[which(weight$year==year_i&weight$sex==as.character("M")& weight$age==age_i),names(weight)==as.character("LngtClass")]){
show_wgt<-weight[which(weight$year==year_i&weight$LngtClass==lencl_i&weight$sex==as.character("M")&weight$age==age_i),]
show_nr<-nr_fish[which(nr_fish$year==year_i&nr_fish$LngtClass==lencl_i&nr_fish$sex==as.character("M")&nr_fish$age==age_i),]
nr_fish_samp_M[i,j]<-sum(show_nr$nr_fish)
wgt_samp_M[i,j]<-mean(show_wgt$weight, na.rm=T)
} else{
nr_fish_samp_M[i,j]<-0
wgt_samp_M[i,j]<-NA
}
}
}
# Number of fish sampled by length class:
## Female
nr_fish_samp_F$tot<-NA
for (i in 1:nrow(nr_fish_samp_F)){
nr_fish_samp_F[i,names(nr_fish_samp_F)==as.character("tot")]<-sum(nr_fish_samp_F[i,names(nr_fish_samp_F)%in%c("a1","a2","a3","a4","a5","a6","a7","a8","a9","a10")])
}
## Male
nr_fish_samp_M$tot<-NA
for (i in 1:nrow(nr_fish_samp_M)){
nr_fish_samp_M[i,names(nr_fish_samp_M)==as.character("tot")]<-sum(nr_fish_samp_M[i,names(nr_fish_samp_M)%in%c("a1","a2","a3","a4","a5","a6","a7","a8","a9","a10")])
}
## F and M combined
Lfreq_samp<-data.frame(year=rep(year, each=length(lencl)),LngtClass=rep(lencl,length(year)), tot_samp=NA)
for (i in 1:nrow(Lfreq_samp)){
year_i<-Lfreq_samp[i,names(Lfreq_samp)==as.character("year")]
lencl_i<-Lfreq_samp[i,names(Lfreq_samp)==as.character("LngtClass")]
F_sum<-nr_fish_samp_F[which(nr_fish_samp_F$year==year_i&nr_fish_samp_F$LngtClass==lencl_i),names(nr_fish_samp_F)==as.character("tot")]
M_sum<-nr_fish_samp_M[which(nr_fish_samp_M$year==year_i&nr_fish_samp_M$LngtClass==lencl_i),names(nr_fish_samp_M)==as.character("tot")]
Lfreq_samp[i,names(Lfreq_samp)==as.character("tot_samp")] <-F_sum+ M_sum
}
# Number of fish sampled by age:
## Make tables
tot_samp_age_F<-data.frame(year=year,a1=NA,a2=NA,a3=NA,a4=NA,a5=NA,a6=NA,a7=NA,a8=NA,a9=NA,a10=NA)
tot_samp_age_M<-data.frame(year=year,a1=NA,a2=NA,a3=NA,a4=NA,a5=NA,a6=NA,a7=NA,a8=NA,a9=NA,a10=NA)
## Female
for (i in 1:nrow(tot_samp_age_F)){
for (j in 2:ncol(tot_samp_age_F)){
year_i<-tot_samp_age_F[i,names(tot_samp_age_F)==as.character("year")]
show_data<- nr_fish_samp_F[which(nr_fish_samp_F$year==year_i),]
tot_samp_age_F[i,j]<-sum(show_data[,names(show_data)==names(tot_samp_age_F)[j]])
}
}
## Male
for (i in 1:nrow(tot_samp_age_M)){
for (j in 2:ncol(tot_samp_age_M)){
year_i<-tot_samp_age_M[i,names(tot_samp_age_M)==as.character("year")]
show_data<- nr_fish_samp_M[which(nr_fish_samp_M$year==year_i),]
tot_samp_age_M[i,j]<-sum(show_data[,names(show_data)==names(tot_samp_age_M)[j]])
}
}
#=====
#===
# Sex Ratios ----
#===
# Make tables
sex_ratio_F<- tot_samp_age_F
sex_ratio_F[,-1]<-NA
# Fill tables
for (i in 1:nrow(sex_ratio_F)){
for (j in 2:ncol(sex_ratio_F)){
sex_ratio_F[i,j]<-tot_samp_age_F[i,j]/(tot_samp_age_F[i,j]+tot_samp_age_M[i,j])
}
}
#=====
#===
# Full survey numbers for raising ----
#===
# Make Tables
nr_fish_sea_M<-nr_fish_samp_M
nr_fish_sea_M[,-(1:2)]<-NA
nr_fish_sea_F<-nr_fish_sea_M
## Female
for (i in 1:nrow(nr_fish_sea_F)){
for (j in 3:(ncol(nr_fish_sea_F)-1)){
if(nr_fish_samp_F$tot[i]>0&!is.na(Lfreq_sea$HLNoLngt[i])){
pct_age_tot<-nr_fish_samp_F[i,j]/nr_fish_samp_F[i,names(nr_fish_samp_F)==as.character("tot")]
F_M_ratio<-nr_fish_samp_F[i,names(nr_fish_samp_F)==as.character("tot")]/Lfreq_samp$tot_samp[i]
Lfreq<-Lfreq_sea$HLNoLngt[i]
nr_fish_sea_F[i,j]<-pct_age_tot*Lfreq* F_M_ratio
}else{
nr_fish_sea_F[i,j]<-0
}
}
}
## Male
for (i in 1:nrow(nr_fish_sea_M)){
for (j in 3:(ncol(nr_fish_sea_M)-1)){
if(nr_fish_samp_M$tot[i]>0&!is.na(Lfreq_sea$HLNoLngt[i])){
pct_age_tot<-nr_fish_samp_M[i,j]/nr_fish_samp_M[i,names(nr_fish_samp_M)==as.character("tot")]
M_F_ratio<-nr_fish_samp_M[i,names(nr_fish_samp_M)==as.character("tot")]/Lfreq_samp$tot_samp[i]
Lfreq<-Lfreq_sea$HLNoLngt[i]
nr_fish_sea_M[i,j]<-pct_age_tot*Lfreq* M_F_ratio
}else{
nr_fish_sea_M[i,j]<-0
}
}
}
# Sum of numbers over age-groups, by length
## Females
for (i in 1:nrow(nr_fish_sea_F)){
nr_fish_sea_F[i,names(nr_fish_sea_F)==as.character("tot")]<-sum(nr_fish_sea_F[i,names(nr_fish_sea_F)%in%c("a1","a2","a3","a4","a5","a6","a7","a8","a9","a10")])
}
## Males
for (i in 1:nrow(nr_fish_sea_M)){
nr_fish_sea_M[i,names( nr_fish_sea_M)==as.character("tot")]<-sum(nr_fish_sea_M[i,names(nr_fish_sea_M)%in%c("a1","a2","a3","a4","a5","a6","a7","a8","a9","a10")])
}
# Sum of numbers over length-groups, by age
## Make tables
sum_nr_fish_sea_F<-tot_samp_age_F
sum_nr_fish_sea_F[,-1]<-NA
sum_nr_fish_sea_M<-sum_nr_fish_sea_F
## Females
for (i in 1:nrow(sum_nr_fish_sea_F)){
for (j in 2:ncol(sum_nr_fish_sea_F)){
year_i<-sum_nr_fish_sea_F[i,names(sum_nr_fish_sea_F)==as.character("year")]
sum_nr_fish_sea_F[i,j]<-sum(nr_fish_sea_F[which(nr_fish_sea_F$year==year_i),names(nr_fish_sea_F)==names(sum_nr_fish_sea_F)[j]])
}
}
## Males
for (i in 1:nrow(sum_nr_fish_sea_M)){
for (j in 2:ncol(sum_nr_fish_sea_M)){
year_i<-sum_nr_fish_sea_M[i,names(sum_nr_fish_sea_M)==as.character("year")]
sum_nr_fish_sea_M[i,j]<-sum(nr_fish_sea_M[which(nr_fish_sea_M$year==year_i),names(nr_fish_sea_M)==names(sum_nr_fish_sea_M)[j]])
}
}
#=====
#===
# Raise proportions to full survey numbers ----
#===
# Make tables
wgt_nr_F<-nr_fish_sea_F
wgt_nr_F[,3:ncol(wgt_nr_F)]<-NA
wgt_nr_M<-wgt_nr_F
## Female
for (i in 1:nrow(wgt_nr_F)){
for (j in 3:(ncol(wgt_nr_F)-1)) {
if(nr_fish_sea_F[i,j]>0&!is.na(wgt_samp_F[i,j])){
wgt<-wgt_samp_F[i,j]
nr<-nr_fish_sea_F[i,j]
wgt_nr_F[i,j]<-wgt*nr
}else{
wgt_nr_F[i,j]<-0
}
}
}
## Male
for (i in 1:nrow(wgt_nr_M)){
for (j in 3:(ncol(wgt_nr_M)-1)) {
if(nr_fish_sea_M[i,j]>0&!is.na(wgt_samp_M[i,j])){
wgt<-wgt_samp_M[i,j]
nr<-nr_fish_sea_M[i,j]
wgt_nr_M[i,j]<-wgt*nr
}else{
wgt_nr_M[i,j]<-0
}
}
}
# Sum weights by age over length classes
## Make Tables
sum_wgt_nr_F<-tot_samp_age_F
sum_wgt_nr_F[,-1]<-NA
sum_wgt_nr_M<-sum_wgt_nr_F
## Females
for (i in 1:nrow(sum_wgt_nr_F)){
for (j in 2:ncol(sum_wgt_nr_F)){
year_i<-sum_wgt_nr_F[i,names(sum_wgt_nr_F)==as.character("year")]
sum_wgt_nr_F[i,j]<-sum(wgt_nr_F[which(wgt_nr_F$year==year_i),names(wgt_nr_F)==names(sum_wgt_nr_F)[j]])
}
}
## Males
for (i in 1:nrow(sum_wgt_nr_M)){
for (j in 2:ncol(sum_wgt_nr_M)){
year_i<-sum_wgt_nr_M[i,names(sum_wgt_nr_M)==as.character("year")]
sum_wgt_nr_M[i,j]<-sum(wgt_nr_M[which(wgt_nr_M$year==year_i),names(wgt_nr_M)==names(sum_wgt_nr_M)[j]])
}
}
#=====
#===
# Stock mean weight at age ----
#===
# Make Tables
WAA_F<-sum_wgt_nr_F
WAA_F[,-1]<-NA
WAA_M<-WAA_F
WAA_combsex<-WAA_M
## Females
for (i in 1:nrow(WAA_F)){
for (j in 2:ncol(WAA_F)){
WAA_F[i,j]<-sum_wgt_nr_F[i,j]/sum_nr_fish_sea_F[i,j]
}
}
## Males
for (i in 1:nrow(WAA_M)){
for (j in 2:ncol(WAA_M)){
WAA_M[i,j]<-sum_wgt_nr_M[i,j]/sum_nr_fish_sea_M[i,j]
}
}
## comb of F and M
for (i in 1:nrow(WAA_combsex)){
for (j in 2:ncol(WAA_combsex)){
if( !is.na(WAA_F[i,j])&!is.na(WAA_M[i,j])){
WAA_combsex[i,j]<-(WAA_F[i,j]*sex_ratio_F[i,j])+(WAA_M[i,j]*(1-sex_ratio_F[i,j]))
}
if(is.na(WAA_F[i,j])&!is.na(WAA_M[i,j])){
WAA_combsex[i,j]<-WAA_M[i,j]
}
if(is.na(WAA_M[i,j])&!is.na(WAA_F[i,j])){
WAA_combsex[i,j]<-WAA_F[i,j]
}
}
}
## Convert weights to kg, not grams
WAA_combsex <- WAA_combsex/1000
WAA_combsex$year <- WAA_combsex$year*1000
WAA_combsex$year <- as.integer(as.character(WAA_combsex$year))
#=====
#===
# Reshape WAA data and calculate means +/- 95%CI for plotting ----
#===
# Wide to long format
swaal <- reshape(WAA_combsex,
varying = colnames(WAA_combsex)[-1],
v.names = "MeanWeight",
timevar = "Age",
times = colnames(WAA_combsex)[-1],
idvar = "year",
direction = "long")
## Calculate mean swaa for 1999:2019 ----
m19WAA <- aggregate(MeanWeight ~ Age,
data = swaal[swaal$year %in% c(1999:2019),],
FUN = mean)
# Calculate std err around the above mean 1999:2019
mw19stdr <- aggregate(MeanWeight ~ Age,
data = swaal[swaal$year %in% c(1999:2019), ],
FUN = function(x){sd(x, na.rm = TRUE)/sqrt(length(na.omit(x)))})
mw19stdr$AgeN <- as.numeric(sub(pattern = "a", replacement = "", x = mw19stdr$Age))
colnames(mw19stdr)[colnames(mw19stdr) %in% c("MeanWeight")] <- "StdErr"
mw19stdr$MeanWeight <- m19WAA$MeanWeight
mw19stdr$ymax <- mw19stdr$MeanWeight + 1.96*mw19stdr$StdErr
mw19stdr$ymin <- mw19stdr$MeanWeight - 1.96*mw19stdr$StdErr
m19WAA$year <- as.factor(rep("Mean 1999-2019", n = nrow(m19WAA)))
mw19stdr <- mw19stdr[order(mw19stdr$AgeN), ]
## Calculate previous 3 year mean ----
m3WAA <- aggregate(MeanWeight ~ Age,
data = swaal[swaal$year %in% c(((DataYear+1)-2):(DataYear+1)),],
FUN = mean)
# Calculate std err around the above 3 year mean
m3wstdr <- aggregate(MeanWeight ~ Age,
data = swaal[swaal$year %in% c(((DataYear+1)-2):(DataYear+1)), ],
FUN = function(x){sd(x, na.rm = TRUE)/sqrt(length(na.omit(x)))})
m3wstdr$AgeN <- as.numeric(sub(pattern = "a", replacement = "", x = m3wstdr$Age))
colnames(m3wstdr)[colnames(m3wstdr) %in% c("MeanWeight")] <- "StdErr"
m3wstdr$MeanWeight <- m3WAA$MeanWeight
m3wstdr$ymax <- m3wstdr$MeanWeight + 1.96*m3wstdr$StdErr
m3wstdr$ymin <- m3wstdr$MeanWeight - 1.96*m3wstdr$StdErr
m3WAA$year <- as.factor(rep("3y Mean", n = nrow(m3WAA)))
m3wstdr <- m3wstdr[order(m3wstdr$AgeN), ]
## Calculate mean of 1999 to current year ----
rmWAA <- aggregate(MeanWeight ~ Age,
data = swaal[swaal$year %in% c(1999:(DataYear+1)),],
FUN = mean)
# Standard error of 1999 -> mean
rmwstdr <- aggregate(MeanWeight ~ Age,
data = swaal[swaal$year %in% c(1999:(DataYear+1)), ],
FUN = function(x){sd(x, na.rm = TRUE)/sqrt(length(na.omit(x)))})
rmwstdr$AgeN <- as.numeric(sub(pattern = "a", replacement = "", x = rmwstdr$Age))
colnames(rmwstdr)[colnames(rmwstdr) %in% c("MeanWeight")] <- "StdErr"
rmwstdr$MeanWeight <- rmWAA$MeanWeight
rmwstdr$ymax <- rmwstdr$MeanWeight + 1.96*rmwstdr$StdErr
rmwstdr$ymin <- rmwstdr$MeanWeight - 1.96*rmwstdr$StdErr
rmWAA$year <- as.factor(rep("Running Mean", n = nrow(rmWAA)))
rmwstdr <- rmwstdr[order(rmwstdr$AgeN), ]
# Combine all years, 2002 -> mean and 15 year mean into one object
swaal <- rbind(swaal, m19WAA, rmWAA, m3WAA)
swaal$AgeN <- as.numeric(sub(pattern = "a", replacement = "", x = swaal$Age))
swaal <- swaal[order(swaal$AgeN), ]
swaal$yearF <- as.factor(as.character(swaal$year))
## Calculate 3y sliding window average for whole timeseries ----
### Calculate the mean ----
ada <- an(3, nrow(WAA_combsex))
sw3sw <- frollmean(WAA_combsex[, -1], n=ada, na.rm = T, align = "right", adaptive = TRUE)
sw3sw <- as.data.frame(do.call(cbind, sw3sw))
sw3sw$year <- WAA_combsex$year
#### Wide to long format
sw3swL <- reshape(sw3sw,
varying = colnames(sw3sw)[-11],
v.names = "MeanWeight",
timevar = "Age",
times = colnames(sw3sw)[-11],
idvar = "year",
direction = "long")
sw3swL$Age <- as.numeric(gsub(pattern = "V", replacement = "", x = sw3swL$Age))
### Calcualte the standard error ----
sw3stdr <- frollapply(WAA_combsex[, -1],
3,
FUN = function(x){sd(x, na.rm = TRUE)/sqrt(length(na.omit(x)))})
sw3stdr <- as.data.frame(do.call(cbind, sw3stdr))
sw3stdr$year <- WAA_combsex$year
#### Wide to long format
sw3stdrL <- reshape(sw3stdr,
varying = colnames(sw3stdr)[-11],
v.names = "StdErrWeight",
timevar = "Age",
times = colnames(sw3stdr)[-11],
idvar = "year",
direction = "long")
sw3stdrL$Age <- as.numeric(gsub(pattern = "V", replacement = "", x = sw3stdrL$Age))
### Calculate 95% confidence intervals ----
sw3swL <- merge(x = sw3swL, y = sw3stdrL, by = c("year", "Age"), all.x = TRUE)
sw3swL$ymax <- sw3swL$MeanWeight + (1.96*sw3swL$StdErrWeight)
sw3swL$ymin <- sw3swL$MeanWeight - (1.96*sw3swL$StdErrWeight)
sw3swL$AgeN <- sw3swL$Age
sw3swL$Age <- as.character(sw3swL$Age)
sw3swL$year <- as.numeric(sw3swL$year)
sw3swL$yearF <- as.factor(as.character(sw3swL$year))
## Calculate 5y sliding window average for whole timeseries ----
### Calculate the mean ----
ada <- an(5, nrow(WAA_combsex))
sw5sw <- frollmean(WAA_combsex[, -1], n = ada, na.rm = T, align = "right", adaptive = TRUE)
sw5sw <- as.data.frame(do.call(cbind, sw5sw))
sw5sw$year <- WAA_combsex$year
#### Wide to long format
sw5swL <- reshape(sw5sw,
varying = colnames(sw5sw)[-11],
v.names = "MeanWeight",
timevar = "Age",
times = colnames(sw5sw)[-11],
idvar = "year",
direction = "long")
sw5swL$Age <- as.numeric(gsub(pattern = "V", replacement = "", x = sw5swL$Age))
### Calcualte the standard error ----
sw5stdr <- frollapply(WAA_combsex[, -1],
5,
FUN = function(x){sd(x, na.rm = TRUE)/sqrt(length(na.omit(x)))})
sw5stdr <- as.data.frame(do.call(cbind, sw5stdr))
sw5stdr$year <- WAA_combsex$year
sw5stdr[sw5stdr$year == 2002, -ncol(sw5stdr)] <- sw5stdr[sw5stdr$year == 2003, -ncol(sw5stdr)]
#### Wide to long format
sw5stdrL <- reshape(sw5stdr,
varying = colnames(sw5stdr)[-11],
v.names = "StdErrWeight",
timevar = "Age",
times = colnames(sw5stdr)[-11],
idvar = "year",
direction = "long")
sw5stdrL$Age <- as.numeric(gsub(pattern = "V", replacement = "", x = sw5stdrL$Age))
### Calculate 95% confidence intervals ----
sw5swL <- merge(x = sw5swL, y = sw5stdrL, by = c("year", "Age"), all.x = TRUE)
sw5swL$ymax <- sw5swL$MeanWeight + (1.96*sw5swL$StdErrWeight)
sw5swL$ymin <- sw5swL$MeanWeight - (1.96*sw5swL$StdErrWeight)
sw5swL$AgeN <- sw5swL$Age
sw5swL$Age <- as.character(sw5swL$Age)
sw5swL$year <- as.numeric(sw5swL$year)
sw5swL$yearF <- as.factor(as.character(sw5swL$year))
## Calculate annual values from a smoothing spline ----
### Build the spline
swaal2 <- swaal[!swaal$year %in% c("Mean 1999-2019", "Running Mean", "3y Mean"),]
swaal2$Age <- as.factor(swaal2$Age)
swaal2$year <- as.numeric(swaal2$year)
msw <- gam(formula = MeanWeight ~ s(year, bs = "cr", by = Age), data = swaal2, family = gaussian(link = "identity"))
### Predict with the smoothed spline
predweights <- predict.gam(msw, newdata = swaal2,type = "response", se.fit = TRUE)
swaal2$smoothMeanWeights <- unname(predweights$fit)
swaal2$seSmoothMeanWeights <- unname(predweights$se.fit)
#=====
#===
# save results
#====
write.csv(WAA_combsex,file="Data/BiologicalData/ple.27.21-32_StockWeightAtAge_1999-DataYear.csv")
# save_as_lowestoft Annaul Estimates
WAA_combsexSAM <- as.data.frame(WAA_combsex)
### save_as_lowestoft Mean 2002-DataYear
mWAA <- data.frame(lapply(colMeans(WAA_combsex[WAA_combsex$year %in% c(2002:(DataYear+1)), !colnames(WAA_combsex) %in% ("year")], na.rm = TRUE), type.convert), stringsAsFactors=FALSE)
save_as_lowestoft(df = mWAA, #[, !colnames(mWAA) %in% "Age"],
file_path = "Assessment Input/sw.dat",
title = "Stock weights at age for Ple.27.21-32. Mean values for combined sexes 2002 to data year, from survey data",
rando_numbers = paste(c(1, 4), collapse = "\t"),
year_range = paste(c(2002, (DataYear+1)), collapse = "\t"),
age_range = paste(c(1, 10), collapse = "\t"),
timeseries_type = 2)
### save_as_lowestoft Raw Annual Estimates
#### old removal of NAs
# for(i in 1:ncol(WAA_combsexSAM)){
# for(j in 1:nrow(WAA_combsexSAM)){
# WAA_combsexSAM[j, i] <- ifelse(is.na(WAA_combsexSAM[j, i]), -9, WAA_combsexSAM[j, i])
# }
# }
#### Make early timeseries "NA"s into plausible value
for(i in 1:ncol(WAA_combsexSAM)){
for(j in 1:nrow(WAA_combsexSAM)){
WAA_combsexSAM[j, i] <- ifelse(is.na(WAA_combsexSAM[j, i]), median(c(WAA_combsexSAM[,i]), na.rm = T), WAA_combsexSAM[j, i])
}
}
save_as_lowestoft(df = WAA_combsexSAM[WAA_combsexSAM$year %in% 2002:(DataYear+1), !colnames(WAA_combsexSAM) %in% "year"],
file_path = "Assessment Input/AlternateModel_AnnualRawBio/sw.dat",
title = "Stock weights at age for Ple.27.21-32. Annual values for combined sexes, from survey data",
rando_numbers = paste(c(1, 4), collapse = "\t"),
year_range = paste(c(2002, (DataYear+1)), collapse = "\t"),
age_range = paste(c(1, 10), collapse = "\t"))
### save_as_lowestoft Three Year Sliding Window
sw3swSAM <- as.data.frame(sw3sw)
# for(i in 1:ncol(sw3swSAM)){
# for(j in 1:nrow(sw3swSAM)){
# sw3swSAM[j, i] <- ifelse(is.na(sw3swSAM[j, i]), -9, sw3swSAM[j, i])
# }
# }
#### Make early timeseries "NA"s into plausible value
for(i in 1:ncol(sw3swSAM)){
for(j in 1:nrow(sw3swSAM)){
sw3swSAM[j, i] <- ifelse(is.na(sw3swSAM[j, i]), median(c(sw3swSAM[,i]), na.rm = T), sw3swSAM[j, i])
}
}
save_as_lowestoft(df = sw3swSAM[sw3swSAM$year %in% 2002:(DataYear+1), !colnames(sw3swSAM) %in% "year"],
file_path = "Assessment Input/AlternateModel_3ysw/sw.dat",
title = "Stock weights at age for Ple.27.21-32. Three year sliding window means for combined sexes, from survey data",
rando_numbers = paste(c(1, 4), collapse = "\t"),
year_range = paste(c(2002, (DataYear+1)), collapse = "\t"),
age_range = paste(c(1, 10), collapse = "\t"))
### save_as_lowestoft Five Year Sliding Window
sw5swSAM <- as.data.frame(sw5sw)
# for(i in 1:ncol(sw5swSAM)){
# for(j in 1:nrow(sw5swSAM)){
# sw5swSAM[j, i] <- ifelse(is.na(sw5swSAM[j, i]), -9, sw5swSAM[j, i])
# }
# }
#### Make early timeseries "NA"s into plausible value
for(i in 1:ncol(sw5swSAM)){
for(j in 1:nrow(sw5swSAM)){
sw5swSAM[j, i] <- ifelse(is.na(sw5swSAM[j, i]), median(c(sw5swSAM[,i]), na.rm = T), sw5swSAM[j, i])
}
}
save_as_lowestoft(df = sw5swSAM[sw5swSAM$year %in% 2002:(DataYear+1), !colnames(sw5swSAM) %in% "year"],
file_path = "Assessment Input/AlternateModel_5ysw/sw.dat",
title = "Stock weights at age for Ple.27.21-32. Five year sliding window means for combined sexes, from survey data",
rando_numbers = paste(c(1, 4), collapse = "\t"),
year_range = paste(c(2002, (DataYear+1)), collapse = "\t"),
age_range = paste(c(1, 10), collapse = "\t"))
### save_as_lowestoft spline-smoothed
swaa2 <- reshape(swaal2[, c("year", "Age", "smoothMeanWeights")],
varying = colnames(WAA_combsex)[-1],
v.names = "smoothMeanWeights",
timevar = "Age",
idvar = "year",
direction = "wide")
# for(i in 2:ncol(swaa2)){
# for(j in 2:nrow(swaa2)){
# swaa2[j, i] <- ifelse(is.na(swaa2[j, i]), -9, swaa2[j, i])
# swaa2[j, i] <- ifelse((swaa2[j, i]>= 0), -9, swaa2[j, i])
# }
# }
#### Make early timeseries "NA"s into plausible value
for(i in 1:ncol(swaa2)){
for(j in 1:nrow(swaa2)){
swaa2[j, i] <- ifelse(is.na(swaa2[j, i]), median(c(swaa2[,i]), na.rm = T), swaa2[j, i])
swaa2[j, i] <- ifelse((swaa2[j, i]>1), 1, swaa2[j, i])
}
}
save_as_lowestoft(df = swaa2[swaa2$year %in% c(2002:(DataYear+1)), ],
file_path = "Assessment Input/AlternateModel_SSbio/sw.dat",
title = "Stock weights at age for Ple.27.21-32. Smooth Spline for combined sexes, from survey data",
rando_numbers = paste(c(1, 4), collapse = "\t"),
year_range = paste(c(2002, (DataYear+1)), collapse = "\t"),
age_range = paste(c(1, 10), collapse = "\t"))
# saveRDS(swaal, file = "Biological Data/Ple.27.21-23_swaal.RDS")
# saveRDS(mw19stdr, file("Biological Data/Ple.27.21-23_SWAA_AveStdErr_1999-2019.RDS"))
# saveRDS(rmwstdr, file("Biological Data/Ple.27.21-23_SWAA_AveStdErr_1999-DataYear.RDS"))
# saveRDS(m3wstdr, file("Biological Data/Ple.27.21-23_SWAA_AveStdErr_3yr.RDS"))
#=====
k <- append(k, c("WAA_combsex", "swaal", "swaal2", "mw19stdr", "rmwstdr", "m3wstdr", "sw3swL", "sw5swL", "sw_annual_raw", "sex_ratio_F"))
# rm(list = ls()[!ls() %in% k])Similar to the maturity ogives, the stock mean weight at age (SWAA) in the assessment has historically, in the ple.27.21-23 stock, been an average of the whole time-series, namely from 1999 - 2023. However, because the weight at age has shifted from being relatively stable to being progressively lower since 2020 (probably due to density dependent competition) we can no longer use the mean of the time series. Over the last years of the ple.27.21-23 stock, WGBFAS agreed to, instead, use the mean for the period where it was stable, and sampling was low, and shift to the annual values for later years. This means the means for years 1999:2019 are applied to the years 1999:2019, while all years from 2020:2023 utilise the actual year’s estimated values from survey samples. Below we investigate what the SWAA values would be according to the stock annex compared to our new approach by first looking at the values in tables and then comparing them graphically.
# === Table of SWAA from stock annex ====
kable(swaal[swaal$year == "Running Mean", c("AgeN", "MeanWeight")], caption = "Static Stock Weight-at-age from mean of timeseries, according to ple.27.21-23 stock annex.")| AgeN | MeanWeight | |
|---|---|---|
| 11 | 1 | 0.0353194 |
| 31 | 2 | 0.0716080 |
| 41 | 3 | 0.1267258 |
| 51 | 4 | 0.1941207 |
| 61 | 5 | 0.2520088 |
| 71 | 6 | 0.3013635 |
| 81 | 7 | 0.3867862 |
| 91 | 8 | 0.4165005 |
| 101 | 9 | 0.4264940 |
| 21 | 10 | 0.5196663 |
# =====# === Data Prep ====
mw19stdr$cat <- "mw1999_2019"
awCurrent <- swaal[swaal$year %in% c(2020:DataYear + 1), ]
awCurrent$cat <- as.character(awCurrent$year)
awCurrent$StdErr <- NA
awCurrent$ymax <- NA
awCurrent$ymin <- NA
years <- rep(1999:2019, 10)[order(rep(1999:2019, 10))]
mw19_19 <- data.frame(year = years, AgeN = rep(1:10, length(1999:2019)), MeanWeight = rep(mw19stdr$MeanWeight,
length(1999:2019)), ymin = rep(mw19stdr$ymin, length(1999:2019)), ymax = rep(mw19stdr$ymax,
length(1999:2019)), StdErr = rep(mw19stdr$StdErr, length(1999:2019)))
swaasepLong <- rbind(mw19_19, awCurrent[, colnames(awCurrent)[colnames(awCurrent) %in%
colnames(mw19_19)]])
swaasepSAM <- dcast(data = swaasepLong[, c("year", "AgeN", "MeanWeight")], formula = year ~
AgeN, value.var = "MeanWeight")
# write.csv(swaasepSAM, file = 'Biological
# Data/Ple.27.21-32_SWAA_Ave1999-2019_Value2020-Current.csv') ==== === Table of
# SWAA for Assessment ====
kable(swaasepSAM, caption = "Stock weight-at-age with mean values applied for 1999:2019 and observed values for 2020 onwards.")| year | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
|---|---|---|---|---|---|---|---|---|---|---|
| 1999 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2000 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2001 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2002 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2003 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2004 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2005 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2006 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2007 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2008 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2009 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2010 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2011 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2012 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2013 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2014 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2015 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2016 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2017 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2018 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2019 | 0.0398838 | 0.0777201 | 0.1363725 | 0.2100388 | 0.2728213 | 0.3256387 | 0.4220236 | 0.4487470 | 0.4495387 | 0.5637859 |
| 2021 | 0.0154406 | 0.0502570 | 0.0856119 | 0.1535145 | 0.1830782 | 0.2259203 | 0.2523265 | 0.3274388 | 0.3429285 | 0.3091177 |
| 2022 | 0.0175870 | 0.0442823 | 0.0906907 | 0.1190161 | 0.1552038 | 0.1841127 | 0.2715534 | 0.2746941 | 0.2639065 | 0.3422108 |
| 2023 | 0.0169497 | 0.0445842 | 0.0743983 | 0.1090106 | 0.1487086 | 0.2000505 | 0.2178145 | 0.2828484 | 0.3866234 | 0.3980744 |
| 2024 | 0.0129716 | 0.0364625 | 0.0626792 | 0.0902232 | 0.1123439 | 0.1461400 | 0.1640149 | 0.2214779 | 0.3267716 | 0.2532022 |
# =====# === Figure stock weight at age (Calculated from DATRAS Exchange Data) ====
ggplotly(ggplot() + geom_line(data = rmwstdr, mapping = aes(x = AgeN, y = MeanWeight),
size = 1.25, colour = ebpal[2]) + geom_ribbon(data = rmwstdr, mapping = aes(x = AgeN,
ymin = ymin, ymax = ymax), fill = ebpal[2], alpha = 0.2) + geom_line(data = mw19stdr,
mapping = aes(x = AgeN, y = MeanWeight), size = 1.25, colour = ebpal[1]) + geom_ribbon(data = mw19stdr,
mapping = aes(x = AgeN, ymin = ymin, ymax = ymax), fill = ebpal[1], alpha = 0.2) +
geom_line(data = swaal[swaal$year %in% c(2020:(DataYear + 1)), ], mapping = aes(x = AgeN,
y = MeanWeight, colour = year)) + theme_clean() + scale_colour_manual(values = ebpal[3:length(ebpal)]) +
scale_x_continuous(breaks = function(x) pretty(x, n = 10)))Figure 2.23: Stock weight at age. The light green line and ribbon are the full time-series mean +/- 95%CI, to be used as fixed for all years according to the stock annex. The light blue line and ribbon is the 1999:2019 average (+/- 95%CI), which is applied to those years. Subsequent years, 2020 to 2024, are included independently in the assessment run and are shown in the remaining colours.
# =====swalba <- melt(setDT(WAA_combsex), id.vars = "year", variable.name = "Age", value.name = "MeanWeight")
ggplotly(ggplot() + geom_line(data = swalba, mapping = aes(x = year, y = MeanWeight)) +
facet_wrap(. ~ Age) + theme_few() + theme(axis.text.x = element_text(angle = 45,
hjust = 1, vjust = 1)) + scale_x_continuous(breaks = function(x) pretty(x, n = 5)) +
scale_y_continuous(breaks = c(0, 0.2, 0.4, 0.6)) + coord_cartesian(ylim = c(0,
0.6)))Figure 2.24: Stock weight at age by year, over time.
ggplotly(ggplot() + geom_line(data = swalba, mapping = aes(x = year, y = MeanWeight,
colour = Age)) + theme_few() + theme(axis.text.x = element_text(angle = 45, hjust = 1,
vjust = 1)) + scale_x_continuous(breaks = function(x) pretty(x, n = 5)) + scale_y_continuous(breaks = c(0,
0.2, 0.4, 0.6, 0.8)) + coord_cartesian(ylim = c(0, 0.8)) + scale_color_manual(values = c("#000000",
ebpal)))Figure 2.25: Stock weight at age by year, over time, all in one panel (FOR JESPER).
ggplotly(ggplot() +
geom_line(data = swaal2,
mapping = aes(x = year,
y = smoothMeanWeights,
colour = Age)) +
geom_ribbon(data = swaal2,
mapping = aes(x = year,
ymin = smoothMeanWeights-(1.96*seSmoothMeanWeights),
ymax = smoothMeanWeights+(1.96*seSmoothMeanWeights),
# colour = Age,
fill = Age),
alpha = 0.15) +
theme_few()+
theme(axis.text.x = element_text(angle = 45, hjust = 1, vjust = 1)) +
scale_x_continuous(breaks = function(x) pretty(x, n = 5)) +
scale_y_continuous(breaks = c(0, 0.2, 0.4, 0.6, 0.8)) +
coord_cartesian(ylim = c(0,0.95))+
scale_color_manual(values = c(ebpal)) +
scale_fill_manual(values = c(ebpal))
)Figure 2.26: Stock weight at age by year, over time, estimated from smoothed splines by age.
swalba <- melt(setDT(WAA_combsex), id.vars = "year", variable.name = "Age", value.name = "MeanWeight")
ggplotly(ggplot() +
geom_line(data = swalba,
mapping = aes(x = year,
y = MeanWeight)) +
geom_line(data = swaal2,
mapping = aes(x = year,
y = smoothMeanWeights,
colour = Age)) +
geom_ribbon(data = swaal2,
mapping = aes(x = year,
ymin = smoothMeanWeights-(1.96*seSmoothMeanWeights),
ymax = smoothMeanWeights+(1.96*seSmoothMeanWeights),
# colour = Age,
fill = Age),
alpha = 0.15) +
facet_wrap(.~Age)+
theme_few()+
theme(axis.text.x = element_text(angle = 45, hjust = 1, vjust = 1)) +
scale_x_continuous(breaks = function(x) pretty(x, n = 5)) +
# scale_y_continuous(breaks = c(0, 0.2, 0.4, 0.6)) +
coord_cartesian(ylim = c(0,0.95)) +
scale_color_manual(values = c(ebpal)) +
scale_fill_manual(values = c(ebpal))
)Figure 2.27: Stock weight at age by year, over time, actual values (black) and predicted from smoother (colours); ribbons = 95% CI.
2.3.5 Stock Length at Age
While not necessary for the assessment, we can use the above aggregations to also get the stock mean length at age.
# === Get numbers of length samples ==== Make tables
len_nr_F <- wgt_nr_F
len_nr_F[, 3:ncol(len_nr_F)] <- NA
len_nr_M <- len_nr_F
# Females
for (i in 1:nrow(len_nr_F)) {
for (j in 3:(ncol(len_nr_F) - 1)) {
if (nr_fish_sea_F[i, j] > 0) {
len_nr_F[i, j] <- nr_fish_sea_F[i, j] * nr_fish_sea_F[i, names(nr_fish_sea_F) ==
as.character("LngtClass")]
} else {
len_nr_F[i, j] <- 0
}
}
}
# Males
for (i in 1:nrow(len_nr_M)) {
for (j in 3:(ncol(len_nr_M) - 1)) {
if (nr_fish_sea_M[i, j] > 0) {
len_nr_M[i, j] <- nr_fish_sea_M[i, j] * nr_fish_sea_M[i, names(nr_fish_sea_M) ==
as.character("LngtClass")]
} else {
len_nr_M[i, j] <- 0
}
}
}
# === Sum of these values over lenghtclasses, by age ==== Make Tables
sum_len_nr_F <- sum_wgt_nr_F
sum_len_nr_F[, -1] <- NA
sum_len_nr_M <- sum_len_nr_F
# Females
for (i in 1:nrow(sum_len_nr_F)) {
for (j in 2:ncol(sum_len_nr_F)) {
year_i <- sum_len_nr_F[i, names(sum_len_nr_F) == as.character("year")]
sum_len_nr_F[i, j] <- sum(len_nr_F[which(len_nr_F$year == year_i), names(len_nr_F) ==
names(sum_len_nr_F)[j]])
}
}
# Males
for (i in 1:nrow(sum_len_nr_M)) {
for (j in 2:ncol(sum_len_nr_M)) {
year_i <- sum_len_nr_M[i, names(sum_len_nr_M) == as.character("year")]
sum_len_nr_M[i, j] <- sum(len_nr_M[which(len_nr_M$year == year_i), names(len_nr_M) ==
names(sum_len_nr_M)[j]])
}
}
# =====
# === Calculate mean length at age ==== Make tables
LAA_F <- WAA_F
LAA_F[, -1] <- NA
LAA_M <- LAA_F
LAA_combsex <- LAA_M
# Females
for (i in 1:nrow(LAA_F)) {
for (j in 2:ncol(LAA_F)) {
LAA_F[i, j] <- sum_len_nr_F[i, j]/sum_nr_fish_sea_F[i, j]
}
}
# Males
for (i in 1:nrow(LAA_M)) {
for (j in 2:ncol(LAA_M)) {
LAA_M[i, j] <- sum_len_nr_M[i, j]/sum_nr_fish_sea_M[i, j]
}
}
# comb of F and M
for (i in 1:nrow(LAA_combsex)) {
for (j in 2:ncol(LAA_combsex)) {
if (!is.na(LAA_F[i, j]) & !is.na(LAA_M[i, j])) {
LAA_combsex[i, j] <- (LAA_F[i, j] * sex_ratio_F[i, j]) + (LAA_M[i, j] *
(1 - sex_ratio_F[i, j]))
}
if (is.na(LAA_F[i, j]) & !is.na(LAA_M[i, j])) {
LAA_combsex[i, j] <- LAA_M[i, j]
}
if (is.na(LAA_M[i, j]) & !is.na(LAA_F[i, j])) {
LAA_combsex[i, j] <- LAA_F[i, j]
}
}
}
# === Clean Data Environment ====
k <- append(k, c("LAA_combsex", "LAA_F", "LAA_M", "WAA_combsex", "WAA_F", "WAA_M",
"swaal", "m19WAA", "rm19WAA"))
rm(list = ls()[!ls() %in% k])
# =====2.3.5.1 Stock Lengths at age
slalba <- melt(setDT(LAA_combsex), id.vars = "year", variable.name = "Age", value.name = "MeanLength")
ggplotly(ggplot() + geom_line(data = slalba, mapping = aes(x = year, y = MeanLength,
colour = Age)) + facet_wrap(. ~ Age) + theme_few() + theme(axis.text.x = element_text(angle = 45,
hjust = 1, vjust = 1)) + scale_x_continuous(breaks = function(x) pretty(x, n = 5)) +
scale_color_manual(values = ebpal))Figure 2.28: Stock lengths at age by year, over time.
2.3.5.2 Raw lengths at age observations
ca_hh_fin_bits_ibts$AgeN <- ca_hh_fin_bits_ibts$Age
ca_hh_fin_bits_ibts$AgeF <- factor(as.character(ca_hh_fin_bits_ibts$AgeN), levels = as.character(c(1:max(ca_hh_fin_bits_ibts$AgeN,
na.rm = T))))
ggplot() + geom_point(data = ca_hh_fin_bits_ibts[!is.na(ca_hh_fin_bits_ibts$AgeF) &
ca_hh_fin_bits_ibts$LngtClass < 110 & ca_hh_fin_bits_ibts$Quarter == 1 & ca_hh_fin_bits_ibts$AgeN <
26, ], mapping = aes(x = AgeF, y = LngtClass)) + theme_few() + scale_color_manual(values = c(ebpal)) +
scale_fill_manual(values = c(ebpal)) + guides(colour = "none", fill = "none")
Figure 2.29: Raw lengths at age from survey
ggplot() + geom_point(data = ca_hh_fin_bits_ibts[!is.na(ca_hh_fin_bits_ibts$Age) &
ca_hh_fin_bits_ibts$LngtClass < 110 & ca_hh_fin_bits_ibts$Quarter == 1, ], mapping = aes(x = Year,
y = LngtClass)) + facet_wrap(facets = "AgeF") + theme_few() + scale_x_continuous(breaks = function(x) pretty(x,
n = 5)) + scale_color_manual(values = c(ebpal)) + scale_fill_manual(values = c(ebpal)) +
guides(colour = "none", fill = "none")
Figure 2.30: Lengths at age, over time.
2.3.6 Natural Mortality Estimates
To try and improve upon the magic numbers that are currently the basis of the assumed natural mortality across ages, we can use the Gislason (et al. 2010) method for estimating natural mortality based on length, \(L_{\infty}\), and \(K\). \[ ln(M) = 0.55 - 1.61ln(L) + 1.44ln(L_{\infty}) + ln(K) \] Where M is natural mortality, \(L\) is length at a given age (and/or age*year), \(L_{\infty}\) is the asymptotal length of the stock, and \(K\) is the shape parameter of the Von Bertalanffy growth curve.
To estimate \(L_{\infty}\) and \(K\), we can fit a linear model to the 1-year staggered lengths (Ford-Walford method) and derive these values from the slope and intercept parameters.
## Create staggered length at age data
LAA_long <- reshape(data = LAA_combsex, varying = colnames(LAA_combsex)[-1], v.names = "MeanLength",
timevar = "Age", times = colnames(LAA_combsex)[-1], idvar = "year", direction = "long")
LAA_long$AgeN <- as.numeric(gsub(pattern = "a", replacement = "", x = LAA_long$Age))
tempshift <- LAA_long
tempshift$AgeN <- tempshift$AgeN + 1
tempshift$year <- tempshift$year + 1
names(tempshift)[3] <- "FirstMeanLength"
temp_df <- merge(LAA_long, tempshift, by = c("AgeN", "year"), all.x = TRUE)
m1 <- lm(MeanLength ~ FirstMeanLength, data = temp_df[!is.na(temp_df$MeanLength) &
!is.na(temp_df$FirstMeanLength), ])
plt_fw <- ggplot(temp_df[!is.na(temp_df$MeanLength) & !is.na(temp_df$FirstMeanLength),
]) + geom_point(mapping = aes(x = FirstMeanLength, y = MeanLength, colour = Age.x)) +
geom_smooth(mapping = aes(x = FirstMeanLength, y = MeanLength), method = "lm") +
scale_colour_manual(values = ebpal) + theme_few()
ggplotly(plt_fw)## `geom_smooth()` using formula = 'y ~ x'From this linear model we can calculate that \(L_{\infty} =\) 34.0148381, and \(K =\) 0.3041184. Now we can calculate the natural mortality estimates for each age across time.
Alternately, we can estimate and from raw survey data using maximum likelihood estimation. we can look at the length at age plots to propose some good starting points for this estimation.
## Initialising values for linf, k and t0
theta <- c(unname(m1$coefficients[1]/(1 - m1$coefficients[2])), unname(-log(m1$coefficients[2])),
0)
## Explicit ages and lengths to be used in optimisation
tempdf <- ca_hh_fin_bits_ibts[!is.na(ca_hh_fin_bits_ibts$Age) & ca_hh_fin_bits_ibts$LngtClass <
110 & ca_hh_fin_bits_ibts$Quarter == 1 & ca_hh_fin_bits_ibts$AgeN < 26, ]
age <- tempdf[, "AgeN"]
lt <- tempdf[, "LngtClass"]
## Optimisation function
SSQ <- function(theta, x) {
Linf <- theta[1]
K <- theta[2]
t0 <- theta[3]
epsilon <- rep(0, length(age))
lpred <- rep(0, length(age))
for (i in 1:length(age)) {
lpred[i] <- Linf * (1 - exp(-K * (age[i] - t0)))
epsilon[i] <- (lt[i] - lpred[i])^2
}
ssq <- sum(epsilon)
return(ssq)
}
out <- optim(theta, fn = SSQ, method = "BFGS", x = age, hessian = TRUE)
out$V <- solve(out$hessian) #solve the hessian
out$S <- sqrt(diag(out$V)) #Standard Error
out$R <- out$V/(out$S %o% out$S) #CorrelationL_inf <- unname(m1$coefficients[1]/(1-m1$coefficients[2]))
K <- unname(-log(m1$coefficients[2]))
t0 <- 0
sd <- 1
vbg <- bbmle::mle2(log(LngtClass) ~ #response variable
dnorm(mean = log(L_inf)+log(1-exp(-K*(AgeN-t0))), sd=sd),
data=tempdf, #data frame
start=list(L_inf=L_inf , K=K, t0=t0, sd=sd))
coef <- data.frame(coeff = names(vbg@fullcoef),
mech = c(L_inf, K, t0, sd),
est = as.numeric(vbg@fullcoef))Now by utilising the raw input data we arrive at similar values for L and k as we acheived with the Ford-Walford method.
tempdf$est_sv <- out$par[1] * (1 - exp(-out$par[2] * (tempdf$AgeN - out$par[3])))
tdf <- data.frame(AgeN = seq(from = 0, to = max(tempdf$AgeN), by = 0.1))
tdf$est_sv <- out$par[1] * (1 - exp(-out$par[2] * (tdf$AgeN - out$par[3])))
ggplot() + geom_point(data = tempdf, mapping = aes(x = AgeF, y = LngtClass)) + geom_line(data = tdf,
mapping = aes(x = AgeN, y = est_sv), colour = ebpal[1], size = 2) + theme_few() +
scale_color_manual(values = c(ebpal)) + scale_fill_manual(values = c(ebpal)) +
guides(colour = "none", fill = "none")
Figure 2.31: Plot estimated growth parameters over lengths at age.
After all of this, these life history parameters do not look acceptable for this species. Therefore we take values from fishbase.se, such that: \(L_{\infty} =\) 52cm and \(K =\) 0.168 (\(t0 =\) -0.72, in this case).
## Select life history parameters
Linf <- 52 #Fishbase for Baltic
K <- 0.168 #Fishbase for Baltic
t0 <- -0.72 #Fishbase for Baltictdf <- data.frame(AgeN = seq(from = 0, to = max(tempdf$AgeN), by = 0.1))
tdf$vbgf <- Linf * (1 - exp(-K * (tdf$AgeN - t0)))
ggplot() + geom_line(data = tdf, mapping = aes(x = AgeN, y = vbgf), colour = ebpal[1],
size = 2) + theme_few() + scale_color_manual(values = c(ebpal)) + scale_fill_manual(values = c(ebpal)) +
guides(colour = "none", fill = "none")
Figure 2.32: Plot of Selected VBGF.
## Calculate annually varying natural mortality
LAA_long$nm <- exp(0.55 - 1.61 * log(LAA_long$MeanLength) + 1.44 * log(Linf) + log(K))
nm_var <- LAA_long[!is.na(LAA_long$nm), ]
## Aggregate to time invariant natural mortality
nm_fix <- aggregate(nm ~ AgeN, data = nm_var, FUN = "mean")
nm_stdr <- aggregate(nm ~ AgeN, data = nm_var, FUN = function(x) {
sd(x, na.rm = TRUE)/sqrt(length(na.omit(x)))
})
names(nm_stdr)[2] <- "StdErr"
nm_stdr <- merge(nm_stdr, nm_fix, by = "AgeN", all.y = T)
nm_stdr$ymax <- nm_stdr$nm + 1.96 * nm_stdr$StdErr
nm_stdr$ymin <- nm_stdr$nm - 1.96 * nm_stdr$StdErr
nm_var$Year <- as.factor(as.character(nm_var$year))
nm_var$Age <- factor(as.character(nm_var$AgeN), levels = unique(as.character(nm_var$AgeN)))plt_nmvar_age <- ggplot(nm_var) + geom_line(mapping = aes(x = year, y = nm, group = Age,
colour = Age)) + facet_wrap(facets = "Age", scales = "free_y") + scale_color_manual(values = ebpal) +
theme_few() + theme(axis.text.x = element_text(angle = 45))
ggplotly(plt_nmvar_age)Figure 2.33: Time and age varying natural mortality based on Gislason. Time faceted by age.
plt_nmvar_year <- ggplot(nm_var) + geom_line(mapping = aes(x = AgeN, y = nm)) + facet_wrap(facets = "Year") +
theme_few()
ggplotly(plt_nmvar_year)Figure 2.34: Time and age varying natural mortality based on Gislason. Ages faceted by time.
nm_sam <- as.data.frame(nm_var[, c("AgeN", "year", "nm")])
nm_sam <- reshape(nm_sam, v.names = "nm", timevar = "AgeN", idvar = "year", direction = "wide")
# Raw annual estimates
nm_SAM <- nm_sam
# for(i in 1:ncol(nm_SAM)){ for(j in 1:nrow(nm_SAM)){ nm_SAM[j, i] <-
# ifelse(is.na(nm_SAM[j, i]), -9, nm_SAM[j, i]) } }
#### Make early 'NA' Values plausible
for (i in 1:ncol(nm_SAM)) {
for (j in 1:nrow(nm_SAM)) {
nm_SAM[j, i] <- ifelse(is.na(nm_SAM[j, i]), median(c(nm_SAM[, i]), na.rm = T),
nm_SAM[j, i])
}
}
save_as_lowestoft(df = nm_SAM[nm_SAM$year %in% c(2002:(DataYear + 1)), colnames(nm_SAM)[-1]],
file_path = "Assessment Input/NaturalMortalityVariants/timeVariantGislasonRaw_nm.dat",
title = "Natural mortality estimates from Gislason method, annual estimations by age and year.",
rando_numbers = paste(c(1, 4), collapse = "\t"), year_range = paste(c(2002, (DataYear +
1)), collapse = "\t"), age_range = paste(c(1, 10), collapse = "\t"), timeseries_type = 1)
## Calculate 3y sliding window average for whole timeseries ---- Calculate the
## mean ----
ada <- an(3, nrow(nm_sam))
nmG3sw <- frollmean(x = nm_sam[, -1], n = ada, na.rm = T, align = "right", adaptive = TRUE)
nmG3sw <- as.data.frame(do.call(cbind, nmG3sw))
#### Retain only the correct years
nmG3sw$year <- nm_sam$year
nmG3sw <- nmG3sw[nmG3sw$year %in% c(2002:(DataYear + 1)), ]
#### Wide to long format
nmG3swL <- reshape(nmG3sw, varying = colnames(nmG3sw)[-11], v.names = "nm", timevar = "Age",
times = colnames(nmG3sw)[-11], idvar = "year", direction = "long")
nmG3swL$Age <- as.numeric(gsub(pattern = "V", replacement = "", x = nmG3swL$Age))
### Calcualte the standard error ----
nmG3stdr <- frollapply(nm_sam[, -1], 3, FUN = function(x) {
sd(x, na.rm = TRUE)/sqrt(length(na.omit(x)))
})
nmG3stdr <- as.data.frame(do.call(cbind, nmG3stdr))
nmG3stdr$year <- nm_sam$year
#### Retain only the correct years
nmG3stdr <- nmG3stdr[nmG3stdr$year %in% c(2002:(DataYear + 1)), ]
#### Wide to long format
nmG3stdrL <- reshape(nmG3stdr, varying = colnames(nmG3stdr)[-11], v.names = "StdErrnm",
timevar = "Age", times = colnames(nmG3stdr)[-11], idvar = "year", direction = "long")
nmG3stdrL$Age <- as.numeric(gsub(pattern = "V", replacement = "", x = nmG3stdrL$Age))
### Calculate 93% confidence intervals ----
nmG3swL <- merge(x = nmG3swL, y = nmG3stdrL, by = c("year", "Age"), all.x = TRUE)
nmG3swL$ymax <- nmG3swL$nm + (1.96 * nmG3swL$StdErrnm)
nmG3swL$ymin <- nmG3swL$nm - (1.96 * nmG3swL$StdErrnm)
nmG3swL$AgeN <- nmG3swL$Age
nmG3swL$Age <- as.character(nmG3swL$Age)
nmG3swL <- nmG3swL[nmG3swL$year %in% c(2002:(DataYear + 1)), ]
nm_SAM <- nmG3sw
# for(i in 1:ncol(nm_SAM)){ for(j in 1:nrow(nm_SAM)){ nm_SAM[j, i] <-
# ifelse(is.na(nm_SAM[j, i]), -9, nm_SAM[j, i]) } }
#### Make early 'NA' Values plausible
for (i in 1:ncol(nm_SAM)) {
for (j in 1:nrow(nm_SAM)) {
nm_SAM[j, i] <- ifelse(is.na(nm_SAM[j, i]), median(c(nm_SAM[, i]), na.rm = T),
nm_SAM[j, i])
}
}
save_as_lowestoft(df = nm_SAM[nm_SAM$year %in% c(2002:(DataYear + 1)), colnames(nm_SAM)[-11]],
file_path = "Assessment Input/NaturalMortalityVariants/timeVariantGislason3ysw_nm.dat",
title = "Natural mortality estimates from Gislason method, five year sliding window means, by age and year.",
rando_numbers = paste(c(1, 4), collapse = "\t"), year_range = paste(c(2002, (DataYear +
1)), collapse = "\t"), age_range = paste(c(1, 10), collapse = "\t"), timeseries_type = 1)
## Calculate 5y sliding window average for whole timeseries ---- Calculate the
## mean ----
ada <- an(5, nrow(nm_sam))
nmG5sw <- frollmean(x = nm_sam[, -1], n = ada, na.rm = T, align = "right", adaptive = TRUE)
nmG5sw <- as.data.frame(do.call(cbind, nmG5sw))
#### Make the first year equal to the first calculable 5y average
nmG5sw$year <- nm_sam$year
# mo5sw[mo5sw$year == 2002, -ncol(mo5sw)] <- mo5sw[mo5sw$year == 2003,
# -ncol(mo5sw)]
nmG5sw <- nmG5sw[nmG5sw$year %in% c(2002:(DataYear + 1)), ]
#### Wide to long format
nmG5swL <- reshape(nmG5sw, varying = colnames(nmG5sw)[-11], v.names = "nm", timevar = "Age",
times = colnames(nmG5sw)[-11], idvar = "year", direction = "long")
nmG5swL$Age <- as.numeric(gsub(pattern = "V", replacement = "", x = nmG5swL$Age))
### Calcualte the standard error ----
nmG5stdr <- frollapply(nm_sam[, -1], 5, FUN = function(x) {
sd(x, na.rm = TRUE)/sqrt(length(na.omit(x)))
})
nmG5stdr <- as.data.frame(do.call(cbind, nmG5stdr))
nmG5stdr$year <- nm_sam$year
#### Make the first year equal to the first calculable 5y average
nmG5stdr[nmG5stdr$year == 2002, -ncol(nmG5stdr)] <- nmG5stdr[nmG5stdr$year == 2003,
-ncol(nmG5stdr)]
nmG5stdr <- nmG5stdr[nmG5stdr$year %in% c(2002:(DataYear + 1)), ]
#### Wide to long format
nmG5stdrL <- reshape(nmG5stdr, varying = colnames(nmG5stdr)[-11], v.names = "StdErrnm",
timevar = "Age", times = colnames(nmG5stdr)[-11], idvar = "year", direction = "long")
nmG5stdrL$Age <- as.numeric(gsub(pattern = "V", replacement = "", x = nmG5stdrL$Age))
### Calculate 95% confidence intervals ----
nmG5swL <- merge(x = nmG5swL, y = nmG5stdrL, by = c("year", "Age"), all.x = TRUE)
nmG5swL$ymax <- nmG5swL$nm + (1.96 * nmG5swL$StdErrnm)
nmG5swL$ymin <- nmG5swL$nm - (1.96 * nmG5swL$StdErrnm)
nmG5swL$AgeN <- nmG5swL$Age
nmG5swL$Age <- as.character(nmG5swL$Age)
nmG5swL <- nmG5swL[nmG5swL$year %in% c(2002:(DataYear + 1)), ]
nm_SAM <- nmG5sw
# for(i in 1:ncol(nm_SAM)){ for(j in 1:nrow(nm_SAM)){ nm_SAM[j, i] <-
# ifelse(is.na(nm_SAM[j, i]), -9, nm_SAM[j, i]) } }
#### Make early 'NA' Values plausible
for (i in 1:ncol(nm_SAM)) {
for (j in 1:nrow(nm_SAM)) {
nm_SAM[j, i] <- ifelse(is.na(nm_SAM[j, i]), median(c(nm_SAM[, i]), na.rm = T),
nm_SAM[j, i])
}
}
save_as_lowestoft(df = nm_SAM[nm_SAM$year %in% c(2002:(DataYear + 1)), colnames(nm_SAM)[-11]],
file_path = "Assessment Input/NaturalMortalityVariants/timeVariantGislason5ysw_nm.dat",
title = "Natural mortality estimates from Gislason method, five year sliding window means, by age and year.",
rando_numbers = paste(c(1, 4), collapse = "\t"), year_range = paste(c(2002, (DataYear +
1)), collapse = "\t"), age_range = paste(c(1, 10), collapse = "\t"), timeseries_type = 1)plt_nmfix <- ggplot(nm_stdr) + geom_line(mapping = aes(x = AgeN, y = nm)) + geom_ribbon(mapping = aes(x = AgeN,
ymax = ymax, ymin = ymin), alpha = 0.3) + theme_few()
ggplotly(plt_nmfix)Figure 2.35: Age varying natural mortality based on Gislason (mean values from above time series +/- 95% confidence intervals.
### save_as_lowestoft Five Year Sliding Window
nm_SAM <- nm_stdr
nm_SAM$year <- rep("all", nrow(nm_SAM))
nm_SAM <- reshape(nm_SAM[, c("AgeN", "nm", "year")],
# varying = as.factor(as.character(nm_stdr$AgeN)),
v.names = "nm",
timevar = "AgeN",
idvar = "year",
direction = "wide")
# for(i in 1:ncol(nm_SAM)){
# for(j in 1:nrow(nm_SAM)){
# nm_SAM[j, i] <- ifelse(is.na(nm_SAM[j, i]), -9, nm_SAM[j, i])
# }
# }
#### Make early "NA" Values plausible
for(i in 1:ncol(nm_SAM)){
for(j in 1:nrow(nm_SAM)){
nm_SAM[j, i] <- ifelse(is.na(nm_SAM[j, i]), median(c(nm_SAM[,i]), na.rm = T), nm_SAM[j, i])
}
}
save_as_lowestoft(df = nm_SAM[, colnames(nm_SAM)[-1]],
file_path = "Assessment Input/NaturalMortalityVariants/timeInvariantGislason_nm.dat",
title = "Natural mortality estimates from Gislason method, timeseries means by age for time-invariant profile.",
rando_numbers = paste(c(1, 4), collapse = "\t"),
year_range = paste(c(2002, (DataYear+1)), collapse = "\t"),
age_range = paste(c(1, 10), collapse = "\t"),
timeseries_type = 2)
#Table for HTML
kable(nm_stdr,
caption = "Mean mortalities by age for time invariant natural mortality in SAM.")| AgeN | StdErr | nm | ymax | ymin |
|---|---|---|---|---|
| 1 | 0.0679126 | 1.3259716 | 1.4590803 | 1.1928628 |
| 2 | 0.0203356 | 0.7813209 | 0.8211788 | 0.7414631 |
| 3 | 0.0134688 | 0.5691832 | 0.5955820 | 0.5427844 |
| 4 | 0.0117310 | 0.4549878 | 0.4779807 | 0.4319950 |
| 5 | 0.0106549 | 0.3967858 | 0.4176693 | 0.3759022 |
| 6 | 0.0099764 | 0.3642710 | 0.3838248 | 0.3447172 |
| 7 | 0.0124551 | 0.3313319 | 0.3557438 | 0.3069200 |
| 8 | 0.0148525 | 0.3312037 | 0.3603146 | 0.3020929 |
| 9 | 0.0121269 | 0.3077322 | 0.3315010 | 0.2839635 |
| 10 | 0.0182600 | 0.3071420 | 0.3429317 | 0.2713523 |
Based on scaling these time-invariant natural mortalities up an down, an optimised likelihood profile found that scaling all ages by a factor of 1.2 produced the best model fit. To see the details of this please see the working report, or script in the appendix.
### save_as_lowestoft Five Year Sliding Window
nm_scaled <- nm_stdr
nm_scaled$year <- rep("all", nrow(nm_scaled))
nm_scaled$nm <- nm_scaled$nm*1.2
nm_scaled <- reshape(nm_scaled[, c("AgeN", "nm", "year")],
# varying = as.factor(as.character(nm_stdr$AgeN)),
v.names = "nm",
timevar = "AgeN",
idvar = "year",
direction = "wide")
# for(i in 1:ncol(nm_scaled)){
# for(j in 1:nrow(nm_scaled)){
# nm_scaled[j, i] <- ifelse(is.na(nm_scaled[j, i]), -9, nm_scaled[j, i])
# }
# }
#### Make early "NA" Values plausible
for(i in 1:ncol(nm_scaled)){
for(j in 1:nrow(nm_scaled)){
nm_scaled[j, i] <- ifelse(is.na(nm_scaled[j, i]), median(c(nm_scaled[,i]), nm_scaled = T), nm_scaled[j, i])
}
}
save_as_lowestoft(df = nm_scaled[, colnames(nm_scaled)[-1]],
file_path = "Assessment Input/NaturalMortalityVariants/scaledTimeInvariantGislason_nm.dat",
title = "Time invariant natural mortality estimates from Gislason method scaled according to optimised likelihood profile runs.",
rando_numbers = paste(c(1, 4), collapse = "\t"),
year_range = paste(c(2002, (DataYear+1)), collapse = "\t"),
age_range = paste(c(1, 10), collapse = "\t"),
timeseries_type = 2)
#Table for HTML
kable(nm_stdr,
caption = "Time invariant natural mortality estimates from Gislason method scaled according to optimised likelihood profile runs.")| AgeN | StdErr | nm | ymax | ymin |
|---|---|---|---|---|
| 1 | 0.0679126 | 1.3259716 | 1.4590803 | 1.1928628 |
| 2 | 0.0203356 | 0.7813209 | 0.8211788 | 0.7414631 |
| 3 | 0.0134688 | 0.5691832 | 0.5955820 | 0.5427844 |
| 4 | 0.0117310 | 0.4549878 | 0.4779807 | 0.4319950 |
| 5 | 0.0106549 | 0.3967858 | 0.4176693 | 0.3759022 |
| 6 | 0.0099764 | 0.3642710 | 0.3838248 | 0.3447172 |
| 7 | 0.0124551 | 0.3313319 | 0.3557438 | 0.3069200 |
| 8 | 0.0148525 | 0.3312037 | 0.3603146 | 0.3020929 |
| 9 | 0.0121269 | 0.3077322 | 0.3315010 | 0.2839635 |
| 10 | 0.0182600 | 0.3071420 | 0.3429317 | 0.2713523 |
2.3.7 Catch weight at age (CWAA)
This section is currently a work in progress - need to figure out the appropriate method to raise intercatch data according to the allocations, manually, to be able to automatically update the section each year (instead of copy-pasting intercatch ouput for multiple tables every year).
cw_df <- cw2.4 Cohort Tracking
We can track the progression of different cohorts through time from both commercial and survey data.
Let’s start with catches.
# === Data Prep ====
catchNumLng <- reshape(data = catchNum, varying = list(colnames(catchNum)[2:length((colnames(catchNum)))]),
v.names = c("Number"), timevar = "Age", direction = "long")
# catchNumLng$Age <- as.factor(as.character(catchNumLng$Age)) catchNumLng$Age
# <- factor(catchNumLng$Age, levels = c('1', '2', '3', '4', '5', '6', '7', '8',
# '9', '10')) catchNumLng$Year <- as.factor(as.character(catchNumLng$Year))
catchNumLng$Year <- as.numeric(as.character(catchNumLng$Year))
catchcohortsLng <- reshape(data = catchcohorts, varying = list(colnames(catchcohorts)[2:length((colnames(catchcohorts)))]),
v.names = c("Age"), timevar = "Cohort", direction = "long")
# catchcohortsLng$Age <- factor(catchcohortsLng$Age, levels = c('1', '2', '3',
# '4', '5', '6', '7', '8', '9', '10')) catchcohortsLng$Age <-
# as.factor(as.character(catchcohortsLng$Age), levels = c('1', '2', '3', '4',
# '5', '6', '7', '8', '9', '10')) catchcohortsLng$Year <-
# as.factor(as.character(catchcohortsLng$Year))
catchcohortsLng$Year <- as.numeric(as.character(catchcohortsLng$Year))
# =====
# === Figure Cohorts over time from catches ====
plt_catchCohorts <- ggplot() + geom_line(data = catchcohortsLng, mapping = aes(x = Year,
y = Age, group = Cohort), linetype = "dotted") + geom_point(data = catchNumLng,
mapping = aes(x = Year, y = Age, size = Number)) + theme_clean() + coord_cartesian(xlim = c(1999,
DataYear)) + theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust = 1))
ggplotly(plt_catchCohorts)Figure 2.36: Bubble plot showing numbers (size of bubble) at age (y-axis), over time (x-axis) in catches. The dotted lines show the progression of co-horts where large and small numbers at age can be tracked year to year.
# ====# === Figure Numbers per age group over time (less intuitive than above cohort
# plot) ====
plt_catchTime <- ggplot() + geom_line(data = catchNumLng, mapping = aes(x = Year,
y = Number/1e+06, group = Age)) + facet_grid(rows = vars(Age), scales = "free_y") +
ylab("Number (millions)") + theme_few() + theme(axis.text.x = element_text(angle = 90,
hjust = 1, vjust = 0.5))
ggplotly(plt_catchTime)Figure 2.37: Numbers per age over time.
# =====Now, let’s do the same with survey estimations of numbers at age.
# === Data Prep ====
survcohorts <- read.csv(file = "Data/ple.27.21-32_CohortMatrix_1998-WorkingYear.csv",
header = T)
survcohorts$Year <- as.factor(as.character(survcohorts$Year))
survcohortsLng <- reshape(data = survcohorts, varying = list(colnames(survcohorts)[2:length((colnames(survcohorts)))]),
v.names = c("Age"), timevar = "Cohort", direction = "long")
# survcohortsLng$Age <- as.factor(as.character(survcohortsLng$Age))
survcohortsLng$Year <- as.factor(as.character(survcohortsLng$Year))
survcohortsLng <- survcohortsLng[survcohortsLng$Age <= 6, ]
# =====
# === Plot =====
plt_survCohorts <- ggplot() + geom_point(data = survTunLng, mapping = aes(x = Year,
y = Age, size = Index)) + geom_line(data = survcohortsLng, mapping = aes(x = Year,
y = Age, group = Cohort), linetype = "dotted") + theme_classic() + facet_grid(rows = vars(Survey)) +
theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust = 1))
ggplotly(plt_survCohorts)Figure 2.38: Bubble plot showing numbers (size of bubble) at age (y-axis), over time (x-axis) estimated from surveys. The dotted lines show the progression of co-horts where large and small numbers at age can be tracked year to year.
k <- append(k, "survcohortsLng")
rm(list = ls()[!ls() %in% k])
# ======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