My R plot of a time series is contradictory with the same plot on a larger time span, why is that?

I am currently working on a univariate time series on R. When I plot the time series over the whole time span, that is, from 1981 to 2022, I get that there is a decreasing trend in the second sub-period (2000 - 2022), whereas when I restrict to this sub period and plot it, it seems constant and even increasing. Why is that ? Where is my mistake ?

  production_2 <-serie_elec$`Production brute d'électricité nucléaire (en GWh)`
  prod_2 <- ts(serie_elec, frequency=12)
  prod_2 <- ts(production_2, start=c(1981,1) , end=c(2022,11), frequency=12)
  summary(prod_2)
  plot.ts(prod_2)

Here is the first plot , that of the whole time series.

 production_periode_2 <-serie_elec$`Production brute d'électricité nucléaire (en GWh)`
 prod_periode_2 <- ts(serie_elec, frequency=12)
 prod_periode_2 <- ts(production_periode_2, start=c(2000,1) , end=c(2022,11), 
 frequency=12)
 summary(prod_periode_2)
 plot.ts(prod_periode_2)

Here is the second plot , that of the time series on the 2nd sub period, from 2000 to 2022.

Here is the structure of both datasets for reproducibility :

  > dput(head(prod_2))
  structure(c(22951.429, 21465.026, 19334.531, 19319.365, 19923.664, 
   21275.248), tsp = c(1981, 1981.41666666667, 12), class = "ts")
  > dput(head(prod_periode_2))
  structure(c(22951.429, 21465.026, 19334.531, 19319.365, 19923.664, 
  21275.248), tsp = c(2000, 2000.41666666667, 12), class = "ts")

Thanks in advance for your help !

You are giving ts() the same data both times:

serie_elec$`Production brute d'électricité nucléaire (en GWh)`

The first time you are saying that that data starts in 1981. The second time you are NOT selecting a subperiod. Instead, you are giving it the exact same data, but this time saying that data starts in 2000.

You should use a subset of that data in your second call to ts().

Another comment, when you assign to the same variable on two consecutive lines, like this:

prod_periode_2 <- ts(serie_elec, frequency=12)
prod_periode_2 <- ts(production_periode_2, start=c(2000,1) , end=c(2022,11), frequency=12)

The first assignment is just overwritten.