Estimate the population reduction of a taxon over the most recent three
generations (or ten years, whichever is longer), following section 4.5 of the
IUCN Red List guidelines. A decline model is fitted to the population
estimates and used to read off the population size at the start and end of the
window, and the reduction is the proportional drop between them. The same
calculation gives the estimated continuing decline of criterion C1 and B when
a different window is set through years.
Arguments
- population
Numeric vector of population sizes (number of mature individuals, or an index that scales with it).
- time
Numeric vector of the years the sizes refer to, the same length as
population.- generation_length
Generation length in years. See
rl_generation_length().- model
Decline pattern,
"exponential"(default) or"linear".- assessment_year
The year taken as the present. Default is the most recent year in
time.- years
Length of the assessment window in years. Default is the longer of three generations or ten years.
- subcriterion
Which criterion A subcriterion sets the thresholds for
category_a:"A2"(default),"A1","A3"or"A4".
Value
A one row tibble with the model and window used, the fitted population
sizes at the start and end of the window (n_start, n_present), the
reduction (a proportion) and reduction_pct, and category_a.
Details
Two decline patterns are supported.
"exponential"a constant proportional rate of decline, fitted as a log-linear regression of population size on time. Appropriate when the rate of loss stays proportional to population size, for example a constant harvest fraction.
"linear"a constant number of individuals lost per year, fitted as a linear regression of population size on time. Appropriate when a fixed amount is removed each year, for example a fixed area of habitat lost.
With exactly two estimates the fit passes through both points, reproducing the two-point formulas in the guidelines. With more estimates the regression smooths natural variation, and the reduction is still read over the most recent window.
The category_a column reports the most threatened band the reduction reaches
for the chosen subcriterion, or NA when it reaches none (including an
increase). The criterion A thresholds are 50/70/90 percent (VU/EN/CR) for
"A1" and 30/50/80 percent for "A2", "A3" and "A4". This is the
magnitude threshold only, not a full assessment.
References
IUCN Standards and Petitions Committee. 2024. Guidelines for Using the IUCN Red List Categories and Criteria. Version 16, section 4.5. https://www.iucnredlist.org/documents/RedListGuidelines.pdf
Examples
# Guidelines example: 20000 in 1961 and 14000 in 1981, generation length 20,
# assessed in 2001 so the three generation window runs 1941 to 2001.
# Exponential decline gives a 65.7 percent reduction.
rl_reduction(population = c(20000, 14000), time = c(1961, 1981),
generation_length = 20, model = "exponential",
assessment_year = 2001)
#> # A tibble: 1 × 11
#> model subcriterion generation_length window_years year_start year_present
#> <chr> <chr> <dbl> <dbl> <dbl> <dbl>
#> 1 exponenti… A2 20 60 1941 2001
#> # ℹ 5 more variables: n_start <dbl>, n_present <dbl>, reduction <dbl>,
#> # reduction_pct <dbl>, category_a <chr>
# The same data under a linear decline gives 69.2 percent.
rl_reduction(population = c(20000, 14000), time = c(1961, 1981),
generation_length = 20, model = "linear",
assessment_year = 2001)
#> # A tibble: 1 × 11
#> model subcriterion generation_length window_years year_start year_present
#> <chr> <chr> <dbl> <dbl> <dbl> <dbl>
#> 1 linear A2 20 60 1941 2001
#> # ℹ 5 more variables: n_start <dbl>, n_present <dbl>, reduction <dbl>,
#> # reduction_pct <dbl>, category_a <chr>