Turns the true abundances of a simulation into what a survey would record. Each individual
is detected in each of visits visits with probability detection, independently, so the
count of a species at a site in a visit is binomial and a species with more individuals is
more likely to be found, as in the models of Royle & Nichols (2003) and Royle (2004). The
record keeps the largest count over visits, or only whether the species was detected. The
same applies to every step of series, with new visits at each step. Use it inside
simulate in abc_table(), with detection drawn from a prior, so that the reference table
carries the same observation error as the data.
Usage
observe_detection(x, detection, visits = 3, record = c("counts", "presence"))Arguments
- x
An
mc_simobject or anmc_data()object with true abundances.- detection
Probability that an individual is detected in a visit: one number, or one per species, named by species.
- visits
Number of visits to each site.
- record
"counts"for the largest count over visits or"presence"for detected at least once (1) or not (0).
Value
An mc_data() object with the recorded comm and series, and the attribute
detection.
References
Royle, J. A., & Nichols, J. D. (2003). Estimating abundance from repeated presence–absence data or point counts. Ecology, 84, 777–790. doi:10.1890/0012-9658(2003)084[0777:EAFRPA]2.0.CO;2
Royle, J. A. (2004). N-mixture models for estimating population size from spatially replicated counts. Biometrics, 60, 108–115. doi:10.1111/j.0006-341X.2004.00142.x
See also
Other inference:
abc_choose(),
abc_cv(),
abc_diagnose(),
abc_fit(),
abc_ppc(),
abc_recovery(),
abc_table(),
fold_summarise(),
spatial_folds()
Examples
tree <- treeshapesim::sim_tree_shape(8)
sim <- simulate_metacommunity(landscape_grid(side = 4), species_traits(tree, 0.5),
competition_matrix(tree, "none"), dispersal = 0.05,
n_steps = 40)
seen <- observe_detection(sim, detection = 0.1, visits = 3)
c(true = sum(sim$abundance > 0), recorded = sum(seen$comm > 0))
#> true recorded
#> 91 66
