Description Usage Arguments Details Value Note References See Also Examples

Creates a smoothed resource surface from a covariate of a mask. Smoothing entails summing the value in each pixel weighted by a detection kernel centred on the focal pixel. The detection kernel represents home-range utilization with spatial scale sigma. The resulting surface is equivalent to the denominator used by Royle et al. (2013) to normalize site-specific detection.

1 2 |

`mask` |
secr habitat mask object (single-session) |

`sigma` |
numeric spatial scale of home range model |

`alpha2` |
numeric coefficient of spatial covariate |

`usecov` |
character name of resource covariate |

`detectfn` |
integer or character code for detection function |

`z` |
numeric shape parameter of home range model |

`inverse` |
logical; if TRUE the reciprocal of smoothed resource is returned |

`scale` |
logical; not used |

`detectfn`

may be uniform (‘UN’) or one of the
cumulative-hazard functions (‘HHN’, ‘HHR’, ‘HEX’,
‘HAN’, ‘HCG’) (or integer codes 4, 14:18; see
detectfn).

The default ‘HHN’ corresponds to a halfnormal function on the hazard scale, or a bivariate circular normal home range.

If `usecov`

is not named then it takes the value 1.0 for all points
on the mask and zero otherwise.

The Rsurface can be used implicitly to normalize detection probability when
fitting a model with detector-specific covariate equal to
`usecov`

(see details, but the process is intricate and not
fully documented).

An object with class c(‘Rsurface’, ‘mask’, ‘data.frame’) and covariate ‘Resource’
(other covariates are retained from the input mask). The attribute
‘scale’ is 1.0 if `scale = FALSE`

; otherwise it is the average of the
resource over the masked area.

Consider a focal pixel **s** and another point in the habitat mask
**x**, with distance *d
= |x-s|*. Weights are given by a kernel *f(d)*. Typically the kernel
will be halfnormal *f(d) =
exp(-d^2/(2 σ^2))* (detectfn = ‘HHN’) or exponential
*exp(-d/σ)* (detectfn =
‘HEX’) (see detectfn for other possibilities).

If *z(x)* represents the covariate value at point
**x**, the summed resource availability at **s** is given by

*R(s) = ∑ f(d) exp(α2 z(x)).*

This corresponds to the denominator of eqn 4 in Royle et al. (2013).

By default, the numerical values reported by `Rsurface`

are not raw
*R* values. If `scale = TRUE`

, values are standardized by
dividing by the mean: *R'(s) = R(s) / (∑ R(s) / n)* where *n* is the number of
pixels. Values of *R'(s)* are centred on 1.0.

If `inverse = TRUE`

, the numeric values are *1 / R'(s)*
or *1 / R(s)* as determined by `scale`

.

Royle, J. A., Chandler, R. B., Sun, C. C. and Fuller, A. K. (2013)
Integrating resource selection information with spatial
capture–recapture. *Methods in Ecology and Evolution* **4**,
520–530.

`mask`

, `plot.Rsurface`

,
`spotHeight`

, details

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 | ```
## create binary covariate (0 outside habitat)
msk <- make.mask(traps(possumCH), buffer = 800)
covariates(msk) <- data.frame(z = as.numeric(pointsInPolygon
(msk,possumarea)))
## derive and plot "resource availability"
Rs <- Rsurface(msk, sigma = 100, usecov = 'z')
plot(Rs, plottype = 'contour', col = topo.colors(10))
lines(possumarea)
## Not run:
spotHeight(Rs, dec = 2)
## End(Not run)
``` |

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.