-
Notifications
You must be signed in to change notification settings - Fork 0
/
README.Rmd
156 lines (116 loc) · 7 KB
/
README.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
---
output: github_document
---
<!-- README.md is generated from README.Rmd. Please edit that file -->
```{r, include = FALSE}
knitr::opts_chunk$set(
collapse = TRUE,
comment = "#>",
fig.path = "man/figures/README-",
out.width = "100%"
)
```
# Implementation of the InSiGHTS framework
<!-- badges: start -->
[](https://github.com/iiasa/insights/actions/workflows/R-CMD-check.yaml)
[](https://lifecycle.r-lib.org/articles/stages.html#experimental)
[](https://creativecommons.org/licenses/by/4.0/)
[](https://app.codecov.io/gh/iiasa/insights?branch=main)<!-- badges: end -->
This R-package provides a IIASA implementation of the InSiGHTS Index of Habitat Availability. The index
captures the amount of suitable habitat within the current or a future range of a species. This range can be taken either from existing range maps (e.g. IUCN) or from estimates obtained through species distribution models.
<img src="man/figures/insights_schematic.png" alt="Schematic" align="right" width="300"/>
In it's basic configuration, the InSiGHTS framework combines the climatic suitability from a SDM with a area of habitat (AOH) refinement to obtain the suitable habitat for each time steps.
The InSiGHTS Index of Habitat Availability can then be defined for any given species $s$ and timestep $t$ as:
$Insights_{s,t} = \frac{AOH_{s,t} - AOH_{s, t_{ref}}}{AOH_{s, t_{ref}}}$, where $t_{ref}$ indicates a reference or starting year.
More information on the InSiGHTS framework can be found in [Pearson et al. 2004](https://doi.org/10.1046/j.1466-822X.2003.00042.x), [Rondini and Visconti 2015](https://doi.org/10.1111/cobi.12532), [Visconti et al. 2016](https://doi.org/10.1111/conl.12159) or [Baisero et al. (2021)](https://doi.org/10.1016/j.oneear.2020.05.015).
The package is part of the [IIASA-BEC](https://iiasa.ac.at/programs/bnr/bec) suite of biodiversity indicators and is coupled with the [ibis.iSDM](https://iiasa.github.io/ibis.iSDM/) species distribution model.
## Installation
You can install the development version of Insights from [GitHub](https://github.com/) with:
``` r
# install.packages("devtools")
devtools::install_github("iiasa/insights")
```
The package depends on the [ibis.iSDM](https://iiasa.github.io/ibis.iSDM/) package, which is currently only available via github.
## Basic usage and examples
```{r Load packages for fitting a model, echo=TRUE, message=FALSE}
# Basic packages for use
library(ibis.iSDM)
library(insights)
library(glmnet)
library(terra)
```
Now we use the **ibis.iSDM** package to train a simple SDM and apply the InSiGHTS on it. The assumption here is that the SDMs
are created using climatic variables (temperature, precipitation, etc.) only so as to create a climatic envelope model. The refinement with
time series of land-use is done posthoc on the resulting prediction.
Note that this also works on any other range estimate provided directly as a SpatRaster.
```{r Train a simple SDM, echo=TRUE, message=FALSE}
# Load test data from ibis.iSDM package
background <- terra::rast(system.file('extdata/europegrid_50km.tif', package='ibis.iSDM',mustWork = TRUE))
virtual_points <- sf::st_read(system.file('extdata/input_data.gpkg', package='ibis.iSDM',mustWork = TRUE),'points',quiet = TRUE)
# Get some future predictors
ll <- list.files(system.file('extdata/predictors_presfuture/',package = "ibis.iSDM",mustWork = TRUE),
full.names = T)
# Load the same files future ones
suppressWarnings(
pred_future <- stars::read_stars(ll) |> stars:::slice.stars('Time', seq(1, 86, by = 10))
)
sf::st_crs(pred_future) <- sf::st_crs(4326)
# Get only climatic predictors and take the first time slot
pred_climate <- pred_future |> stars:::select.stars(bio01, bio12)
predictors <- ibis.iSDM:::stars_to_raster(pred_climate, 1)[[1]]
# Add some pseudo-absence data
virtual_points <- ibis.iSDM::add_pseudoabsence(virtual_points,
field_occurrence = 'Observed',
template = background)
# Now train a small little model
fit <- distribution(background) |> # Prediction domain
add_biodiversity_poipa(virtual_points) |> # Add presence-only point data
add_predictors(predictors) |> # Add simple predictors
engine_glmnet() |> # Use glmnet for estimation
train(verbose = FALSE) |> # Train the model
threshold(method = "perc", value = .33) # Percentile threshold
# --- #
# Now load some fractional land-use layers relevant for the species
# Here we assume the species only occurs in Grassland and Sparse vegetation
lu <- c(
terra::rast(system.file('extdata/Grassland.tif', package='insights',mustWork = TRUE)),
terra::rast(system.file('extdata/Sparsely.vegetated.areas.tif', package='insights',mustWork = TRUE))
) / 10000
# Summarize
out <- insights_fraction(range = fit,lu = lu)
plot(out, col = c("grey90", "#FDE8A9", "#FBD35C", "#D1C34A", "#8EB65C",
"#56AA71", "#59A498", "#5C9EBF", "#5C8BAE", "#597182"),
main = "Suitable habitat")
# Summarize
insights_summary(out)
```
Of course it is also possible to directly supply a multi-dimensional gridded file
using the *stars* package or directly through the ibis.iSDM scenario functionalities
(see example below).
```{r With scenario}
# Create a future scenario
sc <- scenario(fit) |>
add_predictors(env = pred_climate, transform = 'scale', derivates = "none") |>
threshold() |>
project()
# --- #
# Now apply insights using time series of future land use
lu <- pred_future |> stars:::select.stars(primn, secdf)
# Normalize for the sake of an example. Note that fractions are needed!
lu <- ibis.iSDM::predictor_transform(lu, "norm") |> round(2)
out <- insights_fraction(range = sc,
lu = lu)
# Summarize
o <- insights_summary(out)
plot(o$suitability~o$band, type = "b",
main = "InSiGHTS index",
ylab = "Suitable habitat relative to 2015",
xlab = "Year")
```
## Citations
See [CITATION.cff] file
P. Visconti, M. Bakkenes, D. Baisero, T. Brooks, S.H.M. Butchart, L. Joppa, R. Alkemade, M. Di Marco, L. Santini, M. Hoffmann, C. Rondinini
_Projecting global biodiversity indicators under future development scenarios_ Conserv. Lett., 9 (2016), pp. 5-13 [DOI](https://doi.org/10.1111/conl.12159)
C. Rondinini and P. Visconti, _Scenarios of large mammal loss in Europe for the 21st century_ Conserv. Biol., 29 (2015), pp. 1028-1036 [DOI](https://doi.org/10.1111/cobi.12532)
## Acknowledgement <a href="https://iiasa.ac.at"><img src="man/figures/IIASA_PNG_logo_blue.png" alt="IIASA" align="right" width="300"/></a>
**InSiGHTS** is developed and maintained by the [Biodiversity, Ecology and Conservation group](https://iiasa.ac.at/programs/bnr/bec) at the International Institute for Applied Systems Analysis (IIASA), Austria.