ATLSS: Spatially Explicit Species Index (SESI) Models
Introduction
Landscape level breeding potential indices (BPI) and
foraging potential indices (FPI) have been created for particular Everglades
species as part of the Across Trophic Level System Simulation (ATLSS)
Program. These are referred to under
the general name of spatially-explicit species index (SESI) models, and have
been developed as an attempt to provide a basis for quantifying assessment of the effects of different water
regulation plans on these species in the context of Everglades restoration.
SESI models are similar to Habitat Suitability Index (HSI)
models both in that population response is predicted by a set of relationships
between a given species and its environment and in that habitat indices are
quantified by an index value. Thus,
SESI values range from 0.0 to 1.0, with 1.0 indicating the most favorable
conditions. SESI models differ from
most HSI models in the following ways.
(1) They are flexible in that they can focus either on one particular
part of the life cycle, such as breeding or foraging, or on the life cycles as
a whole. (2) They incorporate temporal
changes in the environment, and can change from year to year, reflecting the
unique conditions of any particular year.
Thus, for example, short-term variations in hydrology are incorporated
in the index for a particular year (3) They are based on a 'landscape
structure' which, once established, can be used to model the responses of any
species in the system. (4) They provide a landscape index map rather than just
a single index or set of indices.
For SESI models, the landscape is divided into equal‑sized
spatial cells or pixels (500 x 500 m), each pixel having a suite of values that
correspond to the parameters included in the model. Suitability of each pixel
for the particular index is determined by combining a set of rules. Generally,
these rules are of two types. (1) There are binary (0/1 or yes/no) rules which
invoke known or estimated limits on the suitability of habitat or environmental
conditions concerning a species. (2)
There are quantitative rules ‑ whereby having met the basic requirements
of a species the habitat parameters are given a value that reflects their
relative potential for breeding, foraging, or both. These rules are then combined to create an overall index value
between 0 and 1.
The primary output of a SESI model is a visual
representation of the landscape with color‑coded values assigned to each
pixel. One way to use the model is to
compare different management scenarios.
For example, the application of these models for assessing Everglades
restoration plans involves comparing a baseline scenario that assumes no
restoration with an 'alternative' management plan. The scenarios differ only in hydrologic variables, primarily
water depths across the landscape.
Hydrological data for the base scenario assumes no changes to the
current water management practices. The
alternative scenarios incorporate proposed changes to management
activities. The predicted effects of
one scenario could be compared to those of the second by simply subtracting the
index value for each pixel calculated under base conditions from the value for
the same pixel calculated under the alternative scenario. In practice three maps are generated,
representing the index values under the base scenario, the values under the
alternative scenario, and the difference between the two. For the difference map the values range from
‑1 to 1.
This type of methodology represents that of relative
assessment. Relative assessment of the
suitability of habitat under alternative management plans produces meaningful
results, even when knowledge of ecological details are insufficient to assess
habitat suitability of a pixel in an absolute sense (DeAngelis et al. 1998 Ecosystems 1:64-75, Curnutt et al., 2000
Ecological Applications 10:1849-1860). The emphasis of this approach, therefore, is
on comparing the spatial pattern of differences between two alternative
management plans for suitability, in terms of whether a cell is good for
foraging, breeding, or maintaining the overall life cycle of the
population.
For predictive simulations of the SESI model, projected
daily water level for each cell is provided by the South Florida Water
Management Hydrology Model for a 31-year period, based on historical weather
patterns (1965-1995) but reflecting proposed modifications to water delivery
schedules and infrastructure. The ATLSS high resolution hydrology model is used
to translate the SFWM Model water depths at a 2-mi scale of resolution to finer
resolutions needed by our models.