Change and variability are integral parts of natural ecosystems. Thus,
how can one determine whether the state of a recovering ecosystem falls
within some desirable range or still represents an unacceptable extreme?
Studies of recovering ecosystems are likely to use a quantitative measure
of the abundance of an indicator species as a means of monitoring the recovery
process.
Aquatic systems might rely on density or diversity of diatoms, aquatic
insects, protozoa, or fish as indicators.
Terrestrial systems might rely on particular plant species, small mammals,
birds, or even larger vertebrates as measures.
Inherent problems are natural variation in quantitative, population-level
indices, and potential sampling problems. So, it is important to determine
how much temporal variation there is in these measures and to pick a time
scale on which to study this variation.
How do you pick a time-scale? A basic allometric relation linking body
mass to population processes is that the period length of population cycles
will characteristically scale according to the fourth root of body mass.
So, if the organisms of interest in a monitoring study are very small,
e.g. algae or zooplankton, a weekly or monthly time scale might be as important
as one of a longer term.
Variation of population sizes of indicator species on an annual scale
is much more likely to be of significance for most studies of ecosystem
recovery. Variation on an annual cycle may arise from several sources:
biotic interactions, natural succession, climatic fluctuations, natural
population cycles in both plants and animals, sampling error (pseudoreplication,
see Hurlbert, S.H. 1984. Ecological Monograph 54.).
If variation is the rule, then what is the role of stability? Can natural
ecosystems every be considered to be stable?
Stability, in the sense used by ecological theory, implies that there
are "one or more equilibrium points or limit cycles (1) at which the system
remains when faced with a disturbing force or (2) to which it returns if
perturbed by the force. This quantitative definition differs from a qualitative
one that involves only the presence or absence of a species, but this latter
concept of persistence would probably be of less interest for studies trying
to document the recover of ecosystems. Connell and Sousa (1983, American
Naturalist 121) concluded that the most appropriate way to measure stability
was to consider the degree of variation in population number on a time
scale long enough for at least one complete turnover of all individuals.
In this way, stability of populations is scaled to the life history of
each species, allowing comparisons among different populations. Connell
and Sousa found 49 long-term studies that met their criteria. There analysis
revealed a continuum of temporal variability with no real demarcation that
might separate stable and unstable populations. Considered by taxon, the
results suggested that parasites showed uniformly low variability. Schoener
(1985, American Naturalist 126), using similar criteria, showed that lizard
population sizes were less variable over time than most other taxa for
which data were available. He also made a case for including birds among
those species for which relatively little variation is observed.
Unfortunately, there are no obvious candidates for species that, by
virtue of their relatively low variability in population size, would be
of particular interest as indicator species in studies of recovering ecosystems.
Parasites are not useful indicators since it is variation in their hosts
that is usually of interest in monitoring studies. Even among lizards and
birds there was much variation.
Within species there is significant variation in population size from
different localities. It might be possible to use periodicity and range
of some insect populations on a control site that is near the recovery
site as a comparison.
Monitoring of abundance has two components: attempting to determine
the long-term mean and the degree of variation around the mean. This generally
requires investing a number of years in data collection.
What then is an "appropriate length of time?"
Inouye presents two examples: Christmas bird counts of House Sparrows
from 1947 to 1984 at Wichita, Kansas, and the number of flowers of the
Aspen Sunflower from 1974 to 1984 at the Rocky Mountain Biological Lab.
Significant variation occurred in both data sets.
No. of birds recorded in the census area, when divided by the number
of observers, ranged from 2.8 to 106.8 (mean 36.6). The number of aspen
sunflowers varied from 3 to 3450 (mean 1396.1).
How closely might we have approached these long-term means if we had
only conducted shorter studies? This question can be answered by subsampling
all possible studies of a particular length from the total data set.
Discussion of Figures 1-6 from Inouye.
There is only one possible 38-year-long sample of the house sparrow data, but there are 2 possible subsamples of 37 years, 3 subsamples of 36 years, and 37 subsamples of 2-year duration.
White-tailed Deer Very Successful Transplant from other states
Wild Turkey Very Successful Transplant from Missouri
Mallard Unsuccessful Captive Rearing-Hard Release
Ruffed Grouse Limited Success Transplant from Minn. & Wisc.
Giant Canada Goose Successful Captive rearing-soft release
Prairie Chicken Ssuccessful - Transplant from Kansas
Barn Owl Unsuccessful Captive Rearing-Soft Release
River Otter Limited Success Transplant from Louisiana
Trumpeter Swan Just Starting Captive Rearing-soft release
Peregrine Falcon Limited Success Captive Rearing-soft release
Sandhill Crane Just Starting Habitat enhancement - Passive
INTRODUCTION OF EXOTIC SPECIES
Ringed-neck Pheasant Successful Captive Rearing - Hard Release
Gray Partridge Successful Habitat Change - Passive
A MVP is the threshold number of organisms that ensures, at some defined
level of risk, that a population will persist for a given time interval
at a particular location.
Conventional standards for MVP's for vertebrate animals include:
1. Greater than 90% certainty of long-term (usually centuries) persistence.
2. Population maintenance in nature with no significant demographic
or genetic manipulation.
3. Retention of replacement levels of immediate fitness (vigor, fertility,
fecundity) and sufficient genetic variation to adapt by natural selection
to changing environments.
Based on the last criterion, many population biologists believe that
at least several hundred individuals are necessary for an MVP to be established.
Not all potential release sites will be large enough to support several
hundred individuals at carrying capacity. In such cases, it may be necessary
to manage several small populations as a single metapopulation.
A metapopulation consists of a constellation of small populations that
interact loosely, but experience environmental effects independently, and
have differential probabilities of dispersal, establishment, growth, and
extinction.
Managing several small populations as a metapopulation could involve
providing corridors for the natural dispersal of individuals among populations,
direct relocation, and including captive individuals at zoological parks
as a population.
A simulation model called VORTEX (Lacy R. 1991. VORTEX, Simulation model of stochastic population change. Version 8.0. Brookfield: Chicago Zoological Park.) has been developed and used to evaluate some reintroductions of endangered and threatened species.