REALISTIC MODELS I N POPULATION ECOLOGY
255
into account size, influence zone and interaction. His model yields good
agreement with observations of five stands. Finally, Walters and
Bunnell (1971) have developed a general computer program entitled
FARMS, which models ecological events in a large area. That area is
subdivided and annual computations are performed for each region.
Species-dependent plant production is computed, plant utilization as a
function of animal species and numbers is determined, and for each
animal species agesex distributions and mortality are found. The
program computes the effects of harvesting animals and plants. It
appears to be useful both for solving problems in land utilization and as
a teaching tool.
I X . CONCLUDING REMARKS
Realistic population models, based on the physiological, ecological
and social behavior of individuals in the population, provide insight and
have predictive validity. They are useful in determining optimal
strategies for pest control, harvesting, preservation of species etc. Often
the situations are of such complexity that their mathematical models
can only be solved with the aid of large digital computers. Indeed, in
such cases, the only way to gain insight into the combined effects of
many interacting factors is to employ models and computers. One
should, however, be careful to distinguish the real world and the mathematical model. Only those aspects of the real world which are accurately
represented in the model are reflected therein and, as we are well aware,
nature is exceedingly complex and rich in phenomena.
To date, realistic models have been employed with some measure of
success. The objectives of workers in this field remain quite far in
advance of present achievements; however, I believe the current state of
understanding and mathematical capability are adequate to press
forward vigorously in this very important area. The problems which
remain are very interesting and difficult; the gains accruing from solving
these problems could be great.
ACKNOWLEDQEMENTS
I would like to thank Professor Conrad A. Istock of the University
of Rochester for his most helpful discussions of the manuscript and Dr
Richard Smallwood of the Xerox Palo Alto Research Center for his
interesting comments on modeling. Thanks are also due Dr Guiliana
Lavendel of the Xerox Palo Alto Research Center Library for her
invaluable assistance in locating references, and to Dr Ralph Kimball
255
into account size, influence zone and interaction. His model yields good
agreement with observations of five stands. Finally, Walters and
Bunnell (1971) have developed a general computer program entitled
FARMS, which models ecological events in a large area. That area is
subdivided and annual computations are performed for each region.
Species-dependent plant production is computed, plant utilization as a
function of animal species and numbers is determined, and for each
animal species agesex distributions and mortality are found. The
program computes the effects of harvesting animals and plants. It
appears to be useful both for solving problems in land utilization and as
a teaching tool.
I X . CONCLUDING REMARKS
Realistic population models, based on the physiological, ecological
and social behavior of individuals in the population, provide insight and
have predictive validity. They are useful in determining optimal
strategies for pest control, harvesting, preservation of species etc. Often
the situations are of such complexity that their mathematical models
can only be solved with the aid of large digital computers. Indeed, in
such cases, the only way to gain insight into the combined effects of
many interacting factors is to employ models and computers. One
should, however, be careful to distinguish the real world and the mathematical model. Only those aspects of the real world which are accurately
represented in the model are reflected therein and, as we are well aware,
nature is exceedingly complex and rich in phenomena.
To date, realistic models have been employed with some measure of
success. The objectives of workers in this field remain quite far in
advance of present achievements; however, I believe the current state of
understanding and mathematical capability are adequate to press
forward vigorously in this very important area. The problems which
remain are very interesting and difficult; the gains accruing from solving
these problems could be great.
ACKNOWLEDQEMENTS
I would like to thank Professor Conrad A. Istock of the University
of Rochester for his most helpful discussions of the manuscript and Dr
Richard Smallwood of the Xerox Palo Alto Research Center for his
interesting comments on modeling. Thanks are also due Dr Guiliana
Lavendel of the Xerox Palo Alto Research Center Library for her
invaluable assistance in locating references, and to Dr Ralph Kimball