satellite 
Chaos. 
1allenge 
he self- 
ng data, 
r model 
emulate 
irtificial 
is paper 
between 
] future 
ith tools 
hm for 
ges that 
rming), 
of plant 
employs 
1 source 
1ensions 
escribed 
ction at 
put for 
ehavior. 
op eco- 
ed to be 
that can 
genetic 
genetic 
servable 
ork for 
into an 
Object 
d from 
ed into 
rules are 
  
applied. To work properly, it will require artificial neural 
networks that can process data elements whose 
instantaneous attributes could assume any of several 
conditions, at the same time. Clearly, this concept pushes 
the frontiers of database technology beyond today's 
capabilities. 
Status of Genome Research 
If a genetic algorithm template can be created, static data 
from past research could be integrated with future data. 
All that has been learned about an ecosystem would thus 
find service as active digital information. The intent is to 
develop a digital library beginning with each species' 
genetic attributes and functions, and ending with its 
community interactions and environmental responses. The 
integration of species libraries would form the larger and 
more complex communities and ecosystems in which they 
reside, and would allow dynamic data to be inserted and 
updated, as needed. At present, the template is crudely 
articulated. There are, for example, quantum leaps in 
knowledge required to translate genetic base pairs into 
their control over physiologic responses...but this is where 
the most rapid technology is evolving. 
The human genome project was formulated ten years ago. 
It was considered an almost impossible job because of its 
complexity. ^ Today, this genetic map is nearing 
completion, and the race is on to map other species. 
Techniques are evolving that apply to whole groups of 
related species, eliminating the need to map each one 
separately. These maps are becoming commonplace. 
More importantly, the networks and functions behind the 
genes are gradually being unraveled. 
Although research is currently driven by a desire to 
understand the human complex, the research community 
believes that other forms of life will soon be examined. 
By the end of this century, genome research is expected to 
migrate toward agricultural and pharmacological species, 
as well as to endangered and keystone species. 
The digital information that underlies genetic functions is 
represented by a simple string of proteins symbolized as 
G/A/T/C (guanine, adenine, thymine, cytocine). This 
string is the fundamental data structure of any organism's 
biology. Because of this commonality, base pairs can be 
linked to the physiologic and metabolic functions they 
control, and in theory, at least, every species can be 
integrated into an ecosystem model through these object 
attributes. Once the linkages are mapped, as complicated 
as they are, the data structure permits abstraction to higher 
levels of system integration. 
Status of Ecosystem Research 
Carpenter et al. (1995) say that predicting responses of 
ecosystems to perturbation is among the greatest 
challenges in ecology. Initial research has focused on 
biogeochemistry and chemical stressors of ecosystems, but 
the scope is expanding to include community dynamics 
and ecosystem processes. Most people are familiar with 
simple laboratory ecosystems like an aquarium, and many 
have learned how hard it is to create a self-adaptive 
505 
system using only a few species. Nonetheless, these 
simple systems provide data that cannot be measured 
easily in the field. Time compression and cost savings are 
the usual arguments for these models. Time series 
experiments that would take generations to complete in the 
field can be programmed in the laboratory, and their 
results used to develop rule-based systems. Model 
systems serve not only as links between levels of 
complexity, but also to field experiments containing 
environmental noise. 
Over the next decade, ecologists expect more research on 
model systems, tending toward more complex and realistic 
assemblages. Experiments are moving outdoors to 
controlled environmental facilities. Results from 
experiments on CO,, for example, represent one kind of 
research. There is no doubt that rising concentrations of 
global CO, will change the abundance and composition of 
some plant communities. But, the relationships are not 
linear. CO, response is determined by genetic functions 
that are influenced by nutrient levels, water supply, and 
temperature--all of which are themselves influenced by 
changing CO,. 
Field experiments using uncontrolled natural conditions 
are always subject to question. History shows that one or 
two landmark studies can alter paradigms for decades, 
only to fade on the strength of another study that refocuses 
the research agenda. Complexity is the current focus. 
Before data collected using earlier paradigms are lost they 
should be integrated into digital libraries. 
Landscape Ecology and GIS 
Landscape ecology is a relatively young field of research, 
but one that is familiar to spatial and spectral analysts. It 
is concerned with the spatial dynamics of organisms, 
materials, and energy. It promotes models and theories 
that focus on spatial heterogeneity and an examination of 
data collected at different geographic scales; and, it 
recognizes that landscapes are patchworks of natural and 
cultural elements interacting to form a new class of 
ecosystem based on human supremacy. Figure 1 is a 
typical output from this field of inquiry. 
  
Figure 1. Vegetation Cover Over Terrain Model 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996