Declining Songbird Populations Population biologists are interested in monitoring the status of migratory songbirds in the United States. The warbling vireo is a neotropical migrant that overwinters in Mexico and flies northward into the United States during the summer to breed. Population analysis has shown a decline in the coastal California population of warbling vireos over the last twenty-two years. Biologists examined the level of nest productivity and adult survivorship as indices of the per capita birth and death rates in the population, and found that declines in the population were most likely related to low reproductive success and not low survival rates. In terms of population models, this finding would indicate a population decline resulting fromthe low birth rate leading to an overall decline in numbers. This example illustrates the importance of population analysis for conservation biology. By understanding the relative effects of birth and death on the warbling vireo population size, conservation biologists can develop the best strategies to sustain vireo populations. By determining that it is nest failure and not adult survivorship that is causing population declines, biologists can focus on eradicating factors that limit nest success and thereby increase the per capita birth rate in the population.
Studying Habitats and Biomes Abiotic habitat requirements for a local population or even for an entire species can be determined in the laboratory by testing its range of tolerance for each factor. For example, temperature can be regulated in a laboratory experiment to determine the minimum and maximum survival temperatures as well as an optimal range. The same can be done with humidity, light, shelter, and substrate type: "Substrate preference" refers to the solid or liquid matter in which an organism grows and/or moves-for example, soil or rock. The combination of all ranges of tolerance for abiotic factors should describe a population's actual or potential microhabitat within a community. Furthermore, laboratory experiments can theoretically indicate how much environmental change each population can tolerate before it begins to migrate or die. Methods to study the interaction of populations with one another or even the interaction of individuals within one local population are much more complicated and are difficult or impossible to bring into a laboratory setting. These studies most often require collecting field data on distribution, abundance, food habits or nutrient requirements, reproduction and death rates, and behavior in order to describe the relationships between individuals and populations within a community. Later stages of these field investigations could involve experimental manipulations in which scientists purposely change one factor, then observe the population or community response. Often, natural events such as a fire, drought, or flood can provide a disturbance in lieu of a manipulation caused by man. There are obvious limits to how much scientists should tinker with the biosphere merely to see how it works. Populations and even communities in a local area can be manipulated and observed, but it is not practical or advisable to manipulate whole ecosystems or biomes. To a limited extent, scientists can document apparent changes caused by civilization, pollution, and long-term climatic changes. This information, along with population- and community-level data, can be used to construct a mathematical model of a population or community. The model can then be used to predict the changes that would happen if a certain event were to occur. These predictions merely represent the "best guesses" of scientists, based on the knowledge available. Population ecologists often construct reasonably accurate population models that can predict population fluctuations based on changes in food supply, abiotic factors, or habitat. As models begin to encompass communities, ecosystems, and biomes, however, their knowledge bases and predictive powers decline rapidly. Perhaps the most complicating factor in building and testing these large-scale models is that natural changes seldom occur one at a time. Thus, scientists must attempt to build cumulative- effect models that are capable of incorporating multiple changes into a predicted outcome. The biosphere, then, can be studied at different levels of organization, from the individual level through populations, communities, ecosystems, and biomes to the all-encompassing biosphere. Each level has unique relationships that require different methods of inquiry; in fact, these levels describe many of the subdisciplines within the science of ecology.
