Bats as Indicator Species

Due to the extensive nature of the Amazon Rainforest and the limited budget and number of personnel involved with the project, monitoring will consist of what will be essentially "hot spot" checks. These will consist of blood tests on randomly selected
animals in areas of concern to determine the exact degree to which the ecosystem is being polluted. Additional tests, such as gross anatomical observations for deformities and chemical testing for behavioral abnormalities will be conducted to determine the nonlethal synergistic effects of the polluting chemicals on the amphibians2,3.
"Areas of concern" will consist of areas of the rainforest where it is believed that pollution is, or could become a major problem for the overall ecosystem's health. These include industrial waste dumping sites, farmland drainage areas (pesticide runoff), and any other site deemed threatened by pollution. Monitoring actual population numbers will not be necessary unless the toxins present are severe enough in their effects to cause significant mortality rates.

Due to the extremely varied circumstances that a monitoring program would be subjected to before and during its implementation, several proposals have been developed in order to create the most ideal monitoring system under the given conditions. All of these proposals use the same general implementation system, but their foci differ.

- Broad-Spectrum: Monitoring is done throughout the Amazon rainforest, attempting to gather status information on the widest feasible amount of data over an extremely large area of the rainforest. Monitoring sites and relatively high density, with the largest concentrations of monitoring sites centering around areas subject to a large number of current or anticipated threats from human encroachment.

- Focus-Specific: Monitoring focuses almost exclusively on areas under known threats from human encroachment, and data and subsequent analysis are gathered and performed with specific threats and indicators in mind, such as chemical contamination of the ecosystem by local mining operations. Very few resources are spent on monitoring relatively pristine areas.

- Broad-and-Focused: Monitoring is done with emphasis on looking at both specific threats and broad-range effects. Monitoring resources will be equally distributed between gathering a diverse amount of data on relatively pristine areas of the rainforest and collecting information on the effects and spread of human encroachment in threatened areas.

Before continuing, an important note must be clarified: these monitoring systems are designed to use bats as indicators of the presence of a problem in the area and/or the relative extent to which that problem may be affecting the local ecosystem(s). They are not designed to determine exactly what a new threat could be. In essence, the monitoring system is more a warning sign than a diagnosis tool. In order to determine what the problem is specifically, outside research must be conducted, such as learning what kinds of human or climactic activity may be causing a disruption in the balance of the ecosystem that would alert the monitoring systems of a problem. If remote research is unfruitful, then directly visiting the site may be required to determine the cause of the problem.

The proposals also base many assumptions off of the classification of three distinct zones of the rainforest:

- Heavily Threatened: areas currently experiencing heavy or moderate human encroachment with the possibility of extreme or moderate environmental degradation as a result.

- Moderately Threatened: areas of relatively pristine rainforest that are currently experiencing or will be experiencing small amounts of human encroachment within the near future.

- Pristine: areas that are not experiencing any forms of human encroachment are fairly protected and located away from threatened and heavily threatened areas

These three zones can be subject to many interpretations, allowing them to be modified to fit many different types of situations that may be encountered in the field or circumstances that the monitoring project may find itself under. The moderately threatened and highly-threatened zones are especially open to modification, depending on the type of human encroachment. For example, for one type of study an area with a newly built road may be considered only moderately threatened in comparison with a region that has been in close proximity to a large cattle ranching operation for many years, while under a different study the same area with the new road could be considered highly threatened with respect to a region with only extremely small, low-impact fishing villages.

With multiple proposals that each contain specific types and amounts of data based on easily modified zones of relative risk from human encroachment, these monitoring systems are extremely adaptable to whatever sort of circumstance the program may find itself in.

The primary basis of these systems is the technology used to remotely monitor bat populations in the Amazon rainforest and then to analyze the information acquired. The most efficient and accurate bat detectors pick up the high-frequency calls that bats use for echolocation. Each species has its own distinct call, allowing any bats that are in the vicinity to be easily identified. The detectors can also be used to determine the overall ambience of the bat calls, allowing monitors to figure out how many bats there are in the relative vicinity and calculate changes in their numbers and movements.

Bat detectors first transform the ultrasound calls into audible sound, by which the individual calls can be identified. There are three primary methods of performing this process: [1]

- Heterodyning: the most sensitive real-time method of transforming ultrasound signals, but only a small frequency range can be transformed at one time.

- Frequency division: the ultrasound frequency is divided by a constant factor in real-time, converting extremely high frequencies into lower ones. This method is less precise but it allows for the transformation of an entire ultrasonic frequency range as opposed to a very limited range of frequencies.

- Time expansion: digital techniques are used to store recorded ultrasound frequencies. These sounds are then played back at much slower speeds, stretching out the signals over time, creating an extremely detailed playback of the ultrasonic calls.

Analysis of these calls can then be done using software, utilizing a variety of techniques such as power spectrums, pulse interval and length analyses, spectrograms, and zero-crossing by measuring ultrasound pulse duration, maximum and minimum frequencies, time of the start and end of the pulse, and frequency at the point of maximum amplitude. With these methods, the bat calls can be sorted and specific species identified by comparing the gathered calls from known samples.

By monitoring the relative ambience of the bat calls over set intervals of time, one can determine how active a colony is, whether the population is stable, and whether there are any problems that the bats are encountering. Because bats will go into hibernation when they experience times of hardship, the relative ambient noise of the colony will significantly drop during those times. If the population decreases or increases, the change will be reflected in a proportional decrease or increase in the noise ambience. By carefully watching for these changes, the presence of a problem in the nearby area of the colony can be detected.

Auxiliary equipment is also needed in the implementation of the monitoring systems. This equipment will include at the very least: power sources for the detectors, mounting systems to prop the detectors up in an optimal location for listening, protective casing to prevent damage from the environment and transmitters that will allow each monitoring station to transmit the data it collects to one or to several centralized data processing centers.

Power and transmission pose the greatest challenges to the project technology wise. The canopy of the rainforest is very dense and high above the ground. Possibilities for maintaining power for the detectors over long periods of time include harnessing solar or water to generate power. Transmitters could be mounted high in a tree or in an open space. To conserve power and reduce transmission time, the detectors do not have to be on constantly. They could be activated for small intervals of time, for example, 15 seconds once every 2 hours. The data also does not have to be instantaneously transmitted for analysis either. It can be stored with the detector and after enough has accumulated over a designated amount of time, such as a week or two, all of the data can be transmitted at once to the central processing center.

Installation of the devices would be the most significant and difficult part of implementing the monitoring systems. Prospective colonies to be monitored must first be analyzed and then narrowed down to ones that fit the specific criterion of the project(s). Overall density of the detectors based on the locality of the colonies being monitored would vary between the three different zones of heavily threatened, moderately threatened, and pristine zones. Each site would require only about one or two detectors, since the average bat detector has a range of approximately 30 meters.[2]

The installing of the equipment in the rainforest should be done by a competent group of experts who have the skills to make modifications in set-up if necessary. These experts would most likely include, but are not limited to, people experienced with mechanical, electrical, and software systems, along with a good familiarity with the equipment being employed. These experts can either be trained scientists who are in residence in the area or people brought in from outside the area.

Maintenance would most likely be done by trained scientists and/or engineers who live in the local area, since transporting people from distant regions would be expensive and impractical. One possibility would be to train the native inhabitants in regions that are especially difficult to reach by outsiders to help maintain the monitoring stations.

Data synthesis and analysis would be an enormous task, since the sheer amount of data would probably be extraordinary. The information gathered would certainly have to be analyzed by computers, since the enormous volume of data generated by the monitoring detectors would be far too much to be feasibly done by hand.

Due to the different natures of the three proposals and the adaptability of the three divisional zones and their interpretations, calculating how much the entire project would take would result in a rough estimate at best. A ballpark estimate can be obtained by making the following assumptions:

Estimates on the land coverage of areas that are highly threatened and pristine are based on estimates that 1.89 million hectares of rainforest are cut down in the Amazon per year and that there is approximately 344 million hectares of remaining forest coverage left.[3] All costs stated above are educated guesses, and include the cost of labor to install and maintain the stations.

With the above assumptions, the estimated cost of installing the system comes out to be $8.2-9.6 million and subsequent maintenance and data analysis costs come out to be $1.1-1.2 million per year.

[1] Bat Detectors, Petterson Electronik, <www.batsound.com/psondet.html>

[2] Ibid.

[3] William F. Laurence, �The Future of the Brazilian Amazon,� Science Magazine, 19 January 2001.


Home About Us Members Projects Final Report	Initial \| Fragmentation \| Amphibians \| Bats \| Monitoring \| SIVAM Methods of Monitoring Using Amphibians: Due to the extensive nature of the Amazon Rainforest and the limited budget and number of personnel involved with the project, monitoring will consist of what will be essentially "hot spot" checks. These will consist of blood tests on randomly selected animals in areas of concern to determine the exact degree to which the ecosystem is being polluted. Additional tests, such as gross anatomical observations for deformities and chemical testing for behavioral abnormalities will be conducted to determine the nonlethal synergistic effects of the polluting chemicals on the amphibians2,3. "Areas of concern" will consist of areas of the rainforest where it is believed that pollution is, or could become a major problem for the overall ecosystem's health. These include industrial waste dumping sites, farmland drainage areas (pesticide runoff), and any other site deemed threatened by pollution. Monitoring actual population numbers will not be necessary unless the toxins present are severe enough in their effects to cause significant mortality rates. Using Bats: - Monitoring Proposals: Due to the extremely varied circumstances that a monitoring program would be subjected to before and during its implementation, several proposals have been developed in order to create the most ideal monitoring system under the given conditions. All of these proposals use the same general implementation system, but their foci differ. The three developed proposals are described below: - Broad-Spectrum: Monitoring is done throughout the Amazon rainforest, attempting to gather status information on the widest feasible amount of data over an extremely large area of the rainforest. Monitoring sites and relatively high density, with the largest concentrations of monitoring sites centering around areas subject to a large number of current or anticipated threats from human encroachment. - Focus-Specific: Monitoring focuses almost exclusively on areas under known threats from human encroachment, and data and subsequent analysis are gathered and performed with specific threats and indicators in mind, such as chemical contamination of the ecosystem by local mining operations. Very few resources are spent on monitoring relatively pristine areas. - Broad-and-Focused: Monitoring is done with emphasis on looking at both specific threats and broad-range effects. Monitoring resources will be equally distributed between gathering a diverse amount of data on relatively pristine areas of the rainforest and collecting information on the effects and spread of human encroachment in threatened areas. Before continuing, an important note must be clarified: these monitoring systems are designed to use bats as indicators of the presence of a problem in the area and/or the relative extent to which that problem may be affecting the local ecosystem(s). They are not designed to determine exactly what a new threat could be. In essence, the monitoring system is more a warning sign than a diagnosis tool. In order to determine what the problem is specifically, outside research must be conducted, such as learning what kinds of human or climactic activity may be causing a disruption in the balance of the ecosystem that would alert the monitoring systems of a problem. If remote research is unfruitful, then directly visiting the site may be required to determine the cause of the problem. The proposals also base many assumptions off of the classification of three distinct zones of the rainforest: - Heavily Threatened: areas currently experiencing heavy or moderate human encroachment with the possibility of extreme or moderate environmental degradation as a result. - Moderately Threatened: areas of relatively pristine rainforest that are currently experiencing or will be experiencing small amounts of human encroachment within the near future. - Pristine: areas that are not experiencing any forms of human encroachment are fairly protected and located away from threatened and heavily threatened areas These three zones can be subject to many interpretations, allowing them to be modified to fit many different types of situations that may be encountered in the field or circumstances that the monitoring project may find itself under. The moderately threatened and highly-threatened zones are especially open to modification, depending on the type of human encroachment. For example, for one type of study an area with a newly built road may be considered only moderately threatened in comparison with a region that has been in close proximity to a large cattle ranching operation for many years, while under a different study the same area with the new road could be considered highly threatened with respect to a region with only extremely small, low-impact fishing villages. With multiple proposals that each contain specific types and amounts of data based on easily modified zones of relative risk from human encroachment, these monitoring systems are extremely adaptable to whatever sort of circumstance the program may find itself in. - Technology: The primary basis of these systems is the technology used to remotely monitor bat populations in the Amazon rainforest and then to analyze the information acquired. The most efficient and accurate bat detectors pick up the high-frequency calls that bats use for echolocation. Each species has its own distinct call, allowing any bats that are in the vicinity to be easily identified. The detectors can also be used to determine the overall ambience of the bat calls, allowing monitors to figure out how many bats there are in the relative vicinity and calculate changes in their numbers and movements. Bat detectors first transform the ultrasound calls into audible sound, by which the individual calls can be identified. There are three primary methods of performing this process: [1] - Heterodyning: the most sensitive real-time method of transforming ultrasound signals, but only a small frequency range can be transformed at one time. - Frequency division: the ultrasound frequency is divided by a constant factor in real-time, converting extremely high frequencies into lower ones. This method is less precise but it allows for the transformation of an entire ultrasonic frequency range as opposed to a very limited range of frequencies. - Time expansion: digital techniques are used to store recorded ultrasound frequencies. These sounds are then played back at much slower speeds, stretching out the signals over time, creating an extremely detailed playback of the ultrasonic calls. Analysis of these calls can then be done using software, utilizing a variety of techniques such as power spectrums, pulse interval and length analyses, spectrograms, and zero-crossing by measuring ultrasound pulse duration, maximum and minimum frequencies, time of the start and end of the pulse, and frequency at the point of maximum amplitude. With these methods, the bat calls can be sorted and specific species identified by comparing the gathered calls from known samples. By monitoring the relative ambience of the bat calls over set intervals of time, one can determine how active a colony is, whether the population is stable, and whether there are any problems that the bats are encountering. Because bats will go into hibernation when they experience times of hardship, the relative ambient noise of the colony will significantly drop during those times. If the population decreases or increases, the change will be reflected in a proportional decrease or increase in the noise ambience. By carefully watching for these changes, the presence of a problem in the nearby area of the colony can be detected. Auxiliary equipment is also needed in the implementation of the monitoring systems. This equipment will include at the very least: power sources for the detectors, mounting systems to prop the detectors up in an optimal location for listening, protective casing to prevent damage from the environment and transmitters that will allow each monitoring station to transmit the data it collects to one or to several centralized data processing centers. Power and transmission pose the greatest challenges to the project technology wise. The canopy of the rainforest is very dense and high above the ground. Possibilities for maintaining power for the detectors over long periods of time include harnessing solar or water to generate power. Transmitters could be mounted high in a tree or in an open space. To conserve power and reduce transmission time, the detectors do not have to be on constantly. They could be activated for small intervals of time, for example, 15 seconds once every 2 hours. The data also does not have to be instantaneously transmitted for analysis either. It can be stored with the detector and after enough has accumulated over a designated amount of time, such as a week or two, all of the data can be transmitted at once to the central processing center. - Implementation: Installation of the devices would be the most significant and difficult part of implementing the monitoring systems. Prospective colonies to be monitored must first be analyzed and then narrowed down to ones that fit the specific criterion of the project(s). Overall density of the detectors based on the locality of the colonies being monitored would vary between the three different zones of heavily threatened, moderately threatened, and pristine zones. Each site would require only about one or two detectors, since the average bat detector has a range of approximately 30 meters.[2] The installing of the equipment in the rainforest should be done by a competent group of experts who have the skills to make modifications in set-up if necessary. These experts would most likely include, but are not limited to, people experienced with mechanical, electrical, and software systems, along with a good familiarity with the equipment being employed. These experts can either be trained scientists who are in residence in the area or people brought in from outside the area. Maintenance would most likely be done by trained scientists and/or engineers who live in the local area, since transporting people from distant regions would be expensive and impractical. One possibility would be to train the native inhabitants in regions that are especially difficult to reach by outsiders to help maintain the monitoring stations. Data synthesis and analysis would be an enormous task, since the sheer amount of data would probably be extraordinary. The information gathered would certainly have to be analyzed by computers, since the enormous volume of data generated by the monitoring detectors would be far too much to be feasibly done by hand. - Expenses: Due to the different natures of the three proposals and the adaptability of the three divisional zones and their interpretations, calculating how much the entire project would take would result in a rough estimate at best. A ballpark estimate can be obtained by making the following assumptions: - Highly Threatened Zone Concentration: 1 detector per 15 miles² - Moderately Threatened Zone Concentration: 1 detector per 20 miles² - Pristine Zone Concentration: 1 detector per 200-250 miles² - Area of Highly Threatened Zone: 7000 miles² - Area of Moderately Threatened Zone: 36,000 miles² - Area of Pristine Zone: 1,300,000 miles² - Average Cost of Equipment per Station: $1000 - Average Cost of Installation per Station: $100 - Average Cost of Maintenance per Station per year: $20 - Average Cost of Data Analysis per year: $1 million Estimates on the land coverage of areas that are highly threatened and pristine are based on estimates that 1.89 million hectares of rainforest are cut down in the Amazon per year and that there is approximately 344 million hectares of remaining forest coverage left.[3] All costs stated above are educated guesses, and include the cost of labor to install and maintain the stations. With the above assumptions, the estimated cost of installing the system comes out to be $8.2-9.6 million and subsequent maintenance and data analysis costs come out to be $1.1-1.2 million per year. [1] Bat Detectors, Petterson Electronik, <www.batsound.com/psondet.html> [2] Ibid. [3] William F. Laurence, �The Future of the Brazilian Amazon,� Science Magazine, 19 January 2001.

Methods of Monitoring