コンピュータ支援ビデオとデジタル画像解析を用いた銅水溶液の急性暴露に応答して、魚の遊泳行動を定量化

The overall goal of this procedure is to demonstrate how to quantify behavioral responses in white sturgeon to sub-lethal exposure to aqueous copper. This method can help to answer key questions in the ecotoxicology field, such as how sub-lethal exposure to contaminants affects swimming behavior of fish. This method can provide important information for white sturgeon conservation.

It can also be readily applied to other fish species, amphibians, and invertebrates. The main advantage of this technique is that the response is highly sensitive, provides an early indication of toxicity, and is relevant to organism survival. Initially this method may be challenging if the individual is unfamiliar with making behavioral measurements.

Demonstrating the protocol will be my co-author Holly Puglis. First, prepare two liters of aqueous reagent grade copper two sulfate pentahydrate. Next, equilibrate the diluter system and exposure tanks.

Place the intake tube from the automated syringe dispenser into the stock solution flask. Then, set the automated pump spike volume to one milliliter, and turn the diluter on. The system makes a 50%dilution series of five concentrations and a control.

Through a four-way splitter, these solutions are delivered to the four replicate exposure chambers. Next, turn on the pump for the chiller set to maintain the water bath at 15 degrees Celsius. Now, using the automated timer on the diluter, set it to cycle every 30 minutes.

The water level in the tanks is maintained by a stand pipe drain. Prepare the exposure chambers by putting a four centimeter hole in the side, and covering the hole with 30 micron stainless steel mesh. Be sure to select exposure chambers that hold at least one liter of water per 10 grams of fish.

For example, an 89 day old white sturgeon weighs about 373 milligrams, and can thus be tested in a 28 centimeter by 13 point five centimeter by 25 centimeter square chamber. At the initiation, mid-point, and end of an experiment, collect 24 milliliter test solution sub-samples for chemical analysis. Use sipper straws attached to plastic syringes.

Replace the sipper straw with a zero point four five micron polyethersulfone filter, and dispense four milliliters of test water through the filter. Dispose of this water. Then, filter the remaining 20 milliliters of water into an acid-cleaned polyethylene bottle.

Acidify the sample with 16 molar nitric acid to 1%volume by volume. These samples may then be stored for up to three months. Once the test exposures are confirmed, stock three sturgeon in each replicate test chamber for about 172 milligrams of fish per chamber liter.

Collect the sturgeon using a nonabrasive net, and transfer them to a small bucket. In total, 144 fish are needed to collect data from six test solutions. During the 96 hour exposure, do not feed the fish.

Consistent daily observations must be made to score for mortality and abnormal behavior. The morning is the best time of day to do this, and ideally the observer is blind to the test conditions. Loss of equilibrium is defined as the inability of fish to maintain an upright position within the water column.

Immobility is defined as the inability of fish to move or swim unless prodded. Also record any other abnormalities, such as lethargy, hyperactivity, increases or decreases in respiration, color changes, tremors, spasms, bloated abdomens, position in the water column, and any other unusual swimming patterns. Dead fish should always be removed immediately.

Each day, also measure dissolved oxygen levels with a handheld meter from two water samples. At the end of the 96 hour exposure, make videos for swimming behavior analysis. Position the test arena within an exposure tank.

Be sure to select a test arena that is large enough for the fish to swim freely. Directly above each test arena, set up a handheld video camera on a tripod. Now, from the treated fish, randomly select up to three survivors.

Transfer three into the test arena using a small mesh net. Allow the fish to acclimate to the arena for about 30 minutes. While the fish acclimate, focus the image, minimize the glare distortions from moving water, and maximize the contrast of the fish against the background, adjusting the view so the fish is not obscured.

Once the fish is acclimated, start the recording and collect two minutes of video data. After taking the video, euthanize and dispose of the fish. After converting the video files to the format needed for the digital analysis software, open the tracking software and select new default experiment under the create a new experiment option on the main screen.

Fill in the prompts, and press okay. Then, go to the experiment settings option under set up, and choose from video file under video source. Set the number of arenas to one.

Set the number of subjects per arena to three. Next, select center point detection under tracked features and select the desired units. Now, under set up, select the trial list option and click add videos at the top of the screen.

Choose alphabetical order from the dialogue box. Click browse. Navigate to the folder where the videos are located, highlight the pertinent files, and click open.

Now that the video files are selected, proceed with defining the scope of the experiment in the software. First, click add variable at the top of the screen and enter concentration into the label box. Enter copper concentration in micrograms per liter into the description box.

Next, choose numerical from the drop down list for type. Click on the pre-defined values box, and choose the define individual values option from the dialogue box. Enter the concentrations used for the experiment, clicking add between each number addition.

Then, uncheck the allow other values option and click okay. Choose trial from the drop down menu in the scope box. Enter the appropriate concentration for each trial in the boxes provided.

Click add variable at the top of the screen. Enter replicate into the label box. Enter replicate number into the description box.

Choose numerical from the drop down list for type. Next, click on the pre-defined values box and choose the define individual values option from the dialogue box. Enter one, two, three, and four, and un-check the allow other values option.

Then, click okay. Now, choose the trial option from the drop down menu in the scope box. Enter the appropriate replicate number for each trial.

The next task is to set up the arena spaces. Begin with choosing the arena settings under the set up tab, and name the first setting as trial one. From the dialogue box, select the grab background image option and click browse.

Locate the video file for trial one and click open. After the video appears, click grab on the grab background image dialogue box. Next, click the white circle icon near the top of the screen under arena settings, and manipulate the circle that appears to surround the swimming area.

Then, click the calibration scale icon under arena settings. Use a click-hold-drag motion to select the arena space. Now, under the real world distance box in the calibration distance, enter 10 point five.

Then, click okay. If necessary, adjust the calibration line so that it spans the entire diameter of the circular arena. Click the validate arena settings option on the arena settings box, and address any issues in the settings that were not validated.

Now, repeat this process until all the arenas are defined for each trial. Next, set the fish detection parameters. First, choose detection settings under the set up option from the experiment explore tool bar on the left of the screen.

Then, choose dynamic subtraction from the drop down menu under method in the detection settings one dialogue box. Also, choose a fill color for each subject under subject identification in the dialogue box. Next, choose select video and locate the video for trial one.

In the detection settings one dialogue, fill in the sample rate box under video with five point nine nine four one. In the same box, click settings for the reference image option under detection. Then, click the start learning C option in the reference image dialogue box.

Once the image in the reference image dialogue box appears without animals, click use dynamic reference image under acquisition settings within the dialogue box and close this dialogue box. Now, choose darker from the drop down menu for subject is under detection in the detection settings one dialogue box. For the dark contrast under detection, set the smaller number to 33 and the larger number to 153.

Then, click save changes at the bottom right. Now, confirm that that software is successfully tracking animals, as opposed to shadows or debris. Click the play button on the playback control dialogue box and adjust the numbers for dark contrast as needed.

Be sure to save any adjustments made. The software is now set to collect data from the video, so the next task is to track the fish in each trial. First, choose acquisition under the setup option from the experiment explorer tool bar on the left of the screen.

Next, click track next planned trial in the acquisition settings dialogue box, and confirm that the correct trial, video, and arena settings are all displayed under the settings in the acquisition settings dialogue box. Then, check the detection determines speed option in the acquisition control dialogue box. Click the button with the green circle enclosed in a white square to begin the acquisition process.

Repeat this process until all the trials have been tracked. Now, the tracked objects being the fish can be analyzed. First, click data profiles under the analysis option from the experiment explorer tool bar on the left of the screen.

Then, choose time under the nesting option in the components toolbar. Under the select track time interval setting in the time dialogue box, adjust the to value to two minutes and click okay. Then, drag the nest box between the start box and the result one box in the data profiles area on the right side of the screen.

From the experiment explorer toolbar on the left of the screen, look under the analysis option and click analysis profile. Then, in the dependent variables toolbar, under the distance and time heading click velocity. Click add, and under the distance and time heading in the dependent variables toolbar, click distance moved.

Then, click add on the distance moved dialogue box, and under the individual behavior heading in the dependent variables toolbar, click movement. In the movement dialogue box, look under the outlier filter heading and adjust the averaging interval to one. Then, under the threshold heading, adjust the start velocity to two point zero zero and adjust the stop velocity to one point seven five.

Under the calculate statistics for heading, check both boxes for moving and not moving. Then, click add, and under the results option in the experiment explorer toolbar on the left of the screen click analysis output. Now, click calculate.

Once the dependent variables are calculated, click export at the top of the screen. Choose the destination folder from the export analysis output dialogue box, and select Excel from the file type drop down menu. Then, click okay.

Finally, click save experiment under the file tab at the top of the screen. Close the digital tracking software. Import the data to a spreadsheet file, and analysis it using a commercial statistical analysis software package.

After just 72 hours of exposure to various copper concentrations, two days post-hatch sturgeon showed many observable abnormalities. There was an obvious impact on swimming behavior. Video data aided in defining this behavioral impairment.

In further tests, sturgeon at 30 days post-hatch proved sensitive to 96 hour copper exposures. Swimming speed, the time spent moving, and the distance traveled all significantly decreased with increase in copper concentrations. Swimming paths were also reduced with increase in copper concentrations.

After watching this video, you should have a good understanding on how to quantify behavior in a toxicity test. This technique can provide environmental managers a means of using behaviorally-defined criteria to document environmental injury and guide restoration. When conducting this procedure, it is important to remember to be consistent when recording behavioral observations.

Other behaviors, such as feeding and aggression, can be quantified by following this procedure in order to answer additional questions about the long term impacts of contaminant exposure on survival.