New Method Measures Feeding in Small Invertebrates

Measuring different quantities is a key part of science, but getting exact measurements can be tricky because all devices have some level of error. This issue becomes even more significant when studying tiny creatures like mosquitoes, flies, or ants, which often consume very small amounts of food, sometimes in the nanolitre range. Measuring such minuscule quantities can be very challenging and usually requires costly equipment that is not always suitable for observing animals in their natural behavior.

Feeding behavior is important to learn about because it gives us insights into how animals perceive their environment, their food preferences, and their Cognitive Skills. What and how much an animal eats can also indicate its health and well-being. Moreover, feeding is crucial for research in areas like developing better pest control methods and studying the spread of diseases.

Currently, one common way to measure how much small insects eat is via a method called the capillary feeder assay. In this method, scientists measure how much liquid food is used up from a thin glass tube. This requires the insects to drink from the tube while it is standing vertically, which means this method only works for certain types of animals. To help see how much food is consumed, the liquid is often dyed, but this can change the way the insects behave and affect their consumption rates. Additionally, this method can be easily affected by evaporation or spills, leading to inaccuracies.

Another method involves weighing the insects before and after they eat. This gravimetric approach counts how much weight the insect has gained as a measure of how much food it has consumed. However, weighing small creatures requires very sensitive and expensive scales. It can also be a complicated process, as the insects may need to be put to sleep before and after eating, which can change their behavior.

Some researchers use dyes or chemical tracers to help quantify feeding, but these methods can introduce their own problems. The dyes can alter how food is consumed and may require killing the insect afterward, which is not ideal for studying living animals.

There are some specialized systems designed to tackle these issues. For example, the flyPAD system measures changes in capacitance when an insect touches food, helping to estimate consumption based on how long it interacts with the food. However, this system is also expensive and requires a lot of upkeep.

To improve on these current methods, we developed a new non-invasive system to measure how much small invertebrates eat. Many insects and some other small animals visibly expand their bodies when they eat. By tracking this expansion over time, we can estimate the Volume of food they ingest. Previous techniques have used two-dimensional tracking to make these estimates, but we have upgraded this to a three-dimensional system for better accuracy.

Our method uses an open-source software called DeepLabCut that allows us to analyze videos of the animals feeding. By tracking specific body points in 3D, we can create a model of the animal’s body shape over time and measure any changes in volume as they eat. This approach avoids the errors that come with trying to estimate volume based on flat images.

#Building the System

To use this new feeding measurement system, there are several steps to follow. First, the user needs to set up a recording system that is suitable for the animal they are studying. Typically, this involves setting up two cameras at different angles to capture both top and side views of the animal. The cameras must be synchronized so that they capture images at the same time, allowing for proper 3D reconstruction of the feeding event.

Once the cameras are set up, the next step is to use the DeepLabCut software to identify and track key body points of the animal in the videos. This requires training the software with labeled images to recognize different body parts. Once trained, the software can identify these points in new videos.

From the tracked data, we can calculate volumetric changes over time using different methods. Our system comes with a user-friendly graphical interface that helps researchers select the start and end points of a feeding event, making it easy to analyze the data. This setup allows for both the total amount of food ingested and the rates at which food is consumed to be calculated.

#Setting Up for Successful Tracking

For the physical setup, a basic configuration with two cameras is generally enough. However, more cameras may be needed if the animal moves around a lot. It's important to calibrate the cameras using known reference points to ensure accurate measurements in the three-dimensional space. This calibration will help to ensure that the software can accurately pinpoint the location of the tracked points.

The tracking is done using the DeepLabCut software, which allows researchers to gather precise three-dimensional coordinates of the tracked points. These coordinates are then processed to estimate the volume of food consumed by calculating the volume of the expanding body parts.

#Estimating Volume and Consumption Rates

To measure how much food an insect consumes and the rate of consumption, we apply specific volume estimation methods to the tracked data. By determining how the tracked points change over time, we can gather both the total amount of food eaten and how quickly it was consumed.

The system allows users to visualize the data and verify the feeding events. By moving a slider, the user can see different frames of the video and confirm whether the tracking and volume estimations are accurate. Once the feeding event is identified, the data can be analyzed further to calculate variables like crop load (the amount of food consumed) and consumption rate.

#Validating the New Method

To confirm that our new method works accurately, we tested it with Argentine ants, an invasive species. First, we compared the new method's volume estimates with the traditional weighing method to see how closely they matched. We also fed the ants different concentrations of sucrose solutions to see how this affected their feeding behavior and preferences.

The results showed that our volume estimates were fairly consistent with the weight measures, although there were some differences due to factors like evaporation and how the ants might consume food. We specifically noted that as the concentration of sucrose increased, ants tended to consume more food, but their feeding rates decreased. This finding aligns with what we know about ant behavior: thicker solutions take longer to consume.

In addition to testing with sucrose solutions, we also examined how ants reacted to caffeine-laced solutions. Interestingly, caffeine appeared not to affect how much the ants ate or their feeding speed, suggesting that the ants did not detect it or were indifferent to it.

#Advantages and Challenges of the New System

This new method presents several advantages in studying Feeding Behaviors in small invertebrates. It allows researchers to quantify not only the total food intake but also the rate at which food is consumed. Since it relies on observing how an insect's body expands, it avoids many issues associated with traditional methods, such as the need for chemicals or anesthesia.

The approach is also relatively quick, with average recording times being comparable to previous methods. The pose estimation processing can take some time, but with adequate computational resources, it can function with minimal human involvement.

However, the tracking accuracy can be affected if the animal is not in the optimal position relative to the cameras. This limitation can lead to exclusion of recordings if the data cannot be accurately analyzed. To mitigate this, additional cameras could be added to allow for better angles and help track movement during feeding.

Our experiments showed that while more traditional methods may still be useful in some contexts, they can be time-consuming and may lead to loss of subjects due to the stress caused by anesthesia. In our case, the ants showed higher feeding rates without being put to sleep, thus reinforcing the value of a non-invasive approach.

#Conclusion

The method we developed for measuring feeding in small invertebrates provides researchers with a better way to gain insights into feeding behavior. It is adaptable for various organisms and could help in studies of disease vectors or the management of invasive species.

As the system is less invasive, it can be useful for observing natural behaviors in a wider range of animals. The potential for applying this technology to study not only small insects but also larger animals and medical applications means it could offer valuable data across different fields of research.

By providing a more direct way to measure food intake, this method stands to enhance our understanding of feeding preferences and behaviors, which is crucial for both ecological studies and pest management strategies.