From Predator to Prey: Identifying Non-native Python Species using NIR Reflection

From Predator to Prey: Identifying Non-native Python Species using NIR Reflection

How do you measure reflectance from a 10-foot long live python? Very carefully and at a safe distance!

That’s one of the insights gleaned by Carl Arvidson of Extended Reality Systems (New Smyrna Beach, Fla.) and a team of researchers while using sensing technologies to identify non-native snake species in the Everglades.

Non-native species like the Burmese pythons decimating the Everglades put entire ecosystems at risk. These descendants of abandoned pets enter the wild with no predators, giving them a huge advantage over native wildlife. They destroy their surroundings by consuming native, sometimes endangered indigenous species and natural resources. Being able to identify and remove these non-native species can help prevent an environmental disaster.

Introduction

Estimates as high as 300,000 Burmese pythons have been reported in the Florida Everglades. The python’s natural camouflage and the demanding, marshy environment of the Everglades makes pythons very difficult to detect in the wild. To find and remove pythons, researchers have tried everything from trained dogs to poisoning python prey, without much success. With females laying 40-50 eggs at a time, their numbers have grown rapidly. Female pythons become breeders at 3-5 years old, when they are more than 6 feet long. Because the snakes guard their eggs until they hatch, finding females with their eggs will help to slow the explosive python population growth.

Recently, we asked Arvidson about his work with biologists at the University of Florida, the Florida Fish and Wildlife Conservation Commission and others to develop a new methodology for the detection of invasive snakes in the Everglades. Using modular spectroscopy components that can be taken into the field, Arvidson and his colleagues are using NIR reflectance measurements as part of a larger effort to develop a rapid screening tool for non-native species. He shared additional insights here:

Ocean Optics: What inspired your research?

Arvidson: My former life in the DoD (Department of Defense) world working multisensor and multispectral systems for the military involved some adjacent capabilities that would highlight certain objects not visible to the human eye. After 30 years of working in that market with a large military contractor, the idea of taking those applications and applying them to help solve other problems in the world and having fun doing it seemed like a good challenge to take.

Ocean Optics: Why is this research important?

Arvidson: The inadvertent introduction of pythons in the Everglades has led to a large breeding population. There are now estimates of tens of thousands of pythons entrenched in the Everglades but because they are so well camouflaged they are extremely hard to find; no one really has a clue how many are out there. The pythons captured are now approaching 20 feet long and can be 5-6 feet long in 2-3 years. In one example a captured python was found with two full grown deer in it, and raccoons are almost considered to be popcorn to the larger snakes.

Ocean Optics: How did you decide to use spectroscopy?

Arvidson: The idea is that while the snakes are extremely well adapted in the visible region to be camouflaged to predators (including humans), there is no natural selection in the NIR/SWIR bands since so many animals see into that range. By taking the spectral signature of the pythons and running it against dominant plant species, dirt and water, a multispectral filter and algorithm can be developed to highlight the pythons in their habitat. The data taken is now being analyzed to create the requirements for the filters and software that will be used to develop a multispectral solution.

 

NIR Spectroscopy African rock python

Figure 1. A special spectrometer setup (foreground) was rigged for reflectance measurements of the snakes. As Arvidson explained, “The rule is when a 10-foot snake bites you, don’t pull away because its teeth are curved inward. That’s when we decided on a 6-foot pole. This was a great day with some important data taken despite some unhappy snakes.”

Experimental Procedure

Arvidson did his testing in Homestead, Fla., along with the researchers and two companies that provided locally captured snakes. The team used the combination of a Flame-NIR (950-1650 nm) and Flame-VIS-NIR (350-1000 nm) spectrometer along with a diffuse reflection probe (DR-Probe) to make diffuse reflection measurements of the pythons.

Originally, the team planned to make the measurements by placing the probe directly onto the snake’s skin by hand. But one extremely aggressive African rock python (Figure 1) led them to create a unique sampling rig with the spectrometer and probe mounted to a 6-foot pole, providing some distance between the snake and the researchers. The Burmese python was much more docile. Burmese pythons are the most common invasive snake found in the Everglades. Both snakes were 10-12 feet long and about three years old.

 

NIR reflection spectroscopy from Burmese python

Figure 2. These NIR diffuse reflectance spectra from live pythons and habitat samples will help in development of multispectral sensing systems for identifying pythons in the wild.

Results

The reflection spectrum measured for the Burmese python is shown in Figure 2, along with reflectance spectra for materials found in the python’s natural habitat. Even with these preliminary spectra collected to assess feasibility and determine the optimum setup for the measurements, there are some subtle spectral differences that potentially could be used to distinguish snakes from their surroundings. These preliminary spectra are the first of many that will be collected by Extended Reality Systems (ERS) and their partners (Figure 3).

With this new approach to finding pythons, Arvidson hopes to help reduce the python population in the Everglades and slow or even halt the destruction of the habitat and its native species. Once the team has conquered pythons, they plan to expand their detection technique to a variety of environmental areas dealing with invasive species. We asked Arvidson what comes next:

 

NIR Spectroscopy Python Brevard Zoo

Figure 3. Staff at the Brevard Zoo (Melbourne, Fla.) assist in taking spectral measurements of an African rock python.

Ocean Optics: Will the same technology work for other invasive (noninvasive) species?

Arvidson: Multispectral signatures in the NIR/SWIR offer a technology to identify and locate invasive reptiles and amphibians where thermal bands are not effective because the animals are cold blooded. The NIR bands offer a cost-effective band that CMOS cameras can address, and the SWIR bands are becoming more cost effective as the SWIR military technology becomes commercialized. Additionally, there are other opportunities with noninvasive species such as alligators and Florida crocodiles where population surveys are currently carried out at night with direct human involvement. The introduction of multispectral cameras for those missions would offer a safer, more accurate alternative.

Ocean Optics: What else should people take away from this work?

Arvidson: While the opportunity to develop technologies that can help control the population of pythons in the Everglades may not lead to a large or even profitable business enterprise, what has been gratifying has been the willingness of companies like Ocean Optics to support the effort of helping the environment. Once ERS and its partners have the technology developed and hopefully, are successful we will be front and center on solving a very high-profile problem.

For more information on Extended Reality Systems, contact Carl Arvidson at [email protected].

Next post