Mites from Space: Tracking a Microscopic Pest in Puerto Rico

Category: Identifying Invasive Species Extent & Critical Species Habitat
Project Team: Puerto Rico Agriculture
Team Location: NASA Goddard Space Flight Center – Greenbelt, Maryland

The Puerto Rico Agriculture team used vegetation indices to identify palm stands infested with the red palm mite. Image Credit: Puerto Rico Agriculture team

The Puerto Rico Agriculture team used vegetation indices to identify palm stands infested with the red palm mite. Image Credit: Puerto Rico Agriculture team

Dr. Sara Lubkin
Julia Marrs
Sean McCartney
Alison Thieme

Dr. Michael Cosh (U.S. Department of Agriculture, Agricultural Research Service)
Dr. Ronald Ochoa (U.S. Department of Agriculture, Agricultural Research Service)
Dr. John Bolten (NASA Goddard Space Flight Center)


The rapid spread of the red palm mite, Raoiella indica, has devastated coconut palm, banana, and plantain crops throughout the Americas and the Caribbean since its introduction into the hemisphere in 2003. Red palm mites feed via the stomata on plant hosts’ leaves, using specialized mouthparts, resulting in accelerated water loss, decreased crop yield, and a characteristic pattern of leaf yellowing and senescence that can be remotely detected using multispectral imagery and calculated spectral vegetation indices from Landsat 7 Enhanced Thematic Mapper Plus (ETM+), Landsat 8 Operational Land Imager (OLI), Earth Observing-1 (EO-1) Hyperion, IKONOS platforms, and aerial imagery. In conjunction with the U.S. Department of Agriculture (USDA) Agricultural Resource Service (ARS) and the University of Puerto Rico, the distribution of red palm mite infestation in Puerto Rico was tracked from 2002 to 2016, and maps were produced showing the geographic spread of the red palm mite infestation during this time. The high adaptability of red palm mites to multiple plant hosts, adverse environmental conditions, and diverse dispersal pathways, plus the absence of known biotic countermeasures, makes the mite an urgent environmental and economic threat throughout Southeast Asia, Oceania, and the Americas. Understanding past red palm mite invasions will help to prevent its spread into proximal agricultural areas and limit exposure and resulting crop destruction. The tracking methodology and tools established here can be applied to other geographical areas in an effort to mitigate current mite infestations and to monitor and prevent probable future invasions.

Previous story | Main Page | Next story


Gwen 09-09-2016, 09:57

In the video it seemed like the science adviser said you can see the mites from space. Did he mean landsat can detect the yellowing of the plants due to plant damage or does he actually mean using landsat the actual mite population can be detected (maybe there is more red in the pixel with mites than without mites)

Sara Lubkin 09-09-2016, 11:48

Hi Gwen,
Perhaps a better way to say it would be “detected from space”. So, no. We can’t see individual mites or even populations of mites; we can only see their effect on host plants. But, it’s an interesting thought. How many micron scale mites would it take to be visible on a 30 meter pixel? I’m guessing a really scary amount. 🙂

Kelsey Herndon (Chaco Canyon Cross-Cutting) 08-09-2016, 21:01

Great project and great video! I just have a couple of quick questions: Do you plan on expanding your study area in a continuation project, perhaps to the other islands you mentioned in your video? Would it be possible to implement your methodology in a near-real time mite model? Did you find that a particular index/combination of indices was best/ineffective at identifying mite infestations? How would you distinguish yellowing caused by mites from other causes of yellowing, such as drought?



Sara Lubkin 09-09-2016, 08:55

Hi, Kelsey. One of the struggles with this project was that palm plantations in Puerto Rico are small and tend to be long and narrow. We did some preliminary studies in El Salvador and found that large palm plantations give a much clearer signal of yellowing. So, we would like to try this on a larger scale. The decision tree analysis in Orange provided a list of potentially significant indices and bands. I think a larger scale would allow us to say definitively which indices are best.
I don’t know if we could track early infestations. But, we could combine historical mite maps with other variables such as temperature and moisture for ecoforecasting future infestations.
As far as yellowing, there are two potential causes.RPM and lethal yellowing (LY) disease. LY has a different pattern of yellowing and kills trees, while trees survive RPM. Thanks for your questions!

Alison Thieme 09-09-2016, 13:58

Thanks for your questions Kelsey! To add on to Dr. Lubkin’s comment, no single index was a perfectly able to identify mite infestations. Instead, a combination of quite a few bands and indices were incorporated into a classification tree. The details about the classification tree along with the relationships between indices and the mites’ effects can be found in the technical paper.

In response to your question regarding near-real time monitoring, typically near-real time monitoring is used for looking at large scale issues that require rapid responses such as fires, floods, landslides, or changes in air/water quality. Since the mite populations continue to live on host plants, cannot be rapidly controlled, and are located in small patches of host plants, we felt that understanding the extent of the mite distribution was more relevant to our end-users than creating a near-real time monitoring system.

Dash Cruz 07-09-2016, 11:30

Awesome video guys! I was wondering “has anyone made plans to research the possible socioeconomic impact this mite would have if it spread throughout N. America?”

Sara Lubkin 13-09-2016, 12:34

Thanks, Dash. The USDA is concerned about the potential impact of the mite in California and Florida (it is currently present in Florida) and considers RPM a serious economic threat. Fortunately, the mite doesn’t do well in cold temperatures, so, thanks to some colder than usual East coast winters, the spread in the U.S. has been slow. But, climate change and warmer winters could change the distribution.

The risk analysis study is: Borchert D, 2007. Risk analysis and potential consequences associated with the introduction of the red palm mite Raoiella indica into the United States., USA: USDA-APHIS, 30 pp.

Daryl Ann Winstead (Mekong River Basin Agriculture) 18-08-2016, 13:57

Wow! Very interesting! Great video and project! Did the team happen to conduct some sort of accuracy assessment with the data provided by the project partners? Thanks in advance for your response!

Sara Lubkin 18-08-2016, 14:04

Hi Daryl Ann,
Our partners sent us a list of infested trees with coordinates and dates. We checked the list against a forest service classification to identify large stands of palms. We then used those sites as our regions of interest. In addition to observation dates, we knew from our partner at the USDA that none of those sites were infested before 2004 and that they were all infested after 2006. We used all bands and about 20 vegetation indexes to create before and after training data that were used to create a decision tree and other classifications in Orange.

Daryl Ann Winstead (Mekong River Basin Agriculture) 18-08-2016, 19:22

Thank you for your speedy response! I bet it was very interesting to see that within two years they were infested. Do you know what the typical rate for spread is?

Sara Lubkin 18-08-2016, 19:34

I don’t know the rate, but it seems to be really fast. Alison made that great animation for our video showing just how fast the mite spread through the Americas.

Daryl Ann Winstead (Mekong River Basin Agriculture) 19-08-2016, 12:20

Oh wow! Thanks!

Ronald Ochoa 18-08-2016, 12:20

In 2012 Drs Jorge Pena, Jan Bruin and Maurice W. Sabelis were editors for the volumen Exp Appl Acarol (2012) 57:211–213. This is an special issue that go over behavior, morphology, predators, plant associations and the known distribution at that year for the Red Palm Mite. It is an excellent work, I do recommend to read o see the issue and the articles. Raoiella mites are under study and still many unknowns on there behavior and ecology, the remote sensing technology .

Ronald Ochoa 18-08-2016, 12:29

…The remote sensing technology is a wonderful tool that will help us to understand and battle the impact on palms.

Sara H Lubkin 07-09-2016, 12:37

Thank you, Ron.

Rajkiran 16-08-2016, 18:49

Is there any natural way to get rid of this mite?

Sara Lubkin 18-08-2016, 10:23

Hi Rajkiran. In India, the red palm mite has natural predators. But, these don’t live in the Americas. The USDA is doing research to see if they can identify a species that can control the mite, but there is always risk when introducing a new species into an ecosystem. Other ways to reduce populations of the mite are low temperatures and very wet weather, but that’s not something we can control.

Alison Thieme 12-08-2016, 11:08

Special thanks to Chris Pooley, Dr Gary Bauchan, and Dr Ron Ochoa for the Raoiella indica images.

Tyler 12-08-2016, 09:18

Given the spread of the pest, is there anyway to successfully counteract it, without introducing a larger problem into this hemisphere via a natural predator to it?

Sara Lubkin 13-08-2016, 19:39

Hi Tyler,
Not really. The USDA is looking at natural predators, but so far, only very wet weather or very cold weather have made a difference. Guyana is using strong pesticides to protect their palm plantations, but widespread pesticide application isn’t a good long-term solution.


Leave a Reply