Thousands of Martins Killed by Hail in Brazilian Tornado

Reprinted from: Purple Martin Update 6(1): 2-3
James R. Hill, III
Purple Martin Conservation Association


A Brazilian woman holding two of the thousands of Gray-breasted, Brown- chested, and Purple Martins that were killed by wind and hail in their roost trees in Ribeirão Preto, Brazil.


Last summer I received the following letter, dated 21 June, 1994, from Olga Kotchetkoff Henriques, a Biologist with the Environmental Bureau of the Council from Ribeirão Preto, a city in the southern Brazilian state of São Paulo, where Purple Martins roost during our northern winter.


Dear Mr. James Hill, III,

Last May 14th, 1994, late in the evening (11:20 p.m.), a hailstorm beat down on Ribeirão Preto. Winds in excess of 100 kilometers per hour were reported.

Several building roofs, small houses, and walls collapsed. Many trees fell down, including some of the ones that served as resting places for the martin flocks. These birds were severely affected. We believe that 80% of the martins that had not yet migrated from Ribeirão Preto died (about 20,000).

We are interested in knowing what effect this ecological disaster will have on the flocks of these birds. Our intention is to monitor their populations and behavior, such as their choice of next roosting place, during the following summer (your winter).

We thank you in advance if you could inform us about what you feel in relation to their population and behavior during your next summer. Sincerely yours, Olga Kotchetkoff Henriques, Biologist/Sec. Mun. Meio Ambiente


Below is a copy of the response I wrote from the PMCA:


Dear Olga,


This letter is in response to the letter you sent to us last June, 1994. First, I apologize for taking so long to write back to you. I wanted to show your letter to some of the other ornithologists I travelled to Brazil with in the past. One of these ornithologists is Dr. Eugene Morton of the Smithsonian Institution in Washington, D.C., who I understand wrote back to you separately.

I was in the state of São Paulo, Brazil, in February of 1985, 1986, and 1990, studying these roosts of Purple Martins. On several of those occasions, I visited Ribeirão Preto. You have a spectacular roost of andorinhas [Portuguese for "martins"] in your city. Did you know that these roosts in the trees of your city are made up of three different species of martin? They form mixed-species roosting flocks, whose composition changes over time. You have the "Purple Martin" (Progne subis) that migrates from its home in South America to breed in North America; then there are the "Brown-chested Martin" (Progne tapera) and the "Gray-breasted Martin" (Progne chalybea) that spend their entire annual cycle in the Neotropics. All three species sleep at night in the trees of Brazilian central parks, including the roost(s) in your city.

It is our understanding that Progne subis arrives in these Brazilian roosts from North America in November, molts new feathers during the next several months, then gradually departs for North America by mid-March. Many leave much earlier than this, however. The first martins return to the southern United States around mid-January, the trip from Brazil taking them 4-6 weeks. Therefore, the 20,000 martins killed by the hail storm on May 14th, 1994, in your city could not possibly have been Progne subis. They had to be Progne tapera and/or Progne chalybea. I have enclosed some photocopies of information on the three species of Progne [from the book, Swallows and Martins by Houghton Mifflin].

Please understand, that even if they were P. subis, no one in North America would notice a decrease in the number of martins nesting in their martin housing. The total world population of P. subis has been estimated at about 6 million birds. Fifty percent of the adults and subadults die each year of natural causes, as do about 75% of the young. So it is normal for many millions to die each year, most during the high-risk periods of migration. Still, 20,000 martins killed by a hail storm is an incredibly-spectacular, natural disaster! Did anyone take photographs of this? If so, would it be possible to get copies? These would be very valuable to us. We would like to do a story on this storm kill in our magazine. I thank you in advance for bringing this to our attention and for sending any photographs you can. Sincerely, James R. Hill, III, Exec. Dir., Purple Martin Conserv. Assoc.


Near the end of 1994, we received the following letter (translated by J. Hill) and photos, dated 12 December 1994. This was in response to my previous letter.


Dear Mr. James R. Hill, III,


We communicated with you earlier this year about a powerful tornado that touched down in Ribeirão Preto in May, causing vast damage to the city. Many kinds of animals and plants were affected by the violent storm, and it also caused great financial loss, and overtime work for the cleanup crews. It will take a long time for the trees, animals, and humans to recuperate from this great loss.

I am sending you photographs of the tragedy that victimized thousands of martins (4,000-5,000), which perished when this tornado touched down. This happened during the night when all the martins were roosting in the trees near a local highway. As you can see from the photos, this storm caused great damage to the ecology since these martins had their lives taken in such a brutal manner. The outpouring of sympathy toward the injured martins by the people of our city was quite impressive. Most of the birds were killed outright, but the injured birds were gathered up, dried off, fed, and placed in warm baskets. As soon as they had recuperated, many were released.

The mayor of the city where the tragedy occurred communicated with several scientists in the United States and Canada who study the Purple Martin. I am sending you photos of this tragedy for your study.

At the time of the tragedy (14 May 1994), I believe that these martins were preparing to return to the United States and Canada because here in the Southern Hemisphere it was the beginning of Winter [actually, it was just beyond mid-Fall].

Please send me any material you may have on Purple Martins. I have a great interest in them. I will end this letter, wishing you health and peace, and place myself at your disposal should you need anything further. Cordially yours, Antonio Herrique Ravasi


Antonio Herrique Ravasi sent us 24 photos of this disaster, only two of which we had room to print here. Several show town folk methodically searching amongst the fallen limbs looking for any martins that had survived. Close scrutiny of a photo showing 15 of these rescued martins, shows that at least three were P. tapera, about 8 were P. chalybea, and four of them were clearly P. subis; they had the characteristic dusky-gray foreheads and gray neck collars of female (or subadult male) Purple Martins! It is astonishing to me that any Purple Martins were still in Brazil by this extremely late date. Clearly, over 99% of the entire population should have already migrated and reached North America by May 14th. Subadults, the very last age group to arrive in North America, typically begin arriving in southern Ontario, Canada, by May 14th! Why would some Purple Martins still be in southern Brazil on this extremely late date? Is it possible that some subadult males don't migrate? Further study of these roosts, on a year-round basis, would be very enlightening.


Three species of martin carpet the road in the aftermath of a wind and hail storm that killed thousands as they roosted in the trees of the Brazilian city of Ribeirão Preto, state of São Paulo, on the night of May 14th, 1994.

There were inconsistencies between the two letters we received from Brazil on how many martins died in this storm. It's unclear whether it was 20,000, 16,000, or 4,000-5,000. One Houston newspaper even reported that over 1 million martins had died (a confusion probably arising from the fact that the Portuguese word for thousand is "mil"). I wish that an ornithologist would have salvaged the birds for an accurate species count. Most were clearly the Gray-breasted Martin (P. chalybea). Even if 20,000 Purple Martins had been killed in this storm, I contend that no North American landlord would have noticed a decline because of it. As stated in my letter to Olga Kotchetkoff Henriques, 20,000 is biologically insignificant when you look at the normal demographics of Purple Martins - if there really are 6 million Purple Martins in the world (this is just an estimate), then 3-4 million die each year from natural causes.

Another reason North American landlords would not have seen a decline in the number of birds returning from Brazil (had this storm actually killed 20,000 Purple Martins, which it didn't) is that colony-mates don't migrate or overwinter as a group, they spread out over the entire wintering range. If you assume that there are 200,000 landlords in North America with active colonies, each would have had about an equal chance of having a bird killed by this storm, or only about 1/10 of a bird lost per landlord. So, in conclusion, no landlord should be concerned that they might have lost martins in this storm - 99% of the world's Purple Martin population should have already migrated to North America. If your colony suffered a decline in 1994, it's highly unlikely that it was because of this freak storm.

Copyright 1995 by Purple Martin Conservation Association. All Rights Reserved.

Our members benefit from 4 issues annually, packed full of helpful and fascinating information like the article above. You can become a member and support the work of the PMCA by making a tax-deductable donation.