But that tree was something else. Robin Foster of the Chicago Field Museum was very helpful to me in those days, and he remarked that I could easily tell the leaves of the tree from the leaves of the liana because the tree leaves had wavy margins. The leaves of both plants are relatively large, rough, and have pinnate, parallel secondary veins. Both also have beautifully reddish young leaves. After learning about the wavy margins, I easily spotted it:

For handy reference, here are the leaves of Davilla kunthii:

My trouble was, I couldn’t find any mature trees with those wavy-edged leaves. We had plenty of saplings and I saw saplings elsewhere in the area, but not trees.

So I waited….

Finally, this year, one of the sandpaper trees was mature enough to bloom and bear fruit. Gotcha!

This small tree is Curatella americana L. It is in the family Dilleniaceae, which here in the neotropics consists mostly of plants with sandpapery leaves. I have a description of the family in the post on identifying Davilla, so I’ll just summarize the features here.

Plants in the Family Dilleniaceae have (all these images are from the Curatella tree – click on any thumbnail to see a larger image at Flickr):

• a papery fibrous bark

• rough and sandpapery leaves
• straight, parallel, close-together secondary veins

• narrowly winged petiole (leaf stalk)

• flowers with many stamens

With this assurance that we are in the right family, it’s really easy to discover that what we have is Curatella americana, since that is the only tree listed in the Dilleniaceae family in the Trees, Shrubs, and Palms of Panama database. Further, Gentry mentions that it is the only Dilleniaceae tree in northern South America.

Nevertheless, it’s good to know the characteristics of the genus and species. The features (other than wavy leaf margin) that distinguish Curatella americana from any member of the Davilla genus are its flowers and fruits.

The inflorescences appear on the stem, “… in little branches without leaves.” (Carrasquilla) The flowers are cream-colored, 1-1.5 cm in diameter, with 5 petals and many stamens (Zuchowski).

The fruits are 1-1.5 cm across, with two main compartments, each with 2 dark brown to black seeds with thin white aril (Zuchowski).

Its name

Curatella was named by Swedish naturalist Pehr Loefling (1729-1756), one of the 17 “apostles of Linnaeus.”

It is a name from the Greek meaning work or working. Dave’s Botanary says that Amazon natives used the leaves to polish bows and arrows. From the little I’ve read about Loefling, however, his work in tropical America was in Venezuela, not near or on the Amazon. Still, not to quibble. The name is appropriate, since the leaves can be and are still used as sandpaper.

Its habitat

Curatella americana is one of the savanna trees that survives fire well (Kricher), as you may be able to tell from its crooked and twisted appearance. This type of growth is characteristic of such trees. Fires stimulate its seeds to germinate. (Zuchowski)

Its uses

The tannin in its wood is used for curing hides, and the wood itself, which is hard, is used for posts, firewood and charcoal. In Mexico, the edible seeds are toasted and added to chocolate. The buds may be made into a concoction for asthma; bark or leaf concoctions may be used for bathing wounds. (Zuchowski)

In summary

Curatella americana is a member of the Dilleniaceae family, whose rough, sandpapery leaves have pinnate, parallel secondary veins and whose leaf stalk is winged. Young leaves are often strikingly reddish. Flowers in this family have many stamens. The flowers of Curatella americana are white, with 5 petals, and occur in clusters arising from the stem. Its fruits are hairy and reddish, with two main compartments. It is a tree that survives fire well with its twisted shape, and fires actually stimulate its seeds to germinate. It’s well suited for the savanna, and the people who live with this tree have found uses for its buds, fruits, bark, leaves, and wood.

The weevils that swarmed on the sandpaper plant when it was in bloom have been identified. I received this email from Dr. Henry Stockwell, who kindly performed the identification during his recent stay at the Smithsonian Tropical Research Institute.

The weevils that you sent me are named Solaria curtula, of the subfamily Baridinae. As I mentioned to you on the phone, the larvae have their development in stems near the ground of poor nutritional value and must make an effort to find pollen after they emerge from a brief pupal stage. This is one of the very few weevils that is found on a fern, Dicranopteris. This fern is very common and forms a mat of vegetative growth on bare red-clay slopes. The individual single plant may cover several square meters. It is the first plant to reclaim bare hill-sides after a land-slide. Can’t think of a plant that might be poorer nutritionally.

Someone had suggested to me that the weevils might be laying their eggs in the pistils of the sandpaper plant so that the larvae could develop along with the fruit. Dr. Stockwell said on the phone that such behavior would be extremely unlikely because of the known life cycle of this species, which he then described in the email, above.

I should also add that although I saw the weevils mating during their pollen-eating swarm, I did not see them laying eggs, nor did I see them spending much time near the greenish pistil of the flower, which is where they would have had to lay their eggs.

Thank you, Dr. Stockwell.

There were two massive blooms, separated by a three-week period, and the sequence of events was the same each time. Here’s the chronology. All events – both flower development and insect activity – occurred about 30 minutes earlier in the day for the second bloom than for the first, so ranges of time are given.

Before sunrise – Flowers are just barely open and the stamens are barely exposed.

At sunrise - A few bees arrive.

Thirty minutes to an hour past sunrise – Flowers are fully open and the bees are swarming. (You may need to wait for the movie to load.)

The bees are collecting pollen and no doubt distributing it as well. The petals are as wide open as possible to attract the bees to the flowers, and the stamens (pollen-bearing male parts) and the pistil (female part leading to the ovary) are fully exposed to bee activity.

One and a half hours past sunrise - Petals curl back. Weevils arrive, some bees leave.

The petals curl back before they drop off. The bees act as if this curling were a signal to start leaving.

Two and a half to three hours past sunrise - Petals dropping. Bees mostly gone. Weevils swarming and mating.

It has been suggested that the weevils may deposit their eggs in the pistil so their larvae may mature within the flower’s ovary. When the fruit ripens and the seeds form, the young weevils will eat the seeds. This idea needs to be tested, but the weevil behavior here is suggestive.

Three and a half to four hours past sunrise - Petal drop nearly complete. Sepal cups, stamens and pistil remain. Weevils remain.

If the weevils are indeed preparing to deposit eggs, the absence of petals is not an issue for them. The sepal cups now serve as a platform for walking around.

Five to six hours past sunrise - Sepal cups begin closing. Weevils remain. (Please pardon the label. Weevils are a form of beetle, of course.)

Even as the sepal cups close, the stamens and pistils are still easily visible to the small weevil and, if they are depositing eggs, that can still be done at the last minute.

Eight to ten hours past sunrise – Sepals completely closed. Occasional weevil remains.

If that last weevil is looking for a pistil to deposit its eggs, it’s out of luck.

The closed sepal cups will now act as protection for the developing fruit. Soon after taking this picture, I plucked a sepal pair off the plant and tried separating it. The sepal “resisted” being opened. Once I released my grasp on the cups they sprang quickly back into place. Two days later when I repeated the experiment, the sepal cups were “frozen” in place. I could not separate them at all but had to slice them apart.

To me, this sequence is remarkable. The flower goes through its development, from open bud to closed sepals, within an eight-hour period. The period from the time the bud first opens to the time the petals are fully extended is very short – about an hour and a half. During this time bees find the flowers, swarm to collect the pollen, and leave.

It’s worth noting that insect activity was much greater for the second bloom than for the first. At the height of bee swarm during the first bloom, I estimated that (very roughly) 50 bees were buzzing around at a time. During the second bloom, roughly 1,000 bees were there during the swarm. I failed to come up with a good estimate of weevil numbers, but I had the impression that they, too, were more abundant the second time around. It’s as if the bees and the weevils were alerted to the blooming of this plant the first time around, and the word spread. By the second bloom they were ready to pounce.

If indeed the weevils are laying eggs in order for their young to eat the seeds after they develop, then it’s surely no coincidence that the weevils “let” the bees have first dibs at the pollen. After all, the flowers must be fertilized in order for the fruit to develop.

The image is not as well focused as I would like, but the beetle appeared and then disappeared, never to be seen again. So this is what I got.

The other beetles, the weevils that I saw last year and also two weeks ago, returned, as did the bees. This time all insects were in much greater number – probably a thousand bees at this one plant by 7:30 in the morning. I had a harder time estimating the beetles, but there were easily twice the number that there were two weeks ago.

I do not expect this plant to produce more profuse blooms this season. Now is the time to gather all my notes and try to make sense of what went on.

In a much more humble celebration today, I launched another blog, The Accidental Botanist, where I can ramble on about plant things that are unrelated to our neotropical savanna.

Sunrise (6:49 AM):

As the morning progressed and the flowers opened, it became clear there were two kinds of bees interested in this plant – a large one (circled), and a small bee. (or is it?), below. (The image below, as noted by a corresponden- with an Asian name I can’t translate – is a type of hover fly. I have some other images of the small bee, but they’re not clear enough to post.)

8:00 AM: At least a couple dozen bees were buzzing around. By 8:30 there may have been 50 or more. It was clear that the large bee, at least, was collecting pollen. My only image of this so far is fuzzy, but it’s not hard to see the orange pollen sacs.

8:30 AM: the first few weevils. The weevils are the entire reason I was sitting in the shade comfortably, with my coffee and notebook, watching this sandpaper plant put on its spectacular bloom. I had seen and photographed a weevil swarm last year, and Robin Foster of The Field Museum of Chicago suggested I look at the swarm more closely this year. I’m going to be tracking this plant throughout its blooming period to watch for these weevils.
That they are weevils has been verified by Dr. Henry Stockman who is this January at the Smithsonian Tropical Research Institute here in Panama and kindly looked at my images from last year. He suggested that it is possible that these weevils are eating the pollen – which is a rich protein source. As I watched, I could not see whether they were eating, but they certainly did not seem to be collecting it the way the bees were.
By the time the weevils arrived, some of the petals had already dropped from the flowers. The flowers had gone from opening bud at just before 7 AM to fully open flower, to the commencement of petal-dropping at 8:30 AM. One and a half hours. Astonishing!

9:30 AM: The number of bees was down and the number of weevils was up.

10:30 AM: The petals were seriously dropping off. I could hear the petals landing on the sandpaper-like leaves. They sounded like raindrops. There were even fewer bees, but the weevils were still there.

11:15 AM: Nearly all the petals were gone but several weevils stuck around. When the petals fall off, the cup-like green sepals, which underlie the petals, are left behind with the stamens. The weevils seem unperturbed by the absence of petals and simply crawl around the sepals looking for their pollen meal.

1:30 PM: The sepals were closing in on the stamens. A few weevils were still around, hoping to avoid getting trapped in the sepal jaws, I assume.

4:00 PM: The sepals are completely closed around the stamens. I saw one forlorn weevil crawling around on the sepal surfaces. Some sepals qwew separated enough to show yellow – they’re the ones that were bursting with petals to open the next day.

What a day for a plant! I’m sure Darwin would have loved the show. Remember his study of the “sleeping” leaves of the little machete plant, Erythrina? I wonder what he would have made of the opening and closing jaws of these sepals.

Because I had seen the swarm last year on January 23, I decided to start keeping close tabs on the plant in early January. This morning, the first day of January, I found the plants already budding, and I frantically started taking notes and pictures. [Click on any image for an enlargement, or click on the "hi-res" link (when present) to see the original size at Flickr.]

Some of the buds were even showing some yellow where the petals would soon appear.

hi-res

I had no idea just how soon, though. I took the above picture at 6:49 AM. After uploading it into the computer and seeing the cobweb – which I hadn’t noticed when I was out there – I went back to examine the cobweb more closely. By then, at 10:21 AM, the petals were out.

hi-res

The two images are not exactly at the same angle, but they’re certainly of the same flower cluster. Here’s a different cluster, giving you a better sense of the flowers. I did see one flying insect land on the stamens and crawl around for awhile, but naturally I couldn’t get its picture.

hi-res

By early afternoon, the wind had come up and the petals were gone, leaving only the stamens. A petal life of about 4 hours? How do the pollinators find them in such a short time frame? Or do the pollinators still work after the petals are gone? Oh, and what was that cobweb? There’s another cobweb in the last picture as well, stringing down from a petal to the leaf.

This is going to be an intriguing experience.

UPDATE: 2 January 2009. I checked on the petal opening again this morning – this time I went out there right after breakfast, and, lo, where there had been buds at 6:30 AM there were flowers with petals at 8:30 AM. It looks like I’m just going to have to pull up a chair some morning and watch.

Meantime, Festival of Trees edition #31 is already up at Rock Paper Lizard - and has been since December 31, 2008! I’m just about half-way through it, but there’s lots of good reading, as usual.

UPDATE: 21 January 2009. There have been no blooms on any of the Davilla plants since that second day of January. But there are now hundreds of buds on this plant, so it is quite possible that when they start to bloom, they will bloom all at once – or in a couple of waves of blooms. Last year’s burst of flowers came on 23 January 2008 – so it’s very close to time, but the buds are small and tight. Does it know it’s supposed to bloom soon?

UPDATE: 30 January 2009. Massive bloom today – with bees and weevils and more! I’ll write it up this weekend.

We had already decided to keep as many trees as possible on the property, and to encourage other interesting plants, but when we saw that the leaves of the sandpaper plant also frequently took on a nice reddish color, we doubled our efforts to see that these plants did not get cut down, either by us or by our gardener.

Seeing it flower - But it wasn’t until this year that I had enough time to watch for signs of flowering. On January 12, I first noticed a plant with buds.

And a week later, January 19, I saw my first flower:

It seemed a little strange that there should be only one petal (the yellow tongue hanging down in the picture), so I decided to keep watching. Every now and then I’d see another flower or so but invariably I didn’t have my camera with me. Finally, on January 23, the flowers on one plant were suddenly open and covered with bugs! (Click on either image to enlarge. You’ll love the bugs.)

I’m assuming the bugs are the pollinators. They were certainly busy! After uploading the images to the computer, I wasn’t satisfied with them and decided to take some more, even though the wind had picked up and was gusting. One hour and twenty minutes after taking the last of the above pictures, I found this sight:

The petals were gone – all of them! All that’s left are the sepals and the stamens, along with a few pollinators. I mean, those bugs were busy, but did they knock off all the petals? Or was it the gusting wind, which seems not to have disturbed the flower petals of other plants in the vicinity? Regardless of the cause of the disappearing petals, it’s small wonder I haven’t been able to find good pictures of the flowers on the internet!

The Sandpaper Plant is in the Family Dilleniaceae – Whenever I look up sandpaper tree or sandpaper plant in the index of a plant book, I am directed to a description of a plant in the Family Dilleniaceae. The Family name comes from an old-world tropical tree that Linneaus himself named – Dillenia in honor of his friend, the German botanist and physician Johann Jacob Dillen. There’s a nice illustration of a Dillenia flower in wikipedia.Members of the Dilleniaceae family have papery fibrous reddish bark (Gentry)

Their leaves are often rough and sandpapery (more on this later) with straight, parallel, close-together secondary veins (easier to see on the underside, shown below and more easily seen in the enlarged image – click on it), connected by subperpendicular, strongly parallel tertiary veinlets (Gentry)

Note that the secondary veins are not really straight in this species, but they are definitely parallel, as are the tertiary veinlets.The petiole (leaf stalk) is often narrowly winged and usually more or less continuous with the wedge-shaped leaf base. (Gentry)

The winged petiole is pretty clear here, but the wedge-shaped leaf base (where the leaf and petiole meet) is better seen in the previous illustration.The flowers have many stamens (the pollen-bearing male parts inside the petals). The bugs are all over those stamens.

Is it a tree, a shrub, or a vine? All the books I mentioned earlier that have sandpaper plant in the index take me to a description of a tree, Curatella americana. I’ve been watching the sandpaper plants on our property for three years now and only one of them looks like it will grow into tree. That one isn’t pictured here – it’s not in bloom right now, and it’s also pretty scraggly. Further, the flowers of the tree Curatella americana are described as cream-colored in Costa Rica (Zuchowski) and as light green in Panama (Carrasquilla). The plants here look like shrubs, but there’s one plant in a more shaded area that suggests it could be a vine. The next three images follow the main stem as it reaches up toward a tree branch.

Gentry does say that the family Dilleniaceae is mostly lianas, or vines. The answer to whether this plant is a tree, shrub, or vine came from Robin Foster of the Field Museum of Chicago, who has been of great help to this bumbling amateur. He explained that the photos I sent him were of the genus Davilla, which is “…usually a liana [vine] but when growing in an open area can be shrubby.” He also mentioned that the plant “…will develop yellow fruits formed from those two cup-like calyx lobes, until they finally open when ripe.” He calls the Dilleniaceae the sandpaper family. And how do I tell this liana/shrub Davilla from the tree Curatella found in my reference books? Foster says that Curatella will have a “larger longer leaf with a very wavy margin.” So I’ll keep my eyes open.

What about the sandpaper? Foster calls this the sandpaper family, Carrasquilla [Trees & Shrubs of Panama] says the leaves of Curatella are used as sandpaper, and Zuchowski [Tropical Plants of Costa Rica] says that silica in the leaves of Curatella give it a sandpaper quality. So, do the leaves of Davilla have silica in them as well, and can they be used as sandpaper? The leaf certainly feels rough, so I decided to scan it so I could magnify it and look for silica spikes. Here’s what I found:

The arrows point to just a few of the silica spikes. The leaf is studded with them. Then I tried using a leaf as sandpaper. It can be done. My “before (left) and after (right)” shots were taken pretty casually and so don’t show the difference very well, but it was there, trust me. Clicking on either image will give a slightly larger view, which might help.

However, before you decide you can go into the sandpapering business just by growing a few Davilla plants, be forewarned that I used up the silica from an entire leaf on this section of a small stick. Just be sure you have lots of plants in your area before you take on a sizable sandpapering project!

UPDATE: Thanks to Gordo’s comment (#3 below), I decided to provide a link to the Field Museum’s Tropical Plant guides, which show a few different Davilla species. Find those images at http://tinyurl.com/yry334

Also, here’s an earlier picture of the leaves – better exposure than some of the images that went with the story.