What feature brings a plant – whether a tree, shrub, vine, or garden variety – to your attention?
For me, it’s often the flower. I’ll be aware of a plant in the background, but when it blooms, then I really pay attention.
For the Cojoba tree, though, it was the leaves that made me stop and look. Toward the end of February, while the leaves of all the trees on the west side of our orchard wall were in various shades of green, the leaves of one kind of tree suddenly had a soft, rufous look.
The leaves themselves were attractively arranged, swooping down like giant bird feathers. The foliage made a nice contrast to the other trees, and I decided to watch for any sign of flowering to help me learn more about the tree. I thought I could guess already what family it belonged to, though. Those bird feather leaves were the biggest clue.
One of the easiest ways to start recognizing plant families in the tropics is the method developed by Alwyn Gentry. You decide first whether the leaves are
- simple (only one blade is attached to the stem) or
- compound (more than one blade is attached to a stalk, and that stalk is attached to the stem).
[This distinction is discussed in some detail here.]
Then you decide whether the leaves are
- opposite (leaf stalks or petioles are on opposite sides of the stem) or
- alternate (leaf stalks are found first on one side of the stalk, then the other, as they make their way up the stem).
In the case of Cojoba, the leaves are compound and alternate.
When you have this combination, you then examine the nature of the compound leaves. Are they arranged like bird feathers (pinnate) or like fingers on a hand (palmate), or are they found in groups of three (3-foliate)? If the leaves are pinnate, you need to decide whether they are simply pinnate or bipinnate. We already saw that these leaves are somewhat like giant bird feathers, so we know they are pinnate. In the case of Cojoba, they are simply pinnate.
Finally, count the blades on the leaf stalk to see whether you have an even or an odd number. Most of these leaves have 10 blades, some have fewer, and some have 12. So we have an even number, which means these are even-pinnate leaves.
Here’s our combination of leaf characteristics, then:
Gentry lists only 4 tropical families with this combination of traits, one of which is the palm family. I read his descriptions of the other 3 families, and the one that best fits is the Fabaceae family, or the bean family.
I looked at the Fabaceae family in some detail when talking about the living fence tree, the macano. The bean family is very large and is divided into three subfamilies: the mimosa subfamily, the cassia subfamily, and the pea subfamily. To decide which of these subfamilies this tree belongs to, I needed to wait until it bloomed.
The first flowers that I saw were on March 2, but before I got a decent picture, the rain had “wilted” them. About three weeks later we finally had another bloom.
These round balls of stamens immediately told me that the tree belonged in the mimosa subfamily. [Check here if you don't know why.]
When I went back to Gentry I had every reason to believe that this tree belonged to the genus Inga. Gentry stated that Inga is “…unique in being even-pinnate and with glands between all leaflets.” I went looking for the glands, and there they were (click on the image if you need to see a larger version):
The purpose of these glands is obscure. It’s possible they offer nectar to ants to keep the ants away from flowers, but the secretions from such glands may be resinous, gummy, or aromatic. I don’t know why these glands are here.
Inga is the ice cream bean tree
At any rate, I was very excited to think that this tree might be an Inga. It produces long pods with a juicy, white pulp that tastes a little like vanilla ice cream. To learn which Inga it might be, I turned first to Trees, Shrubs, and Palms of Panama, which is a database maintained by the Smithsonian Tropical Research Institute. It has descriptions and images of most of the recognized plants in the country. Not every plant has every feature pictured, however, and I found nothing that seemed to fit this tree.
Next I turned to my friend Carla, of Heliconia fame. She has a book of all the Ingas in Costa Rica, and she sent me a very good key, but, again, I could not find a match.
It’s not Inga – it’s Cojoba
Finally, I sent off some images to the kind botanist of Tropical Plant Guides, Robin Foster, who has been of great help to me in the past. He identified the tree as Cojoba rufescens. He made me feel a little better, though, when he said that it was often confused with Inga. Indeed, when I returned to Trees, Shrubs, and Palms of Panama to look up Cojoba rufescens, I read this paragraph on how to recognize it:
Compound leaves with an even number of leaflets (no terminal leaflet), and a gland between each leaflet pair, characterizes the entire genus Inga, as well as Abarema and some other related genera. Cojoba rufescens is fairly readily recognized, though, by the undulate leaflets, a consistent trait which is not found in Inga.
Cojoba has undulate leaflets
Here are the wavy edges, which, admittedly, are more obvious in real life than in these scans of the leaflet tops (left image) and bottoms (right image). The first image in this post also gives you an idea of the undulations.
The description also states that the leaflets near the base are reduced in size, which you can see here. Further the trunk may be “leaning and branched near the ground.”
Trunk branched near the ground
Here it is. The ground is hidden from view by other vegetation, but you can see several branches of the trunk here.
And finally, “when there are fruits, Cojoba is unmistakable.” I’ll have to wait to get my own picture of the fruit, which is a bean, but here’s the image from the Smithsonian. The bean has a red pod.
Update: The beans were bright and lovely by August of 2008. Here’s one:
The coral snake bean
And finally we get to the name of the tree: Cojoba rufescens. Dave’s Botanary gave me the name for rufescens, which means “becoming reddish.” I believe that must be in reference to the leaves. But nowhere could I find the etymology of the word Cojoba.
Finally, I turned to the Integrated Taxonomic Information System (ITIS) and its report on Cojoba. The expert botanist listed there kindly replied to my query and said that the Taino Indians of the Caribbean called the dried leaves of this plant Cohiba or Cojoba.
The Cojoba genus is found in the West Indies, Central America and into South America.
The species C. rufescens, though, has not been reported from the West Indies, but is distributed throughout Central America with a few reports from western South America.
Panamanians have two common names for C. rufescens: guabito and coralillo (Carrasquilla). Guaba (guabito means “little guaba”) is the ice cream bean! So even though the red, curly bean of Cojoba looks nothing like the ice cream bean – which is long, brown, and thick – the leaves look enough like the ice cream bean tree to be named after it.
The other common name, though, comes from the bean itself. Coralillo is the Spanish word for coral snake, which has alternate bands of red and black. The twisty bean, with its red pod and black beans, which adhere to the pod even after it opens, must look pretty fierce in the tree!