Plants of Southern California: Pictorial Guide to Some Characteristics Needed for Lupine Identification

This is an incomplete, draft page!!

Major Characteristics Used to Separate Species
Terminology of Flower Parts
Pictures of Flower Parts


Lupines are difficult to identify from photos alone, and often difficult to identify in the field unless you know what to look for. Lupines cannot be keyed out from normal photos taken of a plant since many of the characteristics used to discriminate species are minute features, some of which are hidden behind parts of the flower. Others are characteristics that one would never think to photograph in the field, such as whether the cotyledons of an annual species are fused or not, which is detectable on a blooming plant if you photograph the base of the plant.

This page discusses and shows pictures of some of the features needed for lupine identification. This first draft concentrates on minute features of the flower, since it is often difficult to show someone these features in the field. But modern point and shoot cameras, in macro mode, can easily capture those features, if you know what to photograph. All of my photos below were taken with a Sony T-9 Point and Shoot Camera in macro mode; many other such cameras can take similar photos.

The basic problem in lupine identification is that to a first approximation, most lupines look alike. They all follow the same construction plan, with palmate petioled cauline leaves that have nearly identical leaflets, and a terminal raceme of flowers nearly identical in form that are nearly all blue / lavender / purple. Very few lupines show variation in their habit, leaves, or flowers that can be used to discriminate species. Worse, many taxa themselves are quite variable, which throws yet another monkey wrench into species separation.

This dilemma is illustrated by the set of pictures in Fig. 1. One row contains pictures of a single species; the other row contains pictures of four different species; can you guess which row is which?

Row 1
Row 2

Fig. 1. The caption is at the bottom of this page.

Compounding the identification problem even further, there are 71 species and 101 total taxa (including subspecies and varieties) in the Jepson Manual! (and this is after combining four subspecies each of L. bicolor and L. concinnus into the parent species.) It doesn't help much to consider just the taxa in southern California; Munz contains 48 species and 64 taxa, not including the subspecies of L. bicolor and L. concinnus.

Clearly, in order to sort out all those taxa, a key is going to have to resort to some pretty subtle characters and seemingly-small differences.

As a result, it may seem that too many taxa may have been defined, and this may well be the case for a small number of species. In fact, the Jepson Manual treatment mentions that two species are doubtfully distinct from two other species (L. albifrons and L. excubitus; L. albicaulis and L. andersonii).

However, my experience, from the twenty taxa I've come across, is that even taxa that seem to be pretty close, just in reading the Jepson Manual descriptions, are actually quite distinct in the field. For example, L. arizonicus and L. sparsiflorus have only two easily-observable taxonomic differences (flower color and leaf length to width ratio); are genetically very close; and have essentially what I like to call a magic dividing line at the base of the mountains on the coastal side of the Colorado desert that separates their range . All of these things normally raise my suspicions about whether two taxa are distinct. Yet these are clearly separate species, recognizable at a glance in the field, that cannot hybridize due to incompatibility at the gametophyte level.

Major Characteristics Used to Separate Species

Page incomplete past this point

Keys to separate species try to use characters that have different states in about half of the species, so they can be used up front to divide the species into manageable subgroups. The major characters used in the Jepson Manual key are:

Terminology of Flower Parts

Fig. 2 shows the parts of a lupine flower:

Fig. 2. Labeled parts of a lupine flower. (unlabeled picture)

A lupine flower is just a standard pea flower, with five modified petals. It has an upright banner, usually with a spot at its base to direct insect pollinators to the nectar reward. The nectar, stamens and pistil are tightly packed inside the keel (formed by two mostly-fused modified petals), which itself is covered almost entirely by the two wings of the flower. The two wings are attached near their tip, but the rest of the wings are free.

It seems like an odd arrangement to have the sexual parts so well hidden, but one source says that the lupine flower is highly adapted to accommodate the bodies of bees with a minimal production of nectar.

Pictures of Flower Parts

The most important part of a lupine flower for identification is the keel, yet the keel is almost always completely covered by the wings! There may be a fundamental reason for this; the same source says there may very well be some significance to the exact shape of the flower of a particular lupine species and the bees that frequent its habitat.

Whatever the reason, if you want to identify lupines, you have to check the keel for hairs on its upper margin. To do this, you have to either remove the wings; pull apart the wings where they are joined at their tip, and fold them back; or find a flower (usually one that has finished blooming) that has the keel exposed. Fig. 3 shows a keel with hairs on its upper margin, and Fig. 4 shows one with no hairs, both visible after the wings were removed from the flower.

Fig. 3. A flower of L. excubitus var. hallii with its wings removed to show the ciliate keel. (unlabeled view)

Fig. 4. A flower of L. formosus with its wings removed to show the glabrous keel. (unlabeled view)

The two pictures above also show that the shape of the keel can vary between species. The keel of L. formosus is described as upcurved, which is clearly is compared to the keel of L. excubitus. Both keels are slender and pointed. Some species such as L. truncatus have a keel that is more stout and blunt-tipped.

There is one more important property of the keel, whether it is lobed near the base or not. This is the main discriminant between L. excubitus and L. albifrons. The lobe is visible in both Fig. 3 and Fig. 4; can you find it? If not, look at Fig. 5.

Fig. 5. A flower of L. excubitus var. hallii with its wings removed to show the ciliate keel and the lobe at the base of the keel. (unlabeled view)

I don't yet have a picture showing an unlobed keel from a different species; I'll add that when I acquire one. Page incomplete past this point

If you think the properties are the keel are subtle, wait until you see the following pictures about an even more important characteristic, whether the banner back is hairy or not. This is the most important key characteristic in the Jepson Manual perennial / shrub key, and it is often quite subtle! Page incomplete past this point

The Jepson Manual gives a great tip that the easiest place to check for banner back hairs is in bud.

The caption for Fig. 1: Row 1 contains four different pictures of the same species, L. andersonii, taken from Calphotos pictures. Row 2 contains four different species; from left to right, they are: L. formosus, L. hyacinthinus, L. latifolius, and L. excubitus. I took the first two pictures; the last two are from Calphotos.

Go to:

Copyright © 2010 by Tom Chester
Permission is freely granted to reproduce any or all of this page as long as credit is given to me at this source:
Comments and feedback: Tom Chester
Last update: 1 June 2010