Counting Flowers  

Fig. 1:  Tripped flowers on a raceme.  What do they really mean?

If you want to try to measure the amount of pollination that is taking place in an alfalfa field, it is natural to count numbers of open flowers, tripped flowers, and perhaps buds and wilted flowers on a sample of racemes.  Here is something to think about if you plan to make these measurements.

Your counts will be a snapshot in time of the condition of the raceme.  Counts per se give you no information about the rate at which flowers are opening or being tripped.  Those two rates may be very different.  You need to know both of these rates in order to interpret changes in numbers of open flowers over time, and to make comparisons between fields and between years. 

Open flowers that are not tripped remain open for about a week.  Flowers that are tripped wilt within an hour or so.  Temperature plays a role in these rates, but the difference is great.   In some plant species it has been demonstrated that pollination triggers a peak in ethylene production, and this spike in ethylene production results in senescence of the flower^1,2,3.  Thus there is a likely physiological mechanism for the difference flower duration in pollinated vs. unpollinated alfalfa flowers.

In the systems model that I have created for alfalfa pollination, the relationship between buds, open flowers and tripped flowers can be diagramed this way:

Fig. 3 - Flower development.  The phenology of alfalfa flowering, starting with buds (bottom) through open flower, wilted flower, and seed pod development.  Bottom right is a raceme with all flowers removed; pedicels along the raceme indicate where a flower was attached.  Top right is a raceme with maturing curls. 

The model demonstrates that for a given rate of "buds becoming flowers", Open Flowers decrease as the number of foraging bees increase.  This is because the outflow, ("flowers being visited") increases as the number of foraging bees increase.   In the graph below, the curves represent open flowers over time, with numbers of introduced bees increasing from 1 (no bees), to 9 (35,000 bees).

However, Open Flowers increase as the inflow, rate of buds becoming flowers increases.  Thus increases in the inflow and the outflow have the opposite effect.

For example, Breazeale et al. use multiple linear regression models to compare open flowers and tripped flowers between a number of fields and between shelters within fields.  Thus, they measured the accumulation and it's outflow (two variables that are not independent).  They found more tripped flowers in fields with more open flowers.  Is this because fields with more tripped flowers had a more rapid rate of pollination, or a more rapid rate of bud opening?  We don't know without some measure of the inflow into the open flower accumulation.  However, the model suggests that these fields probably differed more in rates of bud opening than in pollination rates, because increased pollination rates should have reduced the number of open flowers for a given rate of bud opening.  Rates of bud opening probably differ considerably between alfalfa varieties and between fields under different management.

In contrast, tripped flowers decreased with distance of the flowers from a bee shelter.  This difference is likely due to decreased pollination further from the shelter.  Within a field the rates of bud opening probably are not as variable as between fields, but pollinators may be slower to reach areas further from the shelter.  A measure of the inflow into the open flower accumulation is needed to confirm this interpretation.

One way to measure the rate of bud opening would be to make counts throughout the day after similar periods of time, say every 2 or 3 hours.  Count and mark the flowers on a raceme that opened, and flowers that were tripped, in each time period .   Or perhaps you could prepare racemes before a measurement period so that all except buds are removed.  Then wait 2 hours, and count the number of open, tripped and wilted flowers. Temperature, time of day, and variety are likely to affect the rate of flower opening.

If however, you count open flowers at the end of the day, you are counting flowers at a time when no more flowers are opening, and no more pollination is taking place.  You are seeing what is left over after a of a day of pollination.  Hence, counts made at the end of the day tell you something about how well the pollination has been going that day.  (But they don't quantify the amount of pollination or the amount of bloom that day).  Don't expect to see tripped flowers once the bees have stopped foraging.  Flowers that are still open will be available for the bees first thing in the morning.  If pollination is heavy relative to the amount of bloom, only buds and wilted flowers will be left on the raceme. 

I hope that this helps alfalfa seed growers and researchers understand why sampling in the evening works. It should also help researchers of alfalfa pollination design experiments.

How many bees should I introduce to my field? A rule of thumb.

Alfalfa is an indeterminate bloomer, like many crops.  That means that it continues to bloom as long as conditions are appropriate and the fruits do not mature.  Read about the implications of indeterminate blooming in my essay Squash, Beans, and Deadheading Flowers. 

¹ Jones, M. L. and W. R. Woodson.  1997.  Pollination-induced ethylene in carnation.  Plant Physiol.  115:205-212. 

² De Martinis, D., G. Cotti, S. te Lintel Hekker,  F.J.M. Harren, C. Mariani. 2002.  Ethylene response to pollen tube growth in Nicotiana tabacum flowers.  Planta 214: 806–812.