Past Research Highlights

February 2005

Sonoran Desert Annuals on Tumamoc Hill

Larry Venable, University of Arizona, Department of Ecology and Evolutionary Biology

About half of the plant species found on the Desert Laboratory and also the Tucson Mountains in general are annual plants. Two thirds of these are "winter annuals" and 1/3 "summer annuals". By definition, annual plants only get one chance to reproduce (they germinate and die within a span of less than a year). That puts them in a precarious situation in deserts with low and unpredictable precipitation. Sometimes there just isn't enough rain for successful growth and reproduction. Yet annual plants not only persist with their "one shot" reproductive strategy, but they do so in great diversity, as witnessed in periodic great spring blooms. On the surface it might seem that they are all doing about the same thing and thus should be more or less interchangeable --- Winter annuals all germinate sometime in the fall, usually in November and December, lie low in the cold months, grow rapidly in February and March if it rains, and then flower, set seeds and die in March and early April. So why are there so many species (120 at the Desert Laboratory alone!) and why don't some exclude the others over time?

The answer to both of these questions (how do annuals persist with their one shot reproductive strategy and how do so many annual species coexist?) is thought to involve their dormancy and germination strategies. Our lab is responsible for an ongoing 22 year so far project to document the ups and downs of the population dynamics of all of the winter annual species occurring along the northern base of the Desert Laboratory and on the creosote flats to the Northwest of Tumamoc Hill. In this project we have documented the fates of plants germinating each year (death/survival, how many seeds produced) and the fates of seeds in the soil (do they all germinate? if not, do they survive to subsequent years?). The solution to the persistence with a one-shot reproductive strategy is fractional germination. Our data show that not all seeds germinate in any given year, even under ideal germination season conditions. Thus the germination strategies seem to involve a sort of "bet hedging" much as wise investors in the 1990s might have held a fraction of their money in cash or bonds even though it seemed the market might rise forever. The answer to species coexistence seems to be partly that different species have different germination strategies. Different species don't always germinate to the same extent in different years. Rather something about the conditions in different years triggers more germination in some species but not others, only to have the pattern reversed in other years. This allows each species to have a somewhat separate "niche in time".

To better understand the basis of the fractional germination and differential germination patterns that we have observed over the years in the natural dynamics of the winter annuals at the Desert Laboratory we recently published a paper reporting the germination patterns of 8 of the common species of desert winter annuals under controlled laboratory conditions designed to simulate observed weather variation at the Desert Laboratory (Figure 1).

Figure 1. Seeds of species used in this study. From left to right, row one: Plantago patagonica (Plantaginaceae), Pectocarya recurvata (Boraginaceae), Eriastrum diffusum (Polemoneaceae); row two: Schismus barbatus (Poaceae), Evax multicaulis Stylocline micropoides (Asteraceae); row three: Pectocarya heterocarpa long seeds, Pectocarya heterocarpa winged seeds (Boraginaceae), and Eriophyllum lanosum (Asteraceae).

The summer conditions that winter annual seeds experience between the time they are produced in March and April and the onset of the germination season in the fall may affect the pattern of germination. If species respond differently to summer conditions, variation in summer weather may be an important contributor to species-specific germination responses. To explore this we created dry/hot, wet/hot, and dry/cool summer storage conditions by placing seeds outdoors on the ground (summer soil temperatures regularly exceed 150° F) at the Desert Laboratory either exposed to summer monsoons (wet/hot), or shielded from rain by a transparent polyethylene roof (dry/hot) or by storing them on greenhouse benches with the thermostat set to 100° F (dry/cool).

The temperature and day length at the time of germination rains can vary among years and can affect which species germinate more in a given year. For example, some species readily germinate in the early fall while others preferentially germinate in late winter. However, it is impossible to tell from field data if it is the temperature/day length combination at the time of rain that affects germination or if it is the result of the actual passage of time. If the answer differs for different desert annual species, this provides a further mechanism for species to germinate differently in different years. To distinguish these possibilities we conducted germination experiments at 5 different times bracketing the germination season (from August to February). At each time we exposed seed in growth chambers to conditions that simulated averaged temperature and day length field conditions for October (29° C day, 14° C night, 11.5 hours of daylight), November (22° C day, 7° C night, 10.5 hours of daylight) and December (18° C day, 3° C night, 10.25 hours of daylight). So seeds of each of these 8 winter annual species were exposed to all 45 combinations of the three summer treatments, five experiment dates and three sets of simulated conditions.

Large numbers of viable seeds remained dormant under all conditions explored (usually > 50% sometimes > 90%). This confirms the idea that a fraction of desert annual seeds remain dormant in persistent seed banks, even under optimal germination conditions. This then is indeed part of the strategy that lets desert winter annuals persist with their one-shot reproductive strategy in arid lands.

Species differed in their germination responses in many ways. There was a range of variation in the average germination of the 8 species (Table 1). But also they differed in how they respond to the experimental variables and in how the experimental variables interacted with each other. So for example, Eriophyllum lanosum had its highest germination in December conditions, while Pectocarya heterocarpa had highest germination in November conditions. Eriastrum diffusum, Evax mulitcaulis, Plantago patagonica, and Stylocline micropoides had higher germination in both November and December conditions, while Schismus barbatus germinated more in October and November conditions (Table 1 shows the conditions giving maximal germination for each species). When we compared these growth chamber germination patterns to the long-term field patterns of germination, we found reasonably close correspondence.

Table 1. Mean and range of percent germination averaged over all experimental conditions and over experimental conditions likely to be encountered in the field (see the paper for an explanation of this distinction). Ranges are ranges of treatment means. Also, the treatment combination resulting in maximum germination is given. When these are not conditions likely to be encountered in the field, conditions for maximum germination among realistic field conditions are also given in bold.

So this investigation has shed light on the germination biology of a guild of desert annuals. A considerable fraction of seeds do not germinate under favorable conditions either in the field or under controlled growth chamber conditions. Such seed behavior is consistent with predictions of bet hedging for desert annuals. Also, models have shown that independently variable germination fractions (i.e. have species by condition interactions for germination percent) are important for maintaining species diversity. We have shown that such interactions exist in the growth chamber responses of eight common members of a guild of desert annuals. More specifically, we were able to show that oversummering conditions, calendar date of germination, temperature and day length play a part in these differential responses. Furthermore, the growth chamber germination results seem to reflect to a reasonable degree what actually happens in the field. Many of the details that determine the germination biology of plants are important in determining population and community dynamic of plants. Thus, focused studies like the one here, conducted in the larger framework of long-term field studies, can yield a deeper understanding of plant population and community dynamics.

Adondakis, S. and D.L. Venable. 2004. Dormancy and germination in a guild of Sonoran Desert annuals. Ecology 85: 2582-2590.

For a downloadable pdf of the paper, visit Larry Venable's web page at http://eebweb.arizona.edu/faculty/venable/

For more information contact:
Larry Venable (venable@u.arizona.edu)

Return to Past Research Highlights Index

US Department of the Interior | US Geological Survey
URL:
Please direct feedback regarding this page to the webmaster.
Last modified: