Seed Dispersal via Caching – The Story of Antelope Bitterbrush

Generally speaking, individual plants produce an enormous amount of seeds. This may seem like a huge waste of resources, but the reality is that while each seed has the potential to grow into an adult plant that will one day produce seeds of its own, relatively few may achieve this. Some seeds will be eaten before they get a chance to germinate. Others germinate and soon die from lack of water, disease, or herbivory. Those that make it past the seedling stage continue to face similar pressures. Reaching adulthood, then, is a remarkable achievement.

Antelope bitterbrush is a shrub that produces hundreds of seeds per individual. Each seed is about the size of an apple seed. Some seeds may be eaten right away. Others fall to the ground and are ignored. But a large number are collected by rodents and either stored in burrows (larder hoarding) or in shallow depressions in the soil (scatter hoarding). It is through caching that antelope bitterbrush seeds are best dispersed. When rodents fail to return to caches during the winter, the seeds are free to sprout in the spring. Some of the seedlings will dry out and others will be eaten, but a few will survive, making the effort to produce all those seeds worth it in the end.

Fruits forming on antelope bitterbrush (Purshia tridentata)

Antelope bitterbrush (Purshia tridentata) is in the rose family and is often simply referred to as bitterbrush. It occurs in grasslands, shrub steppes, and dry woodlands throughout large sections of western North America. It is a deciduous shrub that generally reaches between three and nine feet tall but can grow up to twelve feet. It has wedge-shaped leaves that are green on top, grayish on bottom, and three-lobed. Flowers are yellow, strongly fragrant, and similar in appearance to others in the rose family. Flowering occurs mid-spring to early summer. Fruits are achenes – single seeds surrounded by papery or leathery coverings. The covering must rot away or be removed by animals before the seed can germinate.

Bitterbrush is an important species for wildlife. It is browsed by mule deer, pronghorn antelope, bighorn sheep, and other ungulates, including livestock. It provides cover for birds, rodents, reptiles, and ungulates. Its seeds are collected by harvester ants and rodents, its foliage is consumed by tent caterpillars and other insects, and its flowers are visited by a suite of pollinators. For all that it offers to the animal kingdom, it also relies on it for pollination and seed dispersal. The flowers of bitterbrush are self-incompatible, and if it wasn’t for ants and rodents, the heavy seeds – left to rely on wind and gravity – would have trouble getting any further than just a few feet from the parent plant.

Antelope bitterbrush (Purshia tridentata) in full bloom – photo credit: wikimedia commons

In a study published in The American Naturalist (February 1993), Stephen Vander Wall reported that yellow pine chipmunks were the primary dispersal agents of bitterbrush seeds in his Sierra Nevada study area. The optimal depth for seedling establishment was between 10-30 millimeters. Seeds that are cached too near the surface risk being pushed out of the ground during freeze and thaw cycles where they can desiccate upon germination. Cached bitterbrush seeds benefit when there are several seeds per cache because, as Vander Wall notes, “clumps of seedlings are better able to push through the soil and can establish from greater depths than single seedlings.”

Another study by Vander Wall, published in Ecology (October 1994), reiterated the importance of seed caching by yellow pine chipmunks in the establishment of bitterbrush seedlings. Seed caches, which consisted of anywhere from two to over a hundred seeds, were located as far as 25 meters from the parent plant. Cached seeds are occasionally moved to another location, but Vander Wall found that even these secondary caches produce seedlings. Of course, not all of the seedlings that sprout grow to maturity. Vander Wall states, “attrition over the years gradually reduces the number of seedlings within clumps.” Yet, more than half of the mature shrubs he observed in his study consisted of two or more individuals, leading him to conclude that “they arose from rodent caches.”

A study published in the Journal of Range Management (January 1996) looked at the herbivory of bitterbrush seedlings by rodents. In the introduction the authors discuss how “rodents [may] not only benefit from antelope bitterbrush seed caches as a future seed source, but also benefit from the sprouting of their caches as they return to graze the cotyledons of germinating seeds.”  In this study, Ord’s kangaroo rats, deer mice, and Great Basin pocket mice were all observed consuming bitterbrush seedlings, preferring them even when millet was offered as an alternative. The two species of mice also dug up seedlings, possibly searching for ungerminated seeds. Despite seed dispersal via caching, an overabundance of rodents can result in few bitterbrush seedlings reaching maturity.

A cluster of antelope bitterbrush seedlings that has been browsed. “Succulent, young seedlings are thought to be important in the diets of rodents during early spring because of the nutrients and water they contain.” — Vander Wall (1994)


Photos of antelope bitterbrush seedling clusters were taken at Idaho Botanical Garden, where numerous clusters are presently on display along the pathways of the native plant gardens and the adjoining natural areas. 

Harvester Ants – Seed Predators and Seed Dispersers

“The abundance of ants is legendary. A worker is less than one-millionth the size of a human being, yet ants taken collectively rival people as dominant organisms on the land. …  When combined, all ants in the world taken together weigh about as much as all human beings.” – Journey to the Ants by Bert Hölldobler and Edward O. Wilson

Considering how abundant and widely distributed ants are, it is easy to imagine the profound role they might play in the ecosystems of which they are a part. In fact, in the epilogue to Hölldobler and Wilson’s popular book about ants (quoted above), they conclude that in a world without ants, “species extinction would increase even more over the present rate, and the land ecosystems would shrivel more rapidly as the considerable services provided by these insects were pulled away.” It is no doubt then that ants, through their myriad interactions with their surroundings, are key players in terrestrial ecosystems.

photo credit:

photo credit:

Harvester ants offer a prime example of the important roles that ants can play. In the process of collecting seeds for consumption, harvester ants can help shape the abundance and distribution of the plants in their immediate environment. They do this by selecting the types and amounts of seeds they collect, by abandoning seeds along their collection routes, and by leaving viable seeds to germinate in and around their nests. Hölldobler and Wilson have this to say about harvester ants:

[The] numerical success [of ants] has allowed them to alter not just their nest environments, but the entire habitats in which they live. Harvesting ants, species that regularly include seeds in their diet, have an especially high impact. They consume a large percentage of the seeds produced by plants of many kinds in nearly all terrestrial habitats, from dense tropical forests to deserts. Their influence is not wholly negative. The mistakes they make by losing seeds along the way also disperse plants and compensate at least in part for the damage caused by their predation.

There are more than 150 species of harvester ants, spanning at least 18 genera. They are found throughout the world (except extreme cold locales) and are particularly common in arid to semi-arid environments. Pogonomyrmex is one the largest genera of harvester ants with nearly 70 species occurring throughout North, Central, and South America. Messor is another large genus of harvester ants that mainly occurs in Europe, Asia, and Africa. Both of these genera build large nests and move massive amounts of soil in the process.

Seed dispersal by harvester ants (also known as diszoochory) is a type of secondary (or Phase II) seed dispersal. It is a case of serendipity, as the dispersal occurs largely by accident. Some plants, on the other hand, have developed a mutualistic relationship with ants, enlisting them to disperse their seeds by way of an elaiosome – a fleshy, nutritious structure attached to seeds that attracts ants. Seeds with such structures are picked up by ants and brought to their nests where the elaiosome is consumed and the seed is left to germinate. This form of ant-mediated dispersal is called myrmecochory and is typically not carried out by harvester ants.

photo credit: wikimedia commons

photo credit: wikimedia commons

Harvester ant colonies have both direct and indirect influences on their surrounding environments; however, there is a dearth of research elucidating the exact details of such influences. A paper published in the Annual Review of Ecology and Systematics in 2000 by MacMahon et. al. reviewed available studies concerning harvester ants and explored our current understanding of the influences that harvester ants (particularly those in the genus Pogonomyrmex) can potentially have on community structure and ecosystem functions. Following are some of the direct influences the authors listed:

  • Removal and consumption of seeds and other materials – The relative abundance of plant species can be affected by the selective removal of seeds. Harvester ants also collect leaves, twigs, pollen, flowers, vertebrate feces, and arthropod body parts.
  • Storage and rejection of seeds – Collected seeds can be dropped during transport, rejected after arriving at the nest, or abandoned in nest granaries. All result in the transport of seeds away from the parent plant and dispersal beyond the plants’ primary dispersal mechanisms.
  • Construction and maintenance of nests – All vegetation and debris is removed from the area immediately surrounding the nest including mature and emerging plants. This area is kept clear for the duration of the life of the colony and, in some cases, can be quite extensive.

Harvester ants can also influence soil properties and soil food webs within and in the vicinity of their nests. They bring large amounts of organic matter down into the soil and redistribute vast amounts of soil particles. Their actions also influence the amount of moisture in the soil surrounding their nests.

This is a mere distillation of the influences that harvester ants might have; see the paper by MacMahon et al. to learn more.

In an effort to better understand how the seed predation and seed dispersal behaviors of harvester ants might influence plant population dynamics, a research team in Spain used data obtained from field research to build a computer model that would predict changes over time. The study site was described as “open and heterogeneous shrubland” and the vegetation was stated to be in “a very early stage in the secondary succession” after being subject to “recurring fires.” The harvester ant colonies involved in the study consisted of three species in the genus Messor. The plant species selected for the study were three native shrubs whose seeds were known to be collected by the harvester ants. Each plant species differed slightly in the amount and size of seeds it produced and in its primary seed dispersal mechanism, which is important because the researchers hypothesized that “the effect of seed predation and seed dispersal may depend on plant attributes.”

Messor bouvieri (photo credit:

Messor bouvieri (photo credit:

Data obtained from simulated scenarios and field observations appeared to support this hypothesis; each shrub species interacted differently with the harvester ants. Coronilla minima benefited from “accidental” seed dispersal. Comparatively, it produces a high amount of large seeds, which are primarily dispersed by gravity. Despite predation, ant-mediated dispersal was an advantage. Dorycnium pentaphyllum produced the highest amount of seeds among the three shrub species; however, seed predation was found to have negative effects on its population dynamics. Its primary seed dispersal mechanism involves ballistics (the mechanical ejection of its seeds), so ant-mediated dispersal may not offer an advantage. Finally, Fumana ericoides, despite its limited primary seed dispersal and its comparatively low production of seeds was not affected by the actions of the harvester ants. The authors concluded that “some unknown factor is driving the population dynamics of this species, more than the action of ants.”

Studies such as this, while leaving many unanswered questions, help us understand the important role that harvester ants play in our world. Harvester ants, and ants in general, are truly among Earth’s most enthralling and influential creatures. Learn more about their complex behaviors and countless interactions with flora and fauna by checking out these three documentaries recommended by ANTfinity.