Mustard musters its troops

North American forests are being invaded. The invading forces use chemical warfare to attack the native inhabitants and to repel counterattacks by hostile enemies. As it turns out, the invader is the humble garlic mustard, Alliaria petiolata, which releases toxic chemical compounds into the soil that reduce the growth rate of many native plant species, and has strong chemical defenses that makes it unpalatable to most herbivores.

garlic mustard field

Garlic mustard invasion. Credit Pam Henderson

Lauren Smith-Ramesh wondered why garlic mustard was not even more successful as an invader. Its chemical arsenal should allow it to overrun an area, but she (and many other researchers before her) observed that garlic mustard invasions often decline after a while. As part of her investigations into garlic mustard’s use of chemicals to inhibit native plants, Smith-Ramesh collected seeds from plants from different populations. While shaking these seeds into bags, she noticed that web-building spiders often colonized the garlic mustard’s seed-bearing structure (silique). Were these spiders somehow behind the garlic mustard’s surprising lack-of-success?


Garlic musard silique with web. Credit Lauren Smith-Ramesh

Spiders can benefit plants in several ways. As important predators in food webs, spiders can kill large numbers of herbivorous insects that might otherwise attack a plant. In addition the decaying corpses of their insect prey can add vital nutrients to soils. Garlic mustard does not enjoy these potential spider-associated benefits, because spiders colonize the garlic mustard after it has already gone into decline, and also because garlic mustard is already well-protected (chemically) against herbivorous insects.


About 60% of the spiders were this species – Theriodiosoma gemmosum. Credit Tom Murray.

Smith-Ramesh first wanted to understand the relationship between seed structures (siliques) and spider abundance. She established three different types of plots that measured 2 X 2 meters: (1) S+, which had mustards with intact siliques, (2) S-, which had mustards with siliques removed, and (3) N, which had no garlic mustard plants at all in 2015. After several months, she collected all spiders from the middle square meter of each plot. Plots with garlic mustard with intact siliques (S+) had, by far, the highest spider density. S- plots had a somewhat higher spider density than N plots, which Smith-Ramesh attributes to spiders wandering in from just outside the S- plots (which tended to have more silique-bearing garlic mustard plants nearby than did the N plots). Based on this experiment Smith-Ramesh concluded that garlic mustard siliques were dramatically increasing spider density.

SR2b use

But did increased spider density in S+ plots reduce the number of herbivorous insects, thereby benefiting nearby native plants? Smith-Ramesh set up insect traps that collected insects over two 48 hour time periods – once in August and again in September – in each of the S+, S- and N plots. Both surveys showed fewest herbivorous insects in the S+ plots. This supports Smith-Ramesh’s hypothesis that native plants are benefitting from higher spider density associated with garlic mustard siliques.

SR 2c use

Next, Smith-Ramesh wanted to know whether the decrease in herbivorous insects benefitted native plant growth. To test this directly, she transplanted three types of native plants into her S+, S- and N plots. One of the species, the Hairy Wood Mint Blephilia hirsuta, enjoyed a 50% biomass boost in S+ plots compared to S- plots. The other two native plants species showed very little effect.


Smith-Ramesh collecting data with three siliques in the foreground. Credit: Lauren Smith-Ramesh.

Garlic mustard plants with intact siliques also benefitted the soils by increasing the amount of available phosphorus by approximately 60%. This phosphorus may have originated with insect carcasses that made their way into the soil and released their nutrients. In theory, soils with higher phosphorus availability could help support the growth of native plants. Smith-Ramesh plans to explore other plant communities that are suffering from different invasive plants, to see whether these invaders are also inadvertently providing resources or conditions that may undermine the success of their invasion.

note: the paper that describes this research is from the journal Ecology. The reference is Smith‐Ramesh, L. M. (2017). Invasive plant alters community and ecosystem dynamics by promoting native predators. Ecology98(3), 751-761. Thanks to the Ecological Society of America for allowing me to use figures from the paper. Copyright © 2017 by the Ecological Society of America. All rights reserved.

Cottonwood genes and spider hostels

Back in my working days, people that I met would sometimes ask me what I did for my research. I usually told them that I studied spider sex, which, while true, was a bit misleading, as my interests were actually slightly broader. But studying spider sex was a good fit for my disposition, because, more than anybody I know, I can stare at something for a very long time and not get bored. And when spiders have sex, there can be very long periods, when, to our eyes, nothing is going on. As it turns out there is a great deal of pheromonal communication going on, and considerable vibrational activity as well, but it is not that easy for humans (even abnormally patient ones) to detect.


A spider and her three egg cases within a web she has built in a Taphrina blister. Credit Matt Barbour

The point is that I have a very soft spot in my heart for spiders, and was delighted to learn about an indirect process that provided a comfortable home for needy spiders. Heather Slinn got interested in her project while an undergraduate summer intern. Her colleague, Matt Barbour, pointed out that when he flipped over blistered black cottonwood leaves (Populus trichocarpa) he often found a spider hanging out in there. This observation led to her to study the relationship between cottonwood trees, cottonwood genetics, pathogens that make leaf blisters and spider occupancy rates.

Taphrina fungi form cup-like blisters in the leaves of Populous trees. But these trees vary in how susceptible they are to leaf blisters. The researchers wanted to answer three questions about this relationship. First, do spiders prefer to live in leaf-blisters as opposed to unblistered leaves? Second, are differences in tree susceptibility to Taphrina a result of genetic differences between the trees? And third can differences in Taphrina-resistance account for differences in spider density?

One of the keys to the experiment was establishing a garden with distinct clones of trees of known genetic makeup (genotypes). Slinn and her colleagues studied five different genotypes, with approximately 40 trees per genotype. They dutifully watered them throughout the summer, and then sampled up to 30 leaves from each tree for blister density, blister size, and spider residency, using a modified shop vac to suck up all of the spiders.


Black Cottonwood garden. Credit Matt Barbour

The researchers discovered that blistered leaves were 35 times more likely to have a spider and/or spider web than were unblistered leaves. Clearly spiders found blistered leaves to be highly attractive homes.

But there were pronounced differences among the five genotypes in blister density, blister size, spider density and the probability that a spider was occupying a blister. Graph A shows that genotypes 1 and 3 (G1 and G3) had the lowest mean density of blisters (about 2 or 3 per meter of plant), while G4 averaged about 20 blisters per meter. Although G3 had relatively few blisters, it did boast the largest blisters (see graph B). The researchers concluded that blister density and size were both under genetic control, but not linked to each other.


Mean (A) blister density (number per meter of plant) and (B) blister size, for the five tree genotypes.

But how did blister density and size influence spider residency? G4, the genotype with the greatest number of blisters per plant, also had the greatest number of spiders (graph C). But on a per blister basis, we can see that the two genotypes with the largest blisters (G2 and G3 – see graph B) also had the highest probability of housing a spider in their blisters (graph D). So when spiders make decisions about where to live, both size and number are important.


Mean (C) spider density and (D) probability a blister hosts a spider, for the five tree genotypes.

It is unclear why the spiders are attracted to blisters.  Some spider species have been shown to be attracted to structural complexity, because that provides more or better attachment points for web strands and egg cases.  Depressed blisters may also give protection from abiotic factors such as wind and precipitation.

Like most good studies, this research raises a host of related questions. Why is there so much genetic variation in Taphrina-resistance within this tree species? Slinn suggests that there may be tradeoff whereby investment into Taphrina-resistance might compromise a plant’s ability to invest in other functions such as cold-resistance, or rapid growth and/or high reproductive rates. A second question is how does spider presence influence other species – for example does hosting a spider reduce the number of herbivorous insects that might attack the tree? A third issue raised by the authors is that plants infected with Taphrina may be weaker and more susceptible to herbivores. In that case, the spiders may be present in blisters because they are attracted to the herbivores that are eating the leaves. And finally, herbivores eating the leaves may transmit other diseases affecting our forests, such as dutch elm disease, chestnut blight and white pine blister rust. Unraveling this complex chain of events will keep researchers busy for many years.

note: the paper that describes this research is from the journal Ecology. The reference is Slinn, H. L., Barbour, M. A., Crawford, K. M., Rodriguez‐Cabal, M. A., & Crutsinger, G. M. (2016). Genetic variation in resistance to leaf fungus indirectly affects spider density. Ecology 98(3): 875-881. Thanks to the Ecological Society of America for allowing me to use figures from the paper. Copyright © 2017 by the Ecological Society of America. All rights reserved.