Sticky plant attraction, a new paper

I could not possibly do as good a job as the summary of this paper written by Elizabeth Preston, here, so I’ll first tell the backstory – and when it comes out, I’ll detail another cool part of it (which is not in the preprint version – instead buried in the appendices).

Heliothis phloxiphaga eating a flower bud of Aquilegia eximia, the serpentine or sticky columbine. Lake County, CA. 

Last summer, I spent most of my time studying Trichostema laxum, which I’ve written about in a previous post. I was trying to test my hypothesis that external chemical defenses are easily washed away by rain (which may be driving the pattern that few are found in wet areas) and spent countless hours doing various pieces of this – observing pollinators (the volatiles washed off might affect pollination positively or negatively), counting insects, leaves, buds, flowers and seeds (which can be done in situ!) and by July, the drought and some jackrabbits made this experiment look rather grim. I still haven’t brought myself to analyze these data, as it ended so depressingly…

As the floral color polymorphism was not in the experimental population, I didn’t notice it until the end of the summer, and gathered some, but not enough, data on it. Lake County, CA. 

I was getting a little frustrated and I wandered around, naturalizing, which is always a remedy for frustration (to me at least). I came upon a columbine – Aquilegia eximia – which I instantly knew held some potential for cool experiments. The first thing I noticed was that it was extremely sticky and covered in dead insects and the second was that it had a bunch of predators on it. I immediately thought to Billy Krimmel and Ian Pearse’s cool paper on tarweeds (doi:10.1111/ele.12032) , in which they demonstrated that the dead insects provided food for predators, which protected the plant from herbivores. I figured initially that I would simply test this in another system.

A Hoplinus nymph (probably the most important predator in the system) approaches an
entrapped fly.

Because I was out in the field without access to a genetics lab to get dead flies, I couldn’t replicate their design – where they added dead fruit flies to plants to supplement carrion – so instead I removed all the carrion from half of my 50 plants, hypothesizing that I would get a decrease in predators and an increase in herbivory (which we did!). I also thought hard about what else to test to add to Billy and Ian’s work. I thought that, perhaps, it would be interesting to test whether the plants attracted the various entrapped insects (mostly small flies, wasps and beetles) somehow. Lots of plants attract insects – pollinators are the most obvious, but volatile signals attract predators, other herbivores and even birds (doi: 10.1111/ele.12177). Having petri dishes, plastic mesh and tanglefoot in my field kit – I made little sticky traps, with a sticky mesh top and a petri dish bottom and I put either columbine stems and leaves (a very small amount) or nothing in them. Collecting them 24 hours later, I found that the dishes with columbine had higher insects than the empty ones (which would demonstrate the ambient rate of insects landing on these traps). The trapped insects were also little flies, wasps and beetles, just like on the plants themselves.

Dead Hymenoptera on columbine (I may be giving up entomologist credentials, but I am not sure whether it is a wasp or an ant alate).

So this became a story – perhaps logically – that the plants were somehow attracting insects to kill and feed to the beneficial predators on their surfaces (retaining their services). I presented this work in a talk at a little student conference at Davis during recruitment weekend and played with several ways to frame the story. The first was to be rather dry – columbines attract insects and control a tritrophic defense (or something along those lines). Instead, I thought long and hard about trying to make a metaphor (socialism – a worker’s paradise for the predatory bugs or a Roman bread and circus type thing, but they didn’t really work) and while I don’t remember how I came up with this – I settled on the sirens, figures of classical mythology who lured sailors to their deaths. I therefore framed it as these poor insects – innocent sailors of the California air – are somehow drawn to their deaths on the columbines. Of course, the columbines put the insects to good use in their defense, leaving open the question – which I am sure classical mythologists lose much sleep over – what did the sirens do with their collection of dead sailors?

Serpentine columbines in flower. Lake County, CA. 

Read more here!

Eric F. LoPresti, Ian Seth Pearse, and Grace K. Charles In press. The siren song of a sticky plant: columbines provision mutualist arthropods by attracting and killing passerby insects. Ecology. http://dx.doi.org/10.1890/15-0342.1

Beginning research: floral polymorphisms in Trichostema laxum

The first steps of any research project are, for me, the most exciting. Therefore, I’ll write a quick post on something I’ve been spending a bit of time on. Last summer, I spent most of the summer trying to wash off chemical defenses on leaves. One plant I chose was Trichostema laxum – a mint endemic to California (it may occur in extreme southern Oregon, too) that occurs pretty commonly on dry serpentine streambeds at my field site in Lake/Napa counties. I’ve already written a quick post mentioning this plant, but I know a LOT more now!

Trichostema laxum, October 2014, McLaughlin Reserve, Napa County. Notice the position of the stigma in relation to the anthers – the style (the stigma’s tube) projects well beyond the anthers. This is the “normal” morph of the plant, referenced in the literature and seen in all herbarium specimens I’ve looked at so far.

While the herbivory and exudate stuff awaits analysis (a dissertation proposal will force me to do that soon!), I discovered the aforementioned flower color polymorphism and took a bunch of baseline data on it, which may be important in the coming months. The flower color polymorphism interests me most because of one population which had a high (~3%) proportion of the white/purple morph – no others had it. Was it just a random neutral mutation that didn’t drift out? If not – how is it maintained?

The four polymorphs of flower color. All from summer 2015, McLaughlin Reserve, Napa/Lake Counties.

The first question was, do the various flower colors differ in fitness from the normal (purple) morph? Trichostema laxum along with most other annual plants in California grasslands and serpentine barrens, is extremely variable in size depending on the microclimatic conditions. Within a population, some individuals can have three orders of magnitude more flowers than others (~10 to ~10,000). I found one small bush-sized individual (probably nearly a meter square) on a gopher mound – clearly the gopher had changed the nutrients or hydrology of that specific location favorably! Therefore, I compared polymorphs to their nearest neighbor of the normal morph, in an attempt to minimize this variability. This was a coarse test (without a huge amount of power), but I found no differences, though large variability among individuals. I will – hopefully – be able to confirm this in the laboratory rather easily. I took a good amount of pollinator data – which also awaits analysis.

A pink morph just barely open (though the stigma is open, so maybe its deformed?).  McLaughlin Reserve, Lake County, CA. 

The next logical step in the investigation was to grow plants in the lab and find out whether the color polymorphisms were heritable – an important consideration in any investigation relating to population-level polymorphisms. Trichostema have a reputation for being a tough genus to grow, in fact, a professor at Davis told me a former grad student planned a project on them, but couldn’t get any germination. I’ve been more fortunate (with help from Danny Barney at the USDA) and got decent germination with a rather simple protocol – laxum may be less picky than its relatives. I grew them all fall – they flowered in November and early December.

One of the first individuals in the lab. Isn’t it cute?

When I started looking closely at the plants in the lab, I found two more polymorphisms. The first was the lower lip patterning. In normal plants, the lower lip – and sometimes the next lowest two petals, have some purple splotches on them. This is likely a nectar guide, leading pollinators to the reward (and often only visible/really cool in the UV). I knew in the field that the completely white morph lacked a nectar guide as it lacks anthocyanin, the red/purple pigment in most plants, completely, so a purple nectar guide would be precluded. But I was surprised to see a purple flowered plant lacking it.

Clean purple lower lip. December 2014, in lab. 

While interesting, this was only found in one plant (though I have seeds of it now). Another polymorphism was also obvious in the captive plants and it solved one of my summer mysteries. During the summer, I wanted to do crosses with the various colored flowers. I didn’t think it would be that hard – an older paper reported that T. laxum was non-selfing and covered plants produced no seed. So I placed pollinator exclosures over a bunch of plants and did crosses by moving pollen from one plant to another. I then covered the plants again, letting them naturally set seed and figuring that the only seed I’d get would be that of the crosses. I pollinated ~10 flowers per plant and since mints have only 4 ovaries per flower, I figured I could get about 40 seeds a plant (probably 30 since my fine motor skills aren’t all that great). When I uncaged the plants and collected the seeds in October, I got quite a surprise – large numbers of seeds. Though not a full complement from any plant (there are MANY reasons for this besides lack of pollen), I got way too many seeds to have been either 1) my pollination, or 2) occasional lapses in the pollinator exclosures. Clearly the plants were self-pollinating somehow.

And here is the solution to the mystery! Where is the stigma? Its pretty much in the middle of the anthers. This one isn’t quite mature yet, but instead of opening after growing far past the anthers (see the first picture), it will open either right in the anthers or ever so slightly beyond. McLaughlin Reserve, Lake County, CA.

Looking closely at the individuals in the lab revealed the reason for this mystery. Some plants, like the first picture in this post, had long styles, which projected the stigma far past the anthers. Others, like the one above, had short styles and the stigma was amidst, or ever so slightly past the anthers. This proximity (I think) allows the plant to self pollinating either directly, or with the slightest bit of wind or insect movement (the “self-pollinating” morph. In the lab, more than half the plants developed into the self-pollinating morph, and while I hadn’t noted it during the season, I was able to go back to the hundreds of pictures I took and found pictures of it in the field. Strangely, they are not in the same proportion – my pictures are primarily of the “normal” variety – which accords with the one paper on the plant, as well as descriptions. I then examined the specimens in the herbarium, all of which were the “normal” morph (and purple, with patterned lower lips). Whether the lab creates the right environment for this morph to develop (whether there is a genetic propensity for it, or it is somewhat environmentally-driven) is unknown now, but I am working on it. Jenny Van Wyk – another grad student at Davis and extremely knowledgeable plant reproductive biologist – have quantified the differences between these morphs and found some really interesting correlates.

Flowers of the two morphs (self-pollinating, top; normal, below), at the same scale (lower lip broken in lower photo). 

Preliminarily – and our sample size is low as of now – the self-pollinating morph has larger flowers (corolla length, display height, style length), produces more pollen (300x more!) and has more, but more dilute, nectar. There is variability within morph, but so far, each plant has fit into one of the two morphs easily. How much this is an artifact of the laboratory setting is unclear, but photos of the self-pollinating morph and the pollinator-excluded plants producing seed point to something interesting happening in the field. Right now we are focused on the laboratory aspect, but we are considering experiments and observational data to be performed/gathered this upcoming year. We’d love to hear what anyone thinks of the system and interesting questions we can ask with it!

I’ll have some photos and interesting observations from my three-week trip to Chile soon, too. Lots of interesting botany, entomology and birdwatching (condors!).