More than meets the eye – hidden variation affects how ants plant seeds of forest wildflowers

By Kirsten M. Prior and Carmela M. Buono

Kirsten’s and Carmela’s study, where they and their colleagues ask if functional variation partitions discretely between Aphaenogaster species or along a continuum in this species complex, can be found here.

Ants are high on the list of good seed-dispersing animals. Many seeds of plants capitalize on dispersal by these small yet effective dispersers by producing seeds with lipid-rich appendages called elaiosomes that are attractive to ants and nutritious for growing colonies. Approximately 11,000 plants possess elaiosomes (“myrmecochores”), with several described hotspots of myrmecochory across the globe. North American (NA) eastern deciduous forests are one of these hotspots, where 30-40% of understory plants have adaptations for dispersal by ants. Ant-dispersed plants are many of the beloved showy spring ephemerals in the forest understory, including Trilliums, bloodroot, wild ginger, and violets.

We used seeds of four common ant-dispersed plant in NA eastern forests in our study: Anemone acutiloba (Sharp-lobed hepatica); Sanguinaria canadensis (Bloodroot); Asaurm canadense (Wild ginger); Trillium grandiflorum (White trillium). Seeds with eliaosomes are shown. Photo credits: K. Prior and C. Buono

Some ants are better at dispersing myrmecochorous seeds than others. Good seed-dispersing ants are attracted to the lipid-rich appendages on seeds. They grab the appendages, carry seeds to their nest, remove the nutritious appendage and feed it to the growing brood. They then deposit the intact seed in a waste midden – often a location conducive to germination. Some ants are poor-seed dispersers in that they interact with but don’t move seeds, rob elaiosomes (remove them but not disperse the seeds), or even eat whole seeds or damage them. In hotspots of myrmecochory, there are usually only one to a handful of good seed-dispersing ants, often called “keystone dispersers.” The common woodland ant, Aphaenogaster sp., is the keystone disperser of understory myrmecochores in NA eastern deciduous forests, as they are responsible for most dispersal.

However, there is not just only one species of the keystone disperser, Aphaenogaster, in NA forests – and as it turns out, it’s a bit complicated! There are multiple described species of Aphaenogaster in the eastern US that interact with seeds – including A. fulva, A. rudis, and A. picea – that we refer to as the Aphaenogaster seed disperser complex (ASDC). A. fulva is distinguishable from the other ASDC taxa based on both consistent differences in diagnostic traits and forming a discrete genetic clade. On the other hand, the relationship between A. rudis and A. picea is more nuanced and uncertain. Overlapping and inconsistent patterns from sequence data suggest that they may not be fully resolved species at all due to incomplete divergence of ongoing hybridization. Unsurprisingly, even with a trained eye, these two named species are challenging to delineate and have overlapping characteristics, especially where they co-occur.

We collected six populations of ants in the ASDC for our study. Here they are pictured in order of populations that are more A. picea like to more A. rudis like based on variation in thorax shape, including the length and direction of the propodeal spine.

Our research team is interested in how the identity of mutualist partners (i.e., seed dispersers) affects the outcomes of ant dispersed–plant communities. How do seemingly minor differences in the behavior of ants, such as how many seeds they move and which seeds they prefer, scale up to affect plant communities? For the ADDC in NA eastern deciduous forests, we predicted that there might not only be differences among named species but also intraspecific variation along a gradient within the ASDC that coincides with population-level differentiation.

In our new paper, “Uncovering how behavioral variation underlying mutualist partner quality is partitioned within a species complex of keystone seed-dispersing ants,” Our team from Binghamton University asked if behaviors relating to seed dispersal differed discretely or continuously along a gradient between named species A. rudis and A. picea. Graduate and undergraduate students from Kirsten Prior’s ecology lab (priorecologylab.com) and Tom Powell’s (powellevolab.com) evolutionary-ecology lab teamed up for this study. First, Carmela Buono (Ph.D. candidate) and undergraduate mentee Will Smisko (Undergraduate Summer Scholar Fellow) collected multiple colonies of six populations of putative A. rudis and A. picea. In arenas with myrmecochore seeds, they ran a behavior experiment to measure foraging rates, seed removal rates, and seed preferences. Next, Carmela and undergraduate mentee Allie Radin came up with the idea to test aggression within populations, among populations (within species), and among species – as our previous work shows that aggression can affect seed dispersal behavior. 

Gabriella Quartuccia (Ph.D. candidate) and undergraduate mentee Andrew Lupinksi (Undergraduate Summer Scholar Fellow) developed an approach to quantify complex variation in the ASDC. Key diagnostic characteristics that delineate putative species are found in the thorax (for example, the length and direction of the propodeal spine). Gabby and Andrew created landmarks on the thorax to create 2D shapes and compared 2D shapes among colonies to uncover how they differed. This morphometric analysis delineated putative species, picking up known differences – such as A. rudis having shorter spines that point more upward. However, they also revealed significant colony-level (and population-level) variation, with some populations of what was initially described as A. rudis being more A. picea-like and vice versa. This approach quantified what was primarily descriptive before – that there are intermediate ants in this species complex!

Carmela Buono (Ph.D. student; first picture) led the behavior experiments with Will Smisko (Undergraduate student; third picture), and Allie Radin (Undergraduate; not pictured). Gabby Quartuccia (Ph.D. student; second picture) led the morphometric analysis with Andrew Lupinsky (Undergraduate student; not pictured). We collected colonies in ant boxes (shown in pictures). Photo credits: C. Buono, G. Quartuccia

What was exciting was when we compared colony morphometrics to colony behavior. We found differences in behavior between named species. However, we also found a relationship between colony morphometrics and behavior, such that colonies with intermediate morphotypes had intermediate behaviors!

This is an exciting finding with implications for uncovering variation in this critical ecosystem function. Our work shows that behavioral differences in ant partners are likely to affect plant communities – but not only between species but also among populations along a gradient in the ASDC. Uncovering how partner variation affects mutualisms is a critical question, yet few studies have considered intraspecific variation – despite its likely importance. Here we show that intraspecific variation is as significant (if not more) than interspecific variation, which should not be surprising in this system given that partners are in incomplete stages of speciation.

This study is an exciting starting point for understanding functional variation in this critical mutualism for our research teams. Gabby and Tom are performing population genomics in the ASDC, and the Prior lab, including Carmela and Ph.D. student Rosey Ines, are measuring variation in traits and setting up experiments to understand how functional differences in the ASDC scale up to affect plant communities. 

Collecting ants in ant boxes. Video Credit: C. Buono

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s