Predation drives divergence in sticklebacks

A new study in Evol­u­tion Let­ters provides exper­i­ment­al evid­ence that pred­a­tion drives changes in threespine stickle­back armor traits. Here, author Dr Diana Ren­nison describes her exper­i­ment and its inter­est­ing results.

For quite some time it has been known that dif­fer­ences in eco­lo­gic­al factors between hab­it­ats drive evol­u­tion­ary diver­gence and can lead to the gen­er­a­tion of new spe­cies. How­ever, most of the time we do not know the sources of nat­ur­al selec­tion that drive the evol­u­tion of spe­cif­ic adapt­ive traits. It is also often the case that trait changes are driv­en by mul­tiple eco­lo­gic­al factors and inter­ac­tions between them, which makes it dif­fi­cult to determ­ine the indi­vidu­al con­tri­bu­tions of each factor. As a res­ult, it remains unclear how spe­cies inter­ac­tions, such as pred­a­tion or resource com­pet­i­tion, con­trib­ute to spe­cies diversity. To invest­ig­ate wheth­er dif­fer­en­tial pred­a­tion could be an import­ant factor that drives diver­si­fic­a­tion of traits and their under­ly­ing genes we exper­i­ment­ally manip­u­lated the pres­ence of a nat­ive pred­at­or of threespine stickleback.

Threespine stickle­back (Gas­ter­o­steus aculeatus) are a small fish dis­trib­uted through­out most coastal regions of the north­ern hemi­sphere. Stickle­back can be found in mar­ine hab­it­ats, where they have exis­ted for mil­lions of years. They can also be found in fresh­wa­ter hab­it­ats, which they col­on­ized 10 – 12,000 years ago after the last ice age. With­in fresh­wa­ter, stickle­back have adap­ted to thou­sands of lake and stream hab­it­ats. Inter­est­ingly, with­in five lakes in Brit­ish Columbia Canada two dis­tinct spe­cies of threespine stickle­back have evolved and coex­ist in adja­cent hab­it­ats. In these lakes there is a benthic spe­cies that is lar­ger, deep bod­ied, and feeds in the near shore envir­on­ment on small insect lar­vae, snails and clams. There is also a smal­ler and more stream­lined lim­net­ic spe­cies, that inhab­its the open water, where it feeds on zooplank­ton. Pre­vi­ous work has shown that com­pet­i­tion for resources between the two spe­cies likely drove them to occupy the two dif­fer­ent hab­it­ats. Once in their new hab­it­ats, benthics and lim­net­ics encountered dis­tinct pred­at­ors. In the near shore, insect pred­at­ors are more com­mon, where­as in the open water ver­teb­rate pred­at­ors (birds and cut­throat trout) are more common.

Benthic and lim­net­ic stickle­back dif­fer in many beha­vi­our­al and mor­pho­lo­gic­al traits. How­ever, one strik­ing dif­fer­ence is in the amount of bony armor that they pos­sess. Lim­net­ics have two long dorsal spines along their back and two spines pro­trud­ing from their pel­vis, they also have 5–9 bony plates along each side of their flank. In con­trast most benthic fish have lost their first dorsal and pel­vic spines, and often have only 0 – 4 bony plates. This armor is thought to pro­tect against pred­a­tion by toothed or gape lim­ited pred­at­ors, such as pred­at­ory fish (e.g. trout) or birds. Thus, the dif­fer­en­tial pred­a­tion they exper­i­ence is hypo­thes­ized to have driv­en the diver­gence of these armor traits. Pre­vi­ous work in threespine stickle­back had iden­ti­fied the genes respons­ible for vari­ation in the spine traits. A muta­tion at the Pitx1 locus con­fers loss of the pel­vic spines and girdle and reduced dorsal spine length res­ults from changes in the Msx2a gene.

rennison_females.jpg
Female lim­net­ic (bot­tom) and benthic (top) threespine stickleback.
rennisonfig2.png
Dif­fer­ences in the bony armor of benthic and lim­net­ic stickle­back. The fish have been stained with the chem­ic­al aliz­ar­in red, which binds to cal­ci­um to high­light bone. On the image the let­ter A indic­ates the first dorsal spine and B indic­ates the pel­vic spine and girdle, which are traits present in lim­net­ic fish and most often absent in benthic fish.

To test the idea that pred­a­tion is the source of selec­tion driv­ing changes in the armor of benthic and lim­net­ic stickle­back, we con­duc­ted a manip­u­lat­ive exper­i­ment. In order to isol­ate and estim­ate the selec­tion due to cut­throat we con­duc­ted the exper­i­ment in arti­fi­cial ponds that did not nat­ur­ally con­tain any fish spe­cies. We used hybrid benthic-lim­net­ic stickle­back as the tar­get of selec­tion and manip­u­lated the pres­ence or absence of cut­throat trout. The reas­on for using hybrid stickle­back was to increase the amount of vari­ation avail­able for selec­tion to act upon, which would make it easi­er for us to detect an effect. Some hybrid fish had very benthic armor, some had very lim­net­ic armor and some had inter­me­di­ate trait values.

In the spring of 2012, we intro­duced the first gen­er­a­tion of hybrid fish (F1s) into the ponds. They imme­di­ately star­ted to breed and cre­ated the second gen­er­a­tion of hybrids (F2s), our tar­gets of selec­tion. At this point we took an ini­tial sample of these second-gen­er­a­tion fish, to determ­ine the start­ing fre­quen­cies of the armor traits and their under­ly­ing genes. We then intro­duced two adult cut­throat trout into half of the exper­i­ment­al ponds, with the remain­ing ponds act­ing as our trout free con­trols. The fol­low­ing spring the second gen­er­a­tion of hybrid fish bred, cre­at­ing the third gen­er­a­tion (F3s). In the fall of 2013, after nearly a year of nat­ur­al selec­tion we sampled the phen­o­types and gen­o­types of this third generation.

rennisonfig3.png
Dr Diana Ren­nison intro­du­cing F1 hybrid stickle­back into the exper­i­ment­al ponds to start the experiment.
rennisionfig4.png
Aer­i­al view of the arti­fi­cial ponds used for the exper­i­ment. Photo by Thor Veen. 

To estim­ate the evol­u­tion that had taken place we com­pared the phen­o­types and gen­o­types of the fish col­lec­ted before (F2s) and after (F3s) one year of dif­fer­en­tial nat­ur­al selec­tion. By com­par­ing the evol­u­tion­ary response of the ponds that con­tained trout to the con­trol ponds we were able to determ­ine wheth­er trout drive the diver­gence of armor. Remark­ably we found that in as little as one year the addi­tion of trout caused a hab­it­at shift of the stickle­backs in the ponds and evol­u­tion­ary diver­gence in armor phen­o­types and in allele fre­quen­cies at the two genes con­trolling spine length.

This study provides one of the first examples in which the evol­u­tion of a phen­o­typ­ic trait has been linked to both the source of selec­tion and the under­ly­ing genes. Our find­ings also expand the known role of biot­ic inter­ac­tions, such as pred­a­tion, on spe­cies diver­si­fic­a­tion and sug­gest that per­turb­a­tions to a food web could have cas­cad­ing effects on the evol­u­tion and per­sist­ence of species.

 

Dr Diana Ren­nison is a cur­rently a postdoc­tor­al fel­low in the Peichel Lab at the Uni­ver­sity of Bern. This fall she will be start­ing her own lab at the Uni­ver­sity of Cali­for­nia San Diego. The ori­gin­al study is freely avail­able to read and down­load from Evol­u­tion Let­ters.