Microsite availability, not floral herbivory, limits recruitment in peripheral native thistle populations

Variation in insect herbivory can drive variation in plant fitness and population dynamics.  However, our ability to predict the ecological contexts in which insect herbivores will reduce plant fitness or population growth is limited.  In theory, populations at the periphery of a plant species’ biogeographic range, are expected to experience reduced herbivory. Further, in montane landscapes, elevation is expected to drive variation in abiotic conditions and variation in plant-insect interactions.  Specifically, less insect herbivory may occur at cooler, higher elevations.  To examine these predictions, we quantified effects of inflorescence- and seed-feeding insect herbivores in populations of the short-lived, monocarpic, perennial forb Cirsium canescens (Platte thistle) in montane grasslands in Colorado, USA. We asked: 1) Does insect flower head herbivory and predispersal seed predation limit Platte thistle lifetime seed production?, 2) Does this insect herbivory limit seedling recruitment?, 3) Does ecological context, including spatial - especially elevational - and temporal variation, affect the outcome of these interactions? We conducted insect exclusion experiments in three years at five sites over 52% of Platte thistle’s elevation range in our region.  We compared both lifetime viable seed production and seedlings recruited between plants with ambient vs. insecticide-reduced levels of flower head herbivory.  Insect herbivory on flower heads significantly reduced Platte thistle lifetime viable seed production at all sites, independent of elevation.  Unexpectedly, however, increasing seed by reducing herbivory did not lead to a proportional increase in seedling recruitment.  The relationship between viable seed production and seedling recruitment per plant was non-linear, decelerating across the range of seed production achieved by both plants exposed to and protected from flower head herbivory.  While elevation altered Platte thistle flowering phenology, it did not influence insect damage, viable seed production, or seedling recruitment.  These results show that flower head- and seed-feeding insect herbivores strongly reduced Platte thistle lifetime viable seed production, a key component of maternal fitness, in these peripheral populations. Yet, the herbivory did not determine population recruitment, suggesting post-dispersal processes limit recruitment here.  Further, elevation did not drive context-dependent variation in the insect herbivore outcomes.

Methods
The plant-based data sets (Plattethistleseedproduction.csv, Plattethistleseedlingestablishment.csv, Plattethistlefirstflowering.csv) are from experiments that examined the effects of insect floral herbivory and pre-dispersal seed predation on the number of viable seeds produced and the number of seedlings established by Platte thistles (Cirsium canescens) along an elevation gradient in Caffee County, Colorado USA.  Specifically, we were interested in addressing 1) how elevation affected Platte thistle viable seed production, 2) how insect herbivory on reproductive adult Platte thistles affected their viable seed production and 3) whether the effect of insect herbivory varied with elevation.  The experiments that contributed to this data set involved an insect exclusion treatment in which insect herbivory either was reduced on plants through insecticide application or the plants were not sprayed with insecticide and served as controls.  Control plants either were entirely unmanipulated or were sprayed with water.  In one experiment (Plattethistleseedlingestablishment.csv), seeds were pulled out of post-anthesis flower heads in the field to estimate viable seed production and then inserted back into the flower head to allow normal dispersal.  This experiment was focused upon quantifying effects of insect floral herbivory and pre-dispersal seed predation on seedling establishment.  Seedling establishment was monitored for two growing seasons after seed dispersal within a 2m radius circle of the parent plant.  In the other experiment (Plattethistleseedproduction.csv), post-anthesis flower heads were collected from the plants and dissected in the lab to quantify viable seed production per flower head.  Both experiments occurred in 2017, 2018, and 2019, using different adult Platte thistle individuals in each year. The species is monocarpic. The experiment was replicated at five sites along the elevation gradient. The plants used in the seedling establishment and seed production experiments were also monitored for flowering phenology (Plattethistlefirstflowering.csv)  

The data that are contained in the data sets titled 'Platteplantcommunitybiomass.csv', 'Baregroundcover.csv', 'Soilhardness.csv', and 'Soilmoisture.csv' all provide environmental context for the two experiments described above.  These environmental context data were collected in an experiment that was immediately adjacent to the two experiments described above at each study site.  This adjacent experiment examined the effects of rodent granivory / seedling herbivory and soil disturbance upon establishment of Platte thistle (Cirsium canescens) seedlings from seeds. Environmental context data were only collected in plots that did not experience the soil disturbance manipulation.