Pollen limitation of Vaccinium myrtillus along an elevational gradient
Data files
May 27, 2025 version files 43.91 KB
Abstract
Plant dependence on pollinator-aided reproduction may decrease with colder and more unstable climates, for example at higher elevations. Correspondingly, plant populations may also be more pollen limited in alpine than in lowlands or under more optimal conditions. However, such expectations have largely been investigated across species, and rarely among populations within a species. The findings of such studies may have implications for predictions of plant sexual reproduction in a changing climate. We combined a bagging experiment of Vaccinium myrtillus (bilberry) flowers with a supplemental pollen experiment at three sites along an elevational gradient at 100, 500 and 900 m a.s.l. in a mild boreal climate, by the Sognefjord, Norway. Our goal was to investigate the variation in reproductive success and pollen limitation in a functionally important plant species, V. myrtillus, and to test whether the outcomes may be dependent on insect flower visitation and climatic context (i.e. temperature related to elevation). Our experimental treatments clearly affected reproductive success of V. myrtillus. Bagging of flowers reduced the fruit set close to zero, whereas supplemental pollination resulted in 30% higher fruit set. However, fruit set effects appeared to be independent of elevation in the current study. Total number of seeds was highest at the optimal mid-site location, but here the proportion of mature seeds was also lowest. Our results clearly showed that the reproductive success of V. myrtillus depends on pollinator visitation and benefits from extra pollen supply, but that these effects appeared to be independent of the climatic context.
Pollen limitation of Vaccinium myrtillus along an elevational gradient
https://doi.org/10.5061/dryad.41ns1rnr1
Description of the data and file structure
To assess the baseline question of the degree to which V. myrtillus reproduction is dependent on pollinator visitation, we first bagged 16 flowers on different individuals in each of the elevational sites in 2019. Individual plants were selected randomly within four 5 × 5 m blocks established in a related study (Hegland and Gillespie 2024). We selected the individual plants randomly within a subsection in each corner of the block to ensure that plants were minimum 3 m apart, that the ramets would be of similar size (10–25 cm tall) and that it would be practically possible to bag a single flower not yet open to avoid cross-pollination having already occurred. The bag was made of fine mesh so that insects could not penetrate, but that light and air were still available to the flower and the leaves within the bag. We collected the fruits and recorded fruit set ca 60–70 days after the bagging treatment. The investigations on the insect-pollination dependency laid the groundwork for designing the pollen-limitation experiment.
To examine how pollen limitation and reproductive success in V. myrtillus may vary along an elevational gradient dependent on climate and/or pollinator abundance, we performed a pollen supplementation experiment in 2021. During peak flowering and when bees were active at each site (Low: 14-15 May, Mid: 29-30 May, and High: 3-4 June), we randomly selected 30 ramets of 12-22 cm height that contained both newly open and unopen flowers. Due to the intensive and short flowering-period at Low because of the warm spring of 2021, we only found 21 ramets that met the selection criteria on the days of experimental application. To ensure cross pollination, pollen was collected from *V. myrtillus *clones at least 15 metres away from the target plants, by pulling out the anthers with tweezers. We then used the anthers to saturate stigmas of newly opened flowers positioned with pollen the at the top branch of the selected ramets. The pollen treatment was performed over two consecutive days with pollen from a different donor each day. Each experimental flower was combined with an internal open-pollinated control flower to assess potential resource allocation at the ramet level, which may be a problem when it is not practically possible to add pollen to all flowers on a plant (Knight et al., 2006), and an external open-pollinated control flower at a ramet ≥ 5 m away to test for the direct experimental effect of pollen limitation. All ramets and flowers were marked to enable relocation during the fruiting season. We counted and collected the ripe fruits ca. 60 to 70 days after the supplemental pollen treatment, and the seeds were counted in the laboratory. We counted all the seeds or ovules and categorised the seeds as a) mature (fully filled), b) incomplete (partly filled) and c) undeveloped (small and not filled) following the protocol of Jacquemart and Thompson (1996). From this we obtained three response variables. 1) Fruit set (proportion of developed fruits) and 2) seed set (proportion of matured seeds, calculated as a/a+b+c) that both could reflect reproductive success, whereas 3) total number of seeds (or ovules, calculated as a+b+c) represented the reproductive investment.
Files and variables
| Variable explanation in Poll suppl exp | |
|---|---|
| flowermark | date of marking and pollen supplementation |
| elevation | high, mid, low: gradient factor, site |
| transect | 1,2,3: plants were selected along these transect in a structured way. |
| code | E for experiment, C for control, I for internal control; 1-X for number along transect: 1,2 or 3; lastly site name |
| id | unique id to sort and tag each plant |
| treatment | hand pollinated, control internal and external. Experimental factor. Control internal only for testing whether fruit set differed when applying extra pollen. Initial test showed no difference and can therefore be excluded from further analysis. |
| paired | specific experimental and control plants that belong together; structuring variable |
| #flowers | always 1, used to calculate fruit set (0 or 1) |
| flower_count_ramet | how many flowers ramet had at start of experiment |
| lenght_ramet(cm) | how high ramet was at start of experiment; potential controlling variable? |
| harvestdate | when berries were harvested. |
| #berries | the fruit set, in this study 0 or 1; response variable for fruit quantity |
| seeds_developed | # of developed/filled seeds, that is germinating relatively good. Response variable for seed number, alternatively divided on total number of seeds |
| seeds_partly_developed | # partly filled seeds, germination rate very low |
| seeds_undeveloped | #little/unfilled filled seeds, not germinating in 49 trials |
| #seeds_total | total # of ovules or seeds, potential variable for reproductive success |
| NA’s for fruits were mainly not retrieved or had broken stems | |
| Control internals were not counted for seeds as the treatment had no effect on fruits (=NA’s) |
| Variable explanation in Bagged | |
|---|---|
| #Berries | number of single flower developing into a fruit |