Data: Plant-pollinator interactions, pollinator body size (inter-tegular distance, ITD) and plant reproductive success (number of seeds produced).

Data collected by Christie J. Webber.

Data collected in 14 experimental flowering plant patches during December 2012–February 2013. Patches were located in a 105 hectare sheep farm pasture in Oxford, North Canterbury, New Zealand (43°19’21”S 172°12’25”E).

Files:

• Data_S1: contains plant-pollinator interactions sampled and pollinator inter-tegular distance (ITD). Data_S1 columns: patch ID where the interaction was recorded, plant species, pollinator ITD (mm), and pollinator family, genus and species.
• Data_S2: contains the number of seeds produced by each of the five flowers of each plant individual from each plant species on each patch. Data_S2 columns: patch ID where the measurement was taken, plant species, plant number (individual sampled), number of seeds.

Dataset used in "Scale-dependent effects of landscape structure on pollinator traits, species interactions and pollination success" by G. Peralta, C.J. Webber, G.L.W. Perry, D.B. Stouffer, D.P. Vázquez and J.M. Tylianakis.

Fourteen experimental plant patches (each in a square fenced enclosure approx. 2 m x 2 m) were set up at increasing distances from each other in a sheep farm pasture in Oxford, North Canterbury, New Zealand, 43°19’21”S 172°12’25”E (Supplementary Material Figs. S1–S3). The minimum and maximum distances from any patch to its nearest neighbour were 113 m and 322 m, and the maximum distance between any two patches was 1483 m. Patches were fenced off and rabbit-proofed with gates and hex netting. 

Each experimental plant patch consisted of 10 plant species. Each patch contained four Achillea millefolium (yarrow), three Agastache foeniculum (anise hyssop), twelve Brassica napus (canola), three Coreopsis tinctoria (coreopsis), two Echium vulgare (blueweed), twelve Fagopyrum esculentum (buckwheat), seven Lobularia maritime (alyssum), three Medicago sativa (lucerne), seven Nepeta grandiflora (catnip) and nine Trifolium repens (clover) plants. Trifolium repens was propagated from cuttings taken from the University of Canterbury (New Zealand) grounds, whereas the remaining nine plant species were grown from seed. All plants were cultivated in the University of Canterbury glasshouses, timed to begin flowering in early December. The plants were potted with slow-release fertiliser and watered daily while out in the field, to control for potential resource limitation. The number of individuals per species placed on a patch differed across species for logistical reasons, but plant species composition was consistent across patches. All plants used were of the same age and size within species. 

Throughout the flowering season of the experimental plants (December 2012–February 2013), six sampling rounds were conducted under warm and dry weather conditions, in temperatures over 14°C, between 10:00 and 17:00 (NZT). At each patch, for one hour per sampling round, all flower-visiting insects observed were captured and the plant species from which they were collected recorded. Insects were later identified to species level or, when identification keys did not exist, grouped into morphospecies with the help of reference samples identified by pollinator experts (Dr. Barry Donovan and Dr. Brad Howlett).

The body size of all insects collected was estimated by measuring their inter-tegular distance (ITD), i.e. distance between wing bases, with digital callipers. 

Plant reproductive success was measured on each experimental patch in seven of the ten experimental plant species (Agastache foeniculum, Brassica napus, Coreopsis tinctoria, Fagopyrum esculentum, Medicago sativa, Nepeta grandiflora, and Trifolium repens). For the remaining three species (Achillea millefolium, Echium vulgare, Lobularia maritime), damage from sheep grazing prevented us from measuring a sufficient proportion of flowers. To estimate plant reproductive success, five flowers on every experimental plant (individual) per patch were tagged and monitored. Only flowers that opened after being transplanted to the field were used for this assessment. Once flowers had closed and had sufficient time to set seed, we collected the flower heads and counted the seeds produced by each flower.