Ornamental flowers are commonly planted in urban and suburban areas to provide foraging resources for pollinator populations. However, their role in supporting broad pollinator biodiversity is not well established as previous studies have been conducted in urban landscapes with pollinator communities that are distinct from those in natural systems. We observed pollinator visitation patterns to five ornamental annual plant genera and their cultivars over multiple years at two semi-natural sites in Pennsylvania to understand their potential for supporting diverse pollinator communities. There was significant variation in visitor abundance and diversity by season and year for many annual ornamental cultivars. Within some genera, cultivars had similar visitor abundance, diversity, and main visitor taxa, while cultivars in other genera varied greatly in these measures. We observed only polylectic (pollen generalist) bee species visiting annual ornamentals, despite the presence of oligolectic (pollen specialist) bee species in the background population. We conclude that the attractiveness of annual ornamental plants likely depends on both cultivar characteristics and environmental context. While their role in supporting complex pollinator populations is limited both based on the number of and dietary breadth of the species they support, ornamental plants may nonetheless provide long-lasting supplemental foraging resources for the generalist pollinator communities characteristic of urban and suburban environments.

This data represents recordings of all pollinator visitors for 10 minutes to ornamental annual plants in a field setting in Central Pennsylvania throughout the summers of 2016/2017. Visitors were classified to morphotaxa and then later sorted into functional groups for statistical analyses. The data was analyzed using Generalized Linear Mixed-Effects Models. Also included are the floral area measurements (cm2) for each individual plant included in the study (plantNum) that were used to standardize the observations. Floral area was calculated in photoshop using a defined pixels/cm metric unique to each photo. This metric was caluclated based on a measuring stick included in each photo held at the crown of the plant. 

The recorded value for each taxa represents the number of independent visits to a plant within each group for a 10 minute time period. If a visitor visited multiple flowers on one plant, this was recorded as a single visit. If the visitor left the observer's plain of vision and then returned (so it could not be confirmed that it was the same individual), this was multiple visits. Genera/species are abbreviated as listed below (upper case here and in data sets). Cultivars are similarly abbreviated (in lower case here- upper case in datasets). 

MAR= Marigold spp.

mto = 'Taishin Orange'

mdo = 'Alumia Deep Orange'

my = 'Alumia Yellow'

mr = 'Alumia Red'

mf = 'Alumia Flame'

ZIN = Zinnia spp.

zsr = 'Zahara Starlight Rose'

zsun = 'Zahara Sunburst'

zras = 'Zahara Raspberry'

zred = 'Zahara Red'

zpc = 'Profusion Cherry'

PEN = Pentas laceolata

pgrl = 'Graffiti Red Lace'

pgw = 'Graffiti White'

plav = 'Starcluster Lavender'

psw = 'Starcluster White'

psr = 'Starcluster Red'

LOB = Lobularia maritima

lebv = 'Easter Bonnet Violet'

lsp = 'Snow Princess'

lwdp = 'Wonderland Deep Purple'

lfk = 'Frosty Knight'

lcrys = 'Clear Srystal Mix'

LAN = Lantana camara

lrg = 'Rose Glow'

lsr = 'Sunrise Rose'

lps = 'Peach Sunrise'

ly = 'Yellow'

lcit = 'Citrus'

NA represents a plant that did not have any flowers. A value of 0 represents a plant in flower that recieved no visitors during the observation period.

For observations, there are two data sets (abundance, diversity) included that differ slightly to maintain accuracy. Since the diversity calculation required accurate morphotaxa identification to a finer taxonomic level than the proportional and overall abundance caluclations did, and we had multiple observers with varying levels of field identification experience, we trimmed this data set accordingly to ensure that only reliable morphotaxa identifications were included. We do not anticipate that these actions changed the interpretation of the data.