Data from: Between-year changes in community composition shape species' roles in an Arctic plant-pollinator network

• Main Authors: Cirtwill, Alyssa R., Roslin, Tomas, Rasmussen, Claus, Olesen, Jens M., Stouffer, Daniel B., Olesen, Jens Mogens
• Format: Dataset
• Language: English
• Published: Dryad 2021
• Subjects: pollination; network structure; turnover; Holocene; Life sciences; medicine and health care; Greenland; High Arctic; envir; geo; Arctic; Zackenberg
• Online Access: https://doi.org/10.5061/dryad.mh0qs

Inter-annual turnover in community composition can affect the richness and functioning of ecological communities. If incoming and outgoing species do not interact with the same partners, ecological functions such as pollination may be disrupted. Here, we explore the extent to which turnover affects species' roles --as defined based on their participation in different motifs positions-- in a series of temporally replicated plant-pollinator networks from high-Arctic Zackenberg, Greenland. We observed substantial turnover in the plant and pollinator assemblages, combined with significant variation in species' roles between networks. Variation in the roles of plants and pollinators tended to increase with the amount of community turnover, although a negative interaction between turnover in the plant and pollinator assemblages complicated this trend for the roles of pollinators. This suggests that increasing turnover in the future will result in changes to the roles of plants and likely those of pollinators. These changing roles may in turn affect the functioning or stability of this pollination network. Dates_of_first_interactionContains tables listing the date of first observed interactions for each insect (I) and plant (P) in 1996, 1997, 2010, and 2011. Also contains tables identifying which insects (I) and plants (P) had tentatively dated interactions in each year, as well as the best-guess dates associated with these observations.Lists_of_interactionsContains lists of plants and pollinators (i.e., pollination networks) used in our analyses. There were no tentatively dated interactions (n.d.) in 2010 or 2011, so these networks were identical regardless of how these interactions were handled. For 1996 and 1997 networks, the yearly networks were identical in all cases and are therefore only included once (in the "n.d. year only" worksheets). For monthly networks, the "n.d. year only" worksheets describe networks which did not include any tentatively dated interactions, the "n.d. best-guess" worksheets describe ...