Breeding systems and floral visitation data for <i>Syzygium guineense</i> subsp. <i>barotsense </i>and <i>Julbernardia paniculata</i>.

Breeding systems data from a pollination exclusion study and floral visitation data from two species, <i>Syzygium guineense</i> subsp. <i>barotsense</i><i> </i>and <i>Julbernardia paniculata</i><i> </i>(<i>S. guineense</i>:<i> </i><i>n</i>=14 trees, <i>J. paniculata</i>:<i> </i><i>n</i>=12 trees), in Northwestern Zambia south of West Lunga National Park, conducted between 2018 and 2019. Trees were chosen to be generally similar with regards to maturity stage, height, diameter at breast height, and were selected within a small geographic area to control for geographic and climatic differences.<br>Three replicate inflorescences (on average 10.57 ± 0.76 <i>SE</i> individual flowers per inflorescence for <i>S. guineense</i> and 2.13 ± 0.19 <i>SE</i> for <i>J. paniculata</i>) with predominantly unopened flowers, were selected for each of four treatments on each tree (3×4 inflorescences per tree, i.e. 3×4×14= 168 treatment inflorescences for <i>S. guineense</i>, and 3×4×12= 144 treatment inflorescences for <i>J. paniculata</i>), with each treatment being applied to a subset of flowers on separate inflorescences (i.e. only one treatment type was applied per treatment inflorescence). The treatments included: (a) a naturally open-pollinated treatment; (b) a pollinator exclusion treatment for which selected inflorescences were bagged with mesh bags (mesh size ∼1 mm²) to prevent pollinators accessing the flowers; (c) a self-pollinated treatment for which treated flowers were bagged before opening and then virgin flowers were pollinated manually with pollen from the same tree and then re-bagged (pollen was taken from a different flower on the tree); (d) a cross-pollinated treatment for which treated flowers were bagged before opening and then virgin flowers were manually treated with cross pollen from a different tree and then re-bagged (for consistency the same pollen donor tree was used for cross pollinations on all experimental trees).<br>Floral observations were conducted during the respective flowering seasons for each species: November 2018 for <i>S. guineense</i> and May 2019 for <i>J. paniculata</i>. A total of 43 × 30-min observations were carried out for <i>S. guineense</i> (21.5 hrs total on 43 different trees), and 41 × 30-min observations for <i>J. paniculata</i> (20.5 hrs total on 41 different trees). Observations were stratified throughout the day to control for diurnal changes in floral visitation and floral visitor behaviour. Floral observation patches within the canopy were reached using the doubled rope tree climbing technique and their size was limited by the number of flowers that could be safely reached enabling the observer to be at least 1.5m away and limited to an area that could be viewed within a single field of view so that no floral visitors would be missed. Floral visitors were identified <i>in situ</i> when possible but when further identification under a microscope was necessary, insects were collected and euthanised.