Paul G. Peterson

Assessing pollination and fruit dispersal in Fuchsia excorticata (Onagraceae)

Abstract We assessed the degree of pollen and fruit‐dispersal limitation in the endemic gynodioe‐cious New Zealand tree fuchsia Fuchsia excorticata over several seasons and at 17 sites throughout New Zealand. The flowers were visited mainly by two endemic honeyeaters (bellbirds and tui), and in some cases very frequently by silvereyes, which also occasionally rob nectar from flowers. We confirmed that hermaphrodites account for more than half the plants in all populations, are fully self‐compatible, and can autonomously self in the absence of pollinators (especially in plants with smaller herkogamy). Fruit production in hermaphrodites and (particularly) females was frequently pollen‐limited (mean Pollen Limitation Indices of 0.17 and 0.40, respectively), and was correlated with visual assessments of pollen loads on the stigma, a useful index of pollinator service. A comparison of the proportion of ripe or overripe fruit on branches exposed to birds versus branches enclosed in wire cages showed that uncaged fruit on Kapiti Island is removed almost as soon as it is ripe but on the mainland it persists for much longer. The proportion of ripe or overripe compared to green fruit is therefore an approximate index of dispersal service. Both indices may be useful to managers concerned with measuring the level of mutualism service provided by native birds.

Pollination performance and vulnerability to pollination breakdown of sixteen native shrub species from New Zealand

Abstract The reproductive biology of 16 native shrub species was studied in 34 populations to identify breeding systems, pollen limitation, local abundance, and population age structures. Seven of the study species are hermaphroditic, seven dioecious, and two gynodioecious. one of the 18 hermaphrodite populations (Alseuosmia macrophylla at Mamaku Plateau) had high levels of self‐incompatibility and pollen limitation and mutualism failure was evident. in the gender‐dimorphic taxa, two populations (Co‐prosma spathulata from Hakarimata and Cyathodes juniperina from Pukemokemoke) had high levels of pollen limitation and insect‐pollinated species consistently ranked higher in a vulnerability assessment compared with wind‐pollinated species. There was no significant relationship between natural fruit set and the distance to the nearest conspecific pollen in any of the study populations. Seedling recruitment was variable but evident in 32 of the 34 study populations and appeared to be related to availability of suitable habitat. The species we studied occur mostly on forest edges where they rely on disturbed soil and high light conditions for establishment, and edges may be important for successful reproduction of some native plants, especially shrubs. We have demonstrated that plants with self‐incompatibility mechanisms and pollinator specialisation are at greater risk from pollen limitation and mutualism failure than self‐compatible or generalist species.

Density dependent prey‐feeding time of Stethorus bifidus (Coleoptera: Coccinellidae) on Tetranychus lintearius (Acari: Tetranychidae)

Abstract The gorse spider mite, Tetranychus lintearius, a biological control agent introduced into New Zealand to control gorse, is often attacked by the endemic coccinellid, Stethorus bifidus. Predation by S. bifidus has been suggested as a reason why T. lintearius colonies collapse. For S. bifidus predation to regulate T. lintearius populations, at least one component of its numerical or functional response must result in an increased proportion of mites being killed as mite density increases. Laboratory experiments showed that feeding time (a sub‐component of the functional response) decreased markedly with increased T. lintearius density. An increase in available prey density from 3–25 mites/177 mm2 led to an exponential decrease in mean feeding time from 870 s to 100 s. Furthermore, despite S. bifidus killing more mites, it extracted progressively smaller proportions of the contents of each mite killed as mite density increased.

Breaking and remaking a seed and seed predator interaction in the introduced range of Scotch Broom (Cytisus scoparius) in New Zealand

1. Introduced plants may initially experience enemy release, but some of those interactions may be reinstated through biological control. These cases provide opportunities to explore the dynamics of broken and remade consumer-resource interactions.