Seed dispersal and seedling recruitment in Miro (Podocarpus ferrugineus, Podocarpaceae) & Puriri (Vitex lucens, Verbenaceae) Andrew Pegman PhD Candidate.

Presentation on theme: "Seed dispersal and seedling recruitment in Miro (Podocarpus ferrugineus, Podocarpaceae) & Puriri (Vitex lucens, Verbenaceae) Andrew Pegman PhD Candidate."— Presentation transcript:

1 Seed dispersal and seedling recruitment in Miro (Podocarpus ferrugineus, Podocarpaceae) & Puriri (Vitex lucens, Verbenaceae) Andrew Pegman PhD Candidate SGGES University of Auckland Supervisors: G. Perry, M. Clout, J. Ogden Funding: UoA and ARC

2 Aims To determine: Local seed shadows at multiple sites in isolated Miro and Puriri trees The potential effects of decline or loss in Kereru (the disperser) numbers on the seed dispersal processes and community structure Whether seedling distributions are concordant with seed shadow patterns Seedling distributions of five Kereru-dispersed tree species Spatial distributions and population dynamics are linked to the shape and scale of seed shadows two components: local vs. long-distance dispersal

3 Field Methods Seed trapping using ‘modified cylindrical plastic traps’ (buckets) under 18 individuals at three sites seed counts, mesocarp intactness, and seed viability assessment total of 3000 buckets deployed! Kereru densities were determined in each site using the ‘distance sampling’ technique Seedling distributions were measured under each tree and seedlings were tagged Presence and abundance of seedlings of Kereru-dispersed species Miro, Puriri, Tarairi, Tawa, and Karaka under the canopy of individuals of these same species

4 Puriri (L) & Miro (R) trees

5 Puriri (L) & Miro (R) fruit

6 Seed-fall traps (pilot study left, main study right)

7 Data analysis Seed shadows: plots of numbers of seeds versus distance, combining trees from each area to obtain SE Log-likelihood methods were used to determine which phenomenological model best ‘fitted’ these data Tested for association between Kereru densities and mean fraction of ‘depulped’ seeds for each species in each area seed shadows and seedling distributions The % of species of seedlings growing under adult individuals of five kereru-dispersed species

8 Waitakere Miro seed shadows
Canopy edge at 3.29-m (m) Note the high peak near the source and the location of the ‘depulped’ seed shadow

9 Hunua Miro seed shadows
Canopy edge at 2.60-m (m) Note the peak near the source and the different location of the ‘depulped’ seed shadow

10 Waitakere Puriri seed shadows
Canopy edge at 5.66-m (m) Note the flatness of the shadows and the location of the ‘depulped’ shadow

11 Wenderholm Puriri seed shadows
Canopy edge at 5.20-m (m) Note the peak and different location of the ‘depulped’ shadow

12 Fitting theoretical dispersal models to empirical data
(& species) AIC (-LL + 2K) K (parameters) Δi Gamma (Miro) 10476 2 Lognormal (Miro) 10909 433 Weibull (Miro) 10945 469 Weibull (Puriri) 5059 Neg. binomial (Puriri) 7331 2272 Poisson (Puriri) 7335 1 2276 Other studies have found support for these models

13 Miro total seed shadows and the Gamma function

14 Puriri total seed shadow and the Weibull function
(m)

15 Seed shadow conclusions
Plants disperse their seeds using non-random distributions: traditionally assumed to be Gaussian or Normal patterns (but rates of plant migration are too high to have resulted from these theoretical models) models which fit Miro and Puriri dispersal are Gamma and Weibull respectively both are leptokurtic which facilitates faster migration (Weibull can approximate Normal though)

16 Kereru densities ?

17 Kereru density v. fraction of depulped seeds in 4 areas
Wenderholm Puriri Hunua Miro Waitakere Miro Waitakere Puriri

18 Kereru density v. fraction of depulped seeds in 4 areas
H M We P W M r = 0.954, p < 0.05 W P

19 Kereru density v. mean number of depulped seeds in 4 areas
H M W M We P W P

20 Log of total seeds v. distance for Miro in Hunua & Waitakere

21 Log of total seeds v. distance for Puriri in Waitakere & Wenderholm

22 Kereru densities: conclusions
High Kereru densities result in a higher fraction of depulped seeds in the seed shadow but not always a higher absolute number Increased Kereru densities extend the seed shadow on a local scale for Miro, but less so for Puriri No other frugivores appear to be taking on the role of dispersal vector: declining Kereru numbers could result in dispersal failure

23 Seedling concordance If seeds are ‘escaping’ from the parent (i.e. negative relationship between seed-rain & seedlings) then the resulting seedlings may have greater survival rates: Miro- some evidence of ‘escape’ of seedlings, but mostly independence. Sometimes seedlings establish under parent. Puriri- no evidence of ‘escape’, 1 example of a positive relationship

24 Hunua Miro: a negative correlation (escape) between seed and seedling distributions
r = at p < 0.05

25 % of trees of each species that have specified seedlings under the canopy, combining all areas

26 Conclusions: seedlings under canopies of different tree species
The most common seedling under any tree is its own species, especially for Karaka Miro is the least likely seedling species to be encountered under other tree species Puriri and Tarairi are common under all tree species Karaka was common under Puriri and Tarairi

27 Any Questions? Tarairi seedling