Abstract
The introduction of marram grass (Ammophila arenaria) to the New Zealand dune ecosystem has brought about rapid changes to dune morphology and plant assemblages . The native sand binder Pingao (Desmoschoenus spiralis) has been widely displaced by the invasive A.arenaria in the dune landscape in New Zealand. However, there is very little known about the seed ecology of Ammophila in New Zealand conditions, nor from other overseas research. Seed ecology research for Desmoschoenus has been limited to the understanding of "best practices" for the harvesting of seed and raising plants in restorative efforts in affected dunelands. Dispersal studies of A. arenaria and D. spiralis have not been undertaken, and this has left a gap in our ecological understanding of the invader and competitive plant paradigm in the New Zealand coastal dune context.
The seeds from both species were collected during the 2004 and 2005 summer periods for eight weeks from Toko Mouth in South Otago. The Toko Mouth site was monitored for seasonal changes in sand movement, wind regime, temperature, and rainfall during the study period. In 2004, which was hot and dry during the pollination and seed maturation periods, 40.08% of fresh A. arenaria were germinable, whereas, only 1.56% of D. spiralis produced germinable seeds, and exhibited possible drought stress, through premature seed dispersal, and pollination failure due to the prolonged hot weather. The 2005 season saw heavy rain during anthesis and seed maturation, and only 15.90% of fresh A. arenaria seeds were germinable, but D. spiralis germination increased to 11.00%. D. spiralis flowers earlier than A. arenaria and is more sensitive to drought. The low viability of seed in 2004 appears due to possible drought stress triggering premature seed dispersal. Pollination success may also have been low. Reduced seed viability in A.arenaria in 2005 appears to relate to cold wet conditions during its later flowering and seed maturation period. Environmental conditions during the different flowering times between the two species appears likely to effect pollination and seed maturation, and this was observed in the subsequent increase or decrease in the production of germinable seeds.
Drying after seed ripening (22°Celsius for 7 days) and stratifying seeds (4°Celsius for 42-63 days) did not improve the germination of either species, indicating little innate dormancy. Only 7.35% and 7.75% of A.arenaria seeds germinated after drying in the 2004 and 2005 samples, and 0.25% and 2.5% of D. spiralis seeds germinated with drying after ripening in 2004 and 2005. Desmoschoenus seed germination was instead further decreased by the application of this method, possibly due to the already poor condition of Desmoschoenus seed produced in hot weather conditions. Stratification of Desmoschoenus in 2004 was 0.23% 1.77% in 2005 when compared to fresh seed, indicating the lack of a cold treatment requirement to break innate dormancy. Cold storage of D.spiralis seeds in seed banks appears part of a natural seasonal cycle, and dune disturbance accompanied with increases in dune temperature is likely the most important trigger for germination of seeds trapped in seed beds. Tetrazolium testing showed that the long term viability of seeds stored at cold temperatures (≤4°Celsius) was reduced in both species, suggesting that seeds may be only capable of short term survival in seed banks (1 or 2 seasons).
In dispersal testing the Phase I (initial dispersal from the parent) seed shadows of both species was a typically skewed leptokurtic distribution, showing that initial dispersal distance from the parent point source is low. However, the opportunities for movement were changed in the Phase II dispersal (movement from the surface) with Ammophila moving distances of up to 20 metres at high wind speeds. Micro-topographical influences altered the opportunities for wider dispersal of Desmoschoenus that remained in aggregated clumps close to the parent point source. The relative rollability of seeds of both species using a rollability apparatus was used for the first time in seed ecology to test the effects of morphological differences on dispersal behaviour in both species. The mean settling time in Ammophila (43.089 seconds) was slower than Desmoschoenus (31.184 seconds) when dry, revealing that the basal hairs of Ammophila acted as a pivot and anchor for the plume to anchor onto the surface. Settling times (Ammophila 32.746 seconds and Desmoschoenus 25.972 seconds) were greatly reduced when the hairs were wet and configured tightly around the plume, thus creating a more streamlined propagule shape. The morphological differences greatly affect the ability of the two species to reach and colonise new areas after Phase II dispersal.
The higher rate of fecundity and wider dispersal rates of A. arenaria gave it a distinct ecological edge over D.spiralis in the dune landscape in terms of invasion and colonisation. However, the movement of seed for ideal germination conditions was dependent on the orientation of the site and prevailing wind regime prevalent to the dune landscape.