Figure 4.4 Germination rate for various population sizes of Melaleuca ericifolia in
Victoria and Tasmania. A – Apsley Marshes and Cades Road; B – Dowd Morass.
4.3.2 Melaleuca parvistaminea
Viability of seeds from the various populations of M. parvistaminea was much higher
than for M. ericifolia, and varied from 70 – 80 % regardless of population class (Table
4.5). Although most larger population classes had better overall germination, there
was overlap in germination rates between the various population size classes, unlike
the case with M. ericifolia (Table 4.4). Lower germination rates were not associated
with disturbance to the sites. Melaleuca parvistaminea seeds were consistently
) than M. ericifolia seeds (15-29
) (Figure 4.5).
Table 4.5 Population size, seed weight and viability of various populations of
Melaleuca parvistaminea in South Gippsland, Victoria, Australia.
Population location Population size (ha) Number of seeds per mg Mean seed weight (µg) Percent germination (%) Rosedale 1-5
4.3.3 Comparison of M. ericifolia and M. parvistaminea
There was a clear distinction in germination rates and seed weights between
populations of M. ericifolia and M. parvistaminea. Average seed weights of M. parvistaminea were always above 30
whereas those of M. ericifolia were
. Even greater differentiation was seen in overall germination of the
two species with M. ericifolia varying from 0–38 % and M. parvistaminea varying
from 70–80% (Figure 4.6).
Figure 4.6 Comparison of germination rates and seed weights of various populations
of M. parvistaminea and M. ericifolia.
4.4.1 Trade-offs between sexual and nonsexual reproduction
The clear dichotomy in germination rates in relation to regeneration characteristics of
M. parvistaminea and M. ericifolia (Figure 4.6) suggests that there is a trade-off in
resource allocation between these two species. This result confirms findings of a
number of other studies that indicate resources are primarily directed to seed
production in obligate seed-regenerating species and to vegetative reproduction in
rootstock-regenerating species (Armstrong 1982; Sutherland and Vickery 1988). Seed
size and germination rates were consistently higher in the obligate seed-regenerating
species M. parvistaminea than in M. ericifolia, and smaller average seed size and
lower overall germination rates were consistent across the range of provenances of the
rootstock-regenerating species M. ericifolia.
Reduced seed production and viability does not imply reduced genetic diversity in
populations (Ellstrand and Roose 1987). Lamont and Weins (2003) undertook a pair-
wise comparative study across a range of generic pairs of species, mainly in Australia,
that exhibited a similar dichotomy in regeneration method and ecological segregation
to M. ericifolia and M. parvistaminea. They found that 30 of the 33 pairs of woody
species tested had lower viable seed production in the resprouting species. They
postulated however, that lowering of potential sexual reproduction in resprouting
plants did not indicate lowering of genetic diversity within a given population but did
not indicate why this would be. Widen et al. (1994) reported no loss of genetic
diversity in a review of 45 mainly herbaceous clonal species from a range of habitats,
including wetlands, in the Northern Hemisphere.
Sexual reproduction after the initial establishment of vegetatively reproducing, clonal
plants may not be important to survival in the short to medium term (Eriksson 1989;
Silverton 1982; Harper 1978). For example, of the 68 clonal species investigated by
Eriksson (1989), 41 were not observed to recruit into existing established populations.
Even amongst clonal species that do regenerate from seed, for example Solidago canadensis, changes in population structure over time may favour those individuals
with the strongest ability to reproduce vegetatively (Hartnett and Bazzaz 1985).
Sexual reproduction would only be needed upon the death of the parent plants or to
colonise new environments.
4.4.2 Other factors influencing sexual reproduction in M. ericifolia
Low fecundity in M. ericifolia and the lack of consistent observation of seedlings
within existing populations in the wild would indicate that plants are diverting
resources to vegetative, rather than sexual, reproduction. Variations in viability, from
0 % to 38 %, across the range of M. ericifolia indicate that there are other processes
occurring that influence sexual reproductive success. The three populations of M. ericifolia that had zero seed viability (Little River; Cape Nelson: Mt Wellington,
Table 4.4) all occurred at the very extremes of the species’ distribution in western
Victoria and southern Tasmania: the Little River population is approximately 70 km
from the nearest M. ericifolia population, the Cape Nelson population 300 km, and the
Mount Wellington population 200 km. At all three sites there is no obvious
morphological variation within the population and, although seed capsules were found
on the plants, none contained seeds with embryos. It may be that these three
individual populations actually represent individual plants that have seeded into the
site and now persist vegetatively. The largest of the three populations, at Cape Nelson,
covered 1-5 ha (see Table 4.1). Three populations that exhibited low viabilities, Dowd
Morass, Cades Road and Apsley Marshes (6, 12 and 14 % respectively) were all
relatively large (10-1000 ha) and were within 50 – 100m of other existing
populations. All three sites are highly disturbed, occur in primarily agricultural
landscapes and suffer from secondary salinisation (Grayson 2003; Ramsar 2006).
Most of the populations of M. ericifolia with germination rates < 10 % were highly
disturbed remnants, usually adjacent to roadsides or in formerly grazed land. It may
be that these populations represent colonisation by relatively few individuals.
Fragmentation and limited re-colonisation of this type adversely affects genetic
diversity, reduces pollination and variously affects hydrology and potentially
increases nutrient levels depending on surrounding land use (Henriquez 2004;
Renison et al.2004; Wooller and Wooller 2004; Cunningham 2000; Morgan 1999)
with consequent effects on germination capacity and seed size.
Although M. ericifolia can grow in brackish swamps, there is a great deal of variation
in the salinity tolerance of different genets (Ladiges et al.1981) and in the salt
sensitivity of various life stages (Robinson et al. 2006; Salter et al. 2007). Prior to
European settlement, all three of the sites listed above (Dowd Morass, Cades Road
Swamp and Apsley Marshes) were freshwater swamps. Present conditions of the
ground and surface water at least at the first two sites, Dowd Morass and Cades Road,
have become highly saline with levels up to half sea water (> 20 g L
, Grayson 2003;
Waterwatch data). Reduced flowering, seed set and viability with increasing salinity
levels is widely recognised as affecting reproductive success in a range of species
including those that naturally occur in salt marshes and brackish situations (Boscaiu et al. 2005; Labidi et al. 2004; Redondo et al. 2004).
Secondary salinised sites pose difficult and perhaps intractable problems for sexual
reproduction in woody taxa. The underlying conditions that produce secondary
salinisation are complex and the impacts usually cover extensive areas. In the case of
Dowd Morass, removal of nearly 90 % of the original flow of fresh water from the
rivers that feed the Morass has allowed intrusion of salt water from the sea (Grayson
2003). Periodic flushing of Dowd Morass with fresh water no longer occurs and
salinity levels have been progressively rising. Major alterations to salinity and water
regime have been recognised as contributing to the collapse of existing vegetation
communities within several major wetland systems in Australia (Kingsford 2000).
Collapse of the existing M. ericifolia dominated swamp, first identified by reduced
seed viability in this study but more recently from rapid and extensive plant death
(Boon et al. 2007), could see complete removal of M. ericifolia from most of this
wetland system and in time, local extinction. Dowd Morass is one of several Ramsar-
listed wetlands surrounding the Gippsland Lakes. The colonial nesting birds, Ibis
(Threskiornis spp.) and Cormorants (Phalacrocoraxspp.),are reliant on M. ericifolia for nesting and roosting at Dowd Morass. The loss of M. ericifolia would have
significant impacts on these two bird species, making the habitat unsuitable for them.
If present changes to vegetation induced by salinisation of Dowd Morass continue,
predictions made by Bird (1962) that complete alteration of the vegetation to a salt
marsh will become true. Loss of the existing vegetation and suitable sites for nesting
of colonies of birds would inevitably lead to the loss of the Ramsar listing of these
It would appear that the least disturbed and potentially most viable populations of M. ericifolia are in East Gippsland, the Bass Strait Islands and some sites in Tasmania.
The viability of seed from these populations varied from 20 % to 38 %. Similar seed
viabilities, ranging from 3-28 %, have been found in the rootstock regenerator
(resprouting) Melaleuca quinquenervia in Queensland and South Florida
(Reyachhetry et al.1998; Browder and Schroeder 1981; Meskimen 1962). Melaleuca quinquenervia isan ecologically equivalent species to M. ericifolia andindeed
replaces it in near-coastal freshwater to brackish wetlands in northern NSW and
Seed viability of most populations of the rootstock-regenerating species M. ericifolia was markedly lower than the co-occurring seed-only regenerator, M. parvistaminea.
While low seed viability would appear to be detrimental to the continued survival of
M. ericifolia, this is at least partially compensated for by extensive vegetative
reproduction. However, very low viability rates were detected for isolated and highly
fragmented populations of M. ericifolia, as well as those affected by secondary
salinisation, these factors inhibiting the ability of M. ericifolia to sexually reproduce
(Eckert 2001). Loss of sexual reproduction potential is particularly critical in sites
such as Dowd Morass and Cades Road that experience salinisation, as this may lead to
eventual loss of the entire paperbark population. In support of this conclusion,
Bowkett and Kirkpatrick (2003) predicted that long-term survival of M. ericifolia in
Tasmania would only occur in the largest populations of the species in the Tamar
Valley which would be most resistant to ecological change.
To conserve and maintain the sexual viability of populations of M. ericifolia the
underlying causes of decreased viability of seed, below that which is naturally
occurring, needs to be addressed. Interference with naturally isolated populations by
introduction of additional genets and therefore genetic diversity (eg. by revegetation
activities) could alter vegetation dynamics of the sites and competitive abilities of M. ericifolia. Increasing the competitive ability of M. ericifolia in this way could threaten
the existing vegetation communities of these sites. Conversely, introduction of genetic
material between sites in highly fragmented remnants could increase the competitive
ability of genetically weakened remnants, provided sexual reproduction takes place,
contributing to the conservation of the species locally.
Chapter 5 Germination characteristics of Melaleuca ericifolia Sm. (Swamp Paperbark) Abstract Seed collected from Dowd Morass, a secondary-salinised Ramsar-listed wetland of
the Gippsland Lakes region in eastern Victoria, showed very low viability (< 6 %),
with less than 50 % of the seeds germinating even under ideal laboratory conditions.
Greatest germination occurred with surface-sown seed, germinated in darkness at a
mean temperature of 20
C and salinity < 2 g L
. At 20
C, maximum germination
occurred at a salinity of 1 g L
; germination fell rapidly at a near constant rate with
increasing salinity. Lower temperatures, while moderating the inhibitory effects of
salinity, markedly reduced germination; higher temperatures increased the inhibitory
effects of salinity and light and reduced overall germination rates. Seeds subjected to
brief inundation with saline water germinated rapidly if flushed by, and subsequently
grown under, freshwater conditions. Specific timing of management interventions,
particularly manipulations of water regime to control salinity regimes, are required if
germination of M. ericifolia on the landscape scale is to be successful. Even so, the
low overall viability of the seed would present difficulties to large-scale, seed-based
Swamp paperbark (Melaleuca ericifolia Sm.) is a small clonal tree in the Family
Myrtaceae which grows in coastal (freshwater and brackish-water) swamps across
southern and eastern Australia, from Tasmania through to northern New South Wales
(Jeanes 1996). Since the distribution and abundance of this species has decreased
markedly with the clearing or draining of wetlands in which it formerly occurred
(Bowkett and Kirkpatrick 2003), a high priority ofnatural-resource management
agencies and non-government organisations throughout Australia is the rehabilitation
of high-value coastal wetlands that contain, or did contain, M. ericifolia and other
Melaleuca species (de Jong, T.J. 1997; de Jong, N. 2000). While most Melaleuca species are reliant on seed to produce new individual plants (e.g., Melaleuca parvistaminea, Melaleuca quinquenervia), M. ericifolia is unusual in that it is
extensively clonal, producing physically independent ramets across time and space
(Ladiges et al. 1981).
The development of the clonal growth form is commonly seen as a response to limited
opportunities for seed to germinate and seedlings to recruit into the population
(Barsoum 2002; Pan and Price 2002; Sachs 2002). The investment in clonality, at the
apparent expense of sexual reproduction, allows plants to survive and persist in
conditions that would be hostile to non-clonal plants (Jurik 1985; van Kluenen et al. 2000; Eckert 2001; Barsoum 2002). The extensive clonal growth exhibited by M. ericifolia might act as a buffer to all but the most challenging environmental
conditions. Major cataclysmic events, including flood, drought, fire and inundation
by sea water, occur periodically in the habitats occupied by the species and are
potentially lethal to both adult and juvenile M. ericifolia (Ladiges et al.1981; Salter
2001; Grayson 2003). While short to medium-term survival is ensured by the clonal
growth form, colonisation of new sites or re-colonisation of existing sites after
mortality of existing M. ericifolia plants is presumed to be reliant on germination
from seed (de Jong N. 2000).
Although M. ericifolia seed seems to have no specific inherent germination inhibitors
apart from containment within woody capsules, it is possible that short-term, salt- and
temperature-induced dormancy may take place, as has been reported for Melaleuca quinquenervia (Serbesoff-King 2003). Viability, however, is highly variable across
the range of M. ericifolia, with a large percentage of unfilled or otherwise damaged
seed within the capsules, a trait shared with M. quinquenervia (Ladiges et al. 1981;
Rayachhetry et al. 1998; J. Salter pers comm.). Seed of M. ericifolia and M. quinquenervia loses viability rapidly once released from the plant, with germinability
greatly reduced after one year (Woodall 1983; Bodle and Van 1999; Rayamajhi et al.
2002; J. Salter pers comm.). Germination of M. ericifolia seed at Coomonderry
Swamp in New South Wales (Australia) has been reported to be episodic, and survival
of the germinants is reliant on a narrow range of site conditions (de Jong N. 2000).
This conclusion is consistent with the analysis of a series of historical air photographs
of M. ericifolia swamps in western Gippsland, which suggest there are gaps of several
decades between successful recruitment events.
Despite its widespread distribution across coastal areas in eastern and southern
Australia and its priority listing for rehabilitation efforts, it is not clear what factors
control the success of seed germination and plant establishment in M. ericifolia. The
coastal environment in which most M. ericifolia grows suggest that salinity will play a
major role, both in terms of absolute effects of salt and the effects of short-term
exposure to saline water (Ladiges et al. 1981). In common with many other taxa,
temperature and light intensity are also likely to be important (Gul and Weber 1999;
Khan and Gulzar 2003). Finally, the depositional nature of coastal wetlands suggests
that burial may play a role; burial has been shown to be a critical factor controlling
germination success in other wetland and terrestrial plant species (Van et al. 1998).
The only published research on the individual effects of key environmental variables -
temperature, light, burial and salinity - on germination in M. ericifolia is by Ladiges et al. (1981). Although interactive effects among these factors have not been
investigated at all, there are documented effects of the interactive effects of
environmental variables, particularly light, salinity and temperature, on the
germination of a range of brackish wetland species (Gul and Weber 1999; Khan and
The aim of the present study was to quantify the primary and interactive effects of key
environmental parameters - salinity, temperature, light, burial and substrate type - on
the germination of M. ericifolia. This information will be useful in providing
explanations for historical patterns in changes to vegetation in coastal wetlands and in
developing better strategies and protocols for rehabilitating these areas at the
landscape scale. In particular, this study will help in explaining the characteristics
needed to determine safe sites for germination and establishment.
Materials and methods 5.2.1 Seed collection
Seed capsules were collected in April 2004 from 25 adult trees scattered throughout
the population of M. ericifolia at Dowd Morass. Individual plants were determined by
visual assessment of growth configuration; a characteristic dome shape. Seed was
collected from widely separated (> 100 m) plants that were presumed to be un-related
clonally. Capsules were stored in paper bags at 20
C for one week. The bags were
lightly shaken to release seed from capsules, and the contents sieved to remove empty
capsules and other detritus. Seed was placed in clean paper bags for a further 3 days
to remove excess moisture and transferred to sealed glass containers and stored at
C in darkness until used. Preliminary germination trials began in late April 2004,
with the main trial carried out in July 2004.