How do Water Regime and Grazing Alter the Reproductive Capacity of Aquatic Plants?
Dr Margaret A. Brock
Botany, Rural Science and Natural Resources, University of New England
Environment Australia, 2000
How does water regime and clipping influence the allocation of plant biomass to reproduction?
Seed bank collection
Soil containing seeds ('seed bank') was collected in September 1997 from four wetlands on the New England Tablelands (altitude 1400m, 30º 04'S 151º 46'E): Dumaresq Dam, Racecourse Lagoon, Barleyfields Lagoon, and Llangothlin Lagoon. The soil was dried to kill the extant vegetation, then mixed in even proportions. This mixed seed bank was used for experiments in the UNE Experimental Wetlands as well as this Tank Trial experiment.
A set of six round fibreglass tanks, 1m in diameter and 1m deep, was used to examine the effect of water regime and grazing on wetland plant establishment and reproduction. A metal mesh frame was fixed above these tanks, from which the pots (filled with seed bank) were hung using chains and hooks. Pots were raised and lowered in order to simulate water level fluctuations. This allowed all water regimes to be imposed within one tank, removing blocking error. The tanks were kept permanently full with tap water by filling them to overflowing weekly. This also flushed the tanks, removing surface scums and algae.
Pots (17cm diameter, 17cm deep) were filled to within 5cm from the top with soil collected from Llangothlin Lagoon in 1991. They were then filled to the top with the mixed seed bank. Twelve pots were placed in each of the six tanks, and left at 'damp' (soil waterlogged but not submerged) for 24 hours. After this 24 hour period, the following 6 water regimes were imposed on the pots (Fig j1).
- Regime A: permanently submerged (pots permanently at bottom of tank),
- Regime B: permanently damp (pots permanently at water surface),
- Regime D: dry to submerged (pots move from dry to bottom of tank over 16 weeks),
- Regime F: damp to submerged (pots move from damp to bottom of tank over 16 weeks),
- Regime G: submerged to damp (pots move from bottom of tank to damp over 16 weeks), and
- Regime L: mimic (pots mimic the water level fluctuations as recorded at Llangothlin Lagoon).
Each water regime was imposed on two pots in each tank. Within each tank, one pot in each regime had the vegetation clipped twice during the experimental period. The vegetation in the other pot was left unclipped. The clipping treatment consisted of cutting the vegetation with scissors to 1-2cm above the soil. This clipping treatment was imposed twice, at weeks 6 and 11 of the experiment. When clipping vegetation, wet weights were taken of the vegetation removed to get an estimate of how much biomass was being removed from the pots. Following clipping, the tanks were flushed with water to remove scraps of floating vegetation.
During the 16 week experimental period, minimum and maximum air and water temperatures, and rainfall were monitored weekly.
Species presence and absence, and abundance data were collected at 3-4 week intervals throughout the experimental period. Abundance was estimated by counting number of ramets, where a ramet was defined as a shoot with roots. The following classes were used:
- 1-3 plants,
- 4-10 plants,
- 10-20 plants, and
- >20 plants.
Vegetative and sexual reproduction were noted for each species, and whether the individuals were seedlings or adult plants.
At the conclusion of the experiment, the above ground biomass of the following species, (from a variety of aquatic functional groups) was harvested:
Submerged: Vallisneria gigantea,
Amphibious fluctuation responders: Myriophyllum variifolium, Limosella australis
Amphibious fluctuation tolerators: Centipeda minima, Cyperus sanguinolentus, Isotoma fluviatilis and Lythrum salicaria.
Reproductive and non-reproductive plants of each species were separated. The number of reproductive units for each species was counted. A reproductive unit was defined separately for each species as follows:
Vallisneria gigantea: one flowering stalk, either male or female
Myriophyllum variifolium: one node with either male, female or cosexual flowers
Limosella australis: one flower
Centipeda minima: one inflorescence
Cyperus sanguinolentus: one flowering spike
Isotoma fluviatilis: one flower
Lythrum salicaria: one flower
The reproductive and non-reproductive plants of each species in each pot were placed in separate paper bags. The remaining vegetation (non-target species) was removed and the above ground biomass placed into a paper bag. The bags containing the harvested biomass were dried in an oven for one week at 80 degrees Celsius. The content of each bag was weighed separately, providing dry weights of the reproductive and non-reproductive plants of each target species, and the remaining non-target vegetation, to 4 decimal places.
The data analysed and presented in this report is the biomass harvest data. Analysis of variance was used to test for statistically significant differences in total dry weight, reproductive dry weight, non-reproductive dry weight, and number of reproductive units per pot between water regimes and clipping treatments. Post-hoc pair-wise comparisons were made using Scheffe's Test.
Allocation of biomass to reproduction for selected species
Amphibious fluctuation-tolerator species
There was little difference in Centipeda biomass between clipped and unclipped pots for any of the water regimes. Centipeda grew in four regimes (B, D, F and G), but reproduced only in regime B (permanently damp) (Fig j2). In this regime, approximately 60% of the Centipeda biomass was reproductive for both clipped and unclipped treatments. The number of reproductive units per pot was higher in clipped treatments than in unclipped treatments, although this difference was not statistically significant.
Cyperus grew mainly in regime B (permanently damp), and reproduced in this regime under both clipped and unclipped treatments (Fig j3). It was also present, but did not reproduce, in regime D (dry to flooded). Almost 100% of Cyperus biomass in regime B was reproductive. The reproductive biomass from the unclipped treatments had a significantly higher biomass than for that of the clipped treatments (PCyperus, suggesting that clipping increases the reproductive output of this species.
Isotoma was present in all water regimes, but never reproduced (Fig j4). The highest biomass of Isotoma was found in the clipped treatment, regime D (dry to flooded).
Lythrum was found in all water regimes and clipping treatments, yet was abundant only in the clipped pots in regime G (flooded to damp) (Fig j5). 100% of the biomass in the clipped regime G pots were reproductive. Lythrum was also reproductive in the unclipped regime G pots, and the clipped regime L pots, but these reproductive plants accounted for only 50-60% of the total biomass. Also, the number of reproductive units produced in these treatments was significantly less than the number produced in regime G (clipped) (P
Amphibious fluctuation-responder species
Limosella was present in all regimes, reproducing usually under damp conditions, and only sometimes under fluctuating water levels (Fig j6). Most biomass of Limosella was found in regime G (flooded to damp), with significantly more biomass found in the clipped treatment as opposed to the unclipped treatment (P
Myriophyllum was present in all water regimes and clipping treatments, but reproduced only under a narrower range of conditions (Fig j7). A higher percentage of Myriophyllum biomass was reproductive under fluctuating water regimes (damp-flooded and flooded-damp) than under stable water regimes (permanently damp and permanently flooded). The highest number of reproductive units was produced in regime G (flooded-damp). While there was no statistically significant difference between the clipped and unclipped treatments in this regime, clipping did appear to reduce the number of reproductive units produced.
Vallisneria grew only in flooded regimes, with the amount of biomass increasing with the amount of time the seed bank had been flooded (Fig j8). The unclipped permanently flooded and mimic regimes had the most biomass of Vallisneria, as well as the highest percentage biomass reproductive, and the most number of reproductive units produced. For this species, clipping reduced the amount of reproductive biomass and number of reproductive units produced, although this was not statistically significant for any regime.
General discussion and conclusions
For amphibious species, water regimes incorporating damp conditions (regimes B and G) were the best for establishing vegetation (ie biomass) and reproducing.
Regime G (flooded to damp) was found to be a treatment under which several target species (Lythrum, Limosella and Myriophyllum) were able to establish populations with a high biomass and produce large numbers of reproductive units (relative to other water regimes). Imposing a clipping treatment on species in this regime further increased the biomass and reproductive output of Lythrum and Limosella, but that of Myriophyllum decreased. In regime G, flooded seed bank with relatively few species present was brought out of the water column and exposed to the air. Thus species which colonise rapidly from the seed bank could take advantage of the damp conditions - known to be favorable for establishment (Casanova and Brock 2000) - and lack of competition to germinate, establish and set seed quickly.
In contrast, permanently damp conditions (regime B) better suited Centipeda and Cyperus for producing biomass and reproductive units. These amphibious fluctuation-tolerator species, while being present under some submerged (but fluctuating) conditions, reproduced only when kept permanently damp. Clipping did not appear to either significantly increase or decrease the number of reproductive units.
In comparison to the amphibious species, the submerged species (Vallisneria) was absent (regime B), or present only in low numbers and not reproductive (regime G) under damp conditions. Reproduction of submerged species such as Vallisneria (and thus addition of seeds to the seed bank) requires flooded conditions such as those provided by regimes A (permanently flooded), F (damp to flooded) and L (mimic). Clipping reduced both the amount of biomass and the number of reproductive units produced for this submerged species.
In the interpretation of these results, the limitations of clipping trials in simulating grazing must be taken into account. Clipping simulates only one aspect of grazing, and does not incorporate other effects such as trampling or preferential grazing. These additional aspects of grazing could alter the trends of wetland plant reproduction observed in this tank trial if similar experiments were carried out in a field situation with exclosure plots.
Above ground biomass of Vallisneria gigantea under different water regimes and clipping treatments. (a) Total dry weight, (b) Reproductive dry weight, (c) Non-reproductive dry weight, (d) % dry weight reproductive, (e) Average number of reproductive units.