Initial decomposition of floating leaf blades of waterlilies: causes, damage types and impacts

Sites

Field research took place in dense, nearly mono-specific stands of waterlilies in three different water bodies located in The Netherlands: Haarsteegse Wiel (HW), Oude Waal (OW) and Voorste Goorven (VG) (Table 1). Three plots were laid out in stands of Nuphar lutea (L.) Sm. (HW and OW in 1977; VG in 1988), two plots in stands of Nymphaea alba L. (OW in 1977; VG in 1988) and one plot in a stand of Nymphaea candida C. Presl in J. et C. Presl (HW in 1977). The plots were accessed with a small zodiac, which was navigated by gently paddling. Otherwise no boating or navigation occurred in the water bodies, which prevented damage of the plants by propellers.

Haarsteegse Wiel, located in the Province of Noord-Brabant, originates from two connected ponds created by dike bursts along the River Meuse. The now isolated water body is eutrophic and has a relatively low alkalinity. The water level fluctuates, depending on precipitation, groundwater seepage and evaporation. Stratification of the water column occurs during summer. The lake bottom consists of sand and an organic layer with increasing thickness towards the littoral zone. The waterlily beds are situated in the wind-sheltered part of the lake.

Oude Waal in the Province of Gelderland is a highly eutrophic oxbow lake in the forelands of the River Waal. Depth during the growing season is shallow, except for three connected breakthrough ponds. The water level is dependent on precipitation, groundwater seepage, overflow of the River Waal in winter or spring, which strongly influences water chemistry and quality, and evaporation. The bottom consisting of clay and sand is covered by an organic layer of varying thickness in the nymphaeid beds.

Finally, Voorste Goorven in the Province of Noord-Brabant is a shallow, oligotrophic, isolated, culturally acidified moorland pond with very low alkalinity. It is surrounded by forests stocking on poorly buffered sandy soils. The hydrology is mainly dependent on precipitation, groundwater seepage and evaporation.

Leaf area

The potential and actual leaf areas were determined to quantify leaf area loss. The potential area refers to the area of the intact leaf. The actual area was defined as the potential area minus the area that was missing. The potential leaf area was calculated by using a quadratic regression to relate it to leaf length (Van der Velde & Peelen-Bexkens, 1983; Klok & Van der Velde, 2017; Table 2). Specifically, undamaged, fully green floating leaves randomly sampled outside the plots were taken to the laboratory where both length and area were measured to establish relationships of the form: (1)${\displaystyle A\left(L\right){=c}_i{L}^2}$

where:

A(L) = potential leaf area at length L (cm²)

L = leaf length from the leaf tip to a basal lobe tip (cm)

c_i = regression coefficient of species i

i = species (Nuphar lutea, Nymphaea alba, Nymphaea candida)

Study design and data collection

Six representative plots of 1 m² were laid out in the center of mono-specific stands, each containing one rhizome apex per plot. A non-destructive method was used to tag all floating leaves individually within the plots (Klok & Van der Velde, 2017). Newly unrolled leaves were tagged with uniquely numbered Rotex tape fixed around the petiole just under the leaf blade. This enabled us to collect data during the full life-span of the leaves. Each plot was bordered by a square perforated PVC tube frame, held approximately 15 cm below the water surface by cork floaters and anchored to the bottom by four bricks. This set-up does not affect the unrolling of floating leaves in the plots. All leaves having their petioles within the frame were counted and measured. A leaf was considered present as long as, after partial degradation and disintegration, tissue of the lamina was connected to the petiole in the case of OW and HW. In VG a leaf was considered ‘gone’ when it was completely brown, dead and submerged, or when it had disappeared.

All leaves within the plots were inspected and measured at weekly intervals during the growing season, typically from April until November. Site visits involved tagging newly unrolled leaves, counting the number of leaves, measuring leaf length from the leaf tip to one of the basal lobe tips and visually estimating different types of initial decomposition expressed as percentage of the potential leaf area of each leaf. Leaves showing several types of damage were harvested outside the plots to be photographed in the laboratory.

Causes and impacts of initial decomposition

The causes of damage classified in the present study are senescence, frost, hailstones, dehydration, mechanical damage, bird scratches, feeding waterfowl (Fulica atra L. and Gallinula chloropus L., Rallidae), pond snails (Lymnaea sp., Lymnaeidae, Gastropoda), water-lily reed beetle (D. crassipes F., Chrysomelidae, Coleoptera), adults and larvae of the water-lily leaf beetle (Galerucella nymphaeae L., Chrysomelidae, Coleoptera), a weevil (Bagous rotundicollis Bohemann, Curculionidae, Coleoptera), larvae of the aquatic moth brown china mark (Elophila nymphaeata (L.), Crambidae, Lepidoptera), larvae of a dung fly (Hydromyza livens (Fabricius), Scathophagidae, Diptera), chironomid larvae (Chironomidae, Diptera), including Endochironomus spp. and T. intextus (Walker), a phytopathogenic fungus (Colletotrichum nymphaeae (Pass.) Aa) and an oomycete (Pythium sp.), and finally microbial decay (Fig. 3, Table 4). In some cases, specific causes could not be identified.

Frost

Frost in early spring can damage the tips of young leaves sticking out of the water. As a result, such leaves can lose up to one third of their area (Fig. 5). However, the effect on the total leaf surface area was less than 5%.

Bird scratches

Scratches are often caused by the claws of birds, mostly coots (F. atra) but also the common moorhen (G. chloropus), as they walk or run over the leaves (Fig. 8). In general, the scratches are straight and affect only the epidermis of the leaf, but angle-shaped cuts due to claws penetrating the leaf tissue also occur. The affected leaf surface area was low, generally below 5%, although a high percentage of leaves was affected, sometimes up to 100% for all plots at HW and OW. In contrast, the plots at VG showed no scratches.

Consumption by pond snails

A major cause of damage on fresh leaf tissue is caused mainly by Lymnaea stagnalis L., to a lesser extent also by other lymnaeids. Pond snails consume folded leaves still under water. Rows of holes can then be seen in the unrolled leaf blades, large near the edge and smaller towards the center of the leaf (Fig. 10). Lymnaeid and other freshwater pulmonate snails show a preference for decaying leaf material, such as areas infected by fungi. Damage by snails was generally an important cause of damage during the whole period for both Nuphar lutea and Nymphaea alba, contributing up to 20% to the total leaf area loss in HW.

Consumption by water-lily reed beetles

Both Nuphar and Nymphaea spp. are host plants of the water-lily reed beetle D. crassipes. The adult beetles live on the upper side of floating leaves where they feed on leaf tissue (upper epidermis, parenchyma and lower epidermis). The leaf areas removed as a result of tissue consumption by these beetles are round to oval. Eggs are deposited in two or three rows on the leaf underside. To this end, the beetle gnaws a round or oval hole in the leaf, then sticks its abdomen through the hole to reach the leaf underside and oviposit (Fig. 11). The percentage of leaf area damaged by reed beetles was minimal.

Consumption by water-lily leaf beetles

The water-lily leaf beetle (G. nymphaeae) completes its full life cycle on the upper surface of floating leaves. Both adult beetles and larvae feed on the upper epidermis and palisade and sponge parenchyma. The larvae create irregular trenches on the surface, leaving the lower epidermis of the leaf intact and depositing their faeces in the trenches. The resulting pattern of leaf tissue damage is easily recognized. The adult beetles consume smaller areas (Fig. 12). Damage was only found in N. alba at VG, where leaves started to be affected in mid-June, rising to 30–40% between August and October and reaching a sharp peak of 60% in mid-October. The percentage of lost leaf area was minimal.

Consumption by weevils

The adults of B. rotundicollis scrape off areas of leaf tissue (ca. 1 cm diameter) from the underside of floating leaves near the margin. Only the lower epidermis and sponge parenchyma are consumed, whereas the palisade parenchyma and upper epidermis remain intact (Fig. 13). Damage by weevils was found only in N. alba at VG, with up to 30% of these leaves being affected. Leaf area loss was minimal.

Consumption and damage by the brown china mark

The caterpillar of the aquatic moth E. nymphaeata damages floating leaves in two ways, by leaf tissue consumption and by cutting out oval leaf patches that they can attach to the underside of a floating leaf to make a shelter. They can also spin two patches together to construct a floating shelter (Fig. 14). The effect of these activities on leaf surface area was low, at most 5%. Nymphaea candida at HW, Nuphar lutea at VG and Nymphaea alba at VG were not damaged by the moth.

Mining by a dung fly

Larvae of the dung fly Hydromyza livens only occurred in Nuphar leaves, where they mine and consume leaf tissue. Eggs are laid at the underside of the leaves. For that purpose the fly goes underwater, following the dichotomous veins on the underside of the leaves till it reaches the midrib to lay an egg. The newly hatched larvae immediately start to mine the leaf tissue. The mine track has a characteristic shape as the larvae first move from the midrib towards the margins of the leaf, then turn to continue mining in parallel to the leaf margin, then turn again towards the midrib and mine further into the petiole where they pupate. This creates a breaking point where the leaf blade can detach and float away (Fig. 15). Overall, the effect of dung flies mining the leaves was less than 8%.

Mining by chironomids

Larvae of some Chironomidae mine their way through the leaf tissue by consuming particular tissue layers while leaving the upper and lower epidermis unaffected for protection. Typical damage on Nuphar leaves is caused by larvae of T. intextus. These larvae mine leaves still folded underwater, resulting in rows of small holes that become visible when the floating leaves unroll at the water surface (Fig. 16). Also observed at the study sites were larvae of C. trifasciatus (Meigen) (Fig. 17), which makes an open mine by removing the upper epidermis while leaving the lower epidermis intact. The species was observed in OW to cause some damage at the leaf margins of Nuphar lutea in the neighbourhood of Nymphoides peltata (Gmel.) O. Kuntze, its main food plant. Overall, however, the impact of these chironomid species on floating leaves was minimal.

Microbial decay

The resistance of a leaf to microbial infection quickly disappears during senescence, facilitating microbial decay (Fig. 4) as indicated by a change in leaf colour from yellow to brown and the softening of leaf tissue by maceration. The affected surface area rose to 15–25%, with an exceptional extent of 60% reached in Nymphaea candida at HW at the very end of the growing season.

Senescence

Newly unrolled leaf blades of waterlilies are fully green and hydrophobic due to a thick epicuticular wax layer. This waxy layer gradually erodes during senescence and as cellulolytic and other bacteria and fungi colonize the leaf tissue (Howard-Williams, Davies & Cross, 1978; Robb et al., 1979; Rogers & Breen, 1981; Barnabas, 1992). Senescence starts shortly after the first leaves are fully grown and continues throughout the growth period. During senescence, an orderly physiological process controlled by the plant itself, the leaves turn from green to yellow, and ultimately to brown. Concomitant microbial decay softens the leaves.

Infection by phytopathogens and microbial decay

In Nuphar both microbial decay and infection by the phytopathogenic oomycete Pythium sp. “type F” were important from the start of the season. In Nymphaea, infection by the phytopathogenic fungus Colletotrichum nymphaeae also started early and increased in importance towards the end of the season. In general, microbial decay and phytopathogenic infection gradually increased in importance, whereas most other causes of damage diminished over time.

Weather conditions

Minor causes of leaf impairment occurring once during spring were frost damage of the first newly unrolled leaves and hailstones. Hailstones hardly caused leaf area loss. High solar radiation and air temperature dehydrated leaves that had been flipped over, with the impact being high in HW and OW but not in the wind-sheltered VG. Prolonged cloudy and wet weather imposes stress on waterlilies by weakening the defense of leaves due to reduced solar radiation, and thus promoting heavy infection and damage by phytopathogens (Van der Aa, 1978). One mechanism is that poor light conditions reduce the content of phenolics with fungistatic properties in the leaf tissue (Vergeer & Van der Velde, 1997), which turns mature leaves vulnerable to infection.

Damage by animals

Causes of damage by insects were similar for Nymphaea and Nuphar with the exception of Hydromyza livens and T. intextus, which appear to be specific for Nuphar (Brock & Van der Velde, 1983; Van der Velde & Hiddink, 1987). Some species such as B. rotundicollis (Van der Velde, Kok & Van Vorstenbosch, 1989) and D. crassipes (Gaevskaya, 1969) exclusively feed on Nymphaeaceae. Other species such as Galerucella nymphaeae and Elophila nymphaeae feed on both floating-leaved and emergent macrophytes (Gaevskaya, 1969; Lammens & Van der Velde, 1978; Pappers et al., 2001). C. trifasciatus primarily causes damage on leaves of Nymphoides peltata (Lammens & Van der Velde, 1978) but was also observed to damage nearby Nuphar lutea leaves (Van der Velde & Hiddink, 1987).

In VG, damage was mainly caused by phytophagous insects consuming floating leaf tissue, particularly herbivorous beetles, fly larvae and mining chironomid larvae. Leaf disintegration was hardly observed in the acidic VG, which was also the site most sheltered against wind and wave action by a surrounding forest. Protection from wind and wave action allowed the water-lily leaf beetle Galerucella nymphaeae to cause extensive damage, because the wind blows them from the leaves and they float away as a result of wave action. Although this species spares the lower epidermis of their tracks, the epidermis becomes vulnerable to microbial attack and thus disappears at a later stage (Wesenberg-Lund, 1943; Roweck, 1988). As observed in the present study, the minor leaf area loss by the beetle and its larvae is succeeded by damage caused by fungi, oomycetes or bacteria (Wallace & O’Hop, 1985). The damaged areas characterized by regular margins made by adult Galerucella nymphaeae are distinct from those made by adult D. crassipes where the margins of damaged areas are rather irregular (Roweck, 1988). Galerucella nymphaeae was absent in the two water bodies frequently exposed to strong wind.

Consumption by snails was restricted to the two hardwater lakes, since they require calcium to build their shells. Snails at those sites prefer consuming microbially colonized, decaying parts of the leaves (Kok, 1993).

Nymphaea candida (HW) showed an increase in scratches by bird claws towards the end of June, which may have been due to young coots. High densities of waterfowl at HW and OW are facilitated by the surrounding meadows where birds graze during winter.

pH and alkalinity

Decomposition of leaves was slowed down at the acidic site (VG). Such water bodies are characterized by a very low alkalinity and high Al concentrations of the water, as well as low pH (Leuven, Van der Velde & Kersten, 1992). A laboratory study in chemostats with synthetic media showed that pH, Al and HCO₃⁻ concentrations clearly influence the decomposition and chemical composition of leaf blades of floating-leaf plants, with low pH and elevated Al concentrations inhibiting and high bicarbonate concentrations (alkalinity) stimulating decomposition (Kok, Meesters & Kempers, 1990). Al is toxic to microorganisms and low pH slows down leaf disintegration by inhibiting cell-wall degradation by microbial pectin-degrading exoenzymes and xylanase (Kok & Van der Velde, 1991). At low pH, tannins accumulate in the slowly decomposing leaf material, microbial colonization is inhibited and maceration of the leaf tissue is reduced, resulting in a low-quality food resource for detritivores (Kok et al., 1992). The occurrence of detritivores is also inhibited by high Al concentrations and low pH (Kok & Van der Velde, 1994). Finally, fungal degradation of major groups of structural carbohydrates is inhibited by low pH (Kok, Haverkamp & Van der Aa, 1992).

Harvested fresh and decaying leaf blades of Nymphaea alba placed in litter bags in the field and in the laboratory showed lower leaf area loss under acidic conditions in a moorland pond (VG) than in a eutrophic, hardwater oxbow lake (OW), and results under laboratory conditions mimicking differences in water chemistry were similar (Brock, Boon & Paffen, 1985). Depending on water chemistry, mass loss was pronounced and organic matter chemical composition changed rapidly during the first 10–30 days, followed by an accumulation of structural plant polymers such as cellulose, hemicellulose and lignin. The disappearance of those fractions was dependent on the water quality of the water body (Brock, Boon & Paffen, 1985).

In conclusion, the present study shows that the decomposition pattern of Nuphar lutea was similar in the two hardwater lakes, and differed from those of Nymphaea alba and N. candida. In the acidic VG, the effect of leaf damage on leaf area loss was minimal for both Nuphar lutea and Nymphaea alba.