View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Federation ResearchOnline Research Reports

Leaf litter decomposition and utilisation by macroinvertebrates in a central Victorian river in Australia

JMCK Jayawardana1, 2, Martin Westbrooke1, Michael Wilson3

1School of Science and Engineering, University of Ballarat, Vic 3353, Australia 2Department of Natural Resources, Faculty of Applied Sciences, Sabaragamuwa University Belihuloya, Sri Lanka. Email: [email protected] 3 Murray Darling Basin Commission, GPO Box 409, Canberra, ACT 2601, Australia

Abstract Th is study suggested that long-term persistence of willow roots can provide a positive eff ect on aquatic com- munities until the new native vegetation cover establishes. Leaf type has no signifi cant infl uence over taxa richness; however their availability and persistence in the system has a signifi cant eff ect on macroinvertebrate abundance in these streams. A combination of Phragmites reeds with Leptospermum and Callistemon shrubs in revegetation programs will ensure a persistent supply of leaf litter to macroinvertebrate communities in these streams thus should be retained until native canopy cover eg. Eucalyptus spp. is re-established. (Th e Victorian Naturalist 127 (4) 2010, 104–114) Keywords: Leaf litter, exotic, willow, Phragmites, Eucalyptus Introduction Th e landscapes of central Victoria are largely vol- Redistribution of Phragmites in rivers managed canic in origin and highly modifi ed through clear- by irrigation has been observed in south- eastern ing for agriculture. Exotic willows Salix fragilis Australia (Roberts 2000). However, quantitative have successfully colonised and become natural- studies on the eff ects of this on river health and ised taxa in many riverbanks in these landscapes aquatic biodiversity are limited (Roberts 2000). (Ladson et al. 1997). It is estimated that willows In addition, Phragmites is sometimes used dur- have spread across approximately 30 000 km of ing revegetation, increasing the likelihood it will the 68 000 km river frontage in Victoria (Lad- be a coloniser following willow removal. Native son et al. 1997). Willows are considered a Weed shrubs such as Leptospermum spp. and Calliste- of National Signifi cance in Australia and willow mon spp. are common in many temperate Aus- management strategies recommend their removal tralian streams and are a signifi cant component as the preferred and generally only management in the riparian zone of streams where willows are option (National Weeds Strategy Executive Com- removed. mittee 2000). Revegetation is encouraged follow- A key diff erence between willows and the na- ing removal. Th erefore, the sequence of riparian tive Australian riparian vegetation is the timing changes is: willow lined streams → open streams and quality of leaf fall (Pidgeon and Cairns 1981; → native vegetation lined streams. Decades will Yeates 1994; Frankenberg 1995). It has been be required to achieve a canopy cover of native argued that a change in the timing or quality species equivalent to the preceding willow cover. of litter fall resulting from a change in riparian Phragmites australis is one of the common native vegetation phenology would aff ect stream en- reeds of open streams and wetlands worldwide ergetics and be refl ected in the composition of (Sainty and Jacobs 1981). Th ere is increasing evi- stream biota (Campbell 1993; Schulze and Walk- dence to suggest that riparian disturbance, altera- er 1997). Th ese were attributed to the diff erences tion of natural hydrologic regime, soil salinisation in breakdown rates, palatability and chemical and increased sedimentation favour invasion and composition of leaf types (Schulze and Walker continued spread of Phragmites in lotic systems 1997; Irons et al. 1988; Campbell and Fuchshu- (McNabb and Batterson 1991; Saltonstall 2003). ber 1995). A study by Yeates and Barmuta (1999) Invasion of Phragmites in open reaches with low supported the idea that willow leaves were more riparian vegetation cover has been observed in palatable and preferred by macroinvertebrates many river catchments in Victoria (CALP 1997). than the leaves of Manna Gum Eucalyptus vimi-

104 Th e Victorian Naturalist Research Reports nalis. However, the availability of leaf litter was camaldulensis and native reed Phragmites australis limited during some seasons in willow-lined were investigated. Th e outcome of this study will streams because of the seasonality of litter-fall provide important information for management pattern of willows. It has been found that small of exotic species and revegetation programs. streams that rely entirely on willows for their leaf inputs might run short of coarse particulate food Study site matter during winter (Pidgeon and Cairns 1981; Moorabool River catchment covers approxi- Cummins et al. 1989). Some authors suggest that mately 148 000 ha. Over 75% of the catch- the greater palatability of willow leaves com- ment is used for agriculture, either grazing or pared to native leaves causes a short term boost broad-acre cropping (Department of Water of macroinvertebrates under willow lined chan- Resources Victoria 1990). Th ere are three areas nels during autumn when willows shed most of of mixed-species forest in the middle reaches their leaves (Yeates and Barmuta 1999). Phrag- of the Moorabool River. Willows are commonly mites australis is lightly grazed in the living state, distributed along the riparian zone, particularly and the greatest part of the primary production in the upper catchment. In the middle sections, ultimately enters detrital systems (Imhof 1973). willow removal programs have been under- Mathews and Kowalczewski (1969) and Th omas taken. Th e study site was located near Mor- risons on the Moorabool River (latitude 370 6’ (1970) reported a faster decomposition rate for 0 tree leaves than emergent macrophytes such as S., longitude 144 2’E., altitude 442 m) where Phragmites. No information exists on decom- willow removal has been carried out (Fig. 1). position and macroinvertebrate use of leaves Th is section of the river is in moderate condi- of common shrubs such as Leptospermum and tion according to criteria of ISC classifi cation Callistemon spp. However, these leaves are scle- (Index of Stream Condition) and has peren- rophyllous and frequently aromatic (Walsh and nial fl ow (Department of Sustainability and Entwisle 1996; Lis-Balchin et al. 2000), suggest- Environment 2005). Th e experiment was con- ing a high content of lignin and herbivore deter- ducted from early April (mid autumn) to the rents. Th ese compounds may act as deterrents end of July (mid winter) 2005. Th e stream fl ow for macroinvertebrate and fungal activity lead- remained relatively constant throughout the ing to the slower decomposition of leaves. experimental period, ranging from 0.01 to 0.15 Another important habitat structure remaining m/sec. Channel depth and width ranged from in streams from which willows have been removed 18 cm to 30 cm and 12 to 14 m. Dissolved oxy- are willow root mats, since stumps of willows are gen concentration and pH ranged from 7.4 to oft en left aft er removing the upper parts. Root 7.9 mg/L and 7.4 to 7.9 during the study period. Temperature and conductivity ranged from 6 to decomposition is an oft en ignored, yet potentially 0 important regulator of carbon and nutrient cycling 14 C and 287 to 245 μS/cm respectively. in terrestrial systems (Ostertag and Hobbie 1999). Materials and Methods It has been observed that willow roots remain in For the experiement, fresh willow, Eucalyptus, streams long aft er the trees have been cut down. Phragmites, Leptospermum and Callistemon Th is suggests that willow roots have slow decom- leaves were collected from plants along with position rates but empirical evidence to support fresh willow roots. Aft er collecting, leaves and this observation is scant. It appears the eff ect of leaf willow roots were air dried for one week until a litter input from such exotic and native vegetation constant weight was attained. Roots were rinsed successional processes may have important infl u- well to remove attached mineral particles before ences on the macroinvertebrate communities and drying. Fift een grams of dry leaves from each energy transfer process of these streams. However, leaf type and roots were put into 5 mm mesh no comprehensive study has been carried out to bags separately and tied to nylon lines in a ran- investigate the eff ects of such vegetation changes dom order. Th is mesh size was selected to allow on leaf litter availability and macroinvertebrate macroinvertebrates to enter the mesh bags for communities in Australia. Th erefore, decompo- feeding and also to reduce leaf loss through the sition rates and macroinvertebrate colonisation net. Each nylon line was attached to six mesh of leaves and roots of exotic willow, Salix fragilis, bags containing fi ve types of leaves and willow leaves of native shrub species Leptospermum lani- roots. Th e nylon lines were attached to bricks gerum and Callistemon paludosus, a native ripar- positioned one metre apart, in a homogeneous ian woody tree the River Red Gum Eucalyptus Vol 127 (4) 2010 105 Research Reports

Fig. 1. Location of study sites. section of the stream bed. Th ree replicate sam- was dried for fi ve days until constant weight was ples from each leaf type and root packs were attained then dry weight was recorded. randomly removed aft er 7, 14, 28, 42, 56, 70, 84, Leaf/root weight loss with time 98 and 112 days. Care was taken to avoid losing Decomposition rate of leaves and roots is based bag contents while removing them from the wa- on ‘mass loss of leaves/roots from initial mass’ ter. Water quality parameters were also recorded as defi ned by Hofsten and Edberg (1972). Leaf when removing leaves. Contents of the mesh pack processing as measured by weight loss of bags were put in plastic trays and macroinverte- packs through time was estimated with a sim- brates were separated and preserved in 70% al- ple exponential decay model (Petersen and cohol for later identifi cation. Th e macroinverte- Cummins 1974) as stated in Equation 1. Mean brate taxa present in each sample were identifi ed per cent of leaf weight remaining was used as to the lowest possible taxonomic level (Pinder the dependent variable. and Brinkhurst 1994; Cranston 1996; Dean and Suter 1996; Smith 1996; CSIRO 1999; Hawking Equation 1 and Th eischinger 1999; Gooderham and Tsyrlin W= W e -k t 2002). Information from a number of sources t 0 (i.e. Log (Wt/W0) = -kt) was used to assign the invertebrate fauna to Wis the weight aft er t days, major functional feeding groups, viz shredders, t W0 is the initial weight and collectors, predators, grazers and fi lter feeders k is the decay coeffi cient. (Merritt et al. 1984; Hauer and Lamberti 1996; Gooderham and Tsyrlin 2002) (Appendix I). Th e decay coeffi cient, half-lives of decay of Remaining leaf matter without invertebrates each leaf type and mean weight remaining also were calculated.

106 Th e Victorian Naturalist Research Reports

Macroinvertebrate taxa richness and abun- Macroinvertebrate colonisation, taxa richness dance and abundance A one way between group analysis of covari- Th ere is a consistent trend of increasing num- ance (ANCOVA) was conducted to compare bers of individuals and taxa in each pack type the eff ect of leaf type on total abundance, taxa over the fi rst two months (Fig. 3). In all pack richness and mean leaf weight remaining. Th e types the number of taxa plateaus or declines length of time packs remained in the stream over the subsequent two months. By contrast, was considered as a covariate in the analysis. the number of individuals trends upward over Preliminary checks were conducted to ensure the two month time period in all pack types ex- that there was no violation of the assumptions cept willow leaves. of normality, linearity, homogeneity of vari- For all pack types, the number of taxa ranged ance, homogeneity of regression slope and reli- from ten at seven days to approximately 25 at able measurement of the covariate. In situations the peak in the experimental period. Th e varia- when homogeneity of variance was not met, tion between pack types in the number of indi- data were log transformed before analysis. viduals was greater. With the exception of wil- low leaves, most pack types had between 50 and Macroinvertebrate community composition 85 individuals per pack at day seven. Th e peak Taxa community composition changes among numbers in Leptospermum, Eucalyptus and leaf types were assessed using ANOSIM. Spe- Phragmites leaves were higher (approximately cies responsible for assemblage diff erences 380, 350 and 320 respectively) than other pack were identifi ed by SIMPER routine in PRIM- types; in decreasing order were Callistemon, ER version 5 package (Plymouth Marine willow leaves and willow roots (approximately Laboratory, Plymouth, UK). Th is process al- (275, 225 and 200 respectively). lowed visualising overall percentage contribu- Results of ANCOVA showed no signifi - tion of each taxon to the average Bray-Curtis cant eff ect of leaf type on the taxa richness, dissimilarity between two groups of leaves. F(5,155)=1.27, p=0.28 and a signifi cant eff ect of Data were log transformed (log (x+1)) before leaf types on taxa abundance F (5,155) = 4.88, analysis and Bray-Curtis dissimilarity meas- P = 0.000. Th ere were signifi cant diff erences ure and 999 permutations were conducted. (p<0.05) between the number of individuals Results in Eucalyptus leaves and willow leaves, Lept- Leaf decomposition ospermum leaves with willow roots, and willow Decay coeffi cient values of Eucalyptus, Phrag- leaves and Callistemon leaves. mites and willow roots were in the ‘slow process- Macroinvertebrate community composition ing group’ (<0.005) according to the criteria of ANOSIM analysis indicated a signifi cant dif- Petersen and Cummins (1974) (Table 1). Callis- ference in macroinvertebrate community temon (-k= 0.007 day-1) and willow leaves (-k= composition between diff erent substrate types 0.008 day-1) could be categorised into ‘medium (leaf/root) (Global R = 0.3, p<001). Main taxa processing group’ (0.005-0.010). Leptosper- responsible for discriminating diff erent treat- mum were in the ‘fast processing group’ (0.010- ment groups are given in Table 3. In majority 0.015). Th e percentage of leaf weight remain- of comparisons, three taxa could be used to ing at the end of the experimental period was discriminate leaf/root pairs i.e. Amphipoda: higher for willow roots and Eucalyptus leaves Paramelitidae: Antipodeus sp., Ephemeroptera: (88.98% and 73.64%) (Fig.2). Percentage leaf Caenidae: sp. and Ephemeroptera: Leptophle- weight remaining was lowest for Leptospermum biidae: Atalophlebia australasica. One excep- (19.18%). Phragmites, willow leaves and Callis- tion is in Phragmites and Leptospermum leaf temon were intermediate (50.62%, 32.33% and packs where, Diptera: Chironomidae: Tanypo- 47.03% respectively). Pair wise comparisons dinae predators are responsible for discrimi- of percentage weight remaining indicated the nating groups. Th ere were few other taxa also following order from highest to lowest; willow responsible for community diff erences among roots and Eucalyptus leaves > Phragmites > Cal- leaf packs (Appendix 1). listemon > willow leaves > Leptospermum leaves (Table 2).

Vol 127 (4) 2010 107 Research Reports

Discussion invertebrate communities infl uence the plant Decomposition of leaf litter litter decomposition (Bärlocher 1990; Sridhar When comparing leaves, Eucalyptus and Phrag- and Bärlocher 1993). Th is has also led to dif- mites could be categorised into slow process- ferences in leaf decomposition rates recorded ing categories according to the criteria of Pe- for some species under diff erent geographic or tersen and Cummins (1974). It was previously catchment conditions. Some of the inconsisten- speculated that Phragmites leaf decomposition cies in the decomposition rates of leaves in the is slower than terrestrial leaf litter. However, present and previous studies may have resulted a comparison of half-lives of leaf decay be- from such diff erences. Further, in many studies tween Eucalyptus (533.15 days) and Phragmites senescent leaves were used to compare decom- (169.05 days) showed a slower decomposition position rates, but in the present experiment of Eucalyptus than Phragmites leaves. Th e de- green leaves and fresh roots were used. Among composition rate recorded for Eucalyptus was factors responsible for leaf decomposition rates, consistent with that recorded for River Red chemical composition of leaves is important. Gum by Schulze and Walker (1997). Leaf de- Eucalyptus leaves have high lignin (15-30% composition rate recorded in the present exper- dry weight) (Cork and Pahl 1984), moderately iment for Phragmites was within a similar range high phenolic (16%) and condensed tannin to that recorded for Phragmites by Menendez et contents (Campbell and Fuchshuber 1995). al. (2001) in streams in Spain under a similar Willow leaves have less lignin and low levels of temperature range. Willow leaves were in the condensed tannin (Pasteels and Rowell-Rahier medium category (half life of decay = 88.86 1992) but high levels of other phenols (Binns days). Th is was consistent with the decomposi- et al. 1968). Willow leaves also have a lower C: tion rate of willow leaves recorded by Essafi et N ratio than do Eucalyptus (S. alba C: N =25.4 al. (1994). In contrast, Gessner et al. (1991) and [Chauvet et al. 1993]; E. obliqua 52.5 [Barmuta Schulze and Walker (1997) recorded faster de- 1978]). Th is may aff ect the palatability of these composition rates for willow leaves. Leptosper- leaves for macroinvertebrates and thus lead to mum and Callistemon leaves were expected to diff erences in decomposition rates. be in the slow decomposing category because Leptospermum and Callistemon leaves lost of their antiseptic characteristics (Williams et most of their initial weight at the end of the al. 1993). However, a faster decomposition rate incubation period. It is possible that leaves are for Leptospermum leaves and moderate rate for more vulnerable to disintegration, rapid loss Callistemon leaves was recorded. from the system or utilisation by biota. Th ere- Breakdown and decomposition of leaf litter fore it can be expected that slower decompos- in aquatic ecosystems are complex processes, ing Eucalyptus and Phragmites contribute a per- infl uenced by many factors. Factors such as sistent low level of nutrient input to the system temperature, pH, nutrient (e.g. N, P) supply, compared to willow, Leptospermum and Callis- activity of fungi and bacteria and structure of temon leaves. Decomposition of willow roots As predicted, the decay coeffi cients calculated for roots and leaves in the present study indi- cated that roots have long processing rates. Th e predicted half-life of decay for willow root mats in the present study is 6931 days. However, it can be expected that the root breakdown may start aft er a conditioning period (which is well beyond the present experimental period) thus making a much shorter half-life. Th erefore, it can be expected that extending the incubation period is necessary for the accurate prediction of root decomposition rates. Further studies on chemical composition of roots and leaves are Fig. 2. Percent initial weight (mean ± SE) remaining desirable for accurate prediction of root de- of leaves/roots with time. composition mechanisms. 108 Th e Victorian Naturalist Research Reports

Table 1. Rates of processing of leaf types and roots

Leaf type R2 K (-day) T50 (days) P category

Willow roots 0.08 0.0001 6931.00 Ns Slow Eucalyptus 0.86 0.001 533.15 0.005 Slow Willow leaves 0.92 0.008 88.86 0.002 Medium Phragmites 0.69 0.004 169.05 0.005 Slow Callistemon 0.90 0.007 105.02 0.000 Medium Leptospermum 0.94 0.013 54.15 0.001 Fast

Table 2. Estimated marginal means for: total number of individuals, taxa richness, leaf/root weight remaining (log) in leaf types and willow roots. Treatment Total no. of individuals Taxa richness Log leaf weight remaining

Willow roots 170.70 ± 13 11.53 ± 0.6 1.12 ± 0.03 Eucalyptus 201.07 ± 14 11.53 ± 0.7 1.09 ± 0.02 Willow leaves 141.44 ± 12 10.70 ± 0.5 0.83 ± 0.01 Phragmites 174.93 ± 13 11.78 ± 0.4 0.95 ± 0.03 Callistemon 140.96 ± 15 9.89 ± 0.6 0.93 ± 0.04 Leptospermum 209.19±17 10.00 ± 0.5 0.77 ± 0.02

Table 3. Main taxa responsible for discriminating leaf /root groups based on the SIMPER analysis. Treatment groups Main taxa responsible for discriminating treatment groups

Willow root & Eucalyptus Paramelitidae, Caenidae Willow roots & Willow leaves Paramelitidae, Caenidae Eucalyptus & Willow leaves Paramelitidae, Caenidae willow roots & Phragmites Caenidae , Leptophlebiidae Eucalyptus & Phragmites Caenidae , Leptophlebiidae Willow leaves & Phragmites Caenidae, Leptophlebiidae Willow roots & Callistemon Paramelitidae, Caenidae Eucalyptus & Callistemon Caenidae, Tanypodinae Willow leaves & Callistemon Caenidae, Leptophlebiidae Phragmites & Callistemon Caenidae, Leptophlebiidae Willow roots & Leptospermum Paramelitidae, Caenidae Eucalyptus & Leptospermum Caenidae, Leptophlebiidae Willow leaves & Leptospermum Caenidae, Leptophlebiidae Phragmites & Leptospermum Leptophlebiidae, Tanypodinae, Caenidae Callistemon & Leptospermum Caenidae, Leptophlebiidae

Macroinvertebrate taxa richness and abundance 1974). Time of peak macroinvertebrate coloni- Macroinvertebrate feeding is an important sation of willow leaves observed in the present component in leaf decomposition processes. study is consistent with the recorded values of In the present study, macroinvertebrate abun- Collier and Winterbourn (1986). Even though dance and colonisation followed general trends diff erences in taxa richness and total number reported in other studies (e.g. Collier and Win- of individuals were expected in diff erent leaf terbourn 1986; Schulze and Walker 1997). In types and roots, taxa richness among leaf types all leaf types, macroinvertebrate taxa increased showed no signifi cant diff erences. However, slowly through the fi rst phase of colonisation there was a signifi cant increase in total number and peaked aft er a rapid increase of taxa num- of individuals in Eucalyptus and Leptospermum bers. Th is trend seems to confi rm a ‘condition- leaves compared to other leaf types and willow ing period’ during which pack material is made roots. Th e highest number of individuals was palatable for macroinvertebrates by microbial recorded for Leptospermum leaf packs, fol- activity (Cummins 1974; Petersen and Cummins lowed by Eucalyptus leaves. Even though it was

Vol 127 (4) 2010 109 Research Reports

Fig. 3. Total number of individuals (mean ± SE) and taxa richness associated with each leaf type/ roots with time. expected that willow leaves would harbour a sation of natural and artifi cial leaves by Cortes higher number of individuals, willow and Cal- et al. (1997) proved that use of leaf litter is prin- listemon had a comparatively lower number of cipally determined by its food value rather than individuals. Th is was inconsistent with the fi nd- the microhabitat and substrate. Palatability of ings of Schulze and Walker (1997) who found the leaves is determined by their chemical com- minor diff erences in diversity and composition position, particularly the secondary chemicals of macroinvertebrates associated with willows and the ratio of hard degradable components. and native Eucalyptus in the River Murray. Feeding choice of leaves by macroinvertebrates Th e comparison of macroinvertebrate coloni- is related to lignin, tannin/phenol and fi bre

110 Th e Victorian Naturalist Research Reports content, the C:N ratio and the type of biofi lm their nutritive value. Th ey concluded that in- (Suberkropp et al. 1975; Arsuffi and Suber- vertebrates were then attracted to the packs as kropp 1984; Lester et al. 1994a). Tannin and a refuge rather than for their palatability. Th e other phenolic compounds have been shown to persistence of invertebrate numbers even aft er be negatively associated with invertebrate leaf losing leaves of Leptospermum may also be at- consumption and growth (Irons et al. 1988). tributed to the refuge provided by fi ne branches Higher lignin and moderately high phenolic of Leptospermum. and condensed tannin contents in Eucalyptus leaves may have a negative eff ect on leaf con- Macroinvertebrate colonisation sumption by macroinvertebrates (Campbell Functional feeding group categories indicated that and Fuchshuber 1995). Willow leaves have early colonisers of all leaf pack types were pre- been found to be more palatable to macroin- dominantly shredding detritivores, particularly vertebrates due to less lignin and low levels of Antipodeus sp. (Paramelitidae) and Austrochilto- condensed tannin (Pasteels and Rowell-Rah- nia sp. (Hyalidae). Petersen and Cummins (1974) nen 1992). Glova and Sagar (1994) and Lester postulated the presence of a hierarchy of leaf spe- et al. (1994a,b) reported an enhanced diversity cies along a processing continuum in woodland of invertebrate taxa associated with willows streams. It follows that shredder species would due to contribution of readily processed litter take advantage of a leaf-processing continuum and or to stimulating production through nutrients probably depend upon the continuum for survival. from leaf leachate. Many authors have found a In contrast to that, a study by Bunn (1986) reported positive correlation between biofi lm and inver- that shredders did not respond to the input of leaf tebrate colonisation (Hax and Golladay 1993; litter in northern Jarrah forest streams in Western Schulze and Walker 1997). A study by Schulze Australia. Similarly Benfi eld et al. (1977) showed a and Walker (1997) showed that increased lack of shredders in leaf processing in a pastureland weight of diatoms in eucalypt leaves compared stream in Virginia, USA. Th ey speculated that the to willow leaves aft er eight weeks’ submersion absence of shredders in those streams is related to attracted many invertebrates. It also showed the absence of a functional leaf-processing contin- that Paratya australiensis preferred Eucalyptus uum. However, the present study indicated that, leaves, which are colonised by micro-organisms. in this river, shredding detritivores are important Increased invertebrate numbers were observed in leaf processing, consistent with the postulation on Eucalyptus leaves in the present experiment, of Petersen and Cummins (1974). Th is suggests possibly because of the growth of biofi lm on that, in this stream system, shredding detritivores those leaves. Similarly Hax and Golladay (1993) are an important functional feeding group in the found that the density and richness of inverte- leaf decomposition process. Shredding detriti- brates were correlated with indices of biofi lm vores gradually declined over time and collectors biomass (ATP, ergosterol, Chlorophyll a) and dominated the leaf packs. Th is can be related to concluded that microbial density was higher on increased fi ne particulate organic matter accumu- wood than leaves because of its greater stability lation with leaf decomposition process over time. and surface complexity. It can be expected that However, in willow roots, shredder numbers were a similar response of biofi lm growth in willow relatively stable and continued to be so until the roots may have attracted many invertebrates in end of the study period. Th e number of shredders the present study. in willow roots was signifi cantly higher than in leaf Aft er about 70 days, most Leptospermum leaf litter, and there was a signifi cant reduction in col- packs lost their leaves and only fi ne branches lectors in willow root mats compared to the leaf remained. It was previously predicted that litter. Th is suggests that shredders or detritivores Leptospermum leaves would decompose more may be using roots as refuge sites rather than food, slowly due to their antiseptic qualities. How- or may be because of the long processing time of ever, no decline in total number of individual roots compared to leaves. invertebrates was observed. A similar trend of An interesting fi nding in this experiment is taxa colonisation has been observed for wil- that even though 39 taxa were recorded at the low leaves by Essafi et al. (1994) who indicated sampling site, only three taxa were predomi- no decrease in invertebrate biomass once wil- nantly responsible for discriminating leaf/root low leaves were skeletonised and lost most of types. Th ey were Amphipoda: Paramelitidae: Antipodeus sp., Ephemeroptera: Caenidae: sp. Vol 127 (4) 2010 111 Research Reports and Ephemeroptera: Leptophlebiidae: Atalo- Acknowledgements phlebia australasica. One exception is in Phrag- Th e International Postgraduate Research Scholar- mites and Leptospermum leaf packs, where Dip- ship fi nancially supported this study. We are espe- cially grateful to Josh Cemera for fi eld assistance and tera: Chironomidae: Tanypodinae predators are Wendy Cloke for technical support. We also thank responsible for discriminating groups Th ere the anonymous reviewers whose suggestions greatly were few other taxa also responsible for com- improved this manuscript. munity diff erences among leaf packs. Th is sug- References gests that despite the diff erences in composition Arsuffi TL and Suberkropp K (1984) Leaf processing capabilities of riparian and in-stream leaf litter input to these of aquatic hyphomycetes, interspecifi c diff erences and infl u- streams, relatively fewer taxa of macroinverte- ence on shredder feeding preference. Oikos 42, 144-154. Bärlocher F (1990) Factors that delay colonization of fresh brates can successfully contribute to processing alder leaves by aquatic hyphomycetes. Archiv für Hydro- and energy dynamics of these streams. biologie 119, 249-255. Barmuta LA (1978) Th e decomposition of Eucalyptus obliqua Conclusion leaf packs in small stream ecosystems. (BSc (Hons.)Th esis, De- partment of Zoology, University of Adelaide, South Australia) Th ese results suggest that leaf litter is an im- Benfi eld EF, Jones DS and Patterson MF (1977) Leaf pack portant food source for shredders in this river. processing in a pastureland stream. Oikos 29, 99-103. Leaves of Eucalyptus followed by those of Phrag- Binns WW, Blunden G and Woods DL (1968) Distribution of leucoanthocyanidins, phenolic glycosides and amino acids mites, persisted longer in the river compared to in leaves of Salix spp. Phytochemistry 7, 1577-1581. willow, Callistemon and Leptospermum. Th ese Boulton AJ and Boon PI (1991) A review of methodology persistent leaves provide a continuous supply of used to measure leaf litter decomposition in lotic environ- ments: Time to turn over an old leaf? Australian Journal of food for communities in those streams. Th ese Marine and Freshwater Research 42, 1- 43. results also indicate that leaf type does not af- Bunn SE (1986) Spatial and temporal variation in the mac- fect taxa richness in this river, though it may roinvertebrate fauna of streams of the northern jarrah for- est, Western Australia: functional organisation. Freshwater aff ect the community composition of macroin- Biology 16, 621-632. vertebrates. Willow root mats may also be im- CALP (1997) Corangamite Regional Catchment Strategy. Co- portant because of the long processing life and lac, Corangamite Catchment and Land Protection Board. Campbell IC (1993) Riparian stream linkages: an Australian the refuge provided for macroinvertebrates. perspective on in-stream issues. In Ecology and Manage- Th eir long-term persistence was found to fa- ment of Riparian Zones in Australia, pp. 21-30. Eds SE Bunn, BJ Pusey, and P Price. (LWRRDC: Canberra) cilitate higher invertebrate numbers. Th is sug- Campbell IC and Fuchshuber L (1995) Polyphenols, con- gests that long-term persistence of willow roots densed tannins and processing rates of tropical and tem- in this system, until the new native vegetation perate leaves in an Australian stream. Journal of North American Benthological Society 14, 174-182. cover establishes may have a positive eff ect on Chauvet E, Giani N and Gessner MO (1993) Breakdown and aquatic communities. invertebrate colonisation of leaf litter in two contrasting Th ere are some limitations in this study that streams: signifi cance of oligochaetes in a large river. Cana- dian Journal of Fisheries and Aquatic Sciences 50, 488-495. need to be addressed in any continuation of the Collier KJ and Winterbourn MJ (1986) Processing of wil- study. Th e study was carried out in a single site low leaves in two suburban streams in Christchurch, New of a selected stream because of time and prac- Zealand. New Zealand Journal of Marine and Freshwater Research 20, 575-582. tical constraints. It is desirable for this experi- Cork SJ and Pahl L (1984) Th e possible infl uence of nutritional ment to be replicated in several sites in several factors on diet and habitat selection by the ringtail possum streams, to enable generalisation of the outcome (Pseudocheirus peregrinus). In Possums and Gliders, pp. 269- 276. Eds A Smith and L Hume (Surrey Beatty: Sydney) on a broader scale. Root and leaf decomposition Cortes RM, Abelho M and Rebelo SB (1997) Th e macroin- is complex, involving several processes such as vertebrate colonization of leaf bags: is there a pattern? Lim- netica 13, 71-75. leaching of soluble components, physical frag- Cranston P (1996) Identifi cation Guide to the Chironomidae mentation, microbial catabolism and feeding of New South Wales. AWT Identifi cation Guide No.1. (AWT by invertebrates (Boulton and Boon 1991). In Pty Ltd: NSW, Australia) CSIRO (1999) Interactive Guide to Australian Aquatic Inver- the present experiment, leaf/root decomposi- tebrates. CD ROM. 2 ed (CSIRO Publishing: Australia) tion rates and macroinvertebrate colonisation Cummins KW (1974) Structure and function of stream eco- have been investigated. For a better estimate systems. Bio Science 24, 631-41. Cummins KW, Wilzbach MA, Gates DM, Perry JB and Talia- of the mechanisms leading to decomposition ferro WB (1989) Shredders and riparian vegetation. Bio process of these materials, it is also important Science 39, 24-30. to estimate chemical composition and micro- Dean JC and Suter PJ (1996) Mayfl y nymphs of Australia: A Guide to genera. Identifi cation Guide No.7. (Co-operative bial decomposition processes of these leaves. Research Centre for Freshwater Ecology, Albury, NSW)

112 Th e Victorian Naturalist Research Reports

Department of Sustainability and Environment (2005) Index Merritt RW, Cummins KW and Burton TM (1984) Th e role of Stream Condition: Th e Second Benchmark of Victorian of aquatic insects in the processing and cycling of nutri- River Condition, Melbourne, Victoria. ents. In Th e Ecology of Aquatic Insects pp 134-163. Eds VH Department of Water Resources Victoria (1990). Water Vic- Resh and DM Rosenberg. (Praeger Scientifi c, New York) toria: an Environmental Handbook. (Victorian Govern- National Weeds Strategy Executive Committee (2000) Na- ment Printing Offi ce: North Melbourne, Victoria) tional Weeds Strategy Web Page. [WWW web page] URL Essafi K, Chergui H, Pattee E and Mathieu J (1994) Th e break- http//:www.weeds.org.au. down of dead leaves buried in the sediment of a permanent Ostertag R and Hobbie SE (1999) Early stages of root and stream in Morocco. Archiv für Hydrobiologie 130, 105-112. leaf decomposition in Hawaiian forests: eff ects of nutrient Frankenberg J (1995) Willows: the species, their biology availability. Oecologia 121, 564-573. and control. In. Willows, Weeds and Fish, pp 1-16. Ed R Pasteels JM and Rowell-Rahier M (1992) Th e chemical ecol- Hodges. (Proceedings of River Basin Management Society, ogy of herbivory on Willows. Proceedings of Royal Society Geelong, Victoria) of Edinburgh 98B, 63-73. Gessner MO, Meyer E and Schwoerbel J (1991) Rapid Petersen RC and Cummins KW (1974) Leaf processing in a processing of fresh leaf litter in an upland stream. Verhan- woodland stream. Freshwater Biology 4, 343-368. delungen International Vereinigung für Th eoretische und Pidgeon RWJ and Cairns SC (1981) Decomposition and col- Ange-wendte Limnologies 24, 1846-1850. onisation by invertebrates of native and exotic leaf material Glova GJ and Sagar PM (1994) Comparison of fi sh and mac- in a small stream in New England (Australia). Hydrobio- roinvertebrate standing stocks in relation to riparian wil- logia 77, 113-127. lows (Salix sp.) in three New Zealand streams. New Zealand Pinder AM and Brinkhurst RO (1994) A Preliminary Guide Journal of Marine and Freshwater Research 28, 255-266. to the Identifi cation of the Microdrile Ologochaeta of Aus- Gooderham J and Tsyrlin E (2002) Th e Waterbug Book: A tralian Inland Waters. Identifi cation Guide No.1. (CRCFE: Guide to the Freshwater Macroinvertebrates of Temperate Albury, NSW) Australia.( CSIRO Publishing: Australia) Read MG and Barmuta LA (1999) Comparison of benthic Hauer FR. and Lamberti GA (1996) Methods in stream ecol- communities adjacent to riparian native eucalypt and intro- ogy. (Academic Press: San Diego, California) duced willow vegetation. Freshwater Biology 42, 359-374. Hawking JH and Th eischinger G (1999) Dragonfl y Larvae Roberts J (2000) Changes in Phragmites australis in south (Odonata): A guide to the Identifi cation of Larvae of Aus- eastern Australia: A habitat assessment. Folia Geobotanica tralian Families and to the Identifi cation and Ecology of Lar- and Phytotaxonomica 35, 353-362. vae from New South Wales. AWT Identifi cation Guide No.3 Sainty GR and Jacobs SWL (1981) Waterplants of New South and CRCFE Identifi cation Guide No. 24. (CRCFE: Albury Wales. Water Resource Commission New South Wales, and AWT Pty Ltd: West Ryde, Australia) Sydney. Hax CL and Golladay SW (1993) Macroinvertebrate colo- Saltonstall K (2003) Recent Research on Phragmites australis nisation and biofi lm development on leaves and wood in a in North America: Implications for management, Univer- boreal river. Freshwater Biology 29, 79-87. sity of Maryland Center for Environmental Science. Pro- Hofsten BV and Edberg N (1972) Estimating the rate of deg- ceedings of the Aquatic Invaders of the Delaware Estuary radation of cellulose fi bres in water. Oikos 23, 29-34. Symposium, Malvern, Pennsylvania, May 20, 2003 pp. Imhof G (1973) Aspects of energy fl ow by diff erent food 12-15. chains in a reed bed. A review. Polshie Archives für Hydro- Schulze DJ and Walker KF (1997) Riparian willow and euca- biologie 20, 165-168. lypt trees and their interactions with littoral invertebrates Irons JG, Oswood MW and Bryant JP (1988) Consumption in the River Murray, South Australia. Regulated Rivers: Re- of leaf detritus by a stream shredder: infl uence of tree spe- search and Management 13, 557-77. cies and nutrient status. Hydrobiologia 160, 53-61. Smith BJ (1996) Identifi cation Keys to the Families and Genera Ladson A, Gerrish G, Garr G and Th exton E (1997) Willows of Bivalve and Gastropod Molluscs found in Australian Inland along Victorian Waterways. Towards a management strat- Waters. Identifi cation Guide No. 6. (CRCFE: Albury, Australia) egy. Waterways Unit. Department of Natural Resources Sridhar KR and Bärlocher F (1993) Seasonal changes in mi- and Environment. crobial colonization of fresh and dried leaves. Archiv für Lester PJ, Mitchell SF and Scott D (1994a) Eff ects of riparian Hydrobiologie 128, 1-12. willow trees (Salix fragilis) on macroinvertebrate densities Suberkropp K, Klug MJ and Cummins KW (1975) Com- in two small Central Otago, New Zealand, streams. New munity processing of leaf litter in a woodland stream. In- Zealand Journal of Marine and Freshwater Research 28, ternationale Vereinigung für Th eoretische und Angewandte 267-276. Limnologie 19, 1653-1658. Lester PJ, Mitchell SF and Scott D (1994b) Willow leaf and Th omas WA (1970) Weight and calcium losses from decom- periphyton chemical composition, and feeding preference posing tree leaves on land and water. Journal of Applied of Olinga feredayi (Trichoptera, Conoesucidae). New Zea- Ecology 7, 237-241. land Journal of Marine and Freshwater Research 28, 13-18. Walsh NG and Entwisle TJ (1996) Flora of Victoria (vol. 3). Lis-Balchin MS, Hart L and Deans SG (2000) Pharmaco- (Inkata Press: Melbourne) logical and antimicrobial studies on diff erent tea-tree Williams LR, Home VN and Lusanzi I (1993) An evalua- oils (Melaleuca alternifolia, Leptospermum scoparium or tion of the contribution of 1,8-cineole and terpinen-4-ol to Manuka and Kunzea ericoides or Kanuka), originating in the overall antimicrobial activity of tea tree oil. Cosmetics, Australia and New Zealand. Phytotherapy Research 14, Aerosols and Toiletries in Australia 7, 25-28. 623-629. Yeates LV (1994) Th e role of invertebrates and microorgan- Mathews CP and Kowalczewski A (1969) Th e disappearance isms in the breakdown of willow and eucalypt leaves in a of leaf litter and its contribution to production in River Tasmanian stream. (B Sc (Hons) Th esis, University of Tas- Th ames. Journal of Ecology 57, 543-552. mania) Mc Nabb CD and Batterson TR (1991) Occurrence of the Yeates LV, and Barmuta L (1999) Th e eff ects of willow and common reed Phragmites australis along road side in Low- Eucalyptus leaves on feeding preference and growth of er Michigan. Michigan Academician 23, 211-220. some Australian aquatic macroinvertebrates. Australian Menendez M, Martinez M, Hernandez O and Comin FA Journal of Ecology 24, 593-598. (2001) Comparison of leaf decomposition in two Mediter- ranean rivers: a large eutrophic river and an oligotrophic stream. International Review of Hydrobiology 86, 475-486. Received 17 September 2009; accepted 4 April 2010

Vol 127 (4) 2010 113 Research Reports

Appendix 1. Macroinvertebrate Functional Feeding Group assignment Class/ Order Family Species Assigned FFG

Amphipoda Hyalidae Austrochiltonia sp. shredder Paramelitidae Antipodeus sp. shredder Isopoda Janiridae Heterias sp. shredder Decapada Atyidae Paratya australiensis shredder Plecoptera Gripopterygidae Leptoperla sp. shredder/grazer Trichoptera Atriplectididae Atriplectides dubios collector Calamoceratidae Anisocentropus sp. shredder Ecnomidae Ecnomus sp. predator Hydropsychidae Cheumatopsyche sp. collector Leptoceridae Oecetis sp. predator Leptoceridae Triplectides sp. shredder Leptoceridae Triplectides volda shredder Limnephilidae Archaeophylax sp. shredder Odontoceridae Marilia sp. collector Coleoptera Elmidae Austrolimnius sp. collector Hydrophilidae Berosus sp. predator Psephenidae Sclerocyphon sp. grazer Hemiptera Corixidae Micronecta sp. predator Diptera Ceratopogonidae sp. collector Chironominae sp. collector Orthocladiinae sp. grazer/collector Tanypodinae sp. predator Odonata Coenagrionidae Ischnura sp. predator Corduliidae sp. predator Isostictidae sp. predator Isostictidae Austrosticta sp. predator Ephemeroptera Caenidae sp. collector Caenidae Tasmanocoenis sp. collector Leptophlebiidae Atalophlebia australis collector Leptophlebiidae Atalophlebia australasica collector Gastropoda Ancylidae Ferrissia sp. grazer Hydrobiidae Potamopyrgus antipodarum grazer Planorbidae Glyptophysa sp. grazer Physidae Physa acuta grazer Planorbidae Gyraulus sp. grazer Bivalvia Corbiculidae Corbicula sp. fi lter feeder Hirudinea Glossiphoniidae sp. predator Oligochaeta Lumbriculidae Lumbriculus variegatus collector Phreodrilidae Antarctodrilus proboscidea collector

Beetles (including weevils) on footpath. Photos by Virgil Hubregtse. See article on page 155. 114 Th e Victorian Naturalist