The Effects of Disturbances on Litter Processing of Native and Non-Native Riparian Species in Tropical Headwater Streams

   

Covich, Alan P.

Fishery and Wildlife Biology, Colorado State University, Ft. Collins, CO 80523 U.S.A.

 

 

Abstract

 

The frequency and intensity of disturbances (associated with hurricanes, floods, and droughts)can determine when different detrital food supplies are available to particular species in stream food webs.  Long-term phenological patterns of leaf and fruit fall among native and non-native riparian species provide spatially and temporally heterogeneous sources of alternative foods for detritivores.  The loss of native species and replacement by non-native species may have unexpected consequences for headwater stream food webs.   Species of freshwater shrimps, crabs, insects, and gastropods are known to consume a wide range of litter inputs but how these food webs function under changing climatic and land-use conditions is unknown, especially in tropical streams.

Recent field studies demonstrate that some Asian species such as bamboo (Bambusa vulgaris ) and Java plum (Syzigium jambos) are well established along relatively steep headwater streams in Hawaii and Puerto Rico.  These non-native riparian trees can spread into disturbed riparian habitats following storm-flow events (associated with hurricanes).  Although invasive, non-native species may provide some of the same resources and ecosystem functions as native species (e.g.,  leaf litter and fruit fall), the non-native resources may not substitute completely for resources supplied by native species.   In one example, riparian bamboo in the Luquillo Experimental Forest, Puerto Rico provides leaf-litter input that serves as important microhabitat for species of freshwater shrimp.  Bamboo alone may not, however, provide an adequate supply of detrital food resources for all species of detritivores.   In some streams, the input of ripe fruit from trees such as Java plum provides a major source of detrital food resources, especially during periods when fruit fall from native species of palms may be limited.  Native riparian species such as Cecropia schreveriana  and Prestoea montana are commonly distributed  in the Luquillo Experimental Forest with both species being early colonizers of steep slopes and eroded stream banks.  After tropical storms with high winds, the large leaves and fronds from these native riparian trees provide important inputs of leaf litter to the stream food web,.  The palm fruits are also important food resources for freshwater crabs and other consumers.  Recent experimental results illustrate some important species-specific linkages between freshwater shrimp and leaf-litter processing.  Current studies are examining how the phenology of litter inputs from these native species differ from those of non-native species and how disturbance frequency may alter these relationships.

 

 

INTRODUCTION

 

Long- term data are needed to determine how freshwater communities change over time as a result of their responses to both natural and cultural disturbances (Lodge et al. 1997).   Food-web responses to different types of disturbances (such as floods, droughts, and nutrient loading) are difficult to predict, especially  in small headwater streams that drain forested catchments where the diversity of riparian inputs can be complex (Covich 1988a, Covich et al. 1999).  The consequences of changes in benthic species distributions are often due to complex connections between numerous sediment-dwelling species and the flow of energy from detrital inputs to associated food webs(Gregory et al. 1991, Dudgeon 1994, Lester et al. 1994,  Friberg and Winterbourn 1997, Naiman and Decamps 1997). 

It is essential to set up comparative long-term studies in different locations to understand how the intensity and frequency of different disturbances alter native riparian species distributions and their effects on benthic species and detrital processing.  Networks of sites can evaluate how natural communities respond  over time to different sets of conditions in temperate and tropical ecosystems.  Some existing sites have spatially specific data on litter production in forest plots where comparisons among tree species can be made along well-mapped riparian zones.  

 

A Tropical Case Study:  Luquillo Experimental Forest, Puerto Rico

Long-term data at landscape scales are needed to evaluate riparian detrital inputs and to determine phenological patterns of different energy sources for detrital-based food webs.  Results from on-going, long-term studies of riparian tree species and stream food webs in the Luquillo Experimental Forest (Caribbean National Forest),  Puerto Rico provide an example of forest-stream connections in a tropical montane ecosystems,  The forest contains 225 tree species and covers steep terrain consisting of volcanoclastic sandstones with a network of five rivers draining 11,000 ha. 

Near the El Verde Field Station, a 16 ha forest grid has two streams flowing through the mapped grid (400 20 m x 20 m plots with 5 m x 5 m subplots).  The grid was surveyed in 1990 and has 91 tree species with over 4,400 identified and mapped trees.  Within 5 m of the streams only a few species dominated the riparian forest community.  Riparian species differed in dominance  between  the two streams as a consequence of past land use more than 50 years ago (Reed 1998).  The most common species included: Prestoea montana; Dacroydes excelsa; Tabebuia heterophylla; Guettarda valenzuela; Coccoloba swartzii; Manilkara bidentata; Sapium laurocerasus; Homalium racemosum; Inga laurina; and Caseria sylvestris.  This grid and its riparian zones are being compared with several other riparian areas outside the grid to evaluate the functional roles of native and nonnative riparian tree species and their effects on benthic food webs.  At other sites, especially at lower elevations around the boundaries of the forest reserve, non-native species such as bamboo and Java plum dominate the riparian forest species composition.

 

Disturbances and Riparian Tree Species

At global and regional scales there is a link between the intensity and frequency of hydrological events and various climatic changes such as El Nino-based droughts and floods or hurricanes (Covich et al. 1991, 1996).  Land-use changes further alter patterns of runoff at the catchment scale during periods of variable precipitation (Covich et al. 1998, Covich et al. In press).  These ecosystem drivers result in changes in tree species distributions and their associated riparian functions.  There are close linkages between riparian trees and associated stream fauna so that any change in the riparian community is likely to alter stream food webs.   The context-dependent function of native freshwater species is often influenced by the invasion of non-native species that alter freshwater ecosystem functions (Covich 1993, Covich et al. 1999).  These invasive freshwater species may be associated  with the geographical range expansions of non-native riparian trees.

Species distributions of native riparian trees often persist even when they are exposed to high winds from hurricanes or bank erosion resulting from flood events because native species are generally well adapted to these local conditions.  However, invasions by non-native riparian species or detritivores  do occur and these new species can sometimes short circuit the energy flow through detrital-based food webs.  A disconnect between the rates of leaf-litter and fruit production and the rates of consumption and processing by native freshwater species can greatly alter energy flow.   Thus, if non-native tree species become established along disturbed stream corridors they may alter habitat quality for those species living in the stream in ways that enhance invasion by non-native consumer species.  Although many introductions are transient, the spread of some non-native riparian species and benthic invertebrates can be invasive and persistent.

Because studies on roles of single species in stream ecosystems are generally lacking (Heard and Richardson 1995, Covich 1996, Covich et al. 1999),  some insight can be gained by documenting how range extensions of riparian and/or benthic species into new habitats may alter ecosystem processing such as rates of decomposition.   Long-term monitoring of ecosystem-level consequences following riparian species introductions is needed to document how interconnected these species may be.   Introductions are actively discouraged unless a habitat is so highly degraded that it cannot be recolonized by native species (Lugo 1994, 1997, Ewel et al. 1999).  Historically,  it is important to determine if native riparian or benthic specieswere lost and  if ecosystem processes were affected.  Often “successful” invaders have life history attributes that predispose them to have large impacts on ecosystems and displace well adapted, native species.  

Complex linkages between invasive species of riparian trees and non-native species of detritivores may lead to a series of long-term changes in stream benthic community composition.  For example, if the sources of riparian detritus change because of a shift in tree species and/or shifts in the species of detritivores, then there may be an accumulation of organic detritus.  This buildup of organic debris will result in a decline in dissolved oxygen during drought periods when slow stream flows lower export of riparian leaf litter and fruit fall.   As another illustration, shade from riparian tress is an important  influence on water temperatures so that displacement of evergreen species by seasonally deciduous species can increase stream water temperatures.  Similarly, removal of an intact riparian tree canopy and replacement by low-growing herbaceous species can greatly increase water temperatures and alter detrital inputs of woody debris, leaf litter, fruit fall, and terrestrial insects. 

 

Benthic Detritivores and Frugivores    

Benthic invertebrates are known to be functionally important in many ecosystems (Hutchinson 1993, Wallace and Webster 1996, Covich et al. 1999).   These bottom-dwelling species are often diverse and abundant in freshwater sediments .  They consume a wide variety of food resources from both internally produced plants (algae and macrophytes) and externally produced organic matter (leaf litter, fruit fall, woody debris, and terrestrial insects) from the overhanging riparian canopy.  Certain species of aquatic insects and decapod crustaceans use specialized mouthparts or feeding appendages to break up large pieces of organic detritus such as leaf litter into smaller fragments.  In the process of feeding, some shredded and suspended fragments are transported downstream (along with fecal pellets) in small, headwater tributaries.  Other species are specialized to filter out variously sized particles and are often located downstream of the shredders (Anderson and Cargill 1987, Wallace and Webster 1996, Wallace et al. 1997).  Such linkages suggest that loss of some pivotal species such as shredders would alter food availability for suspension feeders and thereby alter ecosystem processing of detrital carbon.   

Much less is known about the importance of fruit fall in detrital energy budgets.  For example, in the Luquillo Experimental Forest, Puerto Rico, fruit fall is a major source of energy that varies with elevation and forest type.  On average 600 kg ha-1 yr-1 is reported for the entire forest with a large proportion derived from palm fruits (Lugo and Frangi 1993).   Palms (Prestoea montana)  are a dominant component of the riparian forest (Reed 1998).   Phenological patterns of fruit production include seasonal and inter-annual variation.  Palms often synchronize the months of peak fruit production and trees produce many more fruits in some years than other years.  The significance of these palm fruit mast years for stream consumers, such as the freshwater crab (Epilobocera sinuatifrons), is currently under study.  

While it clear that fruit fall is important to many vertebrate species in neotropical streams and rivers (Kubitzki and Ziburski 1994, Moll and Jansen 1995, Goulding et al. 1995, Araujo-Lima and Goulding 1997, Horn 1997) and in Asian rivers (Dudgeon 1999), little is known about how benthic invertebrate consume fallen fruit from riparian trees. Streams on oceanic islands are thought to derive a large portion of their energy from fruit fall (Resh and DeSzalay 1995).  Chemical studies document that certain riparian trees have secondary compounds that alter fish consumers and may have similar effects on invertebrates.  For example, the “mad fish” (Leptobarbus hoevenii), native to the Mekong River, is known to consume fruit from Hydnocarpus trees.  In doing so the fish’s flesh becomes inedible to its predators while it becomes intoxicated with compounds from this fruit (Banarescu and Coad 1991). 

 

Non-native Riparian Species: Asian Bamboo, Java Plum, and Caribbean Shrimp

Several species of Asian bamboo were intentionally introduced to Puerto Rico some 50 years ago (White and Childers 1945).  These species were selected because they were well adapted for holding sediments in place along roads in steeply sloped hillsides.  Yet, their impact in terms of displacing native riparian species or altering litter inputs to benthic consumers had not been determined.  Recent studies in The Luquillo Experimental Forest have examined the roles of non-native bamboo (Bambusa vulgaris, B. longispiculata, B. tulda, B. tuldoides, and Dendrocalamus strictus) and Java plum (Syzigium jambos) along headwater streams (O’Connor 1998). 

The dominant native riparian species such as palms (Prestoea montana) and tabonuco (Dacroydes excelsa) have relatively slow rates of decomposition (Vogt et al. 1996).  Leaf fall, especially of the palms, is often episodic and associated with wind winds and storms (Reed 1998).   Unlike these native species, bamboo forms large clumps along montane streams in the Luquillo Experimental Forest.  Leaf-fall rates for bamboo in this forest averaged 1.61 g m2 day-1  compared to 1.10 g m2 day-1 for native tree species.  Leaf litter  accumulates in pools that appeared to indicate a low rate of decomposition or relatively high rate of production.     In these same montane streams, another non-native species, Java Plum (Syzigium jambos), forms dense mono-specific stands.  In the first studies, leaf litter-bag included 4 g of dry leaves in fine-mesh bags to prevent access by macroinvertebrate consumers.  These litter bags were placed in stream pools in three headwater streams (water temperatures averaged 21 oC ).  Replicate samples were removed over a six week period.  Both species had similar and rapid rates of decomposition.  Rates of dry-mass loss followed a negative linear pattern with daily weight losses being 0.052 grams per day for bamboo and 0.051 grams per day for S. jambos.  In a second study,  tethered leaf packs were exposed to detritivores (freshwater shrimp).  These results were different from the first study’s results.  There was an increased rate of loss only for S. jambos (0.075 grams per day) while the loss rate observed for bamboo when it was exposed to detritivores remained the same low rate as when it was placed in litter bags (O’Connor 1998).  This difference led to the hypothesis that shrimp would avoid pools with bamboo because of its apparently lower value as detrital food (slow rate of leaf breakdown).

A field survey was conducted    in the Luquillo Experimental Forest to determine the distributions of freshwater shrimp in 24 pools distributed in two watersheds.  Shrimp were sampled using baited wire-mesh traps (overnight for four nights) in 12 pools with bamboo and in 12 pools with native riparian species.  Unexpectedly, more freshwater shrimp (both Atya and Macrobrachium) were found in pools with riparian bamboo when compared to adjacent pools of similar size but where bamboo was absent.  Additional two-choice studies indicated a marked preference for non-native bamboo leaves when shrimp were offered either bamboo or native leaves as cover (O’Connor 1998).  These results indicate that native detritivores distribute themselves in ways that suggest bamboo serves not only as a substitute for native leaf litter but as a preferred resource.   The microhabitat created by bamboo litter in streams appears very well suited for use by these shrimp. It is still be determined if the shrimp use the bamboo leaves only cover (as suggested thus far) or if they do provide an abundant supply of leaf litter.  Nor is it yet determined if Java plum leaves and fruit are both used by shrimp as detrital foods.   

 

Native Riparian Species: Cecropia and Freshwater Shrimp

Another series of experiments and long-term studies in headwater streams of the Luquillo Experimental Forest LTER in Puerto Rico show that benthic macroinvertebrates have species-specific roles in processing organic matter.  Several studies demonstrated that one species of common detritivore, a freshwater shrimp (Xiphocaris elongata), can process leaf litter faster because it has small chelipeds that shred leaf litter into fine fragments.  Another common species of shrimp (Atya lanipes) only scrapes leaf surfaces or filter feeds from suspended organic detritus  (Covich 1988b, Covich and McDowell 1996, Crowl et al. In press).   In these studies a series of pools was manipulated to identify effects of the two different species of shrimp.   Pools were cleared of all naturally occurring leaf litter and macroinvertebrates so that effects of detrital processing by a single decapod species could be measured in response to additions of leaf litter from a single riparian tree species.  Leaf litter from Cecropia schreveriana (an early successional tree that often colonizes disturbed riparian habitats) was added back into the same pools along with either of two naturally co-occurring species of detritvorous shrimp (Atya lanipes or Xiphocaris elongata).   Predatory shrimp (Macrobrachium carcinus, M.  crenulatum) were excluded from the pools with instream fencing to further reduce the number of species interactions that could affect rates of leaf decomposition. 

Rates of particulate export from the pools with either species of shrimp were compared to control pools (Cecropia leaves and microbial colonization but lacking all species of shrimp).  Dissolved nutrients and size fractionation of the leaf litter were also measured in treatment and control pools.   Over the 23 days of the experiment,  Xiphocaris shred the leaf litter and, as a result,  they increased the concentration and rate of downstream transport of suspended fine particulate organic matter as well as the concentrations of both total dissolved nitrogen and dissolved organic carbon.  In contrast,  Atya increased the rate of leaf breakdown relative to controls but their processing resulted in less downstream transport of suspended fine organic particulates.   This difference occurred apparently because Atya shred and scrape leaf surfaces as well as filter out suspended detritus depending upon the flow rate (Covich 1988b).   Thus, a single functional classification for Atya is not as effective as for Xiphocaris.  Although both species of shrimp influenced the rates of leaf litter decomposition, their effects were distinctly different and they are not complete substitutes for one another.  When both of these species co-occur in a single pool  they can function asa multi-species processing chain.  Shredding by one species produces more suspended particles for filter-feeding by a second species when it occurs immediately downstream of the first.  Thus they can alter each other’s effectiveness and such detrital processing chains may be generally important.

In a similar experiment, Crowl and Covich (In prep.) manipulated the shrimp community as in the above experiment except that they added a Xiphocaris and Atya together treatment.  Besides measuring leaf breakdown and export, algal consumption and insect consumption and drift were also quantified. The pattern of leaf breakdown and export was similar to the previous experiment for each shrimp species alone.  Xiphocaris actively shredded leaf material, decreasing the size fraction and increasing the net export out of pools.  Atya did little in terms of breaking down leaves, and except for the smallest size fraction, did not affect leaf export relative to controls. Xiphocaris treatments showed large decreases in insect abundance while insect drift was minor suggesting direct predation by Xiphocaris.   Alternatively, Atya pools showed decreases in insects which equaled the increase in insect export suggesting a behavioral effect on insects rather than a direct effect from predation by Atya.  Atya also significantly reduced the algal biovolume in those pools where it had excluded aquatic insects.  Previous studies (Pringle et al. 1993) demonstrated that Atya remove silt from rock surfaces, dislodge insect larvae,  and enhance algal growth of some species.  We suggest that the decrease in algal biomass due to grazing by Atya is at least partly the reason for the increase in insect drift.  Direct physical contact by Atya sweeping rock surfaces and dislodging aquatic insects, especially mayflies, is also likely. 

If each species of shrimp were acting independently or redundantly, we would expect that the flow of materials out of the pools with both species present would simply be an additive effect from that observed for each species separately.  In other words, algae would be expected to decrease, insects would be both consumed and have high drift rates, and particulates would be exported at a high rate.  In general, much of the organic material in the pools would be lost downstream.  However, our observations show a different pattern with very little material leaving pools with both species of shrimp present.  We conclude that the two shrimp species complement each other and together cause a much less “leaky” upstream ecosystem.  Xiphocaris break leaf material into small size fractions, which are available to filter-feeding Atya.  Because of the increase in suspended organic particulates when Xiphocaris is present, Atya may spend less time grazing thereby making algae more available to the aquatic insects.  As a result of higher algal availability, the insects may spend more time grazing and less time drifting to more profitable patches of downstream algae.  Because of their more sedentary behavior, the insects are likely to be more vulnerable to predation by Xiphocaris.  Overall, the food web linkages become much tighter when both species are present as compared to when each species is alone, suggesting that rather than these two species being independent or redundant, they are complementary.  Thus, the spatial location of these two species within or between pools could alter the effectiveness of overall detrital processing.  When Atya occur downstream of Xiphocaris, growth of the latter could be enhanced by increased availability of suspended fine organic particulates. 

The “processing chain” resulting from different species of shrimp interacting as detritivores within and between pools is similar to that hypothesized in the River Continuum Concept where aquatic insect shredders occur primarily in upstream reaches. These insect shredders are thought to increase the downstream availability of fine organic particulates for collectors and suspension-filter feeders (Cummins et al. 1995, Heard 1995, Heard and Richardson 1995,  Wallace et al. 1997).   More field manipulations are needed to determine how various species of aquatic insects and other benthic invertebrates differ in their individual effects on rates of detrital processing and nutrient cycling.

 

 

CONCLUSIONS

 

Non-native riparian species may provide alternative detrital resources that  complement or substitute for detrital inputs from native riparian species.  Use of additional new resources by stream detritivores may include novel sources of food supplies during periods of time when inputs from native species would be reduced or absent.  However, the complete displacement of native riparian species by non-native species will likely limit the diversity of resource supplies over time and space.  Thus, long-term monitoring of the spread of non-native riparian species and their ecological impacts are essential if the ecological integrity of stream food webs is to be preserved.  

 

 

REFERENCES

 

Anderson, N.H. and A.S. Cargill. 1987. Nutritional ecology of aquatic detrivorous insects. Pp.903-925 IN: F. Slansky,Jr. and J.G. Rodriguez (eds.)  Nutritional ecology of insects, mites and spiders. Wiley, New York.

Araujo-Lima, C. and M. Goulding. 1997. So fruitful a fish: Ecology, conservation, and aquaculture of the Amazon’s tambaqui. Columbia University press, New York.

Banarescu, P. and B.W. Coad. 1991. Cyrinids of Eurasia, Pp. 127-155 IN: I.J. Winfield and J.S. Nelson (eds.) Cyprinid fishes: Systematics, biology, and exploitation. Chapman and Hall, London.

Covich, A.P. 1988a. Geographical and historical comparisons of Neotropical streams: biotic diversity and detrital processing in highly variable habitats. Journal of the North American Benthological Society 7: 361-386.

Covich, A.P. 1988b. Atyid shrimp in the headwaters of the Luquillo Mountains, Puerto Rico: filter feeding in natural and artificial streams. Verhandlungen der Internationale Vereinigung fur Theoretische und Angewandte Limnologie 23: 2108-2113.

Covich, A.P. 1993.  Water and ecosystems,  pp.  40-55,  IN: P.H. Gleick (ed.), Water in crisis. Oxford University Press, Oxford.

Covich, A.P. 1996.   Stream biodiversity and ecosystem processes. Bulletin of the North American Benthological Society 13: 294-303.

Covich, A.P., T.A. Crowl, S.L. Johnson, D. Varza, and D.L. Certain. 1991. Post-Hurricane Hugo increases in atyid shrimp abundances in a Puerto Rican montane stream. Biotropica 23: 448-454.

Covich, A.P., T.A. Crowl, S.L. Johnson, and M. Pyron. 1996. Distribution and abundance of  tropical freshwater shrimp along a stream corridor: response to disturbance. Biotropica 28: 484-492.

Covich, A.P. and W. H. McDowell. 1996.  The stream community, pp.  433-459,   IN: D.P. Reagan and R.B. Waide (eds.), The food web of a tropical rain forest. University of Chicago Press, Chicago.

 283-290.

Covich, A.P., T.A. Crowl, and F.N. Scatena. 1998. Drought effects on pool morphology and stream benthos. Pp.91-96 IN: Proceedings of Third International Symposium on Water Resources, American Water Resources Association, Herdon, Virginia.

Covich, A.P., T.A. Crowl, and F.N. Scatena. In press. Linking habitat stability to floods and droughts: Effects on decapods in montane streams. Verhandlungen Internationale Vereinigung fur Theoretische und Angewandte Limnologie.

Covich, A.P., M.A. Palmer, and T.A. Crowl. 1999. The role of benthic invertebrate species in freshwater ecosystems. BioScience 49: 119-127.

Crowl, T.A., W.H. McDowell, A.P. Covich, and  S.L. Johnson. In press . Freshwater shrimp effects on detrital processing and localized nutrient dynamics in a montane tropical rain forest stream. Ecology.

Cummins, K.W., C.E. Cushing, and G.W. Minshall. 1995. Introduction: an overview of stream ecosystems, pp. 1-8, IN: C.E. Cushing, K.W. Cummins, and G.W. Minshall (eds.), River and stream ecosystems.  Elsevier, Amsterdam.

Dudgeon, D. 1994. The influence of riparian vegetation on macroinvertebrate community structure and function in six New Guinea streams.  Hydrobiologia 294: 65-85.

Dudgeon, D. 1999. Tropical Asian streams. Hong Kong University Press, Hong Kong.

Ewel, J.J., D.J. O’Dowd, J. Bergelson, C.C. Daehler, C.M.  D’Antonio, L.D. Gomez, D.R. Gordon, R.J. Hobbs, A. Holt, K.R. Hopper, C.E. Hughes, M. LaHart, R.R.B. Leakey, W.G. Lee, L.L. Loope, D.H. Lorence, S.M. Louda, A.E. Lugo, P.B. McEvoy, D.M. Richardson, and R.M. Vitousek. 1999. Deliberate introduction of species. BioSceince 49: 619-630.

Friberg, N. and M.J. Winterbourn. 1997. Effects of native and exotic forest on benthic stream biota in New Zealand: a colonization study. Marine and Freshwater Research 48: 267-275.

Goulding, M., N. J. H. Smith, D.J. Mahar. 1995. Floods of fortune: Ecology and economy along the Amazon. Columbia University prss, New York.

Gregory, S.V., F.J. Swanson, W.A. McKee, and K.W. Cummins. 1991.  An ecosystem perspective of riparian zones. BioScience 41: 540-551.

Heard, S.B. 1995.  Short-term dynamics of processing chain systems. Ecological Modeling 80:57-68.

Heard, S.B. and J.S. Richardson. 1995. Shredder-collector facilitation in stream detrital food webs: is there enough evidence? Oikos 72:359-366.

Horn, M.H. 1997. Evidence for dispersal of fig seeds by the fruit-eating characid fish Brycon guatemalensis Regan in a Costa Rican tropical rain forest. Oecologia

109: 259-264.

Hutchinson, G.E. 1993. A treatise on limnology. Volume 4. The zoobenthos.  Wiley, New York.

Kubitzki, K. and A. Ziburski. 1994. Seed dispersal in flood-plain forests of Amazonia. Biotropica 26: 30-43.

Lester, P.J., S.F. Mitchell,  and D. Scott. 1994. Effects of riparian willow trees (Salix fragilis) on macroinvertebrate densities in two small Central Otago, New Zealand, streams. New Zealand Journal of Marine and Freshwater Research 28: 267-276.

Lodge, D.M., R.A. Stein, K.M. Brown, A.P.Covich, C. Bronmark, J.E. Garvey. 1997. Predicting impact of freshwater exotic species on native biodiversity: challenges in spatial and temporal scaling. Australian Journal of Ecology

Lugo, A.E. 1994. Maintaining an open mind on exotic species. Pp. 18-20 IN: Principles of conservation biology. G.K. Meffe and C.R. Carroll (eds.), Sinauer Associates, Massachusetts.

Lugo, A.E. 1997. The apparent paradox of reestablishing species richness on degraded lands with tree monocultures. Forest Ecology and Management 99: 9-19.

Lugo, A.E. and J.L. Frangi. 1993. Fruit fall in the Luquillo Experimental Forest, Puerto Rico. Biotropica 25: 73-84.

Moll, D. and K.P. Jansen. 1995. Evidence for a role in seed dispersal by two tropical herbivorous turtles. Biotropica 27: 121-127.

Naiman, R.J. and H. Decamps. 1997. The ecology of interfaces: riparian zones. Annual Review of Ecology and Systematics 28:621-658

O’Connor, P.J. 1998. Habitat selection in insular tropical streams: macroinvertebrate responses to a riparian invasion by non-indigenous bamboos. M.S. Thesis, Colorado State University, Ft. Collins, Colorado. U.S.A.

Pringle, C.M., G.A. Blake, A.P. Covich, K.M.  Buzby, and A.M. Finley. 1993. Effects of omnivorous shrimp in a montane tropical stream: sediment removal, disturbance of sessile invertebrates, and enhancement of understory algal biomass. Oecologia 93: 1-11.

Reed, A.M. 1998.  Scale-dependent influences of riparian processes on the dominant detritivore of a headwater stream in Puerto Rico.  M.S. Thesis, Colorado State University, Ft. Collins, Colorado. U.S.A.

Resh, V.H. and F.A. De Szalay. 1995. Streams and rivers of Oceania. Pp. 717-736 IN:  C.E. Cushing, K.W. Cummins, and G.W. Minshall (Eds.) River and stream ecosystems. Ecosystems of the World 22. Elsevier, Amsterdam.

Vogt, K.A., D.J. Vogt, P. Boom, A. Covich, F.N. Scatena, H. Asbornsen, J.L. O’Hara, J. Perez, T.G. Siccama, J. Bloomfield, and J.F. Ranciato. 1996.  Litter dynamics along stream, riparian, and upslope areas following Hurricane Hugo, Luquillo Experimental Forest, Puerto Rico. Biotropica 28: 458-470.

Wallace, J.B. and J.R. Webster. 1996. The role of macroinvertebrates in stream ecosystem function. Annual Review of Entomology 41: 115-139

Wallace, J.B., S.L. Eggerton, J. L. Meyer, and J.R. Webster. 1997.  Multiple trophic levels of a forest stream linked to terrestrial litter inputs. Science 277: 102-104.

White, D.G. and N.F. Childers. 1945. Bamboo for controlling soil erosion. Journal of the American Society of Agronomy 37: 839-847.