James V. Robinson

Annual Experimental Evaluation of the Effect of Invasion History on Community Structure

Fifty-four communities were developed from repeated introductions of 28 phytoplankton species in three different orders of invasion at three different rates. Variation in species richness values in these communities was assignable to invasion order, invasion rate, and the timing between interspecific invasions. Invasion rate was most influential, explaining 21.8-78.8% of the variation on any given date; invasion order never explained >4.3%, and timing explained between 19.7 and 74.9%. All three facets of invasion had significant effects. One of the low-rate invasion categories developed a unique community structure that was dominated by Chlamydomonas. Communities in this category were invulnerable to the invasion of many species and as a result had low species richness values. The relative abundance patterns of species in categories having low or moderate in- vasion rates but identical invasion orders clustered more closely to each other than they did to their high invasion rate counterparts. In this sense, invasion order is more influential where immigration rates are relatively low (e.g., on islands) than it is where dispersal from outside sources is high (e.g., on continents). Because invasion history was controlled, the methodology used here provides inter- pretable data concerning the potential importance of chance historical events that occur during community assembly. The data indicate that idiosyncrasies in community structure often may be explained on the basis of the random invasion patterns of organisms to different habitats.

Ecological resistance to the invasion of a freshwater clam, Corbicula fluminea: fish predation effects

SummaryFish predation is shown to have a twenty nine fold effect on the abundance of the invasive freshwater clam, Corbicula fluminea, in a Texas reservoir. This predation has prevented the clam from establishing the high densities commonly reported for it elsewhere. The high magnitude of the fish effect is attributed to Corbicula being an invader to this reservoir and not being able to cope well with the mix of resident fish species. In the absence of fish, colonization of the reservoir by Corbicula is spatially patchy. When fish interact with these clams, they remove sufficient numbers of individuals from dense patches to create the appearance of a spatially uniform distribution.

The effect of temporal environmental heterogeneity on community structure: a replicated experimental study

SummaryPeriodicity, predictability and stochasticity of environmental perturbations are shown to influence the community structure that develops in microcosms. Sets of replicate, microalgal communities were subjected to different temporal patterns of rarefaction and resource resupply and their species-abundance patterns after 120 days of such manipulations were determined. Perturbations having, 1, 7, and 28 day periodicities differentially effected community structure. The predictability of these perturbations had a less profound influence on the communities which developed than the average perturbation periodicity.

Mutual Predation in Assembled Communities of Odonate Species

Larval individuals from six anisopteran species were added to artificial ponds so that each pond contained an equal number and size distribution of each species. Half of these ponds were exposed to an additional anisopteran, Anax junius. Interodonate predation in Anax-free ponds was size dependent, with the smallest species, Perithemis tenera, experiencing the highest mortality. Predation in ponds containing Anax was not size dependent, but was species dependent. Plathemis lydia individuals were not signifi- cantly affected by Anax, probably due to the tendency of Plathemis to dig deeply into the sediment. Pe. tenera was shown to have a different allometric relationship between head width and body volume than the other six dragonfly species. We postulate that this is correlated with its small size and is an adaptation to increase the range of prey available to it.

Microhabitat selection as an antipredator strategy in the aquatic insect Pachydiplax longipennis Burmeister (Odonata: Libellulidae)

SummaryAn investigation of the larval dragonfly fauna associated with the plant, Sagittaria platyphylla, was conducted in a small pond. Despite the presence of several larval anisopteran species in the pond, only Pachydiplax longipennis larvae were found on Sagittaria plants. A study of the microspatial distribution of P. longipennis larvae on S. platyphylla indicated that larvae use the various regions of a plant in a highly non-random fashion. Larvae show a strong preference for the leaf axil area. A generalized predator, the bluegill sunfish (Lepomis macrochirus), was allowed to selectively eat either of two larvae placed in various plant regions. This experiment indicated that larvae in a leaf axil area were significantly less susceptible to bluegill predation than larvae positioned in other plant regions. The microspatial distribution of starved larvae revealed that larvae with high hunger levels occupied the leaf axil area significantly less than well fed larvae, suggesting 1) larvae do not use these regions as feeding sites, and 2) high hunger levels may induce a behavioral shift in habitat use, with starved larvae forced into areas of high predation risk by the need to fulfill nutritional requirements.

A stratified sampling approach to compensating for non-random sedimentation of phytoplankton cells in inverted microscope settling chambers

A pronounced “edge effect” resulting in non-random cell sedimentation is documented for standard-design inverted microscope counting chambers. Comparison of mean cell densities of two green algae from random fields in central and peripheral regions of chambers substantiate a marked peripheral settling bias. The non-random settling pattern is independent of phytoplankton cell size and population density over the ranges encountered in this analysis. If this problem is not compensated for, the magnitude of the bias is such that large errors in estimates of mean cell density/chamber can result. A stratified sampling approach is developed to minimize the error resulting from non-random settling without increasing the amount of time spent enumerating cells. Counting of random fields rather than patterned transects is strongly recommended as the standard procedure in phytoplankton studies.

The ecological role of caudal lamellae loss in the larval damselfly, Ischnura posita (Hagen) (Odonata: Zygoptera)

SummaryDamselfly larvae may autotomize and regenerate any of their 3 caudal lamellae. At least one missing or regenerating lamella was evident in 50.1% of field collected Ischnura posita larvae. Lamellae loss during molting is very infrequent (1 out of 117 recorded molts). Laboratory trials indicate that conspecifics remove lamellae and that this process is density dependent. The percentage of larvae losing lamellae during 24 h trials ranged from 73.5 at the highest density tested to 17.3 at the lowest density. I. posita larvae are cannibalistic. The presence of lamellae reduces an individuals chance of being cannibalized. More than twice as many final instar lamellae-less larvae were cannibalized during 24 h trials than analogous individuals having 3 lamellae at experimental initiation. Costs are also associated with lamellae autotomy. 1) Although individuals without lamellae can swim they are more reluctant to release from a wooden stalk and swim when threatened (9% release) than are larvae with lamellae (29% release). Since swimming is part of their repertoire of anti-predator behaviors this behavioral shift should be detrimental. 2) Caudal lamellae function in O2 uptake. Trials were conducted with larvae having and not having lamellae in an experimental horizontal oxygen gradient system. Relative to larvae without lamellae, those with lamellae preferred deeper depths at PO2 values greater than 70 torr. Many lamellae-less larvae distributed themselves at the water surface throughout the range of PO2 values tested. Differential depth distribution between larvae with and without lamellae is highly significant (P < 0.01).

Testing the invulnerability of laboratory island communities to invasion

SummaryOne hundred communities were developed through the controlled introductions of microorganisms to beakers over a six month period of time. Following three months of development, a series of three separate previously unencountered species were introduced into each community. The persistences of these invaders were monitored and their relationships to invaded community complexity, composition and history of development evaluated. The null hypothesis that colonization success is independent of community complexity cannot be rejected. The rate at which species were introduced during development of these communities, as well as beaker size, influenced the invasion success of Dictyosphaerium, but not Staurastrum or Platydorina. The assembled communities were of two types: those dominated by Ochromonas and those domnated by Paramecium bursaria. Ochromonas dominated communities were invulnerable to the invasion of the above 3 species; no clear pattern of invulnerability was evident for P. bursaria communities.

Effects of a thermal effluent on macroinvertebrates in a central Texas reservoir

We investigated the impact of a thermal effluent from an electricity-generating plant on the macroarthropod community in a central Texas reservoir for 1 yr by comparing the community of a 60-ha pond directly receiving the effluent to an area in the main body of the reservoir relatively unaffected by the effluent. Temperature of the pond averaged 7.2 C warmer than the main reservoir sites. Samples of artificial substrates constructed to mimic macrophytic vegetation indicated that the pond generally had lower macroinvertebrate abundance and reduced taxonomic diversity, though direction and severity of effects varied over time for most taxa. Deleterious effects were most severe in summer when temperatures of 40-42 C in the pond eliminated macroinvertebrates. Although taxa recolonized the pond after the summer defaunation, with some taxa briefly obtaining very high population levels, most taxa maintained lower population levels in the pond than the main reservoir throughout the winter. 32 refs., 3 figs., 1 tab.

The effect of predation on the structure and invasibility of assembled communities

SummaryFifty four microcosmic communities were assembled over 4 months from a 28-species source pool of phytoplankton using nine different invasion patterns each replicated six times. Three communities from each set of replicates then were invaded with a cladoceran that feeds on phytoplankton. All communities were then treated identically for an additional 4 months. In all nine invasion categories species richness was greater in predated communities. Predation opened communities to invasion by increasing the representation of infrequently sampled species at the expense of more common species. Invasion rate was four times more influential than predation and over eleven times more important than either invasion order or the timing pattern of interspecific arrivals in determining species richness in this system of communitites.