Trichoptera: Lepidostomatidae) in an Oregon Coniferous Forest Stream
Total Page：16
File Type：pdf, Size：1020Kb
Ecology, 61(4), 1980, pp. 808-816 (g) 1980 by the Ecological Society of America POPULATION DYNAMICS AND ROLE OF TWO SPECIES OF LEPIDOSTOMA (TRICHOPTERA: LEPIDOSTOMATIDAE) IN AN OREGON CONIFEROUS FOREST STREAM, E. GRAFIUS2 AND N. H. ANDERSON Department of Entomology, Oregon State University, Corvallis, Oregon 97331 USA Abstract. The aim of this study was to gather quantitative field and laboratory data on the role of Lepidostoma cascadense (Milne) and Lepidostoma unicolor (Banks) in the processing of conifer needles in a stream ecosystem. Samples were collected monthly in a riffle-pool section of Mack Creek, Lane County, Oregon, USA. For L. cascadense, maximum larval density was 812/m2 and instantaneous growth rate was 1.5%/d. The larvae grew slowly throughout the winter and pupation occurred in May and June. In contrast, L. unicolor increased very little in size during the winter and grew very rapidly during June and early July, reaching a maximum larval density of 320/m 2 and with an instantaneous growth rate (March through July) of 2.7%/d. Production of L. cascadense was estimated as 0.31 g• m- 2. yr and that of L. unicolor was 0.23 g• m-2. yr-I . In addition to temporal separation in periods of maximum growth, the two species occurred in different microhabitats. L. cascadense larvae were found within the sediments and debris while L. unicolor occurred on the surface of the debris. Laboratory studies were conducted with L. unicolor larvae. Consumption and fecal production rates (measured gravimetrically) increased with higher temperature, greater food density, or longer conditioning time of the food, and decreased with increased size of the larvae. Production and biomass of L. cascadense and L. unicolor in Mack Creek are minor in relation to other insects in the system. However, because of high consumption rates and low assimilation efficiency, the processing of large particulate organic matter by these two species contributed signif- icantly to the food available to collectors in the study area. Key words: aquatic; growth; Lepidostoma; life histories; Oregon; population dynamics; respi- ration; shredders. INTRODUCTION in the processing of more refractory conifer needles (Sedell et al. Examination of larvae collected Small streams with riparian vegetation canopies 1975). by Sedell et al. plus extensive emergence collections tend to receive most of their energy in the form of and larval rearing from Mack Creek revealed that the terrestrial litter fall rather than as autochthonous pri- dominant species involved were Lepidostoma casca- mary production (Hynes 1963, Cummins 1973, Gosz dense (Milne) and Lepidostoma unicolor (Banks). et al. 1978). Litter fall may be dominated by deciduous Data were collected on population dynamics and life leaves, as is common in Eastern U.S. streams, or by histories of these two species in Mack Creek. Labo- the more refractory conifer needles and wood, as is ratory studies of feeding and respiration were con- the case in many streams in the Cascade Range in Oregon. ducted with L. unicolor. Of particular interest were: on The work reported here was part of our studies on (1) the impact of L. cascadense and L. unicolor the role of caddis larvae in the processing of alloch- the stream ecosystem in terms of food consumption, fine particulate organic matter production, and pro- thonous organic material (Anderson and Grafius 1975, duction; (2) mechanisms reducing competition for food Anderson 1976, Anderson et al. 1978). In a previous between the two species; and (3) possible adaptive study (Grafius and Anderson 1979), Lepidostoma quercina Ross was examined as representative of significance of factors such as life history and bio- energetic responses to temperature, food quality, or species feeding on deciduous leaves. L. quercina is a fall-growing species, with its major growth period food density. timed to allow maximum exploitation of deciduous leaf METHODS AND MATERIALS input to the stream. It occurs primarily in streams in the coastal range and Willamette Valley in Oregon Field population studies (Anderson 1976). It was of considerable interest to find Mack Creek is a third-order stream, located in the that members of the same genus are also instrumental H. J. Andrews Experimental Forest in the Cascade Range approximately 85 km east of Eugene, Oregon. The sampling site is at an elevation of 775 m, stream I Manuscript received 28 November 1978; revised 10 Oc- tober 1979; accepted 19 October 1979. width is about 6 m, mean discharge 0.60 m3/s, and 2 Present address: Department of Entomology, Michigan gradient 20%. Water temperatures range from 2° to State University, East Lansing, Michigan 48824 USA. 12°C but remain below 5° much of the year (November August 1980 DYNAMICS OF TWO SPECIES OF LEPIDOSTOMA 809 through May or June). The surrounding vegetation is phomycete) origin (F. Triska, personal communica- old-growth Douglas-fir (Pseudotsuga menziesii [Mirb.] tion) and decomposition in the laboratory followed the Franco), western hemlock (Tsuga heterophylla [Raf.] pattern of color and gross structural changes described Sarg.), and western red cedar (Thuja plicata D. Don.), by Hayes (1965) for conifer needle decomposition on with some vine maple (Acer circinatum Pursh.). An- the forest floor. Well-decomposed needles collected nual precipitation is generally 250-300 cm. from Mack Creek, chosen for their dark color, were Mack Creek, typical of many small streams on the used in some experiments. The latter were presumed western slopes of the Cascade Range, experiences to have been in the stream for at least 9 mo, from the large fluctuations in stream flow due to the seasonal previous autumn when most of the litter entered the nature of the precipitation. Peak winter flows are 1000 stream. to 2000 times summer flows and the stream varies Larvae were kept individually in 3 x 4 cm plastic widely in depth, width, and amount of debris present. ice cube compartments. Consumption was measured The site chosen for intensive study was a pool by initially determining the wet masses of small groups formed above a large natural log jam, and the riffle, of needles (----,three times the expected consumption), above and to one side of the pool. The pool was ap- feeding them to the larvae, and determining the dry proximately 4 m wide by 5 m long and up to 1 m deep, mass of the remainder. A small number of uneaten with organic debris often deposited to a depth of 10 fragments were sometimes produced and these were cm or more. The riffle was 1-2 m wide and 10-12 m sorted by hand and included with the uneaten food. long. The wet masses of other groups of needles were deter- Samples were collected monthly (19 October 1974 mined at the same time, oven-dried and the dry masses to 31 July 1975) with a 0.1-m2 Surber sampler (mesh determined to obtain conversion values from wet to 250 pm). A stratified sampling scheme was used, with dry mass. Using these conversion values, the initial three or four samples each from riffle and pool strata. wet masses of the "fed" needles could be compared Because there was little current flow in the pool, the with dry masses of the same needles after feeding. sampler was used to delineate the area to be sampled Wet masses were taken by blotting the needles on pa- and the substrate and associated insects were scooped per towel and allowing them to air dry for 1 min before by hand into the collection net. Larvae were kept on weighing. Needles could not be dried prior to feeding ice and sorted under a binocular microscope while since microbial components have been shown to be they were still alive. Sorting of fresh samples had two important stimuli to feeding (Kaushik and Hynes 1971, advantages over sorting preserved samples: (1) larvae Kostalos 1972, Grafius 1974). Larvae were killed and were much easier to find among the debris when they the dry masses determined at the end of the respective were still alive and moving, and (2) dry masses could experiments. be determined from fresh rather than preserved spec- In the growth experiments, case measurements imens. Random samples of each available instar were (length and width of the anterior opening) were used oven-dried (60°), cooled in a desiccator, and weighed to estimate initial mass of individual larvae. Samples to the nearest 1 or 10 p,g on a Cahn 4100 Electrobal- of larvae representing a range of case sizes were killed ance. Data were analyzed as stratified samples, and the dry masses determined at the beginning of the grouped according to instar. Instantaneous growth experiment. Dry masses of these larvae were com- rate was estimated as the slope of a regression of nat- pared with an index of their respective case volumes ural logarithm of mean mass vs. time. Instantaneous (length x width2). A regression equation computing mortality rate was estimated as the slope of natural dry mass from case volume was calculated. This equa- logarithm of mean number per square metre vs. time. tion was used to estimate initial masses for larvae used in the experiment. Laboratory studies of feeding and Instantaneous growth rates were calculated as: (nat- growth of L. unicolor ural logarithm final mass — natural logarith initial Food consumption and fecal production were mea- mass) x time-. sured gravimetrically. Douglas-fir needles were col- lected by shaking a tree to dislodge those that were Respiration measurements near abscission. The needles were conditioned in aer- Respiration rates were measured using a Gilson Dif- ated containers of dechlorinated tap water in the lab- ferential Respirometer. Insects were acclimated for at oratory at 13°-15° for 4-5 mo. Pieces of wood or leaves least 2 wk at the appropriate temperature prior to test- from the field were added occasionally to aid in fungal ing and were tested individually or in groups of up to colonization of the needles.