The life history and larval generic character development of Malenka bifurcata (Claassen, 1923) (Plecoptera: Nemouridae) in an Oregon summer-dry stream

Abstract

This is the first report of the life history and ontogenic larval development for any of the 11 North American species of Malenka. Concurrent monthly collections of larvae, and adults from emergence traps, were made over 15 years from an intermittent Oregon stream. Data from 2002–2007 collections were used for analysis; 97 males and 163 females were taken from emergence traps, and the head capsule widths of 192 larvae were measured. Counts of antennal and cercal segments, branching and growth of anterior thoracic (AT) gills, wingpad development, setation changes, and other morphological developments were recorded for an array of different sized larvae. Voltinism was variable, and the life history was complex with spread emergence and larva sizes. Progressive development of all studied characters occurred from the smallest field sampled larvae (0.36 mm hcw) to pre-emergent size. Antennal and cercal segments and gill branches nearly doubled in number, gill branch length doubled and typically the anterior-most branch branched secondarily, wingpads increased to 1.05–1.35 mm, intercalary cercal hairs appeared, and individuals could be sexed in latest instars.

Keywords:

• Plecoptera
• Nemouridae
• Malenka
• life history
• development

Introduction

The Nemouridae is an important and often dominant stonefly family in headwater streams, where they sometimes represent more than 80% of the stonefly assemblage. The life histories of 18 North American species are well known, representing about 25% of the fauna (Harper 1973; review by Stewart and Stark 2002). All of them are univoltine, with greatly variable emergence phenologies, egg development and larval growth patterns, and they occupy diverse stream habitats including intermittent streams.

There have been no detailed studies reported of the life history of any of the 11 species of the genus Malenka, nor determination of when the diagnostic generic or specific characters of late instars first appear or change in relation to instar number or growth parameter such as head capsule width. Such tracking of character appearance and development is important in determining when larvae can first be assigned to a family, genus or species. The few Plecoptera hatchlings that have been described are generally unassignable even to a family because they are consistent in being unpigmented, without ocelli, having antennal: cercal segment formulae of 8–11:3–5, compound eyes of 2–4 ommatidia, gills absent or rudimentary as knobs or stubs, generally few hairs compared with later instars, and three tarsal segments, the first two short and together subequal to the longer third (Harper 1979; Stewart and Stark 2002).

This study became a part of a long-term assessment of the species assemblage and phenology of Trichoptera of an intermittent headwater stream on the outskirts of Corvallis, Oregon, begun in 1992 (Anderson 1997). The monthly sampling also yielded an interesting assemblage of stoneflies, including Malenka bifurcata (Claassen), with no congeners. This assured the association of M. bifurcata larvae, offering the opportunity and objectives to determine its life history over a long term, and to study when important diagnostic morphological characteristics first appear and how they change during development. The larva of M. bifurcata from this stream was recently described by Stewart and Anderson (2008), along with the associated nemourid larvae of Ostrocerca dimicki (Frison) and Soyedina producta (Claassen).

Study stream and methods

The study stream, Oak Burn, is a headwater tributary originating near 60th Street, less than 1 km north of the Corvallis, Oregon, city limits in the foothills of the Oregon Coast Range at an elevation of 150 m. It flows from a seasonal wetland, drains a hillside, and after about 1 km it joins the Witham Branch of Oak Creek. The study reach was about 60 m long, flowing through a draw in a defined channel, with three small cascades and a scour pool (Anderson 1997). The flow interval varies from about late October or early December to mid-July or late August, depending on annual precipitation that averaged 122 cm over the 15-year period 1992–2007. Area climate is typically warm, dry summers and mild, wet winters. During summer and into fall, Oak Burn is typically without surface flow, or with a trickle and a few seep areas remaining wetted (Figure 1). Mid-winter flow is moderate (Figure 2), and there are occasional spates during storm events (Figure 3) (Anderson 1997). The riparian overstory includes Oregon White Oak (Quercus garryana), Oregon ash (Fraxinus latifolia), willow (Salix sp.) and Prunus sp. Thickets of blackberry (Rubus discolor) and poison oak (Rhus diversiloba) have been cleared for access to the stream. Grasses and forbs grow on the banks and into the channel. The dominant grass Brachypodium sylvaticum (Figure 2) is an invasive that has crowded out most of the native vegetation in the last decade. Leaf mats and packs in the stream channel provide food for the nemourids. Stream substrates range from clay and silt to gravel and cobble of basaltic origin.

Figures 1–3. Oak Burn stream, with set emergence trap, in (1) summer-dry condition, (2) mid-winter flow condition, (3) spate condition.

Figure 4. Emergence and growth of Malenka bifurcata, 2006 and 2007, and 2002–2007 pooled data. Dot and triangle plots are average monthly head capsule width measurements; numbers beside plots are n measured; vertical lines through plots are range hcw. Horizontal bars at top are bi-weekly emergence numbers and totals of males and females. P indicates a range bar of pooled 2002–2007 data.

Figures 5–6. (5) The four typical AT gills of 0.36–0.45 mm hcw larvae; (6) one side pair of typical AT gills of pre-emergent larvae. Vertical range bar = 0.20 mm.

Larvae were collected monthly with a small net by disturbing eight, approximately 0.1 m² substrate areas with a small soil scoop upstream during moderate flow. Samples of wet substrates were scooped from the streambed during the dry season, generally June to October or November. The net and substrate samples were washed with a garden hose over a 0.5 mm mesh sieve. The sieve mesh was submerged in a pan of water overnight and live larvae were picked from the debris with forceps the following day. Small larvae that crawled through the sieve were readily collected because they were separated from the debris. However, the procedure of washing the debris into the sieve may not have dislodged all of the clinging nemourids from leaf material, and therefore they probably were undersampled. Larvae were preserved in 70–80% ethanol, and substrates taken during the dry season were replaced on sampled spots to help ameliorate alteration of stream habitat.

Adults were collected in elongate emergence traps designed to fit into the narrow stream channels (Figure 2). They measured 30×50 cm at base and had an upper, inner trough containing ethylene glycol as the trapping fluid (Dieterich and Anderson 1995). In most years, these traps were set continuously in the stream with their base on the substrate. During spates they were sometimes dislodged and had to be reset, and during the dry season their base had gaps not in contact with the substrate. Adults were removed approximately every two weeks, preserved in 70–80% ethanol, and later sexed.

Adults and larvae from 2002–2007 were selected for emergence, growth and voltinism analysis, because of larger numbers of adults recovered from the emergence traps and larvae obtained by the benthos sampling procedure in those years (even though small numbers in some months). Growth was assessed by an increase of head capsule width (hcw). All larvae were measured with a calibrated ocular micrometer on a Wild M-5 stereomicroscope.

First appearance and change (development) of diagnostic generic and other external morphological characters were determined from microscopic study at up to 500X of 40 larvae representing an array of 0.36 to 1.20 mm hcw. Characters studied were: (1) number of antennal segments, (2) number and setation of cercal segments, (3) anterior thoracic (AT) gill presence, number of branches per gill, length of longest branch, and any secondary branching (diagnostic characters in late instar Malenka; Stewart and Stark 2002), (4) marginal pronotal bristle fringe pattern, (5) wingpad presence and backward projecting length and (6) sexual dimorphism.

Eggs were not studied, since the emergence trapped females contained immature, soft eggs without developed and sculptured chorions, and ovipositing females were not collected in the field to attempt incubation. Newly emerged female nemourids fly or crawl into riparian vegetation to feed and mate before ovipositing (Hynes 1942; Harper 1973).

Results and discussion

A total of 97 males and 163 females were taken in the emergence traps, and 192 larvae were recovered from the benthos samples over the 2002–2007 period (Figure 4). This indicated low, but sustained populations of M. bifurcata in the study reach. A sieve washing and close examination for four hours of the residual leaf material of a follow-up April 6, 2008, sample from Oak Burn indicated probable under sampling of M. bifurcata larvae of at least 50% in the 2002–2007 samples. Head-capsule widths of individual field sampled larvae ranged from 0.36 to 1.47 mm, and average monthly size of larvae ranged from 0.50 to 1.18 mm (Figure 4).

Voltinism and life history

Adult emergence and larval hcw of the 2002 to 2007 collections indicated a complex life history with much spread, asynchronous seasonal emergence and larva sizes (Figure 4) making determination of voltinism difficult. This is likely the result of an evolved survival strategy in this unpredictable intermittent stream of some combination of extended emergence, differential egg hatch and differential growth rates, with the precise causal mechanism difficult to determine by field sampling alone. The combination of: (1) overlap of peak spring emergence with both pre-emergent and small (recently recruited) larvae, (2) progressive growth from July through winter of a single cohort without further recruitment, and (3) a minor fall emergence with no short-term recruitment (Figure 4), suggest a variable slow and fast univoltine cycle with direct hatch of eggs from spring females, and overwintering egg diapause from eggs of fall females. This complex cycle fits the description of ‘indeterminate voltinism’ by Stewart and Stark (2002). An advantage of the multiple-year sampling, with low or variable numbers in some months, was that numbers could be pooled for 2002–2007 for comparison of patterns with data from 2006 and 2007 (Figure 4); the pooled data showed a similar pattern to that of individual years.

Some emergence occurred in all months, except August and December. Very small numbers of adults were trapped from November to April, and what appeared to be a minor fall emergence occurred in September–October (Figure 4). The seasonal gap in July and August, with only one male taken, was likely due to the lack of surface water required for emergence (or adults escaping from traps that were stranded above water). Since recruitment and presence of smallest larvae was indicated only for the short period of April–May, the eggs from fall emerging females appear to have undergone a short, overwinter diapause. During April and September, males outnumbered females indicating protandry to insure male presence for mating when most females were emerging and present in riparian vegetation. This would seem particularly important with the low population density indicated for M. bifurcata in these particular years. The actual encounter sites of M. bifurcata and other aspects of mate-finding and the mating system discussed for Plecoptera by Stewart (1994) are unknown.

Since the size of hatchlings was not determined, larvae smaller than 0.36 mm hcw remained unknown and their first recruitment undetected by the sampling procedure used. The smallest larvae were found only from late March through May, suggesting a possible synchronised hatch in February–March. The growth pattern was then slow through the warm, dry summer and accelerated during the wet, cool fall and winter (Figure 4).

Morphological development of larvae

Table 1 summarises morphological characters of the smallest field sampled larvae (0.36–0.45 mm hcw), and their appearance and/or change (development) until pre-emergent size of 0.91–1.47 mm hcw. Antennal segments increased from 28–38 to 45–60; a precise count may not have been achieved because apical segments were beadlike, and it appeared that generation of additional segments was on the apical end since the pedicel was established in the smallest individuals and remained unchanged over the growth cycle. Cercal segments increased from 14–20 to 26–30; the apical segments were longest and proximal ones were less distinctly separated, indicating that generation of additional segments and growth proceeded from the proximal end. Cercal segment intercalary hairs indicated as an important diagnostic character in late instars by Stewart and Stark (2002) were not discernible at 50X magnification until after 0.46 mm hcw size or larger.

Table 1. Development of morphological characters in M. bifurcata larvae.

Head capsule width (hcw) mm	Characters
0.36–0.45	28–38 antennal segments; complete fringe of evenly spaced short pronotal bristles (as in late instars); 3–4 branched AT gills, longest branch 0.18–0.21 mm, no secondary branching; no wingpads; 14–20 cercal segments with apical bristles as long or longer than following segment in basal half of cercus and much shorter in apical half, and no intercalary hairs; no sexual dimorphism.
0.46–0.70	38–45 antennal segments; 4–6 branched AT gills, longest branch 0.24–0.30 mm; first short secondary branch discernible on anterior main branch of some gills; no wingpads; 20–28 cercal segments, intercalary cercal hairs barely discernible at 50X and apical bristles similar to previous size range.
0.71–0.90	45–54 antennal segments; 6–7 branched AT gills, longest branch 0.30–0.36 mm, secondary branching of anterior main branch only in one to all four gills; wingpads absent or short with length 0.03–0.36 mm; 20–28 cercal segments with no change in setation.
0.91–1.47	50–60 antennal segments; 6–8 branched AT gills, longest branch 0.36–0.64 mm, secondary branching of anterior main branch only on most gills observed; wingpads developed with length 1.05–1.35 mm; 26–30 cercal segments, intercalary hairs visible at 50X, apical bristle circlet about 0.5 the length of following segment in basal half of cercus and much shorter in apical half. Females with a medial notch in the 8^th abdominal poststernum.

An important question unanswered in this study is if the anterior thoracic gills of M. bifurcata are present in hatchlings or when they first appear, but we did determine that substantial gill development occurs after 0.36 mm hcw (Table 1, Figures 5 and 6). Each AT gill develops from 3–4 branches with the longest branch being 0.18–0.21 mm and no secondary branching in the 0.36–0.45 mm hcw individuals (Figure 5), to 6–8 branches with the longest branch of 0.36–0.54 mm and secondary branching only of the anterior-most one or two main branches in pre-emergent larvae (Figure 6). Secondary branching of AT gills of late instar Malenka species is a major diagnostic character separating them from the simple-branched gills of Amphinemura late instars (Stewart and Stark 2002). A major question in Plecoptera larva taxonomy is when the diagnostic generic characters first appear and develop; in this case of M. bifurcata, based on gill secondary branching alone, their larvae could be separated from Amphinemura only after a size of about 0.46–0.70 mm hcw. Wingpads were first discernible in larvae of 0.71–0.90 mm hcw, then developed to a posterior projection length of up to 1.35 mm in pre-emergent individuals (Table 1). Sexual dimorphism could be determined by a medial notch in the female 8^th poststernum only in individuals of the 0.91–1.20 mm hcw pre-emergent sizes, as in the congener Malenka flexura (Claassen) (Stewart and Stark 2002).

General discussion

Other stoneflies associated with M. bifurcata in Oak Burn were Sweltsa adamantea Surdick (Chloroperlidae), Isoperla (near fusca Needham and Claassen) (Perlodidae), and four species of Nemouridae: Podmosta obscura (Frison), Ostrocerca dimicki (Frison), Ostrocerca foersteri (Ricker), and Soyedina producta (Claassen).

The interpretation of life history and voltinism from field sampling alone is useful in establishing knowledge about a species, but without associated egg incubation experiments (and if possible the difficult rearing of a complete cohort under simulated natural conditions), must be viewed as tentative. Such is the case with this study of M. bifurcata and most published Plecoptera life histories. For example, without knowledge of the egg stage and size of hatchlings it cannot be precisely determined: (1) whether first instar recruitment has been detected in field samples, and (2) whether early instars in field samples, even when associated with adult emergence, has resulted from direct hatching or short-or long-term diapausing eggs, crucial for determining voltinism. Definitive studies of nemourid life histories that have included field sampling, egg incubation, and in some species laboratory rearing through a complete cycle were reported by Harper (1973), Brittain (1973, 1978) and Dieterich and Anderson (1995). Harper (1973) determined the life histories and voltinism of seven species of Nemouridae and two of Leuctridae in Ontario from seasonal field samples of adults and larvae, and incubated eggs collected from ovipositing females.

The complex life history of M. bifurcata we have interpreted from long-term field sampling is unlike any previously reported for other Nemouridae species (review by Stewart and Stark 2002). Most of the reported life histories of the family from Britain, Europe and North America have been univoltine, mostly fast cycles, with spring emergence, over-summer egg development, recruitment in fall, and fast growth of larvae until the following spring emergence.

Few studies have documented semivoltine cycles: Nemurella pictetii Klapálek (Brittain 1978), Zapada columbiana (Claassen) (Mutch and Pritchard 1982, 1984) and Nemoura arctica Esben-Peterson (Stewart, Hassage, Holder and Oswood 1990). In all of these, two distinct cohorts were present during the winter in cold climates, indicating that longer time was required for larva growth and development. The possible option of direct hatching or short- or long-term egg diapause in M. bifurcata is not surprising given that other studies have confirmed the plasticity and option of univoltinism or semivoltinism in Plecoptera, mainly depending on temperature or water presence conditions (Brittain 1978; Snellen and Stewart 1979; Sandberg and Stewart 2004). Snellen and Stewart (1979) reported the option of direct hatching or long-term diapause for up to three years or more for two species of Zealeuctra in Texas intermittent streams, and Sandberg and Stewart (2004) in detailed long-term incubation studies of six species of Isogenoides showed that individual egg hatches may be genetically programmed to have the combination of short, direct incubation and diapause from a few months to up to at least four years.

There have been no complete studies of the ontogenic changes of major diagnostic larva characters in Plecoptera over the full developmental cycle; most attention has been given to the few studies of early instars. Khoo (1964) noted that second instars of several species differed from hatchlings and were more like later instars. Brittain (1973) described the first instars of Nemoura avicularis Morton, and reported the increase in antennal: cercal segments from them (9:4) through instars two (11:5), three (14:8), and four (20:10 or 11). With this rate of adding antennal and cercal segments, we estimate that M. bifurcata larvae had moulted approximately four times before they were found in our field collections. We did not obtain hatchlings of M. bifurcata, so were unable to determine first presence and morphology of AT gills. But our determination of substantial increase in number and length of branches, and secondary branching as larvae grew from 0.36 mm hcw to pre-emergent size suggest exciting discoveries await further study of gill development in stoneflies that lead to the diagnostic gill characters of late instars described by Shepard and Stewart (1983).