Potential Influence of Suspended Sediments on the Population Dynamics and Behavior of Filter-Feeding Brachycentrus occidentalis (Trichoptera: Brachycentridae) Larvae in a Southeastern Minnesota, USA, Trout Stream

Widespread and long-term degradation of stream environments has occurred due to activities (e.g., urbanization, agriculture, logging, and mining) associated with human culture [1,2]. Degradations have included altered hydrology, instream habitat impairments, channelization, water diversion, impoundment, compromised water quality, reductions and loss of aquatic flora and fauna, and more [3,4]. Despite these myriad negative influences, river and stream habitats often display resilience and continue to function as natural ecosystems, albeit in modified form [2,3,4].

In the midwestern USA, agriculture has been a major component of the landscape for the past 150 years [5]. Immigrants from Europe introduced agricultural practices to the region that dramatically altered the landscapes, resulting in significant changes to both terrestrial and lotic ecosystems throughout the region [6]. In southeastern Minnesota, the most significant among these changes was extensive soil erosion, with heavy soil loss from the uplands carried downslope to accumulate in floodplains and their streams and rivers [5,7]. Severe rain events caused so much soil erosion that once-fertile farmlands were abandoned, valley communities were flooded and buried under meters of eroded soils, and native brook trout (Salvelinus fontinalis) and slimy sculpin (Uranidea cognata) were extirpated from formerly productive streams and rivers [5,6].

Despite these severe impacts to streams and rivers and the loss of some native fishes, many native aquatic faunas persisted, saved from extirpation by protective karstic springs that served as refugia within wooded valleys that were too steep to plant or graze [5,6]. Conversion of abandoned farms and other erosion-prone lands into state forests and wildlife refuges and the creation of state parks helped to reverse some of the degradation [5]. Major recovery efforts began in the 1940s when state and federal conservation officials worked with farmers to implement soil-conserving practices on the farmers’ lands [5,8]. Subsequently, erosion was reduced and water quality improved, allowing for the reintroduction of native trout and sculpin to many streams [9]. However, floodplains remain buried beneath deep layers of previously eroded soils, a legacy of poor past land use [5]. These legacy sediments continue to plague regional streams, as steep eroding stream banks and thick deposits of fine sediments that fill former pool habitats can readily mobilize as suspended sediments during elevated discharge events [10,11,12,13].

Southeastern Minnesota currently has over 150 coldwater trout streams and rivers, encompassing >1100 km of water. These trout streams support extensive recreational angling opportunities, with public lands and purchased angling easements on private lands providing angler access. Trout anglers contribute more than USD 1 billion to the regional economy each year [14], which indicates how important maintaining quality trout angling resources is to the region.

Coldwater trout streams in southeastern Minnesota support a diversity of aquatic macroinvertebrates, ranging from snails, fingernail clams, flatworms, roundworms, and leeches to various crustaceans and a wide variety of insects [11,12,15,16] These organisms range from very abundant to extremely rare, and can include a range of sensitivities that, variously, restrict them to a narrow suite of environments or that allow them to survive widely varying conditions [17]. Some taxa can be very sensitive to environmental pollution [18,19], while at the same time becoming very abundant when conditions are favorable for them [20]. Filter-feeding caddisfly larvae in the genus Brachycentrus are one such taxon, highly sensitive to organic pollution, synthetic pyrethroids, and fine sediment [18,21,22], while being tolerant of other stressors [23,24,25]. They often comprise a dominant component of benthic communities in many streams [20,26]. Several genera of Brachycentrus are found in North America, with Brachycentrus occidentalis Banks 1911 being especially widely distributed across the midwestern and western USA, western Canada, and ranging into Alaska [27].

Brachycentrus occidentalis larvae are a very common cased caddisfly (Trichoptera) in coldwater trout streams [28], with densities often exceeding several hundred individuals/m² [20,29,30,31]. They typically are filter feeders, using fine setae on their outstretched legs to capture seston and other potential food particles from the water column [32]. When filter-feeding, they attach their cases securely to underwater objects with silk to maintain their position in the current. They also can detach their cases and graze algae or other organic materials from submerged surfaces under the proper conditions [33]. Because they often are very abundant, they can account for a significant proportion of the diet of trout [34,35], and at times are consumed preferentially by trout in numbers exceeding the larvae’s proportional abundance in the macroinvertebrate community [35].

Even with recent buffer laws enacted to protect streams from human activities within watersheds, streams continue to be impaired by suspended sediments, either from eroding streambanks or from resuspension of deposited fine stream-bottom sediments [11,36]. These suspended sediments may interfere with the ability of filter-feeding caddisfly larvae to obtain food resources adequate to sustain healthy, sustainable populations [37], potentially reducing their abundance and availability to feeding trout. Consequently, we chose to examine the influence of suspended sediments on the population dynamics and behaviors of Brachycentrus occidentalis larvae within a single stream system, where various stream reaches exhibit significantly differing suspended-sediment loads. We hypothesized that larvae exposed to higher suspended-sediment loading would exhibit lower densities, poorer growth rates, reduced secondary production, and altered feeding behaviors, compared to larvae at stream sites with lower suspended-sediment concentrations. To further quantify feeding behaviors relative to suspended-sediment loads, we conducted a laboratory study to examine feeding behaviors and positioning on rocks when exposed to varying concentrations of suspended sediments under controlled conditions. We also predicted that higher suspended-sediment loading would have negative effects on the entire benthic invertebrate community, leading to reduced taxa richness at stream sites with higher suspended-sediment concentrations. In support of these studies, we gathered habitat and water quality data from each of the stream sites to quantify the environmental conditions to which the caddisfly larvae and the remainder of the aquatic community were exposed during the study period.

During each of the three years examined, Main Burns experienced the highest sediment loadings of the sites examined, and West Burns the lowest. Site location within the watershed, resulting in large differences in discharge among sites, played a significant role in the differing loads. However, Main Burns consistently displayed significantly higher TSS concentrations and turbidities than the other two sites, indicating that higher discharges were not the sole reason for the higher total sediment loads at Main Burns. Total loads varied 2- to 5-fold among years at individual sites, which was likely the result of differences in the seasonal timing, number, and magnitude of rain events among the different years. For example, study sites experienced four rain events >2.5 cm during June 2000, the year when sediment loads were highest at all sites; zero rain events >2.5 cm during June 2001, the year when sediment loads were lowest at all sites; and two rain events >2.5 cm during June 2002, the year with intermediate sediment loads at all sites. Rain events were responsible for 75% or more of the total seasonal loads measured at each of the stream sites, so the timing (especially during June, early in the growing season), number, and severity of rain events probably controlled most of the year-to-year variation in sediment discharges. Methods to prevent soil erosion and/or capture eroded soils before they enter streams are well known and in use throughout the study region and beyond [5,7,36]. Unfortunately, the increasing frequency and intensities of storm events within the study region [43,44,45] may overwhelm even the most ambitious soil management efforts [36], allowing for the continued transport of heavy suspended-sediment loads during periods of high discharge.

Recurring high suspended-sediment loads can result in several negative impacts on bottom substrates in streams and rivers. When fine particles settle out of suspension, they can accumulate to varying degrees, a phenomenon which, in the United States, causes more lotic ecosystem degradation (based on stream distance impacted) than all other factors combined [46]. Light to moderate accumulations on coarse substrates may fill in interstitial spaces among cobbles and gravels, embedding those materials and reducing substrate heterogeneity [2]. These may be resuspended during subsequent high levels of discharge, or become part of the shifting transported bedload [2]. Heavier accumulations may completely smother coarse bottom materials, leading to severe homogenization of the stream or river bottom and/or filling of deeper pool habitats [5,7]. The Main Burns site displayed the sediment characteristics of a stream exposed to frequent high suspended-sediment flows. Course substrates were lacking (except for protective cobble/boulder riprap around bridge abutments and flood dikes), embeddedness was maximal throughout the site, and stream habitats were mostly homogeneous runs with similar water depths. In contrast, East Burns and West Burns sites, with lower suspended-sediment loads, were dominated by coarse substrates, embeddedness was moderate, and the habitats were mixtures of riffles, runs, and pools. Steeper stream gradients and faster current velocities within upper stream reaches apparently lessen deposition and reduce embedding of coarse substrates by fine sediments [22].

Excessive amounts of fine sediments, either as transported materials or as stream-bottom deposits, can have a variety of impacts on stream-dwelling organisms. Primary producers can be impaired by reduced light penetration due to suspended particles or by coating/burying of benthic-dwelling forms; invertebrate populations may be reduced due to increased behavioral drift (caused by reduced light levels), loss of habitat within coarse substrate interstices, or interference with grazing and filter-feeding modes; and fish respiration, feeding efficiency, and spawning may be impacted by a combination of suspended and deposited fine sediments [2,7,47,48]. Such impacts can lead to reductions in productivity throughout the lotic food chain, from primary producers through top-level consumers, reducing overall system productivity to levels well below the associated natural potential [2]. Reduced taxa richness and densities within benthic invertebrate communities, as we observed at Main Burns, are typical observations as coarse substrates become embedded with fine sediments [2,7,48].

Within Burns Valley Creek, suspended and deposited fine sediments had the potential to impact Brachycentrus larvae in several ways. First, suspended sediments can interfere with the food capture and digestion of filter-feeding invertebrates such as Brachycentrus larvae [49,50]. Filter-feeders rely on suspended seston as their major food resource [51,52], but inorganic particles may clog filtering structures and/or reduce digestive efficiencies if ingested along with seston [49,53,54]. Next, deposited fine sediments may reduce the availability of solid attachment sites that filter-feeders must use while feeding, forcing them to compete for the limited number of spaces suitable (i.e., current velocity and water depth) for filtering [52,55]. Finally, filter-feeders may cease feeding in response to high concentrations of suspended particles or other stressful conditions, and wait until conditions improve before resuming feeding [52,56].

Densities and secondary production levels of Brachycentrus larvae were the highest in West Burns Valley Creek, the site with the lowest suspended-sediment loads. With average densities exceeding 1200 individuals/m² and annual production > 11 g/m²/year, Brachycentrus in West Burns likely were at or near their maximum possible productivity within the Burns Valley Creek system. These values are similar to or higher than values reported previously for Brachycentrus elsewhere [20,31,32,39,57,58], as well as those for entire benthic communities in many streams (see review by [58]). By comparison, Brachycentrus annual production was 50 to 75% lower at Main Burns and East Burns sites, both sites with higher suspended-sediment loads than those found at West Burns.

Although East Burns had a suspended-sediment load only 15% higher than at West Burns during the 2001 secondary production estimates, East Burns had much colder water temperatures than the other sites. Water temperature differences can lead to dramatic differences in invertebrate secondary production, affecting not only production of the seston food resources [2], but also filtering rates and digestive efficiencies that can affect the ultimate sizes of presently immature insects [29,52,59]. Brachycentrus occidentalis has exhibited its highest growth rates at temperatures of 16 °C or higher, with filter-feeding peaking between 16 and 18 °C [29]. Water temperatures at West Burns and Main Burns sites were similar to these optimal conditions for B. occidentalis, whereas East Burns typically was several degrees cooler and likely less optimal. Consequently, differences in both suspended sediment concentrations and water temperatures among the stream sites may have led to the large differences observed in secondary production at the different sites. Varying densities of Brachycentrus across several years at the three sites suggest that environmental conditions (e.g., number of rain events, suspended-sediment loads, and thermal variation) at those sites differed from year to year, variously benefiting or impairing secondary production of Brachycentrus within Burns Valley Creek.

High suspended-sediment concentrations are known to increase behavioral drift (downstream movement following intentional release from attachment sites) of many aquatic insects, potentially leading to reduced densities [2,7]. However, other effects of suspended sediments on Brachycentrus behavior, specifically their filter-feeding, are not known. Brachycentrus larvae have been reported to respond to various stressors (e.g., changing water temperatures or food supplies, and toxic substances) by ceasing filter-feeding, withdrawing into their cases, altering their case building, burrowing into the bottom substrates, sealing off their cases, or even abandoning their cases [21,29,60,61]. Our laboratory observations, while admittedly limited in duration and very cautiously extrapolated to longer-term field conditions, suggest that Brachycentrus larvae may cease filter-feeding, withdraw into their cases, and stop adjusting their positions toward more optimal filtering sites when exposed to high suspended-sediment concentrations (e.g., turbidities of 500 NTU). A turbidity of 500 NTU equates with a TSS of approximately 800 mg/L dry weight within our stream system (based on our field relationship; see Results), a value well below the majority of storm-event TSS measurements recorded during our study. These observations, considered together, suggest that Brachycentrus larvae in Burns Valley Creek likely cease filter-feeding when suspended-sediment concentrations are elevated during storm-event runoff. Usually, discharges declined and waters cleared (i.e., turbidities were reduced) within one or two days after a heavy rainfall at East Burns and West Burns sites within the upper watershed, but often not for several days at Main Burns in the lower watershed. If Brachycentrus larvae withdrew into their cases and remained there, not filter-feeding, for multiple days during and after each significant rain event, larval growth, and ultimately secondary production, could be compromised significantly. Five or six significant rain events per summer/autumn growing season could translate to two weeks or more of lost filter-feeding time, a major problem for an aquatic invertebrate living in a coldwater trout stream.