Montane meadows in the Sierra Nevada: comparing ... - Springer Link

Environ Monit Assess (2012) 184:6967–6986 DOI 10.1007/s10661-011-2473-0

Montane meadows in the Sierra Nevada: comparing terrestrial and aquatic assessment methods Sarah E. Purdy & Peter B. Moyle & Kenneth W. Tate

Received: 28 June 2011 / Accepted: 21 November 2011 / Published online: 20 December 2011 # Springer Science+Business Media B.V. 2011

Abstract We surveyed montane meadows in the northern Sierra Nevada and southern Cascades for two field seasons to compare commonly used aquatic and terrestrial-based assessments of meadow condition. We surveyed (1) fish, (2) reptiles, (3) amphibians, (4) aquatic macroinvertebrates, (5) stream geomorphology, (6) physical habitat, and (7) terrestrial vegetation in 79 meadows between the elevations of 1,000 and 3,000 m. From the results of those surveys, we calculated five multi-metric indices based on methods commonly used by researchers and land management agencies. The five indices consisted of (1) fish only, (2) native fish and amphibians, (3) macroinvertebrates, (4) physical habitat, and (5) vegetation. We compared the results of the five indices and found that there were significant differences in the outcomes of the five indices. We found positive correlations between the vegetation index and the physical habitat index, the invertebrate index and the physical habitat index, and the two fish-based indices, but there were significant differences between S. E. Purdy (*) : P. B. Moyle Center for Watershed Sciences, University of California, 1 Shields Avenue, Davis, CA 95616, USA e-mail: [email protected] K. W. Tate Department of Plant Sciences, University of California, 1 Shields Avenue, Davis, CA 95616, USA

indices in both range and means. We concluded that the five indices provided very different interpretations of the condition in a given meadow. While our assessment of meadow condition changed based on which index was used, each provided an assessment of different components important to the overall condition of a meadow system. Utilizing a multimetric approach that accounts for both terrestrial and aquatic habitats provides the best means to accurately assess meadow condition, particularly given the disproportionate importance of these systems in the Sierra Nevada landscape. Keywords Meadows . Wetlands . Rapid habitat assessment . Fish . Invertebrates . Vegetation . Stream channel condition

Introduction Montane meadows are wetland systems that have disproportionate importance compared to their surface area (Kattelmann and Embury 1996; Kondolf et al. 1996). In the Sierra Nevada of California and Nevada, meadows support critical ecosystem services including biodiversity, flood attenuation, sediment filtration, water storage, water quality improvement, and carbon sequestration (Sanders and Flett 1989; Potter 1994; DeSante 1995; Mitsch and Gosselink 2000; Woltemade 2000; Povirk et al. 2001; Hammersmark et al. 2008). In

6968

addition, meadow vegetation has significant direct economic value as forage for grazing livestock (USDA Forest Service 1993; Torell et al. 1996). The majority of meadow systems in the Sierra Nevada have suffered anthropogenic impacts to their soils, hydrologic processes, and biotic integrity (Ratliff 1985; Knapp and Matthews 1996; Castelli et al. 2000; Sarr 2002; Blank et al. 2006; Popp et al. 2009). In particular, streambank erosion and channel incision are widespread and highly detrimental to meadow function; these erosional processes are accelerated by improper livestock grazing, culvert and road crossing placement, mining, logging, recreational activities, and water diversions (Dull 1999). The impacts of improper management are often exacerbated by episodic natural events such as drought, fire, and flood (Leege et al. 1981, Belsky et al. 1999; Wemple et al. 1996; Gucinski et al. 2001). At severe levels, erosion and channel incision cannot be reversed by simply removing the disturbance(s). Once critical thresholds of impact have been reached, the meadows do not recover without active intervention (Ratliff 1985; Schlesinger et al. 1990; US Bureau of Land Management 2007; Chambers et al. 2004; Allen-Diaz et al. 1999; Micheli and Kirchner 2002; Briske et al. 2008). Incision of a meadow lowers the local water table and can be viewed as transition to an alternate stable ecological state (e.g., Briske et al. 2008) from a previous stable state with a high water table supporting meandering streams and diverse wetland vegetation. This transition results in a reduction of stream habitat, loss of hydrologic functions, and changes in community structure in both the aquatic and terrestrial ecosystems (Zimmer and Bachmann 1978; Hammersmark et al. 2008, Cornwell and Brown 2008). Without re-elevation of the water table and restoration of hydrologic connectivity between meadow surface and stream channel, the meadow remains altered, potentially for centuries, and becomes a terrace occupied by upland plant communities (Allen-Diaz et al. 1999; Loheide et al. 2009; Briske et al. 2008). This represents a loss of ecosystem services and economic value, but is preventable and even reversible if management actions are taken before such thresholds are crossed. The key is to identify meadows at risk before this threshold is crossed, so that management actions can be taken. Current methodologies to assess condition of both the terrestrial and aquatic components of montane

Environ Monit Assess (2012) 184:6967–6986

meadows generally do not provide adequate information to determine key factors altering meadow condition or to determine how close the meadow is to crossing a threshold to a different, less desirable, state (e.g., Allen 1989; Belsky et al. 1999; Allen-Diaz 1991; Auble et al. 1994; Chambers et al. 2004; Blank et al. 2006; Dwire et al. 2006). In particular, terrestrial and aquatic components are rarely assessed together, although the two components are highly interdependent. It is clear that meadow condition assessments, which integrate local terrestrial and aquatic conditions, are needed to evaluate montane meadow status. Such evaluations can provide the basis for future monitoring and can help to determine how to balance ecological benefits with economic benefits associated with various land management practices. At present, there are three commonly used types of assessments for meadows and their associated stream systems: vegetation surveys, qualitative habitat assessments, and indices of biotic integrity (IBIs). These assessment tools were not developed specifically for meadow evaluation; instead, they have been typically used for rangeland assessments, high gradient stream assessments, or fish surveys.

Vegetation surveys Vegetation surveys based on the concept of “range condition” have been the standard method used to evaluate meadow condition by most natural resource management agencies, where a meadow in good condition is one that has herbaceous vegetation composition which benefits seasonal grazing by livestock. These surveys use metrics such as plant species composition, vegetative cover, plant rooting depth, community type, and seral status to determine meadow condition (i.e., Ratliff 1985, 1993; Weixelman et al. 1997; Winward 2000). They provide quantitative data, usually through the use of transects and quadrats and allow for accurate re-measurement to determine trends through time. However, these methods require a high degree of plant taxonomic expertise to perform. Simultaneously, the heterogeneous nature of meadow systems makes it difficult to extrapolate the conditions found in transects and quadrats to the larger surroundings. Currently, the predominant Forest Service range assessment method in the Sierra Nevada is the vegetation survey method developed by


Weixelman et al. (2003), based in part on the methods of Winward (2000). Qualitative habitat assessments Two qualitative assessment techniques have been commonly applied to meadows: proper functioning condition (PFC) assessment for the terrestrial/hydrological portions and rapid habitat assessment (RHA) for the aquatic/riparian habitat portions. The PFC assessment was developed jointly by the Bureau of Land Management, USDA Forest Service, and Natural Resource Conservation Service and focuses assessment on 17 metrics such as hydrologic connectivity, balance of sediment deposition and erosion, and vegetation composition required to stabilize deposited sediment. The impetus for developing PFC was the need for an assessment method that was rapid, required minimal expertise, and could distinguish diverse conditions encountered in the field from pristine to highly altered (Prichard et al. 1994, 1996, 1993). Similarly, Barbour et al. (1999) developed RHA protocols as a part of their larger rapid bioassessment protocol for small (wadeable) streams. This ten-metric index focuses predominantly on instream and streambed components such as available habitat for invertebrates and fishes, siltation and erosion, bank stability, riparian width, meander ratios, flow regimes, and access to the floodplain (Appendix 1). The RHA provides a numerical basis to visually determine the condition of the stream habitat. It uses few direct measures of the habitat components but provides guidelines for categorizing each metric into four broad condition categories (poor, marginal, sub-optimal, and optimal) to ease interpretation. While these two assessment protocols are more integrative in their approaches and easier to perform than vegetation surveys, they are qualitative and have been criticized for lack of sensitivity to change, inability to accurately monitor trends over time, and excessive observer variability (Coles-Ritchie et al. 2004). Indices of biotic integrity Karr (1981; Karr et al. 1986; Karr and Chu 1997) developed the concept of the IBI as a means of determining the condition of fish populations in Midwestern rivers. The premise of this method is that

6969

the biological community responds to anthropogenic stressors in a predictable fashion. The metrics used for assessment are diversity, abundance, life history, sensitivity, and other factors that are responding to changes in habitat quality, which are in turn responding to stressors. Barbour et al. (1999) used the IBI concepts in developing the rapid bioassessment protocols used by the Environmental Protection Agency (EPA). The approach incorporates fish, aquatic macroinvertebrates, periphyton, and a qualitative habitat assessment similar to RHA assessment. While IBIs have not previously been developed specifically for Sierra Nevada meadow streams, their application to California streams was demonstrated by Moyle and Marchetti (1999). Purpose Our study compared five rapid assessment methods to determine the condition of montane meadows in the Sierra Nevada. We compared methods that were either previously used in the meadows or which we developed by modifying established methods. We developed three original IBI-based methods to quantify condition of (1) fishes, (2) native fishes and amphibians, and (3) aquatic macroinvertebrates as indicators of aquatic and riparian condition. We employed a modified version of the Weixelman et al. (2003) approach to determine vegetation condition and used the EPA Rapid Bioassessment Protocol (Barbour et al. 1999) for habitat assessment to determine stream channel and overall meadow habitat condition. Our study addressed two questions. First, were all five of the measures of meadow condition in agreement? Secondly, if not, what were the differences among the methods? Our hypothesis was that given the inherent complexity and variability of meadow ecosystems, it is unlikely that a singlemethod approach to assessment adequately captures the condition of the meadow and its components. Rather, a multi-functional approach is necessary to get the best information on the true status of the meadow (Karr 2005, 2006; Pellant et al. 2005). However, constraints imposed on monitoring by time, budget, and expertise require an assessment approach that most efficiently captures meadow condition. This paper shows how some commonly used rapid assessments, modified for montane meadow systems, can

6970

produce quite different results when used independently, but provide useful assessments when used together.

Methods Study site selection and study design Over two field seasons (June through September of 2006 and 2007), we assessed 79 meadows in the northern Sierra Nevada and southern Cascade, ranging from Sierra County in the south to Modoc County in the north, visiting each site a single time. We surveyed only meadows associated with a stream (flowing near baseflow at time of assessment) that had previously been surveyed by USDA Forest Service crews using the protocol developed by Weixelman et al. (2003). We surveyed a broad assortment of meadows over a large geographic range in order to capture the variability present in Sierra Nevada meadow systems, in sites ranging from nearly pristine to highly impacted. Site selection was focused primarily in Plumas, Lassen, and Modoc counties with some sites in Sierra and Nevada Counties. We eliminated sites that did not have flowing water. We chose meadows between the elevations of 1,000 and 3,000 m, which were 1 km in length), we sampled two 50-m reaches to account for habitat heterogeneity. Basic fish sampling procedures followed those of Moyle (2002). We conducted single pass backpack


electrofishing surveys using Smith-Root type 12 backpack electrofisher and systematically sampled all habitats within the stream reach from the lower blocknet to the upper blocknet. Stunned fish were captured by two to three people using dip nets. The fish were kept alive in buckets or live wells until they were identified to species (using Moyle 2002), measured (standard length, mm), and weighed (volumetric displacement), then returned alive to the water near where they were caught. Amphibian and reptile survey We surveyed day-active amphibians and reptiles in the riparian zone, using visual encounter surveys (Crump and Scott 1994). At the beginning of the stream reach, two members of the crew performed a timed survey of the stream banks, stream, and adjacent habitats such as oxbows and ephemeral puddles looking for egg masses, tadpoles, or adult amphibians. We attempted to capture all amphibians and reptiles encountered and used a standard snout to vent length measurement. We identified all reptiles and amphibians to species and recorded length and life stage, though that was not used as a metric in the index (Crump and Scott 1994, p. 91). Amphibians observed or captured during the fish sampling were recorded as incidental observations and contributed to the total abundance score for the site. Macroinvertebrate survey Benthic macroinvertebrates were sampled using modified level 2 protocols from Harrington and Born (2000). We took nine total samples from within each 50-m fish sampling reach, using a D-net, preferentially sampling riffles but also collecting from distinctive habitats throughout the reach. The samples were combined, and each was placed in a white enamel pan and the major debris removed. We sorted and identified live invertebrates to family in the field and returned them to the stream afterwards. We identified the first ∼300 invertebrates in each sample. Invertebrates with questionable identification were preserved in 70% ethanol for later identification in the laboratory. Three complete samples were brought back for traditional laboratory processing to validate field sorting accuracy.


Habitat survey We used the USEPA (Barbour et al. 1999) Habitat Assessment Sheet for low gradient streams to assess the habitat structure and geomorphological conditions of the meadow streams. This assessment is based on ten instream, bank stability, and vegetation parameters, each scoring between 0 (worst) and 20 (best) for a total possible score of 200. Each of the ten metrics is divided into four qualitative categories (“optimal,” “sub-optimal,” “marginal,” and “poor”) with a 5-point spread in each. The categories consist of verbal descriptions of pertinent habitat features that distinguish observable human impacts to the stream and riparian area. The numeric values within each category allow the observer to quantify their observations through scoring each site on a scale of 0–200, combining the scores from each of the ten metrics. Vegetation survey We designed our terrestrial vegetation survey as a modified version of Weixelman et al. (2003). The Weixelman assessment was based on seral status (as a proxy for recovery from past disturbance), depth to rooting frequency of >100 roots/dm2, percentage of bare soil in the meadow, and vegetation functional guilds (i.e., wetland indicator status, growth or rooting habit). Our methods differed in that Weixelman et al. (2003) used line transects and quadrats throughout the meadow, whereas we surveyed only the vegetation within 10 m of the stream banks, similar to Winward’s (2000) “Greenline” method. In the 50-m reach, we surveyed both sides of the stream from the water’s edge to 10 m from the banks, riparian vegetation permitting. We estimated the percent cover for all species within the survey area by breaking the 50-m reach into ten 5-×10-m transects. We walked each plot and noted the species present, making a visual estimate of their percent cover within that 5-×10-m plot, then combined the results to get an overall species list and percent cover for the entire 1,000 m2 survey area. We identified all plants to the lowest possible taxonomic level. Unknown plants were either preserved or photographed for later identification. We assumed multiple canopies within each plot (i.e., a shrub layer with forbs in the understory); therefore, percent coverage did not have to equal 100. We

6971

measured the percent of bare ground exposed (as a measure of disturbance) in each 5-×10-m plot and also noted the percentage of rocks and cryptogams in the survey reach. Vegetation functional guilds were determined following Weixelman et al. (2003) and from the USDA Plants Database on identification, habitat, distribution, growth forms, and function of plants (http://plants.usda.gov/wetland.html). Calculated indices of biotic integrity We used five multi-metric indices to assess the condition of the 74 study meadows. The three indices focusing on fish, amphibians, and macroinvertebrates were original to the project. The vegetation and habitat indices came from previously published assessment methods. The vegetation index was constructed and calculated according to Weixelman et al. (2003) in order to be consistent with the vegetation assessments commonly used by natural resource managers in state and federal agencies, although field data collection differed slightly (Table 4). The physical habitat index created by the EPA and described above provided a commonly used qualitative habitat assessment to compare with the IBIs and the Weixelman vegetation index (Resh et al. 1995; Barbour et al. 1999; Harrington and Born 2000; Ode et al. 2005). The fish-only IBI measures habitat suitability and productivity for fishes regardless of whether the fish was of native or introduced origin. The native fish and amphibian index measures the habitat suitability and productivity of native fishes and amphibians and reflected long-term human impacts to native communities. The invertebrate index measures water quality, habitat productivity and availability, and community structure. The vegetation index measures terrestrial and stream bank vegetation as a reflection disturbance and hydrologic conductivity. The habitat index measures available aquatic habitat, stream bed condition, and disturbance. The methods for building IBIs were originally established by Karr (1981) to evaluate fish populations using metrics such as species richness, functional feeding groups, and life stage. The IBI concept evolved under the premise that combining multiple community metrics that respond to different stressors provides a far more reliable indicator of overall ecosystem integrity than a single criterion. A second premise of IBIs was that a scoring system could be

6972


Table 1 Fish-only IBI showing the three metrics used to obtain a final IBI score, biomass, abundance, and species richness Fish-only IBI

Metric Value 1

3

5

Biomass (g/m3)

100

Abundance (n/50 m) Species richness (n/50 m)

50

0–1

2–3

4+

IBI score=(total points/number of metrics)×20 The metric value at the top indicates the score for each of three ranges of values. The scores for the three metrics are then summed, divided by three (the number of metrics in the IBI) and multiplied by 20 to provide a final IBI score out of 100 possible

devised that was easily interpreted by the public and natural resource managers. To that end, Karr (1981) utilized a large data set of many community criteria from an assortment of rivers with a broad range of conditions. Each metric was scored using a 1, 3, or 5 to indicate a range of values for poor, moderate, or good condition. Metric values were determined subjectively by individuals familiar with stream impairment in the region. Since the initial introduction of the IBI concept, the USEPA (Barbour et al. 1999), California Department of Fish and Game (Ode et al. 2005), and Moyle and Marchetti (1999) have developed more quantitative regional IBIs for some parts of the western USA, but as of yet, there is no published IBI specifically for Sierra Nevada meadow systems. Because our study aimed to analyze differences in rapid bioassessment procedures, the IBIs were built on metrics that have been consistently shown by other studies to be important indicators of stream impairment (e.g., Karr and Chu 1997; Barbour et al. 1999; Harrington and Born 2000; Ode et al. 2005). We set individual metric values for the IBIs we built based on Karr’s (1981) model using scores of 1, 3, and 5, respectively to indicate low, moderate, or high condition values. The breaks between values were determined by visually inspecting frequency histograms of each metric from 2 years of field data (Moyle and Marchetti 1999). We used natural breaks present in the upper and lower ends of the histograms for all of our sites combined to determine the breaks between metric values because they likely represented important ecological thresholds better than an

arbitrary percentage. The scores for all of the metrics in a given IBI were combined and then normalized to get a final IBI score as a proportion of 100 total score. This allowed comparison of IBIs from different meadows using a consistent scoring rubric. Fish-only IBI Due to the low species and functional diversity of fishes encountered in most of the Sierra Nevada, we used only three simple metrics for the fish IBI: biomass per cubic meter of habitat, species richness, and total abundance (Table 1). These metrics indicate how well the stream supported fish regardless of if they were native or alien species. The index assumed that the presence of fish in the stream was an indicator of good condition, i.e., that the habitat was of high enough quality to support fish populations. However, the presence of non-native and hatchery origin fishes is a perturbation to native aquatic communities and represents a departure from the historical condition (Knapp 2005; Eby et al. 2006; Schilling et al. 2009). Therefore, we created a second IBI that focused on native fishes and amphibians and regarded non-native fishes and amphibians as detrimental to the condition of the ecosystem (Fig. 1). Native fish and amphibian IBI The native fish and amphibian IBI used eight metrics that included the presence of native trout, percent native species in the sample, number of native species present, number of age classes of native species, fish abundance, fish taxa richness, number of native amphibians, and amphibian taxa

30

Fish-only IBI Score (20-100)

25 20 15 10 5 0 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

Fig. 1 Score distribution of fish-only IBI. The x-axis represents the index score; the y-axis represents the frequency (n=70)

Environ Monit Assess (2012) 184:6967–6986 Table 2 Fish and amphibian IBI with each of the eight metrics and the ranges of values used to obtain a final IBI score

6973

Fish and amphibian IBI

Metric value 1

3

5

Presence of native trout

None

Native trout only

Percentage of native species (number of native individuals/total) Number of native species present

75%

Fish species richness

0–1

2–3

4+

Number of native amphibians (n/50 m)

0

1-3

4+

Native amphibian taxa richness

1

2

3+

IBI score=(total points/number of metrics)×20

richness. We combined the native fishes and the amphibians into a single index because, while we felt it was critical to represent amphibians in the survey of meadow conditions, their presence on the landscape was so rare that they could not support their own index. However, since the historical literature indicates that native amphibians were once common in the areas that we sampled and presumably co-occurred with native fishes, we combined the two taxa into a single index (Grinnell and Storer 1924) (Table 2, Fig. 2).

Invertebrate IBI The invertebrate IBI consisted of seven metrics: (1) the Hilsenhoff family-level index, (2) the EPT index, (3) percent Plecoptera (stoneflies), (4) percent predators, (5) taxa richness, (6) percent Diptera (true flies), and (7) percent Elmidae (riffle beetles). The Hilsenhoff family-level index (Hilsenhoff 1988) provided a measure of organic pollution based on the tolerance values established for individual taxa and their proportionate representation in the sample (see Table 3 for scoring and interpretation). The formula for calculating the HilP senhoff index is HI ¼ ðxi ti Þ=ðnÞð100Þ, where xi is the number of individuals within a species, ti is the tolerance value of a species, and n=total number of organisms in the sample. The second metric was the EPT index, the percent of Ephemeroptera, Plecoptera, and Trichoptera individuals in a sample; values for this metric should increase with improved site condition. These three taxa are considered to be the

most sensitive to disturbance and the least tolerant of poor water quality; they also have broad array of functional morphologies and habitat use. Therefore, the higher the percentage EPT individuals, the better the water quality and habitat complexity (Barbour et al. 1992). Plecoptera were also used separately because stoneflies are consistently the taxon most intolerant of sedimentation and organic pollution and are not necessarily present even though a stream might have a high EPT index (Surdick and Gaufin 1978). Use of Plecoptera abundance twice in the IBI was justified as a way to increase IBI sensitivity to stream degradation. Taxa richness provided a measure of diversity, another metric expected to increase with improved water quality. Percent predators provided a metric of ecosystem condition by describing how well the community supported top predators. While taxa of the predatory guild have varying responses to water

30 25 20

Native Fish and Amphibian IBI Score (20-100)

15 10 5 0 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

Fig. 2 Score distribution of fish and amphibian IBI. The x-axis represents the index score; the y-axis represents the frequency (n=70)

6974 Table 3 Scoring rubric for the Hilsenhoff family-level index from Hilsenhoff (1988)


Biotic index

Water quality

Degree of organic pollution

0.00–3.50

Excellent

No apparent organic pollution

3.51–4.50

Very good

Possible slight organic pollution

4.51–5.50

Good

Some organic pollution

5.51–6.50

Fair

Fairly significant organic pollution

6.51–7.50

Fairly poor

Significant organic pollution

7.51–8.50

Poor

Very significant organic pollution

8.51–10.00

Very poor

Severe organic pollution

quality, their presence was an indication of an environment capable of supporting a multilevel food web (Gross et al. 2009). Percent Diptera, a highly tolerant taxonomic grouping, generally increases with stream degradation or water quality impairment (Barbour et al. 1999; Harrington and Born 2000). Percent Elmidae, a taxon shown to be particularly responsive to mining effluent, was expected to decrease with decreased water quality (Garcia-Criado and Fernandez-Alaez 2001) (Table 4). Vegetation index The vegetation index developed by Weixelman et al. (2003) measured condition by a combination of seral status (e.g., later seral status indicating better condition than early), functional groupings (e.g., obligate wetland plants indicate better condition than facultative or upland plants), amount of bare ground (bare ground being both vulnerable to erosion, and an indicator of disturbance), and plant rooting depth (an indicator of both soil compaction and seral status). See Weixelman et al. (2003) for metric values Table 4 Invertebrate IBI with each of its seven component metrics and the ranges of values used to obtain the final IBI Score

developed for different meadow types and a more in depth description of methods. Habitat index The EPA habitat assessment index developed by Barbour et al. (1999) was designed to provide an assessment of general habitat conditions. While the original index was based on total possible score of 200, we adjusted the scoring to match our other IBIs on a 100-unit scale (see Appendix 1 for metrics and scoring rubric). Index interpretation We calculated scores for each of the five indices for all sites with complete records for all of the parameters measured (n=70). For the purposes of comparing the IBIs, meadow sites were excluded if they were either fishless or not all components of the survey were conducted (nine sites excluded). This helped to ensure that sites were not penalized for lacking fish because most fishless sites were

Metric value Invertebrate IBI

1

3

5

Hilsenhoff index

6–10

4.25–6

26%

10–30%

50% high function

Percent bare ground

0–4%

5–9%

>9%

19 cm

>45% low function

>55% moderate function

>45% high function

Hydric type meadow

Root depth (>100 roots/dm2) Mesic Type Meadow Seral status/functional guild Percent bare ground

>13%

7–13%

0–6%

18 cm

Seral status/functional guild

>45% low function

>55% moderate function

>45% high function

Percent bare ground

>13%

8-13%

100 roots/dm2)

0-3 cm

4-6 cm

>6 cm

Root depth (>100 roots/dm2) Xeric type meadow

IBI score=(total points/number of metrics)×20

6976


Table 6 Percentage of meadows (n=70) for each of the five meadow ecological condition indices, in the condition categories “poor,” “marginal,” “fair,” and “good”

Score Index

20–40 Poor

41–60 Marginal

61–80 Fair

81–100 Good

Fish-only IBI

8

67

16

9

Native Fish and Amphibian IBI

40

41

16

3

Invertebrate IBI

13

31

36

20

Habitat index

3

13

34

50

Vegetation index

1

17

31

51

Fish-only IBI The fish-only IBI had a mean score of 58.7, (SE= 1.9), a maximum of 100, and a minimum of 20 (Table 6). The majority of the sites (67%) scored in the marginal category, while 9% scored in the good category, 16% scored in the fair category, and 8% scored in the poor category. We captured 6,030 fishes of 23 species over the two field seasons. The most abundant taxon captured was speckled dace (Rhinichthys osculus, 37% of fish captured), followed by trout species (31%), and Paiute sculpin (Cottus beldingi, 17%). Of the trout species, brook trout (Salvelinus fontinalis) were the most numerous (18% of fish captured), followed by brown trout (Salmo trutta, 9%), and rainbow trout (Oncorhynchus mykiss, 5%). Lahontan redsides (Richardsonius egregious) made up 6% of the catch. Total fish biomass ranged from 0.1 to 447 g/m3. Abundance ranged from two to 840 fishes per reach.

Native fish and amphibian IBI The native fish and amphibian index had a mean of 48.5 (SE=1.9), a maximum of 85, and a minimum of 30 25

Invertebrate IBI Score (20-100)

20 15

20. The scores for this index were more heavily weighted toward the lower end of the index with 41% of sites rated in marginal condition, 40% in poor condition, 16% in the fair condition, and only 3% in good condition (Table 6). This IBI was structured to give high scores to sites that mainly contained native fishes and amphibians. Overall, the index indicated that native fish and amphibian are generally not abundant or even present in the meadows we studied. Amphibians in particular were very rare and thus drove the index values down. While many of the survey sites occurred in historic mountain yellow-legged frog (Rana sierrae) and Cascade frog (Rana cascadae) habitats that were shown by the Grinnell surveys to contain many amphibians, none were encountered in either the 2006 or 2007 surveys (Grinnell and Storer 1924). In the 2006 field season, only 25 of the study sites had reptiles or amphibians present. We observed Pacific chorus frogs (Pseudacris regilla) at 16 of the sites. Non-native bullfrogs (Lithobates catesbeianus) occurred at three of the sites. Three percent of the sites contained California toads (Bufo boreas halophilus). Reptile observations included western terrestrial garter snakes (Thamnophis elegans) (seven of the sites), western aquatic garter snakes (Thamnophis couchii) (21 of the sites), gopher snakes (Pituophis catenifer) (one site), alligator lizards (Elgaria coerulea) (one site), and western fence lizards (Sceloporus occidentalis) (one site). In the 2007 field season, we found P. regilla at one site and T. elegans at four of the 11 sites.

10

Invertebrate IBI

5 0 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

Fig. 3 Score distribution of invertebrate IBI. The x-axis represents the index score; the y-axis represents the frequency (n=70)

The invertebrate IBI scores indicate that 20% of sites were rated as being in good condition, 36% in fair condition, 31% in marginal condition, and 13% in poor condition. The invertebrate index had a mean of 65.4 (SE, 2.1), a maximum of 100, and a minimum of


27. The mean taxa richness (families) was 19.4, ranging from a minimum of nine to a maximum of 29 families. The mean EPT index was 0.53, ranging from 0.02 to 0.89. The Hilsenhoff index (Hilsenhoff 1988) had a mean of 3.9 and ranged from 2.2 to 7.4, indicating significant variability in water quality throughout the survey meadows. The ephemeropteran family, Baetidae, was the most abundant taxon, dominating (most abundant) the community in 26 of the sites. The dipteran family, Chironomidae, was the next most abundant family, dominating at 18 of the sites. Other abundant taxa included the dipteran family Simuliidae, and the ephemeropteran families, Heptageniidae and Tricorythidae, dominant in 12, 10, and three of the sites, respectively (Fig. 3). Habitat index The results of the habitat index (using the RHA sheet in Appendix 1) indicate that overall meadow condition is better than that indicated by either the fish or invertebrate indices. According to the habitat index, 51% the sites were in good condition, 31% of the sites were in fair condition, 17% of the sites were in marginal condition, and 1% of the sites were in poor condition (Table 6). The habitat index had a mean of 76.0 (SE, 1.7), a maximum of 97, and a minimum of 25 (Fig. 4). The results were strongly skewed to the right with 82% of the sites rated as in either good or fair condition. Vegetation index The results of the vegetation index indicated that 51% of the meadow sites were in good condition, 31% of 30 25 20

Habitat Index Score (20-100)

6977

the sites were in fair condition, 17% of the sites were in marginal condition, and 1% of the sites were in poor condition (Fig. 5, Table 6). The vegetation index had a mean of 79.0 (SE 1.8), a maximum of 100, and a minimum of 33 (Table 7). The survey sites were predominantly (57) mesic (moist) type meadows, with 13 hydric (wet) meadows, and three xeric (dry meadows). In many cases, moisture class was not consistent throughout the entire meadow; 12 of the meadows were mixed hydric/mesic type, nine were mixed mesic/xeric type, and one site had all three types, hydric/mesic/xeric, represented.

Discussion The five indices did not consistently give meadows the same ecological condition ratings (Fig. 6, Table 6). The vegetation and habitat indices tended to rate meadows as being in better condition than the aquatic indices, especially the native fish and amphibian index. The habitat index was correlated with the invertebrate IBI and the vegetation index, but neither of the fish IBIs correlated to any other index (Fig. 6). The lack of correlation between some indices suggested that each index is responding to different drivers or that they are responding at different temporal or spatial scales or that the meadows themselves are sufficiently heterogeneous that they have different capacities to support ecosystem functions and services (Stoffels et al. 2005). For example, invertebrate communities might respond negatively to pollution of the water by livestock (manifested as increased nitrogen and phosphorus as well as increased turbidity), while the surrounding vegetation might respond positively to the input of 25

Vegetation Index

20

Score (20-100)

15

15

10

10

5

5

0

0 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

Fig. 4 Score distribution for habitat index. The x-axis represents the index score; the y-axis represents the frequency (n=70)

25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

Fig. 5 Score distribution for vegetation index. The x-axis represents the index score; the y-axis represents the frequency (n=70)

6978 Table 7 Summary statistics for each of the five meadow condition indices including mean, standard error, kurtosis, skewness, range, minimum, and maximum


Statistic

Fish-only IBI

Native fishamphibian IBI

Invertebrate IBI

Habitat index

Vegetation index

Mean

58.7

48.5

65.4

76.0

78.9

St.Err.

1.9

1.9

2.1

1.7

1.8

Kurtosis

0.76

−0.65

−0.76

2.3

0.17

Skewness

0.51

0.25

−0.16

−1.31

−0.61

Range

80

65

73

72

67

Minimum

20

20

27

25

33

Maximum

100

85

100

97

100

additional nutrients. The mechanisms that drive structure and organization in invertebrate communities range from regional climatic drivers to microhabitat changes at the reach level and below (Stoffels et al. 2005). Fish, with their greater mobility, respond predominantly to influences outside of the reach level such as flow, temperature, and land-use at the watershed level (Lammert and Allan 1999). However, their presence and distribution within a given reach indicate localized habitat preferences (Lammert and Allan 1999). Overall, the different responses of fish and amphibian indices, invertebrate indices, stream habitat, and vegetation indices represent the differential results of legacy effects, on-going changes (e.g., recovery from anthropogenic effects), watershed effects, variable natural conditions, and management actions.

Fig. 6 Box and whisker plots of means, minima, and maxima of each of the indices

Native fish and amphibians The two fish-based indices indicated the poorest condition of the meadows sampled (Fig. 6, Table 6). There are several factors contributing to the observed low scores. Widespread stocking of both native and non-native hatchery trout over the last century has resulted in either fish being present in historically fishless streams or streams that no longer support the native fish fauna. The streams surveyed in the study were predominantly small first and second order streams with small catchment areas. The native trout in our study area consisted of several subspecies of rainbow trout on the west slope and northern Sierra/ southern Cascades, and Lahontan and Paiute cutthroat trout (Oncorhynchus clarki henshawi and Oncorhynchus clarki seleniris) on the east slope and were only

Box and Whisker Plot of Index Means and Ranges Vegetation Index

Habitat index

Invertebrate IBI

Native Fish and Amphibian IBI

Fish-only IBI

Score (20-100) 20

40

60

80

100


present at a small proportion of the survey sites or in the case of Paiute cutthroat trout, not observed in this study. The streams surveyed in the more northerly meadows (i.e. north of Lake Tahoe) tended to have intact native fish faunas, and none were historically fishless, whereas the streams surveyed in the more southern areas tended to be in areas that were likely either historically fishless (due to steep gradients downstream) or no longer support the historic native fish fauna due to extensive stocking. The dominant trout taxon in many sites was non-native brook trout, a species that maintains large populations in headwater streams, excludes other species, and has high densities of individuals with small body sizes (Moyle 2002; Letcher et al. 2007). The fact that the dominant salmonids throughout the study were not native indicates considerable alteration of the historic fish fauna in meadow systems. Indeed, even “native” trout may have been artificially present in many of our streams as the result of stocking at some point. In any case, the nearly ubiquitous stocking of native and non-native trout throughout the Sierra Nevada has been associated with declines of native amphibian populations, especially those of frogs (Knapp and Matthews 1996, 2000; Knapp 2005), although aerial drift of pesticides and novel fungal diseases may play an increasing role (Davidson et al. 2002). Thus, the marked absence of amphibians in the meadows sampled provides a clear case of legacy effects on meadow-associated taxa, rather than being a result of specific contemporary meadow habitat conditions. However, it may be that the extensive grazing that characterized the late 19th and early 20th century and the associated erosion and incision that occurred on many of the meadows also had negative impacts on amphibian populations prior to the introduction of nonnative fishes (Kinney 1996; Dull 1999). The legacy effects of stream channel changes may continue to affect amphibian populations and confound the effects of fish stocking, thus preventing re-colonization. Invertebrates The invertebrate IBI provided a somewhat more positive assessment of meadow condition than the fish/amphibian-based IBIs. Using invertebrate communities to assess habitat condition is a commonlyused tool; however, most invertebrate indices are designed for high gradient, cold temperature streams

6979

(Harrington and Born 2000). In using invertebrates to assess meadow condition, we took into account the distinctive habitat conditions encountered in many meadow systems. Montane meadows are by definition mostly low gradient systems where the substrate is often sandy or silty, an inherent condition that will naturally limit production of coarser substrateassociated individuals, which are often the species associated with better water and habitat quality in high-gradient systems. Water velocity in meadows is commonly low, and there is frequently little woody or shrubby riparian cover, which creates conditions of high solar radiation and warm temperatures, particularly in low-volume streams with small catchment areas. This will cause the invertebrate community to contain more tolerant taxa, which can give the impression of impairment, but may actually represent the unimpaired community for that type of habitat. The ranges for scoring the invertebrate metrics were designed to capture the range of variability we observed in the field. Managers should expect to see lower index scores for open meadows with little riparian vegetation, small catchment basins, low flows, and sandy or silty substrates, whereas, meadows with heavy riparian vegetation, more complex substrate, larger catchment basins, and higher flows should tend to see higher index scores. While the invertebrate community appeared to be a robust indicator for meadow condition, and differentiated between sites well, there are several limitations to relying solely on using aquatic invertebrates as an indicator of condition of the entire meadow system. The first limitation is that once the stream system has stabilized, even if it has entered an alternative state, invertebrate communities may not reflect historical impacts. For example, our data indicated that a meadow stream that is degraded through active erosion by either substrate scouring or silt deposition will generally be reflected accurately by low scores for the invertebrate IBI. However, meadow streams that have significant gullying but that have stabilized can be recolonized by the original invertebrate community, providing a high index score. However, the meadows themselves in such situations often have a lowered water table and a shift of the vegetation community towards more xeric plants. Therefore, invertebrate sampling does not necessarily reveal legacy effects that may be reflected in the other indices. If the substrate has not been greatly altered in

6980

the incision process, the invertebrate community will generally recover within months to years after an impact. This will indicate the current status of the stream bed and new channel, but will not provide a signal for historic impacts represented by loss of the majority of the non-stream meadow habitat. Therefore, we suggest that invertebrate assessments be coupled with quantitative physical habitat and vegetation assessments that capture channel condition, bank erosion, basic water quality parameters, substrate condition, and vegetation cover, type, and seral stage in order to determine whether a meadow stream is functioning normally for its type or if there is some sort of impairment present. However, the index does provide an accurate assessment of existing instream habitat status. Another limitation is that upstream conditions impact invertebrates (as well as the other aquatic indicators), but upstream conditions cannot easily be causally separated from local habitat conditions. Erosion from an upstream timber harvest area might have a profound impact on the invertebrate community downstream through sedimentation, but the cause of those impacts may not be present in the meadow itself. Invertebrate communities are sensitive to changes in condition on small temporal and spatial scales and respond to a variety of factors, which can make it challenging to tease apart the key factors that shape the community (Lammert and Allan 1999). For example, increased nutrient loading from throughout the watershed can increase primary production in meadow streams and result in invertebrate communities that are less driven by local habitat conditions in the meadow (Jackson et al. 2007). Temperature increases from agricultural return water can also shift community structure toward a more tolerant community (Jackson et al. 2007). Sedimentation favors some taxa over others (Angradi 1999). Determining the predominant influences to an invertebrate community requires collecting additional information on water quality, stream geomorphology, potential upstream factors, temperature, and substrate. However, these data are also important in understanding overall condition of the system, are used in several other indices, and are not overly difficult to obtain. The invertebrate IBI had a greater range and more varied results than either the vegetation index or the habitat index. The invertebrate IBI best described short-term conditions within a meadow stream system


and showed particular sensitivity to the effects of scouring, siltation and sedimentation, organic pollution, and thermal changes, all important components of overall ecological condition and function in meadow systems. Habitat index The habitat index showed that most meadows in the study were either in good condition or had significantly recovered from past degradation. The heavily skewed results of the habitat index resulted from it being designed to be very broad and to take into account the full spectrum of stream conditions from pristine to catastrophically impacted. For meadows, the habitat index measures physical changes to geomorphology—particularly incision and erosion—which is a fairly narrow range of the conditions measured by this index. Even the most altered meadow system will not score as low as a heavily degraded urban stream with considerable channel alteration or rip rap. However, despite this index’s lack of sensitivity for assessing streams in a comparatively natural state, there were measurable differences between entrenched, eroding meadow streams versus meandering streams connected to their floodplains, typical of meadows regarded as being in good condition. This index was significantly correlated with the vegetation index (r=0.60, p