The Herbaceous Lacustrine Macrophytes of Indiana , United States of America

We provide the first checklist of the obligate aquatic macrophytes of Indiana, including the geographical distribution and frequency of occurrence of each taxon. The checklist is composed of 216 taxa distributed among 85 genera within 43 families. Families exhibiting the greatest richness of taxa are the Potamogetonaceae and Cyperaceae. Approximately 50 % of these taxa are widespread, whereas only 3.7 % are restricted to either the northern or southern regions of the state. An identification code is provided for each taxon and coefficients of conservatism (C values) are given for 189 native taxa, including 18 species of characean algae. C values assigned to native taxa range from 0-10, with a median value of 6. The C values of aquatic macrophytes presented here bear greater similarity to those proposed by Rothrock (2004) for the State of Indiana than to those established by Swink and Wilhelm (1994) for the Chicago region.


Introduction
Despite having a diversity of lakes and a long history of botanical studies, Indiana lacks a synoptic account of its aquatic macrophytes.Classic comprehensive floras of The Herbaceous Lacustrine Macrophytes of Indiana, United States of America Lists o f species information and taxonomic assistance on aquatic plants to environmental consultants as well as a variety of federal, state, and local agencies.In some cases, these cooperative efforts have resulted in the discovery of exotic species new to Indiana, such as Egeria densa Planch., 1849, Hydrilla verticillata (L.f.) Royle, 1839 (Keller 2007;Alix et al. 2009), and Myriophyllum aquaticum (Vell.)Verdc., 1973 (Alix et al. unpubl. data).Our overall experiences, coupled with inquiries and feedback we have received, indicate that the absence of a checklist of the lacustrine macrophytes of Indiana is an impediment to efforts to accurately document and monitor the species richness of Indiana lakes.Therefore, the primary objective of the paper is to provide an up-to-date checklist of the non-woody aquatic macrophytes of Indiana's natural lakes and impoundments and to present information on the distribution, frequency of occurrence, and where applicable, the conservation status of each taxon.A second objective of this paper is to assign coefficients of conservatism (C values) to native taxa included in the checklist.The assignment of C values serves as the foundation of the Floristic Quality Assessment (FQA) methodology developed by Swink and Wilhelm (1994) for the Chicago region, which includes seven counties in northwest Indiana.This methodology, or modifications thereof (see Alix and Scribailo 2006b), provides a rapid assessment tool useful in the evaluation of lake quality (for a comprehensive explanation of FQA, see Swink and Wilhelm 1994).The IDNR's LARE program currently funds a variety of lake projects most of which require aquatic macrophyte surveys as an initial step in lake assessment.Although FQA is not typically utilized in the analyses of these macrophyte surveys, C values provided in the current checklist should help facilitate this process.The inclusion of these proposed C values was also thought to be of importance because of our observation that many of the C values assigned by Swink and Wilhelm (1994) to aquatic macrophytes did not seem to represent an appropriate level of conservatism for these taxa when used in the assessment of Indiana lakes.While studies contributing to the development of the current checklist were in progress, Rothrock (2004) published C values for the vascular flora of Indiana, and Rothrock and Homoya (2005) compared the Indiana values with those established by Swink and Wilhelm (1994).Because one of our goals in presenting the current checklist is to provide C values that most accurately reflect the fidelity of taxa relative to lake quality, a third objective of this paper is to assess the similarity between our C values and those of Swink and Wilhelm (1994) and Rothrock (2004), and to explain some possible reasons for the observed similarities and differences in the values.

Checklist
The primary emphasis has been placed on the inclusion of submerged, free-floating, and floating-leaved aquatic macrophytes associated with littoral zone habitats of Indiana lakes.Many obligate aquatic grasses and sedges, though typically found in wetlands, have been excluded since they are rare inhabitants in the littoral zones of Indiana lakes.All woody aquatic plant taxa have been omitted from the checklist since they are not typically included in aquatic macrophyte surveys designed to evaluate lake quality in Indiana.This checklist represents a compilation and synthesis of historical and current information on aquatic macrophytes obtained from in-and out-of-state sources, such as primary and secondary literature and herbarium records, as well as floristic surveys conducted by the authors over a span of 10 years.The framework of the vascular plant portion of the checklist is based on the classic works of Coulter (1900) and Deam (1940), and other relevant publications and databases, such as Crovello et al. (1983), Swink andWilhelm (1994), Yatskievych (2000), and Rothrock (2004).The charophyte section of the checklist is based on Daily's (1945;1953) studies on the Characeae of Indiana.Characean algae are rarely included in the assessment of floristic quality because they typically are not collected or identified to species and have not been previously assigned C values (Alix and Scribailo 1998;2006b).Since the ecological attributes of characean algae greatly contribute to the ecosystem quality and stability of lakes and ponds (see Hutchinson 1975;Jeppesen et al. 1998;Van den Berg et al. 1998;Coops 2002) and members of this group of macrophytes are a major component of the flora of Indiana lakes in both abundance and diversity, we have included C values for the Indiana members of this group.Information from the aforementioned sources has been supplemented with data obtained from voucher specimens curated at the Kriebel Herbarium of Purdue University (PUL), Indiana University (IND), Field Museum of Natural History (F), University of Notre Dame, South Bend (NDG), Herbarium of the University of Illinois, Urbana (ILL), and the Herbarium of the Chicago Academy of Sciences (CACS).
Current information on the state-wide distribution and frequency of occurrence of many of the taxa listed herein is derived from floristic surveys of 92 natural and manmade lakes carried out from 1993 through 2007 across 21 counties and five ecoregions of Indiana (Table 1, Figure 1).Sampling intensity was greater in the Central Corn Belt Plains and the Southern Michigan-Northern Indiana Drift Plains (Figure 1) since these ecoregions contain a majority of Indiana's natural lakes and have a greater diversity of aquatic macrophytes.These surveys utilized both in-boat (i.e.visual inspections and rake-assisted collections) and in-water sampling techniques, such as snorkeling and SCUBA.

Systematics
Taxonomy and nomenclature of vascular aquatic macrophytes follow familial treatments of the Flora of North America Editorial Committee (1994;1997;2000;2002a;b;2003;2005;2006;2007) with the following exceptions: Apiaceae, Brassicaceae, Lythraceae, Menyanthaceae, Onagraceae, Primulaceae, and Scrophulariaceae (Gleason and Cronquist 1991), Haloragaceae (Aiken 1981), Lentibulariaceae (Taylor 1989), and Plantaginaceae (The Angiosperm Phylogeny Group 2003).Taxonomic treatment of the Characeae follows Daily (1953) with nomenclatural revisions where necessary (see Wood 1965), and that of the Ricciaceae follows Mayfield et al. (1983).Infrageneric designations within Nuphar (Nymphaeaceae) follow the recent monograph by Padgett (2007).Surnames of nomenclatural authorities have been abbreviated following the rules recommended by Brummitt and Powell (1992) and are from the International Plant Names Database (2004).The terms 'taxon' and 'taxa' are commonly used throughout the text in reference to specific or infraspecific taxonomic ranks.

Taxon identification codes
Taxon identification codes (TICs) were created to formally standardize truncations of scientific names of aquatic plant taxa included in the checklist (Table 2).These codes can be used for database entry, the customization of data dictionaries used with global positioning systems (GPS), and shorthand field data entry forms.Indiana TICs have been derived from methods similar to those outlined in Taft et al. (1997).Each TIC for a given taxon consists of the first three letters of the genus followed by the first three letters of the specific epithet (e.g.Potamogeton epihydrus Raf., 1811 = POTEPI).The TIC of a taxon classified at the subspecific or varietal taxonomic rank is made up of the first three letters of the genus, followed by the first two letters of the specific epithet, and ends with the first letter of the infraspecific name (e.g.Potamogeton pusillus L., 1753 subsp.pusillus = POTPUP and P. pusillus subsp.tenuissimus tenuissimus (Mert.and Koch) R. R. Haynes and Hellq., 1996 = POTPUT).To avoid intergeneric code duplication, the third letter in the respective TICs is replaced with the first letter that is different in the spellings of the genera.For example, this type of code duplication would occur between the genera Wolffia and Wolffiella; however, the former genus is represented as WOA and the latter as WOE.Similarly, infrageneric code duplication is avoided by replacing the sixth letter in the respective TICs with the first letter that is different in the spellings of the specific epithets.As an example, infrageneric code duplication would occur with Lemna minor L., 1753 andLemna minuta Kunth. in Humb. et al., 1816 (i.e. both TICs would result in LEMMIN); however, by substituting the sixth letter (N) in each of the codes with the first different letter within their specific epithets results in LEMMIO and LEMMIU, respectively.

Assignment of C values
In regions and states, where the FQA methodology of Swink and Wilhelm (1994) has been adopted or further developed as an assessment tool, the assignment of C values often represents a cooperative effort among professional botanists.Typically, a committee or panel is formed, whose members are familiar with the ecological attributes of taxa within their local flora.Level of invasiveness, sensitivity to disturbance, patterns of occurrence independent of rarity, and fidelity to pre-settlement conditions are some of the key attributes upon which professional botanists base their judgments and assignments of C values (see Swink and Wilhelm 1979;1994;Taft et al. 1997;Nichols 1999;Rothrock 2004).In this study, each author independently assigned C values (AMI C values) to all native taxa included in the checklist (Table 2) based on the following parameters: C values of 0 or 1 are assigned to widespread and common taxa believed to or have been shown to have broad ecological tolerances, often occurring in the most degraded lake habitats and having no apparent fidelity to high quality lake areas, though they frequently may occur in the latter; C values of 2 or 3 are assigned to taxa, which are believed to or have been shown to have little fidelity to high quality lake areas and often occurring in a wide variety of lake habitats; C values from 4 to 6 are assigned to taxa, which are believed to or have been shown to have moderate fidelity to high quality lake areas and often capable of withstanding moderate levels of disturbance; C values of 7 or 8 are assigned to taxa, which are believed to or have been shown to have fidelity to high quality natural areas and are often capable of withstanding minor levels of disturbance; C values of 9 or 10 are assigned to taxa, which are believed to or have been shown to have high fidelity to high quality lake areas and are often intolerant of disturbance and typically restricted to high quality lake habitats.
The above approach resulted in two lists of preliminary C values that were exchanged between each author for review and assessment, which resulted in disagreement on only 10 % of the AMI C values of the taxa included in the checklist.When the difference between two preliminary C values for a given taxon was greater than 1, that taxon was assigned the average of the two values.This method is similar to that outlined in Swink and Wilhelm (1994).When the difference between the C values was 1, the more conservative (i.e.higher) value was assigned to that taxon.The AMI C values were finalized and are provided in Table 2.

Analyses of C values
Two datasets were created for separate, but identical analyses: 1) a set of C values of taxa common between those given in Table 2 and those from Rothrock (2004), referred to herein as SI C values; 2) a set of C values of taxa common between those given in Table 2 and those from Swink and Wilhelm (1994), referred to herein as CR C values.Taxa absent from one source, but present in another (e.g.AMI C values of charophytes) were excluded from these datasets and subsequent analyses.Cumulative frequency distributions of common sets of C values (i.e.AMI vs. SI and AMI vs. CR) and plots of the divergence of AMI C values from SI and CR C values were constructed.The cumulative frequency distributions of common sets of C values were compared by Kolmogorov-Smirnov two-sample tests using SYSTAT ® version 9.1.Differences between frequency distributions were further analyzed using Mann-Whitney U tests utilizing normal approximation and a constant (Zar 1974).Nonparametric two-sample tests were conducted since these data do not meet the assumptions of normality required for the application of analogous parametric statistical tests.P values less than 0.05 are considered significant.

Assignment of C Values
One hundred and eighty-nine native aquatic macrophytes were assigned C values, which included 163 angiosperms from 36 families, 18 species of charophytes represented by three genera, two liverwort species, and six seedless vascular plants from four families (Table 2).C values ranged from 0 to 10, with a median C value of 6; only one taxon, Phragmites australis (Cav.)Trin.ex Steud., 1841 subsp.americanus Saltonst., P. M. Peterson, and Soreng, 2004, was assigned a C value of 0. Seventy percent of taxa were assigned a C value from 5 to 10 (Figure 3A).The frequency distribution of C values is skewed to the left primarily due to a large number of taxa assigned C values of 10 (Figure 3A).Although no conscious emphasis was placed on assigning higher C values to state-listed taxa, 66 % of these taxa have a C value ranging from 8 to 10.

Analysis of C Values
Of the 189 taxa assigned AMI C values (Table 2), 164 taxa are in common with those of Rothrock (2004) and 142 taxa are in common with those of Swink and Wilhelm (1994) Although it is not the purpose of this paper to explain all of the observed differences in C values between common taxa within the AMI, SI, and CR datasets, a representative example can provide some insight as to why some of these disparities may exist.The family Potamogetonaceae contains the largest number of state-listed and total taxa included in the checklist.Of the 21 pondweed taxa in common between our checklist and Rothrock (2004), nine taxa have C values two or more integrals lower than the latter author's values, whereas six have an identical value of 10 and only two are higher.Similarly, of the 19 pondweed taxa in common between our checklist and Swink and Wilhelm (1994), 12 taxa have C values two or more integrals lower than the latter author's values.Five have an identical value of 10 and none are higher.The greater number of C values of 10 assigned by both Rothrock (2004) and Swink and Wilhelm (1994) to pondweeds indicate that they consider a number of taxa to be of higher fidelity to habitats similar to those of presettlement conditions than we suggest.These differences are likely attributable to at least two factors.First, collections of aquatic macrophytes are both historically and currently rare in Indiana, leading to the impression of an apparent rarity and narrow fidelity of some taxa, such as pondweeds.This impression has contributed to the assignment of inflated C values for these taxa and others.Second, a shortage of adequate habitat data on the aquatic macrophytes of Indiana has led to a reliance on information of this type from adjacent states where these taxa do not necessarily exhibit similar presettlement affinities.Observations from our extensive aquatic plant surveys of lakes have indicated that many taxa, pondweeds in particular, are more common and distributed over lakes of a wider range of water quality and disturbance than is suggested by the higher C values assigned by Rothrock (2004) and Swink and Wilhelm (1994).It is also important to note that the general tendency for CR C values to be significantly higher than those of both the SI and AMI C values may be a reflection of the fact that a greater proportion of taxa will appear to have higher fidelity when the region for which FQA is developed is geographically smaller.The results of this study underscore the importance of further ecological studies of the aquatic macrophytes of Indiana.These studies, particularly if they were coupled with the collection of habitat data, would provide additional information on the nature of habitat fidelity for some taxa and contribute to the refinement of their C values, thus improving the effectiveness of FQA (Swink and Wilhelm 1994) for the evaluation of lake quality.The presentation of this checklist will hopefully provide a tool useful in the facilitation of further floristic studies on Indiana lakes.
Table 2. Checklist of obligate aquatic macrophytes of lacustrine habitats in Indiana.Taxa are arranged in a generally accepted systematic sequence by phylum and alphabetically by family, genus, species, and when applicable, subspecific and varietal ranks, respectively.For each taxon at or below the species rank, a common name, C value(s), taxon identification code (TIC), distributional range, and frequency of occurrence (F) have been included.A state rank (SR) has been provided for each state-listed taxon.C value: CR = Chicago Region (Swink and Wilhelm 1994); AMI = aquatic macrophytes of Indiana (proposed here); SI = State of Indiana (Rothrock 2004).A black circle (•) indicates that a taxon is considered non-native by the author(s), a black dagger ( †) indicates that a C value was not assigned to a hybrid taxon to a hybrid, and a horizontal bar (-) indicates that the taxon is not listed by the author(s).

Figure 1 .
Figure 1.Maps of Indiana.Right, counties of Indiana.Left, level III ecoregions of Indiana (adapted from Omernik and Gallant 1988).Values in the legend boxes represent the total number of water bodies surveyed from a given ecoregion used in the assembly of the checklist and in the assignment of C values.
. The frequency distributions of AMI and SI C values (Figure 3B) are not significantly different as indicated by the Kolmogorov-Smirnov two-sample test (D max = 0.116; two-tailed P = 0.221).In contrast, the frequency distributions of AMI and CR C values (Figure 3C) are significantly different (D max = 0.317; two-tailed P <0.0001).The frequency distribution of CR C values is clearly skewed further to the left than that of the AMI C values (Figure 3C), indicating that Swink and Wilhelm (1994) assigned much higher C values to a majority of the taxa listed in Table2.In fact, only seven taxa out of the 142 in common between the AMI checklist and the Chicago region have been assigned a C value <4, whereas over 75 % of the remaining taxa have been assigned a C value ≥6 (Figure3C).A taxon by taxon analysis of the divergence of AMI C values from those of SI and CR indicates that 91 % and 36 % of the aquatic plant taxa listed in Table2have been assigned the same C values, respectively (Figure4A).The overall mean divergences of AMI C values from those of SI and CR are 0.8 and 1.9, respectively.As a whole, AMI C values are approximately two C values lower than CR C values (Mann-Whitney, U = 9427; Z = 3.03; one-tailed P <0.001).Only 25 % of aquatic plant taxa diverged from SI C values by 2 or more integrals, whereas 57 % of taxa diverged by at least 2 integrals from CR C values (Figure4B).The largest differences in AMI C values from SI C values are observed in taxa assigned a C value of 7 and 8 where the C values have a mean divergence of 1.3 and 1.4, respectively (Figure4B).In contrast, the greatest mean divergence (3.4) of AMI C values from CR C values is observed in taxa assigned a C value of 3 (Figure4B).

Figure 2 .
Figure 2. Frequency distributions of aquatic plants of Indiana: A) grouped by state status; B) grouped by frequency of occurrence.Frequency represents the number of taxa within the same group.N = total number of taxa within all groups.

Figure 4 .
Figure 4. Frequency distributions of the divergence of AMI C values from SI and CR C values: A) overall divergence; B) mean divergence by C value.Frequency represents the number of taxa within a group.

Figure 3 .
Figure 3. Frequency distributions of Indiana C values of aquatic plants: A) Indiana distribution; B) comparison of AMI C values with SI C values (Rothrock 2004); C) comparison of AMI C values with CR C values(Swink and Wilhelm 1994).Frequency represents the number of taxa within a group.
Range: Z = statewide; N = north; S = south; E = east; W = west; C = central.Frequency of occurrence: A = abundant; C = common; O = occasional; R = rare.Each designation is largely defined by the presence/absence of a taxon across multiple aquatic habitats.State ranks: X = state extirpated; E = state endangered; T = state threatened; R = state rare; WL = watch list (from Indiana Natural Heritage Program 2007).

Table 1 .
Summary of Indiana lakes and reservoirs surveyed between 1993 and 2007, including county, survey year (in parentheses), and level III ecoregion (Omernik and Gallant 1988).Abbreviations: CCBP = Central Corn Belt Plains; ECBP = Eastern Corn Belt Plains; IP = Interior Plateau; IRVH = Interior River Valley and Hills; SMNIDP = Southern Michigan-Northern Indiana Drift Plains.

Table 2 . (Continued).
Alix and Scribailo | The Herbaceous Lacustrine Macrophytes of Indiana, United States of America