U.S. Naonal Vegetaon Classificaon: Advancing the Descripon and Management of the Naon’s Ecosystems Revising the USNVC using quantave methods and plot data: A case study using longleaf pine (Pinus palustris) vegetaon of the southeastern Robert K. Peet (University of ) , Kyle A. Palmquist (University of Wyoming), and Michael Schafale (North Carolina Natural Heritage Program)

BACKGROUND AND GOALS A CASE STUDY - REVISING USNVC LONGLEAF PINE GROUPS (G154, G009, G190, G596) • The USNVC is a hierarchical vegetaon STEPS 1-3 Cluster Dendrogram STEP 4 classificaon system for the US (Fig. 1). 5 • The Federal Geographic Data Commiee A) A) A) B) B) B) C) C) C) 4 mandates that USNVC Alliances and G154 G154Xeric G154

Associaons be based on analysis of 3

publicly-available plot data. Height G009 G009 G009 2 • To date, few types have been defined A) NC-SC: NMS 1&2 B) NC-SC: NMS 1&3 C) SC-GA: NMS 1&2

6 G596 G596 G596 4 using this standard. CEC.BCa.B 3584 6 7844

1 Long

Lat 4 3590 7127

• Longleaf pine (Pinus palustris) woodlands 2 G190 G190 G190 Fig. 1. USNVC classificaon scheme. 3592 Wet 4 4492

2 Sand

0 3589 7125 − 9 067 − APAL − 1 067 − APAL 067 − OCAL 3 067 − OCAL 2 001 − 01 0203 001 − 01 0206 001 − 05 0207 001 − 05 0208 001 − 05 0206 044 − 02 0605 044 − 09 0607 044 − 04 0610 044 − 09 0605 112 − 04 1381 112 − 04 1382 128 − 04 1482 044 − 04 0611 044 − 09 0606 112 − 04 1380 003 − 01 0084 044 − 02 0604 003 − 03 0086 044 − 07 0614 044 − 07 0608 075 − 01 1006 028 − 01 0059 028 − 01 0060 003 − 01 0108 003 − 05 0122 003 − 01 0109 089 − 01 1233 089 − 04 1235 089 − 01 1235 002 − 01 0042 088 − 07 1206 003 − 02 0082 003 − 03 0085 044 − 07 0613 044 − 09 0603 044 − 09 0602 003 − 06 0101 016 − 01 0078 016 − 01 0079 002 − 01 0003 002 − 02 0003 002 − 02 0004 002 − 04 0011 002 − 02 0040 002 − 01 0008 002 − 06 0035 002 − 01 0033 002 − 05 0035 002 − 04 0001 002 − 04 0004 002 − 01 0035 002 − 02 0032 002 − 02 0030 002 − 05 0001 002 − 05 0030 002 − 05 0020 002 − 02 0011 002 − 05 0022 002 − 04 0032 003 − 01 0077 003 − 03 0106 064 − 04 0958 003 − 07 0075 016 − 01 0077 003 − 01 0085 003 − 07 0078 075 − 04 1007 003 − 01 0080 003 − 02 0079 003 − 06 0078 003 − 04 0078 003 − 06 0102 075 − 04 1008 002 − 01 0009 002 − 04 0006 002 − 04 0035 002 − 03 0034 002 − 03 0035 002 − 01 0037 002 − 04 0036 002 − 05 0039 002 − 04 0005 003 − 02 0083 003 − 01 0054 003 − 02 0054 003 − 01 0060 003 − 02 0064 003 − 02 0062 003 − 02 0063 003 − 07 0053 116 − 08 0003 116 − 08 0005 003 − 03 0052 003 − 05 0055 003 − 02 0052 003 − 05 0052 003 − 05 0051 003 − 04 0051 003 − 04 0052 003 − 04 0054 003 − 06 0051 002 − 01 0004 002 − 02 0006 002 − 05 0009 002 − 02 0038 002 − 02 0034 128 − 11 1470 128 − 04 1472 128 − 05 1474 128 − 05 1476 111 − 02 1351 002 − 02 0035 002 − 05 0034 002 − 03 0031 128 − 04 1483 002 − 02 0039 002 − 04 0037 002 − 05 0040 002 − 04 0002 002 − 04 0033 002 − 06 0031 002 − 05 0010 002 − 01 0006 002 − 04 0009 002 − 02 0007 002 − 02 0002 002 − 05 0002 002 − 05 0005 002 − 01 0041 002 − 05 0033 002 − 04 0038 002 − 02 0008 002 − 04 0010 002 − 03 0002 002 − 01 0040 002 − 02 0009 028 − 01 0128 003 − 01 0044 003 − 05 0101 003 − 01 0100 003 − 02 0100 003 − 04 0100 003 − 05 0100 002 − 01 0031 002 − 04 0040 002 − 02 0033 002 − 05 0004 002 − 05 0032 003 − 04 0044 002 − 02 0001 002 − 02 0031 002 − 05 0031 002 − 06 0034 002 − 03 0003 002 − 05 0012 003 − 01 0103 003 − 04 0102 003 − 02 0102 082 − 04 1106 002 − 02 0036 002 − 04 0031 002 − 05 0036 002 − 06 0030 002 − 03 0033 002 − 04 0030 002 − 04 0003 002 − 05 0007 002 − 05 0008 003 − 02 0101 003 − 04 0101 002 − 01 0014 110 − 04 1300 063 − 02 0929 002 − 02 0037 003 − 02 0053 003 − 05 0058 044 − 04 0608 088 − 04 1203 064 − 09 0955 016 − 01 0036 016 − 01 0037 044 − 06 0608 044 − 07 0610 044 − 09 0604 003 − 01 0058 003 − 04 0058 003 − 05 0060 003 − 06 0052 003 − 02 0058 003 − 04 0059 116 − 08 0001 003 − 02 0059 003 − 01 0083 003 − 02 0076 003 − 05 0063 003 − 06 0055 002 − 01 0030 003 − 05 0104 027 − 04 0561 003 − 01 0112 088 − 02 1203 088 − 04 1200 002 − 01 0043 002 − 02 0041 003 − 05 0110 003 − 01 0107 003 − 03 0103 003 − 04 0105 003 − 07 0100 003 − 02 0105 016 − 01 0002 016 − 01 0007 016 − 01 0010 016 − 01 0011 016 − 01 0014 016 − 01 0015 027 − 07 0456 044 − 07 0605 027 − 07 0457 044 − 07 0612 044 − 02 0602 044 − 09 0601 016 − 01 0025 044 − 07 0607 016 − 01 0039 044 − 04 0605 016 − 01 0001 044 − 06 0603 034 − 01 0222 034 − 01 0262 016 − 01 0104 016 − 01 0105 016 − 01 0106 016 − 01 0107 002 − 01 0005 002 − 01 0032 002 − 04 0013 002 − 01 0036 002 − 04 0041 002 − 05 0037 002 − 05 0041 002 − 01 0039 002 − 04 0008 002 − 04 0007 044 − 07 0602 002 − 04 0043 082 − 09 1105 111 − 08 1350 003 − 02 0121 110 − 01 1300 110 − 04 1301 117 − 06 1400 112 − 05 1384 112 − 05 1385 003 − 01 0056 003 − 05 0061 003 − 03 0053 117 − 04 1400 003 − 03 0054 003 − 03 0082 003 − 01 0057 116 − 08 0004 003 − 02 0057 003 − 04 0056 003 − 05 0057 003 − 02 0055 003 − 05 0056 003 − 01 0059 003 − 05 0059 11 068 − FRMA 6 067 − OSCE 3 − 1 067 − CROA 003 − 0R 0071 003 − 0R 0072 003 − 0R 0073 015 − 0R 0002 015 − 0R 0009 015 − 0R 0006 015 − 0R 0019 015 − 0R 0010 015 − 0R 0020 015 − 0R 0023 015 − 0R 0036 015 − 0R 0038 015 − 0R 0037 015 − 0R 0039 015 − 0R 0018 015 − 0R 0047 015 − 0R 0001 015 − 0R 0003 015 − 0R 0005 015 − 0R 0007 015 − 0R 0008 015 − 0R 0029 015 − 0R 0030 015 − 0R 0031 015 − 0R 0033 015 − 0R 0034 015 − 0R 0004 003 − 0R 0069 003 − 0R 0070 015 − 0R 0011 015 − 0R 0012 015 − 0R 0016 015 − 0R 0017 015 − 0R 0024 015 − 0R 0025 015 − 0R 0027 015 − 0R 0050 015 − 0R 0055 015 − 0R 0044 015 − 0R 0046 015 − 0R 0048 015 − 0R 0058 015 − 0R 0059 015 − 0R 0066 015 − 0R 0013 015 − 0R 0021 015 − 0R 0056 015 − 0R 0057 015 − 0R 0063 015 − 0R 0065 015 − 0R 0064 015 − 0R 0060 015 − 0R 0062 015 − 0R 0061 015 − 0R 0040 015 − 0R 0041 015 − 0R 0042 015 − 0R 0043 015 − 0R 0014 015 − 0R 0015 015 − 0R 0049 015 − 0R 0022 015 − 0R 0026 015 − 0R 0045 015 − 0R 0051 015 − 0R 0052 015 − 0R 0054 015 − 0R 0053 015 − 0R 0028 015 − 0R 0032 015 − 0R 0067 015 − 0R 0068 − 32 067 − APAL − 62 067 − APAL − 49 067 − APAL − 53 067 − APAL − 54 067 − APAL − 31 067 − APAL − 15 067 − APAL − 59 067 − APAL − 60 067 − APAL − 77 067 − APAL − 83 067 − APAL − 76 067 − APAL − 29 067 − APAL − 73 067 − APAL − 40 067 − APAL − 51 067 − APAL − 81 067 − APAL − 82 067 − APAL − 10 067 − APAL − 52 067 − APAL 067 − OCAL 23 067 − OCAL 51 067 − OCAL 16 067 − OCAL 30 067 − OCAL 61 067 − OCAL 31 067 − OCAL 11 067 − OCAL 24 067 − OCAL 33 067 − OSCE 16 067 − OSCE 37 067 − OSCE 18 067 − OSCE 27 067 − OSCE 35 067 − OSCE 29 067 − OSCE 10 067 − OSCE 23 067 − OSCE 17 067 − OSCE 44 067 − OSCE 45 067 − OSCE 46 067 − OSCE 47 − 33 067 − CROA − 21 067 − CROA − 22 067 − CROA are represented by ~120 Associaons and 068 − FRMA 105 068 − FRMA 102 0

NC-SC SC-GA FL-GA 3577 2 Sand.B 3584 0 0 20 040 20600408020601004080 60100800 10020 040 20600408020601004080 60100800 100 500 100500 15010050 150100 150 veg.dist adequate descripon of these units is crical agnes (*, "flexible") 3586

-2 3590

Sand.B 3 NMDS Fig. 2. Plots first sorted into coarse types based on geography (NC-SC, SC-GA, FL-GA), not 2 NMDS 2 NMDS % Silt % Silt % Silt % Sand % Sand% Sand RichnessRichness at 1000Richness m2at 1000 atm2 1000 m2 0 -2 Lat for effecve conservaon and management. 3592 according to Groups, due to high geographic turnover in species composion. As a result, pH.B -4 3589

-4 Fig. 3. Silt and sand %, along with species richness differ across USNVC we had to use both the clustering soluon and expert opinion to parse plots into Groups. 3577 -2 GOALS: Revise the USNVC longleaf pine Groups, Alliances, and Associaons using a Groups. Boxplots show the range and median values for plots of each Group. r100r1 -6

3586 -6

consistent analycal approach to document vegetaon-environment relaonships. -4 Provide an example of use of plot data and quantave methods to revise the A CASE STUDY - REVISING ASSOCIATIONS OF G154: XERIC LONGLEAF PINE WOODLAND-6 -4 -2 0 2 4 6 -8 -6 -4 -2 0 2 4 6 -6 -4 -2 0 2 4 6 NMDS 1 NMDS 1 NMDS 1 established qualitavely-defined vegetaon types. STEPS 1-4 STEP 6 Association 7127 7125 7844 4492 A) NC-SC: NMS 1&2 B) NC-SC:A) NC-SC: NMS 1&3 NMS 1&2 D) SC-GA:C) SC-GA: NMSB) NC-SC: NMS1&3 1&2 NMS 1&3 E) FL-GA: NMSC) SC-GA: 1&2 NMS 1&2 Tree speciesF) FL-GA: NMS 1&3 Const Cov Const Cov Const Cov Const Cov 6 6 4 4 6 6 CEC.BCa.B CEC.BCa.B Pinus palustris6 Long 4490 100% 5 100% 6 86% 6 100% 6 10 Long 3584 Long 4 7844 3584 7844

Lat 4 Lat 4 Lat OM ANALYTICAL APPROACH 3590 r17127 3590 Silt.B7127 Quercus hemisphaericaCa 7141 100% 2 ------

pH 4 CEC.B 2 2 2 3592 r1004 4492 3592 4 4492 Quercus incanaCa.B CECSand.B 3583 ------79% 4 86% 3 2

2 2 Sand Sand DATA: 1101 longleaf pine vegetaon plots (100 to 1000 m ) distributed from to 5 7133 3589 Silt.B 7125 3589 7125 Quercus laevis2 pH.B 100% 7 100% 6 100% 6 100% 6

0 0 r1

3577 2 3577 Sand.B pH2 Sand.B 7132 . For each plot: presence of all vascular species and cover class codes. Soil 0 0 r100 Quercus margarettae 100% 5 ------71% 6 0 3584 3586 35840 3586 4490 7137 0 7141 Shrub speciesr100 r1 Const Cov Const Cov Const Cov Const Cov -2 3590 -2 3590 pH.B

Sand.B 3 NMDS Sand.B 3 NMDS NMDS 3 NMDS 2 NMDS 3 NMDS NMDS 2 NMDS 2 NMDS 2 NMDS 2 NMDS 0 0 properes for 90% of plots. -2 -2 Lat Lat 3592 3592 7844 Sand Ca.B 3583 Gaylussacia dumosa-2 -- -- 43% 6 71% 4 -- --

-2 Sand.B pH.B pH.B Silt.B -4 3589 -4 3589 7127 7133 Licania michauxii ------100% 3 -4 -4 CLASSIFICATION METHODS: Ulized to define Groups, Alliances, and Associaons. -4 -5 Lat 3577 3577 -2 4492 -2 Sand.B 7132 Serenoa repens ------71% 4 Long

-4 r100 r100 r1 r1 -6 -6 -6 3586 -6 3586 7125 -6 Lat Long 7137 Herb species Const Cov Const Cov Const Cov Const Cov Agglomerave hierarchical clustering with flexible β and -4 -4 [Andropogon + Schizachyrium] 100% 2 100% 4 100% 4 -- -- mulple cluster numbers (2 to the number of exisng types)-6 -4 -2 0 2 4 6 -8 -6 -4-6 -2-4 0-2 2 0 4 2 6 4 6 -4 -6 -2 -8-4 0-6-2 -42 0 -24 2 0 642 46 6 -10 -5-6 0-4 -25 0 10 2 4Aristida beyrichiana6 -5 0 5 ------64% 7 100% 5 NMDS 1 NMDS NMDS1 1 NMDSNMDS 1 NMDS1 1 NMDS 1 NMDS 1 Aristida purpurascensNMDS 1 100% 2 -- -- 71% 2 -- -- Aureolaria pectinata ------64% 2 86% 2 STEP 1 Fig. 4. Idenfied 178 plots and 17 Associaons (each idenfied with a unique 4 digit code Cnidoscolus stimulosus -- -- 71% 2 71% 2 71% 2 Eriogonum tomentosum ------64% 2 86% 2 Presence-absence data Abundance data D) SC-GA: NMS 1&3 called a CEGL) that represent xeric and subxeric longleaf pine vegetaon (G154) within three E) FL-GA:D) SC-GA: NMS 1&2 NMS 1&3 F) FL-GA:E) FL-GA: NMS 1&3 NMS 1&2 F) FL-GA: NMS 1&3 geographic bins that correspond to Alliances (NC-SC, SC-GA, FL-GA). Euphorbia ipecacuanhae -- -- 71% 2 57% 2 -- -- 6 Long 4490 6 Long 4490 10 10 4 4 Liatris tenuifolia ------71% 2 Lat OM Lat OM r1 r1Silt.B Ca Silt.B 7141 Ca Pityopsis graminifolia7141 -- -- 71% 2 93% 2 71% 2

pH pH 4 CEC.B 4 CEC.B r100 r100 Ca.B CECSand.B 3583 Ca.B CECSand.B 3583 2 2 megalocarpa -- -- 71% 2 ------5 5 7133 7133 2 pH.B 2 pH.B Silt.B STEPr1 Silt.B 5 r1 Stipulicida setacea ------86% 2 -- -- Cluster validaon: optpart and silhouee width STEP 2 pH pH 7132 7132 r100 r100 0 0 0 CEGL0 N Current Quantitative Association4490 Relationship7137 4490 Previous Qualitative Association7137 Table 2. Mean species constancy 0 pH.B 7141 r1000 r1 pH.B 7141 r100 r1 NMDS 3 NMDS 2 NMDS 3 NMDS 3 NMDS 2 NMDS 3 NMDS 7844 Sand 7127 Ca.B3 7844Pinus palustris / Quercus laevis / Chrysoma Sand3583 -2 Ca.B 3583≠ Chrysoma pauciflosculosa -2 - (Clinopodium coccineum ) Dwarf-shrubland (const) and cover class code (cov)

-2 Sand.B -2 Sand.B 7127 7127 7133 Silt.B 7133 Silt.B

pauciflosculosa / Aristida purpurascens-4 Woodland -4

-5 -5 Lat Lat 4492 4492 Sand.B 7132 Sand.B 7132 for Associaons in the SC-GA Look for agreement between presence-absence & Long Long

-4 7125-4 7 Pinus palustris / Quercus laevis – Quercus geminata ≠ Pinus palustris / Quercus laevis / Aristida purpurascens / Stipulicida setacea 7125 LatSTEP 3 Long7125 Lat 7137 -6 Long 7137 -6 Alliance. Indicator species are abundance-weighted clustering soluons / Schizachyrium scoparium Woodland (Rhynchospora megalocarpa, Selaginella acanthonota) Woodland highlighted in gray. Species are -4 -2 0 2 4 6 7844-10 14-4-5Pinus palustris-2 0 0 5 2 / 10Quercus laevis4 6 / -5Schizachyrium -10 0 -5 05 5 = 10 Pinus palustris-5 / Quercus laevis0 5 - Quercus incana / Aristida beyrichiana - NMDS 1 NMDS NMDS1 1 NMDS NMDS1 1 NMDS 1 separated according to their scoparium – Stipulicida setacea Woodland Baptisia perfoliata Woodland growth form (tree, shrub, herb). 4492 7 Pinus palustris / Quercus laevis – Q. margarettae / ~ Pinus palustris / Quercus laevis – Quercus incana / Quercus margarettiae / Constant and indicator species Explore environmental-vegetaon relaonships using STEP 4 Licania michauxii / Aristida beyrichiana Woodland Licania michauxii / Aristida beyrichiana Woodland NMDS and mean soil properes for each cluster Table 1. Two clusters within the SC-GA Alliance represent modest (=) and significant (~) levels of refinement of exisng Associaons (7844, were used to define Associaons. 4492). Two are new Associaons (7127, 7125) that have some composional affinies (≠) with previously defined types.

Interpret clusters, map them onto exisng types in the CHALLENGES AND OPPORTUNITIES USNVC or designate new types STEP 5 Challenge: Exisng USNVC units vary in their origin, level of clarity and focus, and the amount of data, analysis, and experience behind them. Challenge: In many regions, publicly-available vegetaon plot data are limited and are largely unavailable from disturbed or managed lands. Where plot data are available, variaon in field protocols and taxonomic standards makes data integraon difficult. Generate constancy tables for each type and idenfy Challenge: Quantavely defined units oen do not map cleanly on to exisng types. This discordance may indicate a need to correct the boundaries of types, diagnosc species using indicator species analysis STEP 6 but it may simply be the result of connuous variaon in vegetaon. Opportunity: The USNVC documents ecological context using a common language. Definion of vegetaon types using quantave data leads to consistent, data-driven descripons that are useful for the management, conservaon, and basic science sectors. Acknowledgements: We thank the US Geologic Survey for financial support and Susan Carr, Alan Weakley, and Thomas Wentworth for parcipaon in the analysis. This work would not have been Opportunity: The broader user community is empowered to submit formal proposals for revision of vegetaon types, thereby leading to improvement of the possible without the many volunteers who parcipated in Carolina Vegetaon Survey PULSE events. USNVC and recognion for these scholarly contribuons.