State-and-Transition Thinking and NMDS for Laws Type E Revegetation (draft)

Author
Affiliation

Inyo County Water Department

Inyo County Water Department, Bishop, California, USA

Published

July 25, 2026

Abstract

Laws Type E transfer parcels were removed from irrigated agriculture and are managed toward precipitation-dependent native communities under the 2003 revegetation plan. Compliance monitoring (cover, composition, species caps) does not by itself describe ecological states or pathways. We outline how state-and-transition models (STMs) and non-metric multidimensional scaling (NMDS) — methods already used in ICWD’s Owens Valley STM work — apply to these sites. The central conceptual addition is an explicit abandoned-agriculture (ABAG) state: bare or eroded ground after irrigation ceases and topsoil is lost to wind, which stock ecological-site STMs for gravelly/loamy arid sites usually omit. NMDS of parcel–year relative species composition places reference scrub parcels and revegetation parcels in a common Bray–Curtis space so recovery toward reference-like composition can be visualized alongside plan goals. This paper is the citable companion to the STM and NMDS pages on the revegetation-projects site.

Keywords

state-and-transition model, NMDS, Type E transfer, abandoned agriculture, ecological site, gravelly loam, Owens Valley, open science

Draft / sandbox — not an official determination

This manuscript is published in the spirit of open science: methods, drafts, and incomplete analyses stay visible while tools and labeling catch up. Treat it as a sandbox, not adopted County or LADWP policy, not a Technical Group finding, and not a certified compliance product.

Some tables, figures, or joins may be placeholders or sparsely labeled while pipelines are prototyped. Prefer the living repository over any single screenshot; challenge numbers by reading the code and sources.

Introduction

Revegetation success under multi-party water agreements is often scored with percent cover and species lists. Those metrics answer compliance questions; they do not fully answer what state the site is in or which transitions are plausible. State-and-transition models (STMs) organize plant communities, soils, and drivers on an ecological site; non-metric multidimensional scaling (NMDS) tests whether field compositions form coherent clusters that match those states (McCune and Grace 2002; Bestelmeyer et al. 2017).

ICWD maintains an Owens Valley STM program (repository stm) that treats Green Book vegetation types as management-relevant states within NRCS ecological sites and uses NMDS on line-point composition to evaluate type separation. This paper applies that framing to Laws Type E transfer parcels (LAW090, LAW094, LAW095, LAW118/129) and their reference parcels.

Ecological site and the missing state

The 2003 plan’s LAW118/129 planting guidance cites an NRCS ecological site described as “gravely loam,” including cover limits that motivate ATPO capping. Majority-component SSURGO attribution places LAW090, LAW094, LAW095, LAW118, and LAW129 on Gravelly Loam 5-8” P.Z. (R029XG009CA; NV Provisional twin R029XY087NV). That site is Tier 3 in ICWD’s stm catalog (no draft STM page yet); upland pages such as Loamy 5-8” P.Z. remain the closest narrative template. Those models emphasize grazing, drought, shrub dominance, and annual invasion among native or lightly disturbed phases.

Laws revegetation parcels are mapped in the Green Book as Type A · Barren Lands – Abag. Their history is irrigated agriculture followed by retirement from irrigation, often with topsoil loss. That condition is not a standard box on stock gravelly/loamy STMs. We therefore propose an explicit ABAG (abandoned agriculture) state: precip-only barren/eroded land that is the typical start of Type E transfer revegetation, distinct from native barren or intact shrub states.

Irrigated ag (E) → stop irrigation / erosion → ABAG → assisted early seral → reference-like Type A scrub

Failure pathways (chronic weeds, repeated drought, abandonment of maintenance) can return early seral stands toward ABAG; the plan’s option to revert to irrigated agriculture is a separate management branch.

NMDS for reference vs revegetation

We ordinate parcel–year samples built from Laws revegetation workbooks and reference long tables. Species covers are aggregated within parcel–year, row-normalized to relative composition, rare species dropped, and Bray–Curtis dissimilarities mapped with metaMDS (same core steps as stm/code/nmds_functions.R). Absolute cover is reserved for compliance dashboards; NMDS asks whether composition of revegetation parcels approaches the reference cloud.

Live analysis: NMDS page. Conceptual STM: STM page.

Relation to plan goals and caps

Cover and composition goals, ATTO/ERNA caps, and rest periods remain the compliance layer. STM/NMDS are the ecological narrative layer: ABAG → early seral → reference-like scrub. Caps should eventually be interpreted as constraints on transitions (e.g., leaving assisted irrigation only when composition persists in a dry year), not as substitutes for state definitions. Public summary of shared direction vs the open cap question: goals & caps summary.

Open work

  1. Draft a Gravelly Loam 5-8” STM page in stm (R029XG009CA / R029XY087NV) and cross-link; ESD IDs for Type E reveg are already in data/processed/laws_type_e_esd.csv.
  2. Join ESD to reference parcels on the map popup (reveg majority-component known).
  3. Extend LAW118/129 multi-year samples for trajectories.
  4. Represent bare ground explicitly in ordinations so ABAG barrenness is not invisible under relative cover alone.
  5. Optional PERMANOVA / envfit tests mirroring stm/analysis/nmds_gb_types.qmd.

Data and code availability

Source and Quarto site: inyo-gov/revegetation-projects. STM/NMDS helpers: code/nmds_laws.R. Upstream methods: inyo-gov/stm.

References

Bestelmeyer, Brandon T., Andrew Ash, Joel R. Brown, Bulgamaa Densambuu, María Fernández-Giménez, Jamin Johanson, Max Levi, et al. 2017. “State and Transition Models: Theory, Applications, and Challenges.” In Rangeland Systems, edited by David D. Briske, 303–45. Springer.
McCune, Bruce, and James B. Grace. 2002. Analysis of Ecological Communities. Gleneden Beach, Oregon: MjM Software Design.