One-ended spanning subforests and treeability of groups

Imagine you are trying to understand the shape and structure of a massive, invisible city. This city is made of "groups" (mathematical structures that describe symmetry and movement), and the streets are the connections between different points in the city.

The paper you asked about is like a team of urban planners (mathematicians) who have discovered a new way to map these cities. They found that for many complex, non-smooth cities, you can draw a special kind of map that turns the chaotic web of streets into a perfect, branching tree.

Here is the breakdown of their discovery using simple analogies:

1. The Big Goal: Turning Chaos into Trees

In mathematics, a "tree" is a shape with no loops. If you walk down a tree, you never come back to where you started unless you turn around.

The Problem: Many groups (cities) have loops everywhere. It's like a subway system where you can take a train in a circle and end up back at the start. This makes them hard to analyze.
The Solution: The authors proved that for a huge new class of these "cities," you can always find a way to remove just enough connections to break all the loops, leaving behind a giant, infinite tree structure.
Why it matters: Trees are much easier to study than messy webs. If you can turn a complex group into a tree, you can solve problems about it much faster.

2. The Secret Weapon: The "One-Ended" Forest

The authors didn't just find any tree; they found something very specific called a "one-ended spanning subforest."

The Analogy: Imagine a forest where every tree has a single, infinite trunk that goes on forever in one direction (like a giant palm tree reaching for the sky).
The "One-Ended" part: In math, a "two-ended" tree looks like a long straight road (you can go left forever or right forever). A "one-ended" tree looks like a river flowing into the ocean; it has a start, but only one direction to go infinitely.
The Discovery: They proved that for certain groups, you can always cut the city's streets to create a forest where every tree flows in only one direction. This is the "magic key" that unlocks the ability to turn the whole group into a tree.

3. The Three New Cities They Mapped

The paper shows that this "tree-making" trick works for three specific types of groups that were previously mysterious:

Planar Cities (Groups with Planar Cayley Graphs):
- Imagine: A city that can be drawn on a flat piece of paper without any roads crossing each other (like a map of a subway system where lines never intersect).
- The Result: If a group's map is flat and non-crossing, it can always be turned into a tree. This includes the fundamental groups of surfaces like donuts (tori) or pretzels.
Elementarily Free Groups:
- Imagine: These are groups that "act" exactly like free groups (the simplest kind of groups, like a collection of independent strings) in every logical way, even if they look complicated on the outside.
- The Result: Even though they look complex, they are secretly just trees in disguise.
The Hyperbolic Plane Group (Isom(H²)):
- Imagine: The group of all possible movements (rotations, slides) you can do on a saddle-shaped surface (hyperbolic geometry).
- The Result: This is a massive, continuous group (like a fluid). The authors proved that every subgroup of this group can also be mapped into a tree. This is a huge deal because it's the first time we've seen such a complex, non-smooth group behave so simply.

4. The "Dual" Trick: Looking at the Back of the Map

How did they do it? They used a clever trick called duality.

The Analogy: Imagine you have a map of a city (the graph). Now, imagine drawing a new map where every neighborhood (face) of the city becomes a point, and every road becomes a connection between neighborhoods. This is the "dual" map.
The Magic: The authors realized that if you look at the dual map, finding a "one-ended tree" there is actually the same as finding a "tree" in the original city. It's like solving a puzzle by looking at its shadow instead of the object itself. They used this shadow to build the tree structure for the original group.

5. The 3D Manifold Mystery

The paper also tackles 3D shapes (like the universe we might live in, if it were a closed box).

The Question: If you have a 3D shape that is "aspherical" (meaning it has no holes that can't be shrunk), what is the complexity of its symmetry group?
The Answer: They proved a "Dichotomy" (a split into two options):
1. The group is "Amenable" (simple, like a flat sheet).
2. The group has a specific complexity level called "Ergodic Dimension 2."
- Translation: You can't have a 3D shape with a group that is "in between" simple and complex. It's either simple, or it hits a specific ceiling of complexity.

Summary

This paper is a breakthrough because it takes groups that were thought to be too messy, too curved, or too complex to simplify, and shows that they all hide a simple tree structure inside them.

They did this by inventing a new way to look at the "shadows" (dual graphs) of these groups and finding a specific type of "one-way forest" within them. This allows mathematicians to apply the simple rules of trees to solve difficult problems about the geometry of the universe, 3D shapes, and abstract algebra.

1. Problem Statement and Context

The paper addresses fundamental questions in measured group theory and the theory of Borel equivalence relations, specifically concerning the treeability of groups.

Treeability: A countable group $\Gamma$ is treeable if it admits a free probability measure-preserving (p.m.p.) action whose orbit equivalence relation can be generated by a Borel graph that is a forest (acyclic).
Strong Treeability: A group is strongly treeable if all its free p.m.p. actions are treeable. This is a much stronger condition. While free groups and amenable groups are known to be strongly treeable, the status of many other non-amenable groups (like surface groups) remained open for two decades.
The Core Question: The authors aim to identify new classes of non-amenable groups that are strongly treeable and to resolve the long-standing question of whether surface groups (fundamental groups of closed hyperbolic surfaces) are strongly treeable.
Higher Dimensions: The paper also investigates the ergodic dimension of groups, defined as the smallest geometric dimension of a contractible complex on which the group acts freely in a p.m.p. setting. For 3-manifold groups, the question is whether they are amenable or have strong ergodic dimension 2.

2. Methodology and Technical Tools

The authors develop a novel geometric-combinatorial approach to constructing treeings, moving beyond standard algebraic methods.

A. One-Ended Spanning Subforests

The central technical innovation is the construction of one-ended spanning subforests in Borel graphs.

Definition: A spanning subforest where every connected component is a tree with exactly one "end" (a way to go to infinity).
Characterization (Theorem 2.1): For a locally finite, measure-preserving, aperiodic Borel graph $G$ , a Borel a.e. one-ended spanning subforest exists if and only if $G$ is $\mu$ -nowhere two-ended (i.e., the set of vertices belonging to two-ended components has measure zero).
Construction Techniques:
- Quadratic Growth: For graphs with at least quadratic growth, they construct such subforests using a sequence of partial functions based on maximal separated sets (Theorem 2.6).
- Hyperfinite Graphs: For $\mu$ -hyperfinite graphs, they use a "shrinking" argument to reduce the problem to finding subforests in one-ended components (Lemma 2.9).
- The "One Percent Theorem" (Theorem 2.14): A dual statement to the classic "99% theorem" of hyperfiniteness, showing that if a graph is nowhere hyperfinite, one can find a small (measure $<\epsilon$ ) complete section that is also nowhere hyperfinite.

B. Planar Graphs and Duality

The paper leverages the duality between planar graphs and their duals to solve the treeability problem.

The Strategy: Instead of finding a treeing directly in a planar graph $G$ , they construct a one-ended spanning subforest in the planar dual graph $G^*$ .
Theorem 3.6 & 3.7: If $G$ is a locally finite Borel planar graph and is $\mu$ -nowhere two-ended, it admits a Borel a.e. one-ended spanning subforest.
Mechanism: Using a 2-basis (a specific collection of cycles), they define a dual graph. By removing edges in the primal graph corresponding to edges in a one-ended subforest of the dual, they obtain an acyclic (tree-like) subgraph in the primal that preserves connectivity.

C. Measure Strong Free Factors

The authors introduce and utilize the concept of measure strong free factors.

A subgroup $\Lambda \le \Gamma$ is a measure strong free factor if, for any free Borel action of $\Gamma$ , the equivalence relation splits as a free product of the relation generated by $\Lambda$ and a treeable relation.
They prove that this property is preserved under various group constructions (free products, HNN extensions, amalgamations over finite subgroups) and is inherited by subgroups (Lemma 5.17).

3. Key Contributions and Results

A. Strong Treeability of New Classes

The paper proves that the following groups are measure strongly treeable (meaning all their free Borel actions are treeable, not just p.m.p. ones):

Surface Groups: $\pi_1(\Sigma)$ for closed hyperbolic surfaces. This resolves a 20-year-old open question (Theorem 1).
Planar Cayley Graphs: Any finitely generated group admitting a planar Cayley graph (Theorem 4.4).
Elementarily Free Groups: Groups sharing the same first-order theory as free groups (Theorem 5.1). These are described as fundamental groups of "IFL towers" (Inductive Free Level towers).
Isom( $H^2$ ) and Subgroups: The group of isometries of the hyperbolic plane, $PSL_2(\mathbb{R})$ , $SL_2(\mathbb{R})$ , and all their closed subgroups (Theorem 6).

B. Fixed Price Consequences

A major consequence of strong treeability is the Fixed Price Conjecture.

If a group is strongly treeable, it has a fixed price (the cost of the group is independent of the specific free p.m.p. action).
The paper provides the first new examples of groups with fixed price $>1$ since Gaboriau's 2000 work, specifically cocompact Fuchsian triangle groups.

C. Ergodic Dimension Dichotomy for 3-Manifolds

The authors establish a sharp dichotomy for the ergodic dimension of fundamental groups of closed aspherical 3-manifolds (Theorem 3):

Either the group is amenable (ergodic dimension 1), or
It has strong ergodic dimension 2.
This result is proven without relying on the Geometrization Theorem. They also generalize this to show that for any compact aspherical $n$ -manifold, the ergodic dimension is at most $n-1$ (Theorem 5).

D. Classification of One-Ended Subforests

The paper provides a complete classification of locally finite p.m.p. graphs that admit Borel a.e. one-ended spanning subforests: they are exactly those that are $\mu$ -nowhere two-ended (Theorem 2.1).

4. Significance and Impact

Resolution of Open Problems: The paper settles the strong treeability of surface groups and elementarily free groups, which were previously unknown.
New Techniques: The method of using one-ended spanning subforests in planar duals is a powerful new tool in measured group theory, bridging combinatorial graph theory (percolation, spanning forests) with ergodic theory.
Generalization to Locally Compact Groups: The results extend from countable discrete groups to locally compact second countable (lcsc) groups (like $SL_2(\mathbb{R})$ ), proving they are measure strongly treeable and have fixed price. This is a significant step in the theory of treeability for continuous groups.
Fixed Price: By establishing strong treeability for these groups, the authors confirm they have fixed price, contributing to the understanding of the "cost" of groups and their von Neumann algebras.
Structural Insights: The work on "measure strong free factors" and IFL towers provides a deeper structural understanding of elementarily free groups, showing they are built up from simpler treeable components in a way that preserves strong treeability.

In summary, this paper significantly advances the theory of treeable groups by introducing a robust geometric method (dual subforests) to prove strong treeability for a wide range of non-amenable groups, thereby resolving long-standing conjectures and expanding the known landscape of groups with fixed price.