Sufficient conditions for spanning k-trees in tough graphs
Pith reviewed 2026-05-07 06:39 UTC · model grok-4.3
The pith
A connected graph with toughness at least t/(t(k-2)+1) contains a spanning k-tree if its order meets a lower bound or its spectral radius (or signless Laplacian spectral radius) is large enough.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
For integers k ≥ 3 and t ≥ 1, if G is connected and satisfies 1/(k-2) > τ(G) ≥ t/(t(k-2)+1), then G contains a spanning k-tree whenever |V(G)| is at least a stated function of k and t, or the spectral radius ρ(G) meets a corresponding lower bound depending on k and t, or the signless Laplacian spectral radius q(G) meets its own lower bound.
What carries the argument
The parameterized toughness threshold t/(t(k-2)+1), which sits between 1/(k-1) and 1/(k-2), combined with either an order lower bound or an extremal spectral-radius condition to force the existence of a spanning subgraph that is a tree of maximum degree k.
If this is right
- When t equals 1 the toughness bound reduces to 1/(k-1) and the spectral conditions recover the earlier results of Liu, Fan and Shu.
- The order-based condition supplies a purely combinatorial sufficient criterion that does not rely on eigenvalues.
- Analogous statements hold simultaneously for the adjacency spectral radius and the signless Laplacian spectral radius.
- The new bounds apply to graphs that are less tough than the classical 1/(k-2) threshold of Win but still satisfy the intermediate toughness floor.
Where Pith is reading between the lines
- As t grows, the toughness floor t/(t(k-2)+1) increases toward 1/(k-2), suggesting a possible sequence of successively stronger sufficient conditions that approach Win's original threshold.
- The spectral bounds are likely sharp for certain families of graphs that achieve equality in the toughness definition, so explicit extremal constructions could be examined to test tightness.
- The same technique might extend to other bounded-degree spanning subgraphs or to k-trees with additional connectivity requirements.
Load-bearing premise
The graph is connected and its toughness is at least t/(t(k-2)+1) for the chosen integer t.
What would settle it
A connected graph whose toughness equals exactly t/(t(k-2)+1), whose order lies below the stated bound, and whose spectral radius lies below the stated bound, yet still contains no spanning k-tree, would falsify the claim.
read the original abstract
The toughness of a graph $G$, denoted by $\tau(G)$, is defined by $\tau(G)=$min $\{\frac{|S|}{c(G-S)}:S\subseteq V(G)$ and $c(G-S)\geq2\}$. A graph $G$ is said to be $\tau$-tough if $\tau(G)\geq \tau$. Let $k\geq2$ be an integer. A tree $T$ is called a $k$-tree if $d_{T}(v)\leq k$ for each $v\in V(T)$, that is, the maximum degree of a $k$-tree is at most $k$. A $k$-tree $T$ is a spanning $k$-tree if $T$ is a spanning subgraph of a connected graph $G$. In 1989, Win [Graphs Combin. 5 (1989) 201--205] proved that if $\tau(G)\geq\frac{1}{k-2}$, where $k\geq3$, then $G$ contains a spanning $k$-tree. Liu, Fan and Shu [Discrete Math. 348 (2025) 114593] provided a tight sufficient condition based on the spectral condition for connected $\frac{1}{k}$-tough and $\frac{1}{k-1}$-tough graphs to contain a spanning $k$-tree, where $k\geq3$ is an integer. A natural and interesting problem arises: Can the value of $\tau$ be refined? When $\frac{1}{k-2}>\tau\geq\frac{1}{k-1}$, we initially establish a lower bound on the size to ensure that a connected $\frac{t}{t(k-2)+1}$-tough graph $G$ contains a spanning $k$-tree, where $k\geq3$ and $t\geq1$ are integers. Meanwhile, we provide two sufficient conditions in terms of spectral radius and signless Laplacian spectral radius for a connected $\frac{t}{t(k-2)+1}$-tough graph $G$ to contain a spanning $k$-tree, where $k\geq3$ and $t\geq1$ are integers. When $t=1$, we obtain the result $\eta=1$ from Liu, Fan and Shu.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper claims to provide sufficient conditions for spanning k-trees in connected graphs with a parameterized toughness τ = t/(t(k-2)+1) for integers k≥3, t≥1. Specifically, such graphs contain a spanning k-tree if their order is at least some bound or if their spectral radius or signless Laplacian spectral radius exceeds a certain threshold. This refines Win's theorem for toughness 1/(k-2) and matches Liu et al. at t=1.
Significance. If correct, this work is significant as it fills the gap in toughness values between 1/(k-1) and 1/(k-2) by using a discrete family parameterized by t, and incorporates modern spectral graph theory tools to give alternative sufficient conditions. It builds directly on prior results and could inspire further refinements in toughness-based extremal problems.
minor comments (4)
- [Abstract] The sentence 'When 1/(k-2)>τ≥1/(k-1), we initially establish a lower bound on the size...' could be clarified by specifying what the lower bound is or referring to the theorem number.
- [Abstract] The mention of 'the result η=1' is unclear without context; ensure that any auxiliary parameter η is defined in the main text and its relation to the toughness is explained.
- [Introduction] Add a brief discussion on the sharpness of the new toughness threshold and whether there are known examples showing that smaller toughness does not guarantee the spanning k-tree even with large order.
- [Main theorems] Verify that the spectral conditions are stated with explicit constants or bounds in terms of k and t.
Simulated Author's Rebuttal
We thank the referee for the positive assessment of our manuscript and the recommendation for minor revision. No specific major comments were provided in the report, so we have no point-by-point responses to address. We will implement minor revisions to improve clarity, notation, and presentation in the next version of the manuscript.
Circularity Check
No significant circularity; extends independent prior theorems
full rationale
The derivation extends Win's 1989 toughness threshold and the 2025 Liu-Fan-Shu spectral conditions to the discrete family of intermediate toughness values τ = t/(t(k-2)+1) for integer t ≥ 1. All load-bearing steps cite these external results as black-box inputs, then apply standard toughness-to-degree-component arguments followed by spectral-radius comparison against known extremal non-k-tree graphs. When t=1 the statement recovers the cited Liu-Fan-Shu theorem verbatim rather than re-deriving it; no equation defines a parameter from data internal to this paper and then renames that parameter as a prediction. The argument chain therefore remains self-contained against the cited external benchmarks.
Axiom & Free-Parameter Ledger
axioms (1)
- standard math Basic properties of finite undirected graphs, trees, and the toughness function as defined in the paper
Reference graph
Works this paper leans on
- [1]
-
[2]
Chv´ atal, Tough graphs and Hamiltonian circuits, Discrete Math
V. Chv´ atal, Tough graphs and Hamiltonian circuits, Discrete Math. 5 (1973) 215–228
work page 1973
-
[3]
Das, Maximizing the sum of the squares of the degrees of a graph, Discrete Math
K. Das, Maximizing the sum of the squares of the degrees of a graph, Discrete Math. 285 (2004) 57–66
work page 2004
- [4]
-
[5]
M.N. Ellingham, X.Y. Zha, Toughness, trees, and walks, J. Graph Theory. 33 (2000) 125–137
work page 2000
-
[6]
D.D. Fan, S. Goryainov, X.Y. Huang, H.Q. Lin, The spanningk-trees, perfect matchings and spectral radius of graphs, Linear Multilinear Algebra. 70 (2022) 7264–7275
work page 2022
- [7]
-
[8]
Hong, A bound on the spectral radius of graphs, Linear Algebra Appl
Y. Hong, A bound on the spectral radius of graphs, Linear Algebra Appl. 108 (1988) 135–139
work page 1988
-
[9]
Johansson, An El-Zah´ ar type condition ensuring path-factors, J
R. Johansson, An El-Zah´ ar type condition ensuring path-factors, J. Graph Theory. 28 (1998) 39–42
work page 1998
-
[10]
M. Johnson, D. Paulusma, C. Wood, Path factors and parallel knock-out schemes of almost claw-free graphs, Discrete Math. 310 (2010) 1413–1423
work page 2010
- [11]
-
[12]
S. Kim, S. O, J. Park, H. Ree, An odd [1, b]-factor in regular graphs from eigenvalues, Discrete Math. 343 (2020) 111906
work page 2020
-
[13]
M. Kano, A. Saito, Star-factors with large components, Discrete Math. 312 (2012) 2005– 2008
work page 2012
-
[14]
Q. Li, K.Q. Feng, On the largest eigenvalue of a graph, Acta Math. Appl. Sinica. 2 (1979) 167–175
work page 1979
-
[15]
R.F. Liu, A. Fan, J.L. Shu, Spectral extremal problems on factors in tough graphs, and beyond, Discrete Math. 348 (2025) 114593
work page 2025
- [16]
- [17]
- [18]
-
[19]
Lv, A degree condition for graphs being fractional (a, b, k)-critical covered, Filomat
X.Y. Lv, A degree condition for graphs being fractional (a, b, k)-critical covered, Filomat. 37 (2023) 3315–3320
work page 2023
-
[20]
Matsuda, Fan-type results for the existence of [a, b]-factors, Discrete Math
H. Matsuda, Fan-type results for the existence of [a, b]-factors, Discrete Math. 306 (2006) 688–693
work page 2006
-
[21]
B. Ning, J. Ge, Spectral radius and Hamiltonian properties of graphs, Linear Multilinear Algebra. 63 (2015) 1520–1530
work page 2015
- [22]
- [23]
-
[24]
Win, On a connection between the existence ofk-trees and the toughness of a graph, Graphs Combin
S. Win, On a connection between the existence ofk-trees and the toughness of a graph, Graphs Combin. 5 (1989) 201–205
work page 1989
-
[25]
Zhou, A neighborhood union condition for fractional (a, b, k)-critical covered graphs, Discrete Appl
S.Z. Zhou, A neighborhood union condition for fractional (a, b, k)-critical covered graphs, Discrete Appl. Math. 323 (2022) 343–348
work page 2022
- [26]
- [27]
-
[28]
S.Z. Zhou, Y. Xu, Z.R. Sun, Some results about star-factors in graphs, Contrib. Discrete Math. 19 (2024) 154–162
work page 2024
- [29]
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.