Soil disturbance as a restoration measure in dry sandy grasslands

Abstract

Severely disturbed habitats such as military training grounds, gravel pits and sand pits contribute to the species diversity of the agricultural landscape in Europe. They host a number of red-listed species not found elsewhere, illustrating that many plant species are threatened by extinction due to too little soil disturbance. Implementing a suitable disturbance regime is therefore crucial to ensure species-rich environments. We have reviewed the literature on soil disturbance as a restoration measure in dry sandy grasslands, with a special focus on xeric sand calcareous grasslands as these are severely threatened. Our objective was to elucidate the relations between diversity and disturbance regimes, and to determine how disturbance can be used to counteract acidification, to reduce nutrient availability and to create gaps in the vegetation. Our findings indicate that the current disturbance regime should be based on the historical disturbance regime, the productivity of the habitat and the propagule supply, in order to promote diversity at a landscape scale. Based on earlier studies and on the diversity/disturbance theory, we propose a conceptual model that can be used to determine the appropriate soil disturbance regime for restoration purposes. Our analysis highlights the importance of considering soil productivity, soil chemistry and dispersal limitations when choosing restoration measures and disturbance regimes for the conservation of biodiversity.

Appendix

Detailed model description

The model has 16 competing plant species governed by Lotka–Volterra dynamics (see below). The first 10 species (i = 1…10) are considered ‘target’ species, adapted to poor nutrient conditions but dispersal limited, and species 11–16 are non-target, dependent on higher levels of soil nutrients but with a more rapid colonisation at appropriate soil conditions. In addition to plant population dynamics, soil nutrients are modelled as a single dynamic variable U, which adds up to a system of 17 differential equations:

$$ \left\{ {\begin{array}{*{20}l} \frac{{dN_{\text{i}} (t)}}{dt} = r_{\text{i}} (t)\left( {1 - \frac{{N_{\text{tot}} (t)}}{{K_{\text{i}} }}} \right)\;,\;i = 1,\ldots,16 \\ \frac{dU}{dt} = cN_{\text{tot}} (t) - \mu_{\text{U}} U(t) \hfill \\ \end{array} } \right., $$

(1)

where

$$ N_{\text{tot}} (t) = \sum\limits_{i = 1}^{16} {N_{\text{i}} (t)} $$

(2)

N _i(t) is abundance (biomass) of population i at time t. r _i and K _i are species specific intrinsic growth rates and carrying capacities, respectively, chosen such that there is a trade-off between fast growth (a high r) and a high carrying capacity (K) (see below). The soil nutrient U increases at a rate c per unit total plant biomass (N _tot) and leaks out of the system at a rate μ _U per mass unit.

All target species (i = 1..10) have a constant, but unique, intrinsic growth rates r _i, whereas the r-values of non-target species increase with the amount of available soil nutrients according to

$$ r_{i} (t) = \left\{ {\begin{array}{*{20}c} {r_{{ 0 , {\text{i}}}} ,} \hfill \\ {r_{{ 0 , {\text{i}}}} + kU(t),} \hfill \\ \end{array} } \right.\;\begin{array}{*{20}c} {i = 1..10} \hfill \\ {i = 11..16} \hfill \\ \end{array} $$

(3)

All biomasses are kept at or above a minimal level corresponding to a constant rain of seeds. The minimal level is set to 10⁻⁴ for target species and 10⁻³ for non-target species, in accordance with the assumption that target species are more dispersal-limited than non-target species.

A ‘ploughing’ disturbance is implemented such that at regular intervals all biomasses are set to their minimal levels, but the soil nutrient content (U) is left intact. ‘Top soil removal’ means the same thing except the soil nutrient content is divided by a factor 100.

Numerical solutions to the differential equations were found by discretizing time (Euler method, Δt = 0.01).

Parameter values, target species (i = 1..10):

• K _1–10 = 0.05, 0.10, 0.15, 0.20, 0.25, 0.30, 0.35, 0.40, 0.45, 0.50

• r _{0, i} = 3–4.5 K _i, i = 1..10

Parameter values, non-target species (i = 11..16):

• K _11–16 = 0.25, 0.40, 0.55, 0.70, 0.85, 1

• r _{0, i} = 1–4.5 K _i, i = 11..16

• k = 20

Parameter values, soil nutrient U:

• c = 0.15

• μ _U = 0.3