Post-fire seed dispersal of a wind-dispersed shrub declined with distance to seed source, yet had high levels of unexplained variation

Abstract

Plant-population recovery across large disturbance areas is often seed-limited. An understanding of seed dispersal patterns is fundamental for determining natural-regeneration potential. However, forecasting seed dispersal rates across heterogeneous landscapes remains a challenge. Our objectives were to determine (i) the landscape patterning of post-disturbance seed dispersal, and underlying sources of variation and the scale at which they operate, and (ii) how the natural seed dispersal patterns relate to a seed augmentation strategy. Vertical seed trapping experiments were replicated across 2 years and five burned and/or managed landscapes in sagebrush steppe. Multi-scale sampling and hierarchical Bayesian models were used to determine the scale of spatial variation in seed dispersal. We then integrated an empirical and mechanistic dispersal kernel for wind-dispersed species to project rates of seed dispersal and compared natural seed arrival to typical post-fire aerial seeding rates. Seeds were captured across the range of tested dispersal distances, up to a maximum distance of 26 m from seed-source plants, although dispersal to the furthest traps was variable. Seed dispersal was better explained by transect heterogeneity than by patch or site heterogeneity (transects were nested within patch within site). The number of seeds captured varied from a modelled mean of ~13 m^-2 adjacent to patches of seed-producing plants, to nearly none at 10 m from patches, standardized over a 49-day period. Maximum seed dispersal distances on average were estimated to be 16 m according to a novel modelling approach using a 'latent' variable for dispersal distance based on seed trapping heights. Surprisingly, statistical representation of wind did not improve model fit and seed rain was not related to the large variation in total available seed of adjacent patches. The models predicted severe seed limitations were likely on typical burned areas, especially compared to the mean 95-250 seeds per m² that previous literature suggested were required to generate sagebrush recovery. More broadly, our Bayesian data fusion approach could be applied to other cases that require quantitative estimates of long-distance seed dispersal across heterogeneous landscapes.

Keywords: Data fusion; ecological forecasting; hierarchical Bayesian models; natural regeneration; seed dispersal.

Figure 1.

Locations of fires (outlines) and trapping sites (dots) for dispersal study shown as an inset map on western USA and 2011 sagebrush cover (%) from the National Land Cover Database (NLCD) (Rigge et al. 2020).

Figure 2.

Relationship of mean trap abundance (bottom panel) and variability (RSE, top panel) of the density of seeds captured (per 0.05 m² of vertical trap area) relative to the distance of seed traps from seed source patch. Seed density is standardized by the number of days in each collection interval period shown as the mean per trap ± the standard error (bottom). Alkie was excluded due to seed crop failure and no seeds trapped.

Figure 3.

Box plots of the estimated total available seeds per patch (fecundity × number of reproductive plants) across sites (top) and number of seeds across traps of all distances caught per 0.05 m² trap area standardized by 49 days deployed (bottom). The graphs do not include under-crown traps. The unit of measure for the top graph is a patch (n = 19) and the unit of measure for the bottom graph is a trap (seed counts aggregated across heights, n = 273). Alkie was excluded due to seed crop failure and no seeds trapped.

Figure 4.

Posterior distributions intervals for parameters of the landscape negative binomial seed density model with intercepts and slopes varied by transect. The centre circle of each distribution shows the median, the thick bars show the 50 % credible interval and the thin lines show the 90 % credible interval. Predictors were scaled prior to analysis so that parameter values represent relative effect size of each predictor on trapped seed density. (A) Global parameters, (B) varying intercepts by transect, (C) varying slope of the height parameter by transect, (D) varying slope of the distance parameter by transect and (E) varying slope of the height:distance parameter by transect.

Figure 5.

Mean number of trapped seeds per m² area predicted from the landscape model with slope varied by transect, showing the interacting effects of trapped height and trapped distance on seed density. The shaded ribbons show the 90 % credible intervals.

Figure 6.

Simulated median seed dispersal (seeds per m²) estimated using the seed dispersal model with transect-level variation in dispersal kernel (1000 simulations), based on global parameters for the p and u parameters of the 2Dt kernel, and assuming an average of 30 000 seeds per reproductive plant, and 25 individuals per patch. The grey ribbon shows the 90 % quantiles of the simulations.