FD1-3: The resilience of the landscape mosaic is controlled by disturbance regime and seed availability.

Tree recruitment in permanent plots (Walker)

See Vegetation sampling methods.

Comparison of pre-fire and post-fire tree species composition (Chapin)

We will establish permanent plots in several road-accessible burns (aged 10-50 years) in interior Alaska to compare pre-fire tree composition (based on densities of fire-killed trees) and post-fire tree composition (seedlings, saplings, and resprouts)(Johnstone and Chapin 2003). If a successional trajectory is resilient, we expect that the same tree species will be present both before (as adults) and after fire (as seedlings and resprouts). Failure of a species to regenerate or arrival of a new tree species that was absent before the fire suggests a shift to a new successional trajectory.

We will select these plots to span a wide range of environments, fire severities, and distance to seed source. In each study site, we will sample five randomly positioned 2 x 50 m2 transects (0.05 ha) within a 0.25 ha area in each stand and count the total number of seedlings or saplings of each tree species. We will record observations of dead post-fire seedlings as we encounter them. For dead seedlings, we assess whether mortality was caused by herbivory or some other factor by looking for evidence of browsed stem tips or compensatory branching.

We document the composition and basal diameters of all trees in the pre-fire community from standing or fallen deadwood that had been rooted within a sampling transect prior to the fire. Genus determinations of dead trees are based on cone, bark and branching morphology. Dead trees of uncertain identity will be allocated to species categories according to the proportional composition of identified individuals. We may need to pool black and white spruce because some black spruce individuals sometimes lack cones and cannot be readily differentiated from dead white spruce trees. We will not count trees that were dead at the time of the fire, judging from deep charring patterns on the bole.

The age of the pre-fire stand will be determined from annual ring counts of 5-10 basal disks or tree cores (30 cm above the uppermost roots) obtained from a mixture of dead pine and spruce trees systematically sampled to represent the largest, and hence oldest, trees in the stand.

We will permanently mark the positions of these plots, so they can be revisited in the future (perhaps decadally) to document subsequent vegetation change. We will document potential controlling factors (slope, aspect, parent material, organic-matter depth, thaw depth, and soil moisture).

Use of demographic data to parameterize matrix population models (Lloyd)

We will use stand demographic data to parameterize a series of matrix population models describing black and white spruce population dynamics along a transect from the interior to arctic treeline. Stand demography will be assessed in permanently marked study plots along a latitudinal transect from the Yukon River to treeline in the Brooks Range. Plots have previously been established north of Coldfoot, Alaska (Table 1; Lloyd and Fastie, unpublished data).

New plots will be established at 2-3 sites immediately north of these, to capture the northernmost individuals of black spruce, and at approximately 5 sites south of these (between Coldfoot, Alaska, and the Yukon River) to capture populations in the center of the species’ distribution.

We will establish a minimum of 3 permanently marked study plots at each site. Study plots will be marked with metal posts and all trees will be tagged to facilitate future remeasurements. Within each plot, we will core all trees with a basal diameter >5 cm, and estimate age from counts of annual bud scars on trees smaller than that threshold. We will estimate reproductive output by a categorical count of the number of cones. We will collect cones from a randomly selected subset of trees to estimate seed viability. Basal cross sections will be collected from any dead trees, and from fire-scarred trees outside the study plots. Internode counts on dead seedlings will be used to estimate seedling mortality rates.

Tree cores and cross sections will be mounted onto wooden supports and sanded with progressively finer sand paper (to 400 grit) to reveal the annual rings. Rings will be measured and cross-dated, and tree ages will be estimated after correcting for time to grow to core height and distance to pith (in cases where the pith is missed). Tree age data will be used in conjunction with data on cone production and seed viability to parameterize a matrix population model for populations in three areas: (a) the northernmost individuals (north of the study sites listed above), (b) the northernmost discrete populations (the group of sites listed in Table 1), and (c) populations in the region between the Yukon River and Coldfoot. Survivorship probabilities for the stage-based model will be estimated from both the age structure of live trees and seedlings and the age structure of dead trees and seedlings. Fecundity will be estimated from cone counts and seed viability data. Models will be constructed and analyzed using Matlab.

Two types of models will be parameterized. First, basic Lefkovitch matrices will be used to estimate population growth rate (λ), and elasticity analyses will be used to determine the sensitivity of λ to different demographic characteristics. Second, the resilience of populations to varying disturbance regimes will be tested with models that combine ‘non-fire’ matrices (which describe population dynamics in the absence of disturbance) and ‘fire’ matrices (which describe population dynamics in the years following a fire; A. Wilson, unpublished thesis). Population dynamics are modeled by employing the non-fire matrix in non-fire years, and then interspersing the fire matrix at intervals simulating the desired disturbance regime. For example, the effects of a fire frequency of 100 years would be estimated by using the non-fire matrix for years 1-99 and the fire matrix in year 100. This approach allows estimation of population stability under different disturbance regimes: resilient populations should be viable in a wide variety of disturbance regimes.