Stuart B. Weiss

GLM versus CCA spatial modeling of plant species distribution

Despite the variety of statistical methods available for static modeling of plant distribution, few studies directly compare methods on a common data set. In this paper, the predictive power of Generalized Linear Models (GLM) versus Canonical Correspondence Analysis (CCA) models of plant distribution in the Spring Mountains of Nevada, USA, are compared. Results show that GLM models give better predictions than CCA models because a species-specific subset of explanatory variables can be selected in GLM, while in CCA, all species are modeled using the same set of composite environmental variables (axes). Although both techniques can be readily ported to a Geographical Information System (GIS), CCA models are more readily implemented for many species at once. Predictions from both techniques rank the species models in the same order of quality; i.e. a species whose distribution is well modeled by GLM is also well modeled by CCA and vice-versa. In both cases, species for which model predictions have the poorest accuracy are either disturbance or fire related, or species for which too few observations were available to calibrate and evaluate the model. Each technique has its advantages and drawbacks. In general GLM will provide better species specific-models, but CCA will provide a broader overview of multiple species, diversity, and plant communities.

SUN, SLOPE, AND BUTTERFLIES: TOPOGRAPHIC DETERMINANTS OF HABITAT QUALITY FOR EUPHYDRYAS EDITHA'

Thermal environments in a large, topographically diverse serpentine soil- based grassland were quantified and ranked using a computer model of clear sky insolation and shading on different slopes to determine the effects of microclimate on the rates of development of each of the life stages of the butterfly Euphydryas editha bayensis. Larvae developed to pupation earlier on warm slopes than on progressively cooler slopes. Avail- ability of sunlight can be limiting for larvae, which bask in direct sun to raise body temperature. Larvae can disperse > 10 m/d, allowing them to transfer between microcli- mates. Pupae on warmer slopes also developed faster than those on cooler slopes. Microclimate also affects the phenology of host plants of larvae and nectar sources of adults. Larval and pupal development and host-plant phenology determine the phase relationship between adult butterfly flight and host-plant senescence, which in turn deter- mines mortality rates of prediapause larvae. Adult females that eclosed early in the season could have their offspring survive on almost all slopes; survivorship of offspring from adults that eclosed in the middle of the flight season was restricted to cooler slopes in the habitat. Some butterflies eclosed too late for their offspring to survive on any slope. Topographic diversity on several scales is a prime indicator of habitat quality for this butterfly. Areas of high local topographic diversity on a scale of tens of metres appear particularly important for long-term population persistence under variable climatic condi- tions.

Leaf size, specific leaf area and microhabitat distribution of chaparral woody plants: contrasting patterns in species level and community level analyses

We examined variation in leaf size and specific leaf area (SLA) in relation to the distribution of 22 chaparral shrub species on small-scale gradients of aspect and elevation. Potential incident solar radiation (insolation) was estimated from a geographic information system to quantify microclimate affinities of these species across north- and south-facing slopes. At the community level, leaf size and SLA both declined with increasing insolation, based on average trait values for the species found in plots along the gradient. However, leaf size and SLA were not significantly correlated across species, suggesting that these two traits are decoupled and associated with different aspects of performance along this environmental gradient. For individual species, SLA was negatively correlated with species distributions along the insolation gradient, and was significantly lower in evergreen versus deciduous species. Leaf size exhibited a negative but non-significant trend in relation to insolation distribution of individual species. At the community level, variance in leaf size increased with increasing insolation. For individual species, there was a greater range of leaf size on south-facing slopes, while there was an absence of small-leaved species on north-facing slopes. These results demonstrate that analyses of plant functional traits along environmental gradients based on community level averages may obscure important aspects of trait variation and distribution among the constituent species.

Three-dimensional Structure of an Old-growth Pseudotsuga-Tsuga Canopy and Its Implications for Radiation Balance, Microclimate, and Gas Exchange

We describe the three-dimensional structure of an old-growth Douglas-fir/western hemlock forest in the central Cascades of southern Washington, USA. We concentrate on the vertical distribution of foliage, crowns, external surface area, wood biomass, and several components of canopy volume. In addition, we estimate the spatial variation of some aspects of structure, including the topography of the outer surface, and of microclimate, including the within-canopy transmittance of photosynthetically active radiation (PAR). The crowns of large stems, especially of Douglas-fir, dominate the structure and many aspects of spatial variation. The mean vertical profile of canopy surfaces, estimated by five methods, generally showed a single maximum in the lower to middle third of the canopy, although the height of that maximum varied by method. The stand leaf area index was around 9 m2 m−2, but also varied according to method (from 6.3 to 12.3). Because of the deep narrow crowns and numerous gaps, the outer canopy surface is extremely complex, with a surface area more than 12 times that of the ground below. The large volume included below the outer canopy surface is very porous, with spaces of several qualitatively distinct environments. Our measurements are consistent with emerging concepts about the structure of old-growth forests, where a high degree of complexity is generated by diverse structural features. These structural characteristics have implications for various ecosystem functions. The height and large volume of the stand indicate a large storage component for microclimatic variables. The high biomass influences the dynamics of those variables, retarding rates of change. The complexity of the canopy outer surface influences radiation balance, particularly in reducing short-wave reflectance. The bottom-heaviness of the foliage profile indicates much radiation absorption and gas exchange activity in the lower canopy. The high porosity contributes to flat gradients of most microclimate variables. Most stand respiration occurs within the canopy and is distributed over a broad vertical range.

Estimating female reproductive success of a threatened butterfly: influence of emergence time and hostplant phenology

We estimated lifetime reproductive success of Euphydryas editha bayensis (Nymphalidae), a federally listed threatened butterfly, based on age-specific fecundity and both adult and offspring survival. Our results indicate that the relative timing of adult emergence and larval hosplant senescence strongly influenced reproductive success of females. For 1992, we estimated that only 8–21% of the eggs laid by females emerging on the 1st day of the 4-week flight season would produce larvae that reach diapause. This figure dropped to 1–5% for females emerging 7 days into the flight season. Within our entire sample, we estimated that 64–88% of the females produced offspring with less than a 2% probability of reaching diapause. These estimates are particularly striking given that they are based on only one source of larval mortality — prediapause starvation due to hostplant senescence. This dependence of reproductive success on the relative timing of female emergence and hostplant senescence may reduce effective population size and render E. editha bayensis especially vulnerable to local extinction events.

Adult emergence phenology in checkerspot butterflies: the effects of macroclimate, topoclimate, and population history

The prediction of adult emergence times in insect populations can be greatly complicated by microclimatic gradients, especially in circumstances where distributions of juveniles along those gradients vary from year to year. To investigate adult emergence patterns in topographically heterogeneous habitats, we built a model of postdiapause development of the Bay checkerspot butterfly, Euphydryas editha bayensis. The model uses slope-specific insolation as the rate-controlling variable, and accounts for both solar exposure of the habitat and cloud cover. Instar-specific larval mass gains per unit of insolation were determined from mark-recapture experiments. A small correction for daily low temperatures was used to calibrate the model to five years of field data on larval mass. The model predicted mean mass of 90% of larval samples within 4 clear days over a 70–120 day growing season. The magnitude of spatial variation in emergence times across habitat slopes is greater than annual variation in emergence times due to yearly weather conditions. Historical variation (yearly shifts in larval distributions across slopes) is an important determinant of mean population emergence dates. All of these factors need to be considered in understanding adult emergence phenology in this butterfly and in other insects inhabiting heterogeneous thermal environments. Such an understanding can be useful in managing insect populations for both pest control and conservation.

Ecological studies and the conservation of the bay checkerspot butterfly, Euphydryas editha bayensis

Abstract The historical decline of the bay checkerspot butterfly Euphydryas editha bayensis is discussed and pertinent ecological information presented. The distribution of the butterfly has become highly restricted as its habitat has been destroyed by development and its numbers reduced by severe drought. Habitat loss and fragmentation disrupts the ability of habitat patches to support populations as key topographic features are eliminated. The continued survival of the butterfly is dependent on a reservoir population which provides colonists for smaller habitat patches in the vicinity. The political struggle to gain federal protection for the butterfly is reviewed.

Growth and dispersal of larvae of the checkerspot butterfly Euphydryas editha

We investigated growth and dispersal of postdiapause Euphydryas editha larvae in a serpentine soil-based grassland, using groups of individually marked and weighed larvae. Larval weight gain was highly correlated with insolation at ground level which is determined by slope exposure and date. Maximum dispersal distances were in excess of 10 m d-l1; dispersal greater than 5 m d-~ resulted in depressed growth rates during the period of dispersal. Larvae which disperse such distances can transfer from cool to warm slopes and can pupate up to two weeks sooner than larvae remaining on cool slopes, even after paying the time and metabolic costs of dispersal. We discuss the implications of larval dispersal on individual fitness and population persistence.

Light-transmission Profiles in an Old-growth Forest Canopy:Simulations of Photosynthetically Active Radiation by Using Spatially Explicit Radiative Transfer Models

Light interception is a driving variable for many key ecosystem processes in forests. Canopy gaps, as natural irregularities, are common features of Pacific Northwest conifer forests and have profound importance on the within-canopy light environment. We used two spatially explicit radiative transfer models (OLTREE and SolTran) to understand better the vertical profile distribution of light penetration in an old-growth forest. Canopy access at the Wind River Canopy Crane Research Facility provided an opportunity to apply these models in a tall, old-growth, Douglas-fir–western hemlock forest. Both models required three-dimensional descriptions for every crown (location, orientation, and size) in a 4-ha area. Crowns were then simulated as foliage-filled ellipsoids through which light is attenuated following Beer’s law. We simulated vertical profiles (2-m height intervals) of transmitted photosynthetically active radiation (PAR) in 16 gaps previously measured by Parker (1997). Point-by-point comparisons (n = 480) between measured and modeled results showed little agreement because small errors in crown location yielded large local differences in PAR transmittance. However, average gap profiles (n = 16) of PAR transmittance showed excellent agreement (r2 = 0.94) between simulated and measured values. SolTran was used to simulate vertical profiles of daily PAR flux at different seasons for the whole 4-ha canopy, not just gaps. Overall, our results show that both models produced excellent simulations of spatially averaged vertical profiles of PAR transmission in the old-growth forest and are suitable for further investigations at other space and time scales.

Landscape-Level Phenology of a Threatened Butterfly: A GIS-Based Modeling Approach

ABSTRACT Phenology of organismal development varies between growing seasons according to the weather and also varies within growing seasons across topoclimatic gradients. Combining these factors is necessary to predict landscape-level patterns of phenology and their consequences for population dynamics. We developed a model on a Geographic Information System (GIS) that predicts phenology of adult emergence of the threatened Bay checkerspot butterfly across complex terrain under variable weather. Physiological time was modeled by accumulated slope-specific direct insolation. Insolation sums through growing seasons were calculated for each cell of a digital terrain model (skipping over cloudy days) until a threshold for adult emergence was reached. Emergence times of adult butterflies for a given year were then mapped out across a 100-ha area. To generate predicted emergence curves for the population in a given year, these maps ofemergence times were then modified by incorporating microdistributions of postdiapause larvae. Different larval microdistributions changed both the magnitude and shape of emergence curves under the same yearly weather and could change mean population-wide emergence dates by 11 days. Reproductive success in this butterfly is strongly dependent on the timing of adult emergence, and these models provide insights into the effects of weather, topography, and population history on population dynamics. Because adult emergence phenology is often a key component of reproductive success for insects, understanding the components of phenological variation in space and time in complex terrain may provide insights into population dynamics for management of pests and conservation of rare species.