population density for a five mile radius around Bacons Bridge within Dorchester county

This file contains breeding bird censuses using 50 m radius counting circles at locations on Plum Island, Massachusetts in the Parker River National Wildlife Refuge, Massachusetts.

BackgroundAssessing exposure to infections in early childhood is of interest in many epidemiological investigations. Because exposure to infections is difficult to measure directly, epidemiological studies have used surrogate measures available from routine data such as birth order and population density. However, the association between population density and exposure to infections is unclear. We assessed whether neighbourhood child population density is associated with respiratory infections in infants.MethodsWith the Basel-Bern lung infant development study (BILD), a prospective Swiss cohort study of healthy neonates, respiratory symptoms and infections were assessed by weekly telephone interviews with the mother throughout the first year of life. Using population census data, we calculated neighbourhood child density as the number of children < 16 years of age living within a 250 m radius around the residence of each child. We used negative binomial regression models to assess associations between neighbourhood child density and the number of weeks with respiratory infections and adjusted for potential confounders including the number of older siblings, day-care attendance and duration of breastfeeding. We investigated possible interactions between neighbourhood child population density and older siblings assuming that older siblings mix with other children in the neighbourhood.ResultsThe analyses included 487 infants. We found no evidence of an association between quintiles of neighbourhood child density and number of respiratory symptoms (p = 0.59, incidence rate ratios comparing highest to lowest quintile: 1.15, 95%-confidence interval: 0.90–1.47). There was no evidence of interaction with older siblings (p = 0.44). Results were similar in crude and in fully adjusted models.ConclusionsOur study suggests that in Switzerland neighbourhood child density is a poor proxy for exposure to infections in infancy.

Spatial population concentration (SPC) is a new measure indicating how close people live together. This measure intuitively captures how many people live on average within a given radius of every person in a specific area. This dataset provides a guidance note along with spatial population concentration for 50km and 100km radius.

population density for a five mile radius around Bacons Bridge within Dorchester county

NEXIS population density exposure is a web map service displaying the number of people per NEXIS residential building within a neighbourhood radius. Population density is calculated by the number of people within 10sqkm, 5sqkm, 1sqkm, 500sqm and 100sqm.

Zip Codes Outside of 100-mile Radius of Nearest Medicare-Participating Dermatologist.

CC: Carrying capacity.The carrying capacity of pods was calculated by fitting the maximum number of pods, including their radius distance, into the convex hull area encompassing the entire population. The carrying capacity of whales is the number of pods multiplied by the mean pod size.

Abstract

Malawi Conditional Cash Transfer Program (CCT) is a randomized cash transfer intervention targeting young women in Zomba region. The program provides incentives to current schoolgirls and recent dropouts to stay in or return to school. The incentives include average payment of US$10 a month conditional on satisfactory school attendance and direct payment of secondary school fees.

The CCT program started at the beginning of the Malawian school year in January 2008 and continued until November 2009. The impact evaluation study was designed to evaluate the impact of the program on various demographic and health outcomes of its target population, such as nutritional health, sexual behavior, fertility, and subsequent HIV risk.

Baseline data collection was administered from September 2007 to January 2008. The research targeted girls and young women, between the ages of 13 and 22, who were never married. Overall, 3,810 girls and young women were surveyed in the first round. The follow-up survey was carried out from October 2008 to February 2009. The third round was conducted between March and September 2010, after Malawi Conditional Cash Transfer Program was completed. The fourth round started in April 2012 and will continue until September 2012.

Datasets from the baseline round are documented here.

Enumeration Areas (EAs) in the study district of Zomba were selected from the universe of EAs produced by the National Statistics Office of Malawi from the 1998 Census. 176 enumeration areas were randomly sampled out of a total of 550 EAs using three strata: urban areas, rural areas near Zomba Town, and rural areas far from Zomba Town.

Baseline schoolgirls in treatment enumeration areas were randomly assigned to receive either conditional or unconditional transfers, or no transfers at all. A multi-topic questionnaire was administered to the heads of households, where the selected sample respondents resided, as well as to girls and young women.

Geographic coverage

Zomba district.

Zomba district in the Southern region was chosen as the site for this study for several reasons. First, it has a large enough population within a small enough geographic area rendering field work logistics easier and keeping transport costs lower. Zomba is a highly populated district, but distances from the district capital (Zomba Town) are relatively small. Second, characteristic of Southern Malawi, Zomba has a high rate of school dropouts and low educational attainment. Third, unlike many other districts, Zomba has the advantage of having a true urban center as well as rural areas. As the study sample was stratified to get representative samples from urban areas (Zomba town), rural areas near Zomba town, and distant rural areas in the district, we can analyze the heterogeneity of the impacts by urban/rural areas. Finally, while Southern Malawi, which includes Zomba, is poorer, has lower levels of education, and higher rates of HIV than Central and Northern Malawi, these differences are relative considering that Malawi is one of the poorest countries in the world with one of the highest rates of HIV prevalence.

Analysis unit

• Households;
• Girls and young women.

Universe

The survey covers never married girls and young women between the ages of 13 and 22 in Zomba district.

Kind of data

Sample survey data [ssd]

Sampling procedure

First, 176 enumeration areas (EA) were randomly sampled out of a total of 550 EAs using three strata in the study district of Zomba. Each of these 176 EAs were then randomly assigned treatment or control status. The three strata are urban, rural areas near Zomba Town, and rural areas far from Zomba Town. Rural areas were defined as being near if they were within a 16-kilometer radius of Zomba Town. Researchers did not sample any EAs in TA Mbiza due to safety concerns (112 EAs).

Enumeration areas (EAs) in Zomba were selected from the universe of EAs produced by the National Statistics Office of Malawi from the 1998 Census. The sample of EAs was stratified by distance to the nearest township or trading centre. Of the 550 EAs in Zomba, 50 are in Zomba town and an additional 30 are classified as urban (township or trading center), while the remaining 470 are rural (population areas, or PAs). The stratified random sample of 176 EAs consisted of 29 EAs in Zomba town, eight trading centers in Zomba rural, 111 population areas within 16 kilometers of Zomba town, and 28 EAs more than 16 kilometers from Zomba town.

After selecting sample EAs, all households were listed in the 176 sample EAs using a short two-stage listing procedure. The first form, Form A, asked each household the following question: "Are there any never-married girls in this household who are between the ages of 13 and 22?" This form allowed the field teams to quickly identify households with members fitting into the sampling frame, thus significantly reducing the costs of listing. If the answer received on Form A was a "yes", then Form B was filled to list members of the household to collect data on age, marital status, current schooling status, etc.

From this researchers could categorize the target population into two main groups: those who were out of school at baseline (baseline dropouts) and those who were in school at baseline (baseline schoolgirls). These two groups comprise the basis of our sampling frame. In each EA, enumerators sampled all eligible dropouts and 75%-100% of all eligible school girls, where the percentage depended on the age of the baseline schoolgirl. This sampling procedure led to a total sample size of 3,810 (in the first round, and 3,805 in follow-up rounds) with an average of 5.1 dropouts and 16.7 schoolgirls per EA.

Mode of data collection

Face-to-face [f2f]

Research instrument

The annual household survey consists of a multi-topic questionnaire administered to the households in which the selected sample respondents reside. The survey consists of two parts: one that is administered to the head of the household and another that is administered to the core respondent - the sampled girl from the target population. The former collects information on the household roster, dwelling characteristics, household assets and durables, shocks and consumption. The core respondent survey provides information about her family background, her education and labor market participation, her health, her dating patterns, sexual behavior, marital expectations, knowledge of HIV/AIDS, her social networks, as well as her own consumption of girl-specific goods (such as soaps, mobile phone airtime, clothing, braids, sodas and alcoholic drinks, etc.).

Experimental public transit transport performance statistics by 200 metre grids for a subset of urban centres in Great Britain, with the following fields (Note: These data are experimental, please see the Methods and Known Limitations/Caveats Sections for more details).AttributeDescriptionidUnique IdentifierpopulationGlobal Human Settlement Layer population estimate downsampled to 200 metre (represents the total population across adjacent 100 metre cells)access_popThe total population that can reach the destination cell within 45 minutes using the public transit network (origins within 11.25 kilometres of the destination cell)proxim_popThe total population within an 11.25 kilometre radius of the destination celltrans_perfThe transport performance of the 200 metre cell. The percentage ratio of accessible to proximal populationcity_nmName of the urban centrecountry_nmName of the country that the urban centre belongs toMethods:

For more information please visit:

· Python Package: https://github.com/datasciencecampus/transport-network-performance

· Docker Image: https://github.com/datasciencecampus/transport-performance-docker

Known Limitations/Caveats:

These data are experimental – see the ONS guidance on experimental statistics for more details. They are being published at this early stage to involve potential users and stakeholders in assessing their quality and suitability. The known caveats and limitations of these experimental statistics are summarised below.

Urban Centre and Population Estimates:

· Population estimates are derived from data using a hybrid method of satellite imagery and national censuses. The alignment of national census boundaries to gridded estimates introduce measurement errors, particularly in newer housing and built-up developments. See section 2.5 of the GHSL technical report release 2023A for more details.

Public Transit Schedule Data (GTFS):

· Does not include effects due to delays (such as congestion and diversions).

· Common GTFS issues are resolved during preprocessing where possible, including removing trips with unrealistic fast travel between stops, cleaning IDs, cleaning arrival/departure times, route name deduplication, dropping stops with no stop times, removing undefined parent stations, and dropping trips, shapes, and routes with no stops. Certain GTFS cleaning steps were not possible in all instances, and in those cases the impacted steps were skipped. Additional work is required to further support GTFS validation and cleaning.

Transport Network Routing:

· “Trapped” centroids: the centroid of destination cells on very rare occasions falls on a private road/pathway. Routing to these cells cannot be performed. This greatly decreases the transport performance in comparison with the neighbouring cells. Potential solutions include interpolation based on neighbouring cells or snapping to the nearest public OSM node (and adjusting the travel time accordingly). Further development to adapt the method for this consideration is necessary.

Please also visit the Python package and Docker Image GitHub issues pages for more details.

How to Contribute:

We hope that the public, other public sector organisations, and National Statistics Institutions can collaborate and build on these data, to help improve the international comparability of statistics and enable higher frequency and more timely comparisons. We welcome feedback and contribution either through GitHub or by contacting datacampus@ons.gov.uk.

Experimental public transit transport performance statistics by 200 metre grids for a subset of urban centres in France, with the following fields (Note: These data are experimental, please see the Methods and Known Limitations/Caveats Sections for more details).AttributeDescriptionidUnique IdentifierpopulationGlobal Human Settlement Layer population estimate downsampled to 200 metre (represents the total population across adjacent 100 metre cells)access_popThe total population that can reach the destination cell within 45 minutes using the public transit network (origins within 11.25 kilometres of the destination cell)proxim_popThe total population within an 11.25 kilometre radius of the destination celltrans_perfThe transport performance of the 200 metre cell. The percentage ratio of accessible to proximal populationcity_nmName of the urban centrecountry_nmName of the country that the urban centre belongs toMethods:

For more information please visit:

· Python Package: https://github.com/datasciencecampus/transport-network-performance

· Docker Image: https://github.com/datasciencecampus/transport-performance-docker

Known Limitations/Caveats:

These data are experimental – see the ONS guidance on experimental statistics for more details. They are being published at this early stage to involve potential users and stakeholders in assessing their quality and suitability. The known caveats and limitations of these experimental statistics are summarised below.

Urban Centre and Population Estimates:

· Population estimates are derived from data using a hybrid method of satellite imagery and national censuses. The alignment of national census boundaries to gridded estimates introduce measurement errors, particularly in newer housing and built-up developments. See section 2.5 of the GHSL technical report release 2023A for more details.

Public Transit Schedule Data (GTFS):

· Does not include effects due to delays (such as congestion and diversions).

· Common GTFS issues are resolved during preprocessing where possible, including removing trips with unrealistic fast travel between stops, cleaning IDs, cleaning arrival/departure times, route name deduplication, dropping stops with no stop times, removing undefined parent stations, and dropping trips, shapes, and routes with no stops. Certain GTFS cleaning steps were not possible in all instances, and in those cases the impacted steps were skipped. Additional work is required to further support GTFS validation and cleaning.

Transport Network Routing:

· “Trapped” centroids: the centroid of destination cells on very rare occasions falls on a private road/pathway. Routing to these cells cannot be performed. This greatly decreases the transport performance in comparison with the neighbouring cells. Potential solutions include interpolation based on neighbouring cells or snapping to the nearest public OSM node (and adjusting the travel time accordingly). Further development to adapt the method for this consideration is necessary.

Please also visit the Python package and Docker Image GitHub issues pages for more details.

How to Contribute:

We hope that the public, other public sector organisations, and National Statistics Institutions can collaborate and build on these data, to help improve the international comparability of statistics and enable higher frequency and more timely comparisons. We welcome feedback and contribution either through GitHub or by contacting datacampus@ons.gov.uk.

Our model is a full-annual-cycle population model {hostetler2015full} that tracks groups of bat surviving through four seasons: breeding season/summer, fall migration, non-breeding/winter, and spring migration. Our state variables are groups of bats that use a specific maternity colony/breeding site and hibernaculum/non-breeding site. Bats are also accounted for by life stages (juveniles/first-year breeders versus adults) and seasonal habitats (breeding versus non-breeding) during each year, This leads to four states variable (here depicted in vector notation): the population of juveniles during the non-breeding season, the population of adults during the non-breeding season, the population of juveniles during the breeding season, and the population of adults during the breeding season, Each vector's elements depict a specific migratory pathway, e.g., is comprised of elements, {non-breeding sites}, {breeding sites}The variables may be summed by either breeding site or non-breeding site to calculate the total population using a specific geographic location. Within our code, we account for this using an index column for breeding sites and an index column for non-breeding sides within the data table. Our choice of state variables caused the time step (i.e. (t)) to be 1 year. However, we recorded the population of each group during the breeding and non-breeding season as an artifact of our state-variable choice. We choose these state variables partially for their biological information and partially to simplify programming. We ran our simulation for 30 years because the USFWS currently issues Indiana Bat take permits for 30 years. Our model covers the range of the Indiana Bat, which is approximately the eastern half of the contiguous United States (Figure \ref{fig:BatInput}). The boundaries of our range was based upon the United States boundary, the NatureServe Range map, and observations of the species. The maximum migration distance was 500-km, which was based upon field observations reported in the literature \citep{gardner2002seasonal, winhold2006aspects}. The landscape was covered with approximately 33,000, 6475-ha grid cells and the grid size was based upon management considerations. The U.S.~Fish and Wildlife Service considers a 2.5 mile radius around a known maternity colony to be its summer habitat range and all of the hibernaculum within a 2.5 miles radius to be a single management unit. Hence the choice of 5-by-5 square grids (25 miles(^2) or 6475 ha). Each group of bats within the model has a summer and winter grid cell as well as a pathway connecting the cells. It is possible for a group to be in the cell for both seasons, but improbable for females (which we modeled). The straight line between summer and winter cells were buffered with different distances (1-km, 2-km, 10-km, 20-km, 100-km, and 200-km) as part of the turbine sensitivity and uncertainty analysis. We dropped the largest two buffer sizes during the model development processes because they were biologically unrealistic and including them caused all populations to go extinct all of the time. Note a 1-km buffer would be a 2-km wide path. An example of two pathways are included in Figure \ref{fig:BatPath}. The buffers accounts for bats not migrating in a straight line. If we had precise locations for all summer maternity colonies, other approaches such as Circuitscape \citep{hanks2013circuit} could have been used to model migration routes and this would have reduced migration uncertainty.

These data were compiled to help understand how climate change may impact dryland pinyon-juniper ecosystems in coming decades, and how resource management might be able to minimize those impacts. Objective(s) of our study were to model the demographic rates of PJ woodlands to estimate the areas that may decline in the future vs. those that will be stable. We quantified populations growth rates across broad geographic areas, and identified the relative roles of recruitment and mortality in driving potential future changes in population viability in 5 tree species that are major components of these dry forests. We used this demographic model to project pinyon-juniper population stability under future climate conditions, assess how robust these projected changes are, and to identify where on the landscape management strategies that decrease tree competition would effectively resist population decline. These data represent estimated recruitment, mortality and population growth across the distribution of five common pinyon-juniper species across the US Southwest. These data were collected by the US Forest service in their monitoring program, which is a systematic survey of forested regions across the entire US. Our data is from western US states, including AZ, CA, CO, ID, MT, NM, ND, NV, OR, SD, TX, UT, and was collected between 2000-2007, depending on state census collection times. These data were collected by the Forest Inventory and Analysis program of the USDA US Forest Service. Within each established plot, all adult trees greater than 12.7 cm (5 in.) diameter at breast height (DBH) are assigned unique tags and tracked within four, 7.32 m (24 ft.) radius subplots. All saplings <12.7 cm & > 2.54 cm (1 in.) DBH are assigned unique tags and tracked within four, 2.07 m (6.8 ft.) radius microplots within the larger adult plots. Finally, seedlings <2.54 cm DBH are counted within the same microplots as the saplings. Two censuses were conducted 10 years apart in each plot. These data can be used to inform how tree species have unique responses to changing climate conditions and how management actions, like tree density reduction, may effectively resist transformation away from pinyon-juniper woodland to other ecosystem types.

At the 70th session of the United Nations in 2015 September 25 the General Assembly resolution "Transforming our world: The 2030 Agenda for Sustainable Development" was adopted. One of the indicators of sustainable development prepared at the detailed level of 1 km grids: 9.1.1. Share of the rural population living within a 2 km radius of a year-round road.

Number of trap-sites, detection area (D.A.) angle and radius, trap-rates, and density estimates for each stratum, used for the calculation of roe deer density estimated by REM in the Italian Apennines.

Geographic proximity to service centres and population centres is an important determinant of socio-economic and health outcomes. Consequently, it is a relevant dimension in the analysis and delivery of policies and programs. To measure this dimension, Statistics Canada developed an Index of Remoteness of communities. For each populated community (census subdivision), the index is determined by its distance to all the population centres defined by Statistics Canada in a given travel radius, as well as their population size.

Aim: Invasive plants may evolve a suite of distinctive traits during spread in the new range. Among these traits, dispersal ability is an important trait determining the invasion speed of exotic plants. There is evidence that higher dispersal ability is favored at the invasion front, where population density may be low. However, no study has explicitly tested how planting density in a common garden affects the dispersal ability of invasive plants. Location: Hainan island of China. Methods: In this study, using 27 populations of an invasive plant, Mikania micrantha, which is expanding its range on Hainan island of China, we examine how three dispersal-related traits (i.e., dispersal ability, fruit mass, and pappus radius) change with distance from invasion centre and field population density, and how planting density in a common garden affects dispersal traits. Results: Dispersal traits did not change with distance from the invasion centre and field population cover either in the natural..., , , # This is the raw data showing the population information, treatments, and dispersal traits of the 27 Mikania micrantha populations.

Description of the data and file structure

In the first page, data of the common garden experiment are shown. These include block, planting density treatments, pappus radius, fruit mass, and area-mass ratio (AMR). A 'Novalue' cell indicates that the corresponding plants did not set any fruits. In the second page, data in the natural environment are shown. These include pappus radius, fruit mass, and area-mass ratio (AMR). In each page, the population origin (western or eastern), distance from invasion center, field cover, and frequency of field occurance of each of the 27 populations are also shown.

JFMA cases and the covariates tested and used in our modelThis data has all variables used in the statistical model as they entered the generalized linear model and the generalized linear mixed model. The variables included are (in the order they appear): year, kebeleID, JFMA total cases, log expected cases, scaled log ratio of SOND cases to the expected SOND cases, scaled DJF mean temperature in degree Celsius, scaled DJF total rainfall in mm, scaled population density from overlapping circles of 5km radius, scaled population density from overlapping circles of 10km radius, scaled weighted distance to roads, scaled inverse square distance to perennial water bodies, scaled average soil water holding capacity, scaled average slope, scaled average NDVI, scaled SST anomalies from the Nino 3.4 region, and IRS status (0/1).covariates_std.csvCount of neighboring kebelesThis data set contains the count of kebeles neighboring each kebele. This file should be used in combination with the Nieghbo...

In a time of global change, having an understanding of the nature of biotic and abiotic factors that drive a species’ range may be the sharpest tool in the arsenal of conservation and management of threatened species. However, such information is lacking for most tropical and epiphytic species due to the complexity of life history, the roles of stochastic events, and the diversity of habitat across the span of a distribution. In this study, we conducted repeated censuses across the core and peripheral range of Trichocentrum undulatum, a threatened orchid that is found throughout the island of Cuba (species core range) and southern Florida (the northern peripheral range). We used demographic matrix modeling as well as stochastic simulations to investigate the impacts of herbivory, hurricanes, and logging (in Cuba) on projected population growth rates (? and ?s) among sites. Methods Field methods Censuses took place between 2013 and 2021. The longest census period was that of the Peripheral population with a total of nine years (2013–2021). All four populations in Cuba used in demographic modeling that were censused more than once: Core 1 site (2016–2019, four years), Core 2 site (2018–2019, two years), Core 3 (2016 and 2018 two years), and Core 4 (2018–2019, two years) (Appendix S1: Table S1). In November 2017, Hurricane Irma hit parts of Cuba and southern Florida, impacting the Peripheral population. The Core 5 population (censused on 2016 and 2018) was small (N=17) with low survival on the second census due to logging. Three additional populations in Cuba were visited only once, Core 6, Core 7, and Core 8 (Table 1). Sites with one census or with a small sample size (Core 5) were not included in the life history and matrix model analyses of this paper due to the lack of population transition information, but they were included in the analysis on the correlation between herbivory and fruit rate, as well as the use of mortality observations from logging for modeling. All Cuban sites were located between Western and Central Cuba, spanning four provinces: Mayabeque (Core 1), Pinar del Rio (Core 2 and Core 6), Matanzas (Core 3 and Core 5), and Sancti Spiritus (Core 4, Core 7, Core 8). At each population of T. undulatum presented in this study, individuals were studied within ~1-km strips where T. undulatum occurrence was deemed representative of the site, mostly occurring along informal forest trails. Once an individual of T. undulatum was located, all trees within a 5-m radius were searched for additional individuals. Since tagging was not permitted, we used a combination of information to track individual plants for the repeated censuses. These include the host species, height of the orchid, DBH of the host tree, and hand-drawn maps. Individual plants were also marked by GPS at the Everglades Peripheral site. If a host tree was found bearing more than one T. undulatum, then we systematically recorded the orchids in order from the lowest to highest as well as used the previous years’ observations in future censuses for individualized notes and size records. We recorded plant size and reproductive variables during each census including: the number of leaves, length of the longest leaf (cm), number of inflorescence stalks, number of flowers, and the number of mature fruits. We also noted any presence of herbivory, such as signs of being bored by M. miamensis, and whether an inflorescence was partially or completely affected by the fly, and whether there was other herbivory, such as D. boisduvalii on leaves. We used logistic regression analysis to examine the effects of year (at the Peripheral site) and sites (all sites) on the presence or absence of inflorescence herbivory at all the sites. Cross tabulation and chi-square analysis were done to examine the associations between whether a plant was able to fruit and the presence of floral herbivory by M. miamensis. The herbivory was scored as either complete or partial. During the orchid population scouting expeditions, we came across a small population in the Matanzas province (Core 5, within 10 km of the Core 3 site) and recorded the demographic information. Although the sampled population was small (N = 17), we were able to observe logging impacts at the site and recorded logging-associated mortality on the subsequent return to the site. Matrix modeling Definition of size-stage classes To assess the life stage transitions and population structures for each plant for each population’s census period we first defined the stage classes for the species. The categorization for each plant’s stage class depended on both its size and reproductive capabilities, a method deemed appropriate for plants (Lefkovitch 1965, Cochran and Ellner 1992). A size index score was calculated for each plant by taking the total number of observed leaves and adding the length of the longest leaf, an indication of accumulated biomass (Borrero et al. 2016). The smallest plant size that attempted to produce an inflorescence is considered the minimum size for an adult plant. Plants were classified by stage based on their size index and flowering capacity as the following: (1) seedlings (or new recruits), i.e., new and small plants with a size index score of less than 6, (2) juveniles, i.e., plants with a size index score of less than 15 with no observed history of flowering, (3) adults, plants with size index scores of 15 or greater. Adult plants of this size or larger are capable of flowering but may not produce an inflorescence in a given year. The orchid’s population matrix models were constructed based on these stages. In general, orchid seedlings are notoriously difficult to observe and easily overlooked in the field due to the small size of protocorms. A newly found juvenile on a subsequent site visit (not the first year) may therefore be considered having previously been a seedling in the preceding year. In this study, we use the discovered “seedlings” as indicatory of recruitment for the populations. Adult plants are able to shrink or transition into the smaller juvenile stage class, but a juvenile cannot shrink to the seedling stage. Matrix elements and population vital rates calculations Annual transition probabilities for every stage class were calculated. A total of 16 site- and year-specific matrices were constructed. When seedling or juvenile sample sizes were < 9, the transitions were estimated using the nearest year or site matrix elements as a proxy. Due to the length of the study and variety of vegetation types with a generally large population size at each site, transition substitutions were made with the average stage transition from all years at the site as priors. If the sample size of the averaged stage was still too small, the averaged transition from a different population located at the same vegetation type was used. We avoided using transition values from populations found in different vegetation types to conserve potential environmental differences. A total of 20% (27/135) of the matrix elements were estimated in this fashion, the majority being seedling stage transitions (19/27) and noted in the Appendices alongside population size (Appendix S1: Table S1). The fertility element transitions from reproductive adults to seedlings were calculated as the number of seedlings produced (and that survived to the census) per adult plant. Deterministic modeling analysis We used integral projection models (IPM) to project the long-term population growth rates for each time period and population. The finite population growth rate (?), stochastic long-term growth rate (?s), and the elasticity were projected for each matrices using R Popbio Package 2.4.4 (Stubben and Milligan 2007, Caswell 2001). The elasticity matrices were summarized by placing each element into one of three categories: fecundity (transition from reproductive adults to seedling stage), growth (all transitions to new and more advanced stage, excluding the fecundity), and stasis (plants that transitioned into the same or a less advanced stage on subsequent census) (Liu et al. 2005). Life table response experiments (LTREs) were conducted to identify the stage transitions that had the greatest effects on observed differences in population growth between select sites and years (i.e., pre-post hurricane impact and site comparisons of same vegetation type). Due to the frequent disturbances that epiphytes in general experience as well as our species’ distribution in hurricane-prone areas, we ran transient dynamic models that assume that the populations censused were not at stable stage distributions (Stott et al. 2011). We calculated three indices for short-term transient dynamics to capture the variation during a 15-year transition period: reactivity, maximum amplification, and amplified inertia. Reactivity measures a population’s growth in a single measured timestep relative to the stable-stage growth, during the simulated transition period. Maximum amplification and amplified inertia are the maximum of future population density and the maximum long-term population density, respectively, relative to a stable-stage population that began at the same initial density (Stott et al. 2011). For these analyses, we used a mean matrix for Core 1, Core 2 Core 3, and Core 4 sites and the population structure of their last census. For the Peripheral site, we averaged the last three matrices post-hurricane disturbance and used the most-recent population structure. We standardized the indices across sites with the assumption of initial population density equal to 1 (Stott et al. 2011). Analysis was done using R Popdemo version 1.3-0 (Stott et al. 2012b). Stochastic simulation We created matrices to simulate the effects of episodic recruitment, hurricane impacts, herbivory, and logging (Appendix S1: Table S2). The Peripheral population is the longest-running site with nine years of censuses (eight

Characteristics of included infants from the BILD cohort.