Many studies have quantified the distribution of heterozygosity and relatedness in natural populations, but few have examined the demographic processes driving these patterns. In this study, we take a novel approach by studying how population structure affects both pairwise identity and the distribution of heterozygosity in a natural population of the self-incompatible plant Antirrhinum majus. Excess variance in heterozygosity between individuals is due to identity disequilibrium, which reflects the variance in inbreeding between individuals; it is measured by the statistic g(2). We calculated g(2) together with F-ST and pairwise relatedness (F-ij) using 91 SNPs in 22,353 individuals collected over 11 years. We find that pairwise F-ij declines rapidly over short spatial scales, and the excess variance in heterozygosity between individuals reflects significant variation in inbreeding. Additionally, we detect an excess of individuals with around half the average heterozygosity, indicating either selfing or matings between close relatives. We use 2 types of simulation to ask whether variation in heterozygosity is consistent with fine-scale spatial population structure. First, by simulating offspring using parents drawn from a range of spatial scales, we show that the known pollen dispersal kernel explains g(2). Second, we simulate a 1,000-generation pedigree using the known dispersal and spatial distribution and find that the resulting g(2) is consistent with that observed from the field data. In contrast, a simulated population with uniform density underestimates g(2), indicating that heterogeneous density promotes identity disequilibrium. Our study shows that heterogeneous density and leptokurtic dispersal can together explain the distribution of heterozygosity.

13 The distribution of heterozygosity in a population is commonly used to quantify 14 inbreeding depression through the use of heterozygosity-fitness correlations (HFCs), but 15 the demographic processes shaping variability in inbreeding are not well understood. For 16 11 years, we measured heterozygosity and six fitness proxies in a population of the self17 incompatible plant Antirrhinum majus. Using a panel of 91 SNPs in 22,353 individuals, 18 we find that relatedness declines rapidly over short spatial scales. Excess variance in 19 heterozygosity between individuals (identity disequilibrium, g2) reflects significant 20 variation in inbreeding. We use two types of simulation to ask whether variation in 21 heterozygosity is consistent with fine-scale spatial patterns of mating and population 22 structure. First, we simulate offspring from matings with fathers at different distances, 23 showing that the pollen dispersal kernel affects expected g2. Second, we simulate a 100024 generation pedigree using the known dispersal and spatial distribution and find that g2 is 25 consistent with that observed. In contrast, a simulated population with uniform density 26 underestimates g2, indicating that heterogeneous density promotes identity 27 disequilibrium. Finally, we estimate inbreeding depression through HFCs. Only 28 flowering stem count increases with heterozygosity. Our study shows that heterogeneous 29 density and leptokurtic dispersal can together explain pairwise FST and the distribution of 30 heterozygosity. 31

Inbreeding depression can be estimated by correlating heterozygosity with fitness components, but such heterozygosity-fitness correlations are typically weak. For over ten years, we studied a population of the self-incompatible plant, Antirrhinum majus, measuring heterozygosity and fitness proxies from 22,353 plants. Using a panel of 91 SNPs, we find that relatedness declines rapidly over short spatial scales. Individual heterozygosity varies more between individuals than expected, reflecting identity disequilibrium (g2) due to variation in inbreeding – a prerequisite for detecting inbreeding depression. We use two types of simulations to ask whether the heterozygosity distribution is consistent with spatially structured mating. First, we simulate offspring from matings with fathers at different distances and find that the distribution of heterozygosity in the field data is consistent with the measured pollen dispersal kernel. Second, we simulate a 1000-generation pedigree using the known spatial distribution, and find that identity disequilibrium, though highly variable between simulations, is consistent with that observed. Finally, we estimate inbreeding depression through the relationships between heterozygosity and six fitness proxies. Only the number of flowering stems is predicted by heterozygosity. Our approach provides a novel example of how long-term studies can elucidate population structure and fitness variation in the wild.

The biotic interactions hypothesis posits that biotic interactions are more important drivers of adaptation closer to the equator, evidenced by "stronger" contemporary interactions (e.g., greater interaction rates) and/or patterns of trait evolution consistent with a history of stronger interactions. Support for the hypothesis is mixed, but few studies span tropical and temperate regions while experimentally controlling for evolutionary history. Here, we integrate field observations and common garden experiments to quantify the relative importance of pollination and herbivory in a pair of tropical-temperate congeneric perennial herbs. Phytolacca rivinoides and P. americana are pioneer species native to the Neotropics and the eastern United States, respectively. We compared plant-pollinator and plant-herbivore interactions between three tropical populations of P. rivinoides from Costa Rica and three temperate populations of P. americana from its northern range edge in Michigan and Ohio. For some metrics of interaction importance, we also included three subtropical populations of P. americana from its southern range edge in Florida. This approach confounds species and region but allows us, uniquely, to measure complementary proxies of interaction importance across a tropical-temperate range in one system. To test the prediction that lower-latitude plants are more reliant on insect pollinators, we quantified floral display and reward, insect visitation rates, and self-pollination ability (autogamy). To test the prediction that lower-latitude plants experience more herbivore pressure, we quantified herbivory rates, herbivore abundance, and leaf palatability. We found evidence supporting the biotic interactions hypothesis for most comparisons between P. rivinoides and north-temperate P. americana (floral display, insect visitation, autogamy, herbivory, herbivore abundance, and young-leaf palatability). Results for subtropical P. americana populations, however, were typically not intermediate between P. rivinoides and north-temperate P. americana, as would be predicted by a linear latitudinal gradient in interaction importance. Subtropical young-leaf palatability was intermediate, but subtropical mature leaves were the least palatable, and pollination-related traits did not differ between temperate and subtropical regions. These nonlinear patterns of interaction importance suggest future work to link interaction importance to climatic or biotic thresholds. In sum, we found that the biotic interactions hypothesis was more consistently supported at the larger spatial scale of our study.

The longstanding biotic interactions hypothesis predicts that herbivore pressure declines with latitude, but the evidence is mixed. To address gaps in previous studies, we measured herbivory and defence in the same system, quantified defence with bioassays, and considered effects of leaf age. We quantified herbivory and defence of young and mature leaves along a continental gradient in eastern North America in the native herb Phytolacca americana L. Herbivory in the field declined with latitude and was strongly correlated with lepidopteran abundance. Laboratory bioassays revealed that leaf palatability was positively correlated with latitude of origin. Young leaves were more damaged than mature leaves at lower latitudes in the field, but less palatable in bioassays. Both defence and palatability displayed non-linear latitudinal patterns, suggesting potential mechanisms based on biological or climatic thresholds. In sum, observational and experimental studies find patterns consistent with high herbivore pressure and stronger plant defences at lower latitudes.

Both herbivory and plant defenses against herbivores have been predicted to increase toward tropical regions. Early tests of this latitudinal herbivory-defense hypothesis (LHDH) were supportive, but accumulating evidence has been mixed. We argue that the lack of clarity might be due to heterogeneity in methodology and problems with study design and interpretation. We suggest possible solutions. Latitudinal studies need to carefully consider spatial and phylogenetic scale, to link plant defense measurements to herbivore performance, and to incorporate additional concepts from plant defense theory such as tolerance and induced defense. In addition, we call for consistent measures of herbivory to standardize comparisons across biomes. Improving methodology in future studies of LHDH should resolve much of the current controversy.

An indirect consequence of plant invasions is the disruption of native plant-pollinator interactions. We examined effects of invasive baby's breath (Gypsophila paniculata) and spotted knapweed (Centaurea maculosa) on floral visitors to federally threatened Pitcher's thistle (Cirsium pitcheri) in Lake Michigan dunes. In sweep net surveys, abundances of pollinator taxa were five times higher in invaded than in naturally invader-free sites. However, plot-level G. paniculata removal treatments increased pollinator visits to C. pitcheri relative to invaded plots and restored visitation to levels found in naturally uninvaded plots. Invader removal also increased native plant species richness, which was positively correlated with pollinator visitation to C. pitcheri, suggesting an indirect effect on pollinators mediated through invader-altered plant composition. In temporary floral arrays, the rate of pollinator visitation to C. pitcheri was not affected by neighbor plant species identity. However, compared with native Monarda punctata, invasive C. maculosa attracted more total pollinators to the array but reduced the proportion of total visits that were to C. pitcheri and increased pollinator movements between plant species. While both G. paniculata and C. maculosa appear to act as magnet species by attracting more pollinators at the plot level, these invaders have the potential to reduce reproduction of C. pitcheri by decreasing pollinator visits and increasing interspecific pollen transfer.