Seeing spots: Measuring, quantifying heritability, and assessing fitness consequences of coat pattern traits in a wild population of giraffes (Giraffa camelopardalis)

Polymorphic phenotypes of mammalian coat color have been important to the study of genetics and evolution, but little is known about the heritability and fitness consequences of variation in complex coat pattern traits in wild populations. Understanding the current evolution of coat patterns requires reliably measuring traits, quantifying heritability of the traits, and identifying the fitness consequences of specific phenotypes. Giraffe coat markings are highly variable and it has been hypothesized that variation in coat patterns most likely affects fitness by camouflaging neonates against predators. We quantified spot pattern traits of wild Masai giraffes using image analysis software, determined whether spot pattern traits were heritable, and assessed whether variation in heritable spot pattern traits was related to fitness as measured by juvenile survival. The methods we described comprise a framework for objective quantification of mammalian coat pattern traits based on photographic coat pattern data. We demonstrated that characteristics of giraffe coat spot shape are heritable. We did not find evidence for juvenile survival consequences of variation in spot traits, suggesting that spot traits are currently not under strong directional, disruptive, or stabilizing selection for neonate camouflage in our study population, but our sample size could not detect small differences in survival. Spot trait variation also may be more relevant to other components of fitness, such as adult survival or fecundity. We hope this case study will inspire further investigations of coat pattern traits.


7
Quantification of Spot Patterns 1 1 8 We analysed spot traits of each animal within the shoulder and rib area by cropping all 1 1 9 images to an analysis rectangle that fit horizontally between the anterior edge of the rear leg 1 2 0 and the chest, and vertically between the back and where the skin folded beneath the posterior 1 2 1 edge of the foreleg (Fig 1). We quantified spot characteristics of each animal's pattern using analysis rectangle (Fig 1). We excluded particles cut off by the edge of the analysis rectangle 1 3 0 to avoid the influence of incomplete spots, and we also excluded spots whose area was We characterized each animal's spot pattern traits within the analysis rectangle using 1 3 3 the following 11 metrics available in ImageJ: number of spots; mean spot size (area); mean 1 3 4 spot perimeter; mean angle between the primary axis of an ellipse fit over the spot and the x-   unable to identify fathers, and our sample did not include any maternal siblings or half-1 6 4 siblings.

6 5
We identified mother-calf pairs by observing extended suckling behavior. Wild 1 6 6 female giraffes very rarely suckle a calf that is not their own (Pratt and Anderson 1979). We 1 6 7 examined all identification photographs for individuals in known mother-calf pairs, and 1 6 8 selected the best-quality photograph for each animal based on focus, clarity, perpendicularity 1 6 9 to the camera, and unobstructed view of the torso. We found 31 known mother-calf pairs with 1 7 0 high-quality photographs of both animals.

7 1
We predicted spot pattern traits of a calf would be correlated with those of its mother 1 7 2 but not with a random cow. For comparison of spot characteristics between known mother-1 7 3 calf pairs, we created a null expectation set of random cow-calf pairs using the same 1 7 4 photographs by assigning a random mother to each calf (without replacement and without 1 7 5 pairing a mother with her own calf). We tested our prediction for each spot characteristic  lack of fit (߯ ଶ ଶ = 97, P = 0.01), but we felt this was largely due to lack of age effects in the 2 0 5 goodness-of-fit tested model, whereas age effects were included in our model selection and We were able to quantify 11 spot traits using ImageJ, and found the traits with greatest 2 1 1 individual heterogeneity as measured by the CV were number of spots and area of spots 2 1 2 (these two traits were negatively correlated), and mode shade (

1 6
We found no spot pattern traits that had significant PO regression coefficients 2 1 7 between calves and random cows, but two characters, circularity and solidity (tortuousness) 2 1 8 (Fig 2) were significantly correlated between calves and their mothers indicating heritability 2 1 9 (  0.80 with r-squared = 0.125, the sample size would need to have been 56 mother-calf pairs.

5
Our survival analysis of 258 calves first encountered as neonates indicated there was 2 2 6 no evidence that individual covariates of spot traits significantly affected survival during the 2 2 7 first season of life, but model selection uncertainty was high (

3 3
Our photographic capture-recapture dataset spanning multiple years enabled us to objectively 2 3 4 and reliably quantify coat pattern traits of wild giraffes using image analysis software, and 2 3 5 demonstrate that giraffe coat pattern traits of spot shape are heritable from mother to calf. The 2 3 6 methods we described should serve as a framework for objective quantification of 2 3 7 mammalian coat pattern traits, and could also be useful for taxonomic or phenotypic new, robust dataset to address taxonomic and evolutionary hypotheses.

6 0
Our analyses highlighted two aspects of giraffe spots that were most heritable and 2 6 1 which may have adaptive significance. Circularity describes how close the spot is to a perfect 2 6 2 circle, and solidity describes how smooth and entire the edges are versus tortuous, ruffled, 2 6 3 lobed, or incised. These two characteristics could form the basis for quantifying spot patterns 2 6 4 of giraffes across Africa, and gives field workers a new quantitative lexicon for describing 2 6 5 spots (Fig 2). Our mode shade measurement was a crude metric, and color is greatly affected 2 6 6 by lighting conditions, so we suggest standardization of photographic methods to control for 2 6 7 lighting if color is to be analyzed in future studies.

6 8
One possible explanation for the lack of juvenile survival effects from spot variation 2 6 9 in our analysis is the recent reduction in large predator density in our study area (Packer et al.

7 4
Alternatively, the possibility remains that spot traits may serve adaptive functions such as inbreeding avoidance also could play a role in the evolution of spot patterns in giraffes 2 8 0 (Beecher 1982;Tibbetts and Dale 2007;Sherman et al. 1997 certain environments, or our measurements of spot traits may not accurately reflect the 2 9 0 aspects of spot patterns most important to adaptation for neonatal crypsis. Other aspects of 2 9 1 spot variation may prove to be more relevant to fitness, such as social effects of individual 2 9 2 recognition or kin recognition, or thermoregulation, and deserve further investigation. We 2 9 3 may also have simply lacked sufficient sample sizes necessary to detect small adaptation or 2 9 4 fitness effects. Small and very difficult to detect differences in fitness may have profound 2 9 5 influence on phenotypic evolution compounded over thousands of generations (Orr 2009).

9 6
Patterned coats of mammals are hypothesized to be formed by two distinct processes: 2 9 7 a spatially oriented developmental mechanism that creates a species-specific pattern of skin 2 9 8 cell differentiation and a pigmentation-oriented mechanism that uses information from the  We have deposited the primary data underlying these analyses as follows: Utilities for performing goodness of fit tests and manipulating CApture-REcapture data.

5 9
The blue rectangle shows the area analysed using ImageJ to characterize spot pattern 4 6 0 traits.